Release PHP-Parser 3.0.3

Conflicts:
	lib/PhpParser/NodeVisitor.php

.gitignore

@@ -0,0 +1,4 @@
+vendor/
+composer.lock
+grammar/kmyacc.exe
+grammar/y.output

.travis.yml

@@ -1,14 +1,31 @@
 language: php
 
+sudo: false
+
+cache:
+  directories:
+    - $HOME/.composer/cache
+
 php:
-  - 5.3
-  - 5.4
   - 5.5
   - 5.6
   - 7.0
+  - nightly
   - hhvm
 
+install:
+  - if [ $TRAVIS_PHP_VERSION = '5.6' ]; then composer require satooshi/php-coveralls '~1.0'; fi
+  - composer install --prefer-dist
+
 matrix:
   allow_failures:
-    - php: 7.0
+    - php: nightly
   fast_finish: true
+
+script:
+  - if [ $TRAVIS_PHP_VERSION = '5.6' ]; then vendor/bin/phpunit --coverage-clover build/logs/clover.xml; else vendor/bin/phpunit; fi
+  - if [ $TRAVIS_PHP_VERSION = '7.0' ]; then test_old/run-php-src.sh; fi
+
+after_success:
+  if [ $TRAVIS_PHP_VERSION = '5.6' ]; then php vendor/bin/coveralls; fi
+

CHANGELOG.md

@@ -1,197 +1,316 @@
-Version 1.2.2-dev
+Version 3.0.4-dev
 -----------------
 
 Nothing yet.
 
-Version 1.2.1 (2015-03-24)
+Version 3.0.3 (2017-02-03)
 --------------------------
 
-* Fixed registration of the aliases introduced in 1.2.0. Previously the old class names could not be used in
-  `instanceof` checks under some circumstances.
+### Fixed
 
-Version 1.2.0 (2015-03-22)
+* In `"$foo[0]"` the `0` is now parsed as an `LNumber` rather than `String`. (#325)
+* Ensure integers and floats are always pretty printed preserving semantics, even if the particular
+  value can only be manually constructed.
+* Throw a `LogicException` when trying to pretty-print an `Error` node. Previously this resulted in
+  an undefined method exception or fatal error.
+
+### Added
+
+* [PHP 7.1] Added support for negative interpolated offsets: `"$foo[-1]"`
+* Added `preserveOriginalNames` option to `NameResolver`. If this option is enabled, an
+  `originalName` attribute, containing the unresolved name, will be added to each resolved name.
+* Added `php-parse --with-positions` option, which dumps nodes with position information.
+
+### Deprecated
+
+* The XML serializer has been deprecated. In particular, the classes `Serializer\XML`,
+  `Unserializer\XML`, as well as the interfaces `Serializer` and `Unserializer` are deprecated.
+
+Version 3.0.2 (2016-12-06)
+--------------------------
+
+### Fixed
+
+* Fixed name resolution of nullable types. (#324)
+* Fixed pretty-printing of nullable types.
+
+Version 3.0.1 (2016-12-01)
+--------------------------
+
+### Fixed
+
+* Fixed handling of nested `list()`s: If the nested list was unkeyed, it was directly included in
+  the list items. If it was keyed, it was wrapped in `ArrayItem`. Now nested `List_` nodes are
+  always wrapped in `ArrayItem`s. (#321)
+
+Version 3.0.0 (2016-11-30)
+--------------------------
+
+### Added
+
+* Added support for dumping node positions in the NodeDumper through the `dumpPositions` option.
+* Added error recovery support for `$`, `new`, `Foo::`.
+
+Version 3.0.0-beta2 (2016-10-29)
+--------------------------------
+
+This release primarily improves our support for error recovery.
+
+### Added
+
+* Added `Node::setDocComment()` method.
+* Added `Error::getMessageWithColumnInfo()` method.
+* Added support for recovery from lexer errors.
+* Added support for recovering from "special" errors (i.e. non-syntax parse errors).
+* Added precise location information for lexer errors.
+* Added `ErrorHandler` interface, and `ErrorHandler\Throwing` and `ErrorHandler\Collecting` as
+  specific implementations. These provide a general mechanism for handling error recovery.
+* Added optional `ErrorHandler` argument to `Parser::parse()`, `Lexer::startLexing()` and
+  `NameResolver::__construct()`.
+* The `NameResolver` now adds a `namespacedName` attribute on name nodes that cannot be statically
+  resolved (unqualified unaliased function or constant names in namespaces).
+  
+### Fixed
+
+* Fixed attribute assignment for `GroupUse` prefix and variables in interpolated strings.
+
+### Changed
+
+* The constants on `NameTraverserInterface` have been moved into the `NameTraverser` class.
+* Due to the error handling changes, the `Parser` interface and `Lexer` API have changed.
+* The emulative lexer now directly postprocesses tokens, instead of using `~__EMU__~` sequences.
+  This changes the protected API of the lexer.
+* The `Name::slice()` method now returns `null` for empty slices, previously `new Name([])` was
+  used. `Name::concat()` now also supports concatenation with `null`.
+
+### Removed
+
+* Removed `Name::append()` and `Name::prepend()`. These mutable methods have been superseded by
+  the immutable `Name::concat()`.
+* Removed `Error::getRawLine()` and `Error::setRawLine()`. These methods have been superseded by
+  `Error::getStartLine()` and `Error::setStartLine()`.
+* Removed support for node cloning in the `NodeTraverser`.
+* Removed `$separator` argument from `Name::toString()`.
+* Removed `throw_on_error` parser option and `Parser::getErrors()` method. Use the `ErrorHandler`
+  mechanism instead.
+
+Version 3.0.0-beta1 (2016-09-16)
+--------------------------------
+
+### Added
+
+* [7.1] Function/method and parameter builders now support PHP 7.1 type hints (void, iterable and
+  nullable types).
+* Nodes and Comments now implement `JsonSerializable`. The node kind is stored in a `nodeType`
+  property.
+* The `InlineHTML` node now has an `hasLeadingNewline` attribute, that specifies whether the
+  preceding closing tag contained a newline. The pretty printer honors this attribute.
+* Partial parsing of `$obj->` (with missing property name) is now supported in error recovery mode.
+* The error recovery mode is now exposed in the `php-parse` script through the `--with-recovery`
+  or `-r` flags.
+
+The following changes are also part of PHP-Parser 2.1.1:
+
+* The PHP 7 parser will now generate a parse error for `$var =& new Obj` assignments.
+* Comments on free-standing code blocks will now be retained as comments on the first statement in
+  the code block.
+
+Version 3.0.0-alpha1 (2016-07-25)
+---------------------------------
+
+### Added
+
+* [7.1] Added support for `void` and `iterable` types. These will now be represented as strings
+  (instead of `Name` instances) similar to other builtin types.
+* [7.1] Added support for class constant visibility. The `ClassConst` node now has a `flags` subnode
+  holding the visibility modifier, as well as `isPublic()`, `isProtected()` and `isPrivate()`
+  methods. The constructor changed to accept the additional subnode.
+* [7.1] Added support for nullable types. These are represented using a new `NullableType` node
+  with a single `type` subnode.
+* [7.1] Added support for short array destructuring syntax. This means that `Array` nodes may now
+  appear as the left-hand-side of assignments and foreach value targets. Additionally the array
+  items may now contain `null` values if elements are skipped.
+* [7.1] Added support for keys in list() destructuring. The `List` subnode `vars` has been renamed
+  to `items` and now contains `ArrayItem`s instead of plain variables.
+* [7.1] Added support for multi-catch. The `Catch` subnode `type` has been renamed to `types` and
+  is now an array of `Name`s.
+* `Name::slice()` now supports lengths and negative offsets. This brings it in line with
+  `array_slice()` functionality.
+
+### Changed
+
+Due to PHP 7.1 support additions described above, the node structure changed as follows:
+
+* `void` and `iterable` types are now stored as strings if the PHP 7 parser is used.
+* The `ClassConst` constructor changed to accept an additional `flags` subnode.
+* The `Array` subnode `items` may now contain `null` elements (destructuring).
+* The `List` subnode `vars` has been renamed to `items` and now contains `ArrayItem`s instead of
+  plain variables.
+* The `Catch` subnode `type` has been renamed to `types` and is now an array of `Name`s.
+
+Additionally the following changes were made:
+
+* The `type` subnode on `Class`, `ClassMethod` and `Property` has been renamed to `flags`. The
+  `type` subnode has retained for backwards compatibility and is populated to the same value as
+  `flags`. However, writes to `type` will not update `flags`.
+* The `TryCatch` subnode `finallyStmts` has been replaced with a `finally` subnode that holds an
+  explicit `Finally` node. This allows for more accurate attribute assignment.
+* The `Trait` constructor now has the same form as the `Class` and `Interface` constructors: It
+  takes an array of subnodes. Unlike classes/interfaces, traits can only have a `stmts` subnode.
+* The `NodeDumper` now prints class/method/property/constant modifiers, as well as the include and
+  use type in a textual representation, instead of only showing the number.
+* All methods on `PrettyPrinter\Standard` are now protected. Previoulsy most of them were public.
+
+### Removed
+
+* Removed support for running on PHP 5.4. It is however still possible to parse PHP 5.2-5.4 code
+  while running on a newer version.
+* The deprecated `Comment::setLine()` and `Comment::setText()` methods have been removed.
+* The deprecated `Name::set()`, `Name::setFirst()` and `Name::setLast()` methods have been removed.
+
+Version 2.1.1 (2016-09-16)
 --------------------------
 
 ### Changed
 
-* To ensure compatibility with PHP 7, the following node classes have been renamed:
+* The pretty printer will now escape all control characters in the range `\x00-\x1F` inside double
+  quoted strings. If no special escape sequence is available, an octal escape will be used.
+* The quality of the error recovery has been improved. In particular unterminated expressions should
+  be handled more gracefully.
+* The PHP 7 parser will now generate a parse error for `$var =& new Obj` assignments.
+* Comments on free-standing code blocks will no be retained as comments on the first statement in
+  the code block.
 
-        OLD                             => NEW
-        PhpParser\Node\Expr\Cast\Bool   => PhpParser\Node\Expr\Cast\Bool_
-        PhpParser\Node\Expr\Cast\Int    => PhpParser\Node\Expr\Cast\Int_
-        PhpParser\Node\Expr\Cast\Object => PhpParser\Node\Expr\Cast\Object_
-        PhpParser\Node\Expr\Cast\String => PhpParser\Node\Expr\Cast\String_
-        PhpParser\Node\Scalar\String    => PhpParser\Node\Scalar\String_
+Version 2.1.0 (2016-04-19)
+--------------------------
 
-  **The previous class names are still supported as aliases.** However it is strongly encouraged to use the new names
-  in order to make your code compatible with PHP 7.
+### Fixed
 
-* Subnodes are now stored using real properties instead of an array. This improves performance and memory usage of the
-  initial parse and subsequent node tree operations. The `NodeAbstract` class still supports the old way of specifying
-  subnodes, however this is *deprecated*. In any case properties that are assigned to a node after creation will no
-  longer be considered as subnodes.
-
-* Methods and property declarations will no longer set the `Stmt\Class_::MODIFIER_PUBLIC` flag if no visibility is
-  explicitly given. However the `isPublic()` method will continue to return true. This allows you to distinguish whether
-  a method/property is explicitly or implicitly public and control the pretty printer output more precisely.
-
-* The `Stmt\Class_`, `Stmt\Interface_` and `Stmt\Trait_` nodes now inherit from `Stmt\ClassLike`, which provides a
-  `getMethods()` method. Previously this method was only available on `Stmt\Class_`.
-
-* Support including the `bootstrap.php` file multiple times.
-
-* Make documentation and tests part of the release tarball again.
-
-* Improve support for HHVM and PHP 7.
+* Properly support `B""` strings (with uppercase `B`) in a number of places.
+* Fixed reformatting of indented parts in a certain non-standard comment style.
 
 ### Added
 
-* Added support for PHP 7 return type declarations. This adds an additional `returnType` subnode to `Stmt\Function_`,
-  `Stmt\ClassMethod` and `Expr\Closure`.
+* Added `dumpComments` option to node dumper, to enable dumping of comments associated with nodes.
+* Added `Stmt\Nop` node, that is used to collect comments located at the end of a block or at the
+  end of a file (without a following node with which they could otherwise be associated).
+* Added `kind` attribute to `Expr\Exit` to distinguish between `exit` and `die`.
+* Added `kind` attribute to `Scalar\LNumber` to distinguish between decimal, binary, octal and
+  hexadecimal numbers.
+* Added `kind` attribtue to `Expr\Array` to distinguish between `array()` and `[]`.
+* Added `kind` attribute to `Scalar\String` and `Scalar\Encapsed` to distinguish between
+  single-quoted, double-quoted, heredoc and nowdoc string.
+* Added `docLabel` attribute to `Scalar\String` and `Scalar\Encapsed`, if it is a heredoc or
+  nowdoc string.
+* Added start file offset information to `Comment` nodes.
+* Added `setReturnType()` method to function and method builders.
+* Added `-h` and `--help` options to `php-parse` script.
 
-* Added support for the PHP 7 null coalesce operator `??`. The operator is represented by `Expr\BinaryOp\Coalesce`.
+### Changed
 
-* Added support for the PHP 7 spaceship operator `<=>`. The operator is represented by `Expr\BinaryOp\Spaceship`.
+* Invalid octal literals now throw a parse error in PHP 7 mode.
+* The pretty printer takes all the new attributes mentioned in the previous section into account.
+* The protected `AbstractPrettyPrinter::pComments()` method no longer returns a trailing newline.
+* The bundled autoloader supports library files being stored in a different directory than
+  `PhpParser` for easier downstream distribution.
 
-* Added use builder.
+### Deprecated
 
-* Added global namespace support to the namespace builder.
+* The `Comment::setLine()` and `Comment::setText()` methods have been deprecated. Construct new
+  objects instead.
 
-* Added a constructor flag to `NodeTraverser`, which disables cloning of nodes.
+### Removed
 
-Version 1.1.0 (2015-01-18)
+* The internal (but public) method `Scalar\LNumber::parse()` has been removed. A non-internal
+  `LNumber::fromString()` method has been added instead.
+
+Version 2.0.1 (2016-02-28)
 --------------------------
 
-* Methods that do not specify an explicit visibility (e.g. `function method()`) will now have the `MODIFIER_PUBLIC`
-  flag set. This also means that their `isPublic()` method will return true.
+### Fixed
 
-* Declaring a property as abstract or final is now an error.
+* `declare() {}` and `declare();` are not semantically equivalent and will now result in different
+  ASTs. The format case will have an empty `stmts` array, while the latter will set `stmts` to
+  `null`.
+* Magic constants are now supported as semi-reserved keywords.
+* A shebang line like `#!/usr/bin/env php` is now allowed at the start of a namespaced file.
+  Previously this generated an exception.
+* The `prettyPrintFile()` method will not strip a trailing `?>` from the raw data that follows a
+  `__halt_compiler()` statement.
+* The `prettyPrintFile()` method will not strip an opening `<?php` if the file starts with a
+  comment followed by InlineHTML.
 
-* The `Lexer` and `Lexer\Emulative` classes now accept an `$options` array in their constructors. Currently only the
-  `usedAttributes` option is supported, which determines which attributes will be added to AST nodes. In particular
-  it is now possible to add information on the token and file positions corresponding to a node. For more details see
-  the [Lexer component](https://github.com/nikic/PHP-Parser/blob/master/doc/component/Lexer.markdown) documentation.
-
-* Node visitors can now return `NodeTraverser::DONT_TRAVERSE_CHILDREN` from `enterNode()` in order to skip all children
-  of the current node, for all visitors.
-
-* Added builders for traits and namespaces.
-
-* The class, interface, trait, function, method and property builders now support adding doc comments using the
-  `setDocComment()` method.
-
-* Added support for fully-qualified and namespace-relative names in builders. No longer allow use of name component
-  arrays.
-
-* Do not add documentation and tests to distribution archive files.
-
-Version 1.0.2 (2014-11-04)
+Version 2.0.0 (2015-12-04)
 --------------------------
 
-* The `NameResolver` visitor now also resolves names in trait adaptations (aliases and precedence declarations).
+### Changed
 
-* Remove stray whitespace when pretty-printing trait adaptations that only change visibility.
+* String parts of encapsed strings are now represented using `Scalar\EncapsStringPart` nodes.
+  Previously raw strings were used. This affects the `parts` child of `Scalar\Encaps` and
+  `Expr\ShellExec`. The change has been done to allow assignment of attributes to encapsed string
+  parts.
 
-Version 1.0.1 (2014-10-14)
---------------------------
-
-* Disallow `new` expressions without a class name. Previously `new;` was accidentally considered to be valid code.
-
-* Support T_ONUMBER token used by HHVM.
-
-* Add ability to directly pass code to the `php-parse.php` script.
-
-* Prevent truncation of `var_dump()` output in the `php-parse.php` script if XDebug is used.
-
-Version 1.0.0 (2014-09-12)
---------------------------
-
-* [BC] Removed deprecated `Template` and `TemplateLoader` classes.
-
-* Fixed XML unserializer to properly work with new namespaced node names.
-
-Version 1.0.0-beta2 (2014-08-31)
+Version 2.0.0-beta1 (2015-10-21)
 --------------------------------
 
-* [PHP 5.6] Updated support for constant scalar expressions to comply with latest changes. This means that arrays
-  and array dimension fetches are now supported as well.
+### Fixed
 
-* [PHP 5.6] Direct array dereferencing of constants is supported now, i.e. both `FOO[0]` and `Foo::BAR[0]` are valid
-  now.
+* Fixed issue with too many newlines being stripped at the end of heredoc/nowdoc strings in some
+  cases. (#227)
 
-* Fixed handling of special class names (`self`, `parent` and `static`) in the name resolver to be case insensitive.
-  Additionally the name resolver now enforces that special class names are only used as unqualified names, e.g. `\self`
-  is considered invalid.
+### Changed
 
-* The case of references to the `static` class name is now preserved. Previously `static` was always lowercased,
-  regardless of the case used in the source code.
+* Update group use support to be in line with recent PHP 7.0 builds.
+* Renamed `php-parse.php` to `php-parse` and registered it as a composer bin.
+* Use composer PSR-4 autoloader instead of custom autoloader.
+* Specify phpunit as a dev dependency.
 
-* The autoloader now only requires a file if it exists. This allows usages like
-  `class_exists('PhpParser\NotExistingClass')`.
+### Added
 
-* Added experimental `bin/php-parse.php` script, which is intended to help exploring and debugging the node tree.
+* Added `shortArraySyntax` option to pretty printer, to print all arrays using short syntax.
 
-* Separated the parser implemention (in `lib/PhpParser/ParserAbstract.php`) and the generated data (in
-  `lib/PhpParser/Parser.php`). Furthermore the parser now uses meaningful variable names and contains comments
-  explaining their usage.
+Version 2.0.0-alpha1 (2015-07-14)
+---------------------------------
 
-Version 1.0.0-beta1 (2014-03-27)
---------------------------------
+A more detailed description of backwards incompatible changes can be found in the
+[upgrading guide](UPGRADE-2.0.md).
 
-* [BC] PHP-Parser now requires PHP 5.3 or newer to run. It is however still possible to *parse* PHP 5.2 source code,
-  while running on a newer version.
+### Removed
 
-* [BC] The library has been moved to use namespaces with the `PhpParser` vendor prefix. However, the old names using
-  underscores are still available as aliases, as such most code should continue running on the new version without
-  further changes.
+* Removed support for running on PHP 5.3. It is however still possible to parse PHP 5.2 and PHP 5.3
+  code while running on a newer version.
+* Removed legacy class name aliases. This includes the old non-namespaced class names and the old
+  names for classes that were renamed for PHP 7 compatibility.
+* Removed support for legacy node format. All nodes must have a `getSubNodeNames()` method now.
 
-  However, code performing dispatch operations on `Node::getType()` may be affected by some of the name changes. For
-  example a `+` node will now return type `Expr_BinaryOp_Plus` instead of `Expr_Plus`. In particular this may affect
-  custom pretty printers.
+### Added
 
-  Due to conflicts with reserved keywords, some class names now end with an underscore, e.g. `PHPParser_Node_Stmt_Class`
-  is now `PhpParser\Node\Stmt\Class_`. (But as usual, the old name is still available)
+* Added support for remaining PHP 7 features that were not present in 1.x:
+  * Group use declarations. These are represented using `Stmt\GroupUse` nodes. Furthermore a `type`
+    attribute was added to `Stmt\UseUse` to handle mixed group use declarations.
+  * Uniform variable syntax.
+  * Generalized yield operator.
+  * Scalar type declarations. These are presented using `'bool'`, `'int'`, `'float'` and `'string'`
+    as the type. The PHP 5 parser also accepts these, however they'll be `Name` instances there.
+  * Unicode escape sequences.
+* Added `PhpParser\ParserFactory` class, which should be used to create parser instances.
+* Added `Name::concat()` which concatenates two names.
+* Added `Name->slice()` which takes a subslice of a name.
 
-* [PHP 5.6] Added support for the power operator `**` (node `Expr\BinaryOp\Pow`) and the compound power assignment
-  operator `**=` (node `Expr\AssignOp\Pow`).
+### Changed
 
-* [PHP 5.6] Added support for variadic functions: `Param` nodes now have `variadic` as a boolean subnode.
-
-* [PHP 5.6] Added support for argument unpacking: `Arg` nodes now have `unpack` as a boolean subnode.
-
-* [PHP 5.6] Added support for aliasing of functions and constants. `Stmt\Use_` nodes now have an integral `type`
-  subnode, which is one of `Stmt\Use_::TYPE_NORMAL` (`use`), `Stmt\Use_::TYPE_FUNCTION` (`use function`) or
-  `Stmt\Use_::TYPE_CONSTANT` (`use const`).
-
-  The `NameResolver` now also supports resolution of such aliases.
-
-* [PHP 5.6] Added support for constant scalar expressions. This means that certain expressions are now allowed as the
-  initializer for constants, properties, parameters, static variables, etc.
-
-* [BC] Improved pretty printing of empty statements lists, which are now printed as `{\n}` instead of `{\n    \n}`.
-  This changes the behavior of the protected `PrettyPrinterAbstract::pStmts()` method, so custom pretty printing code
-  making use it of may need to be adjusted.
-
-* Changed the order of some subnodes to be consistent with their order in the sour code. For example `Stmt\If->cond`
-  will now appear before `Stmt\If->stmts` etc.
-
-* Added `Scalar\MagicConstant->getName()`, which returns the name of the magic constant (e.g. `__CLASS__`).
-
-**The following changes are also included in 0.9.5**:
-
-* [BC] Deprecated `PHPParser_Template` and `PHPParser_TemplateLoader`. This functionality does not belong in the main project
-  and - as far as I know - nobody is using it.
-
-* Add `NodeTraverser::removeVisitor()` method, which removes a visitor from the node traverser. This also modifies the
-  corresponding `NodeTraverserInterface`.
-
-* Fix alias resolution in `NameResolver`: Class names are now correctly handled as case-insensitive.
-
-* The undefined variable error, which is used to the lexer to reset the error state, will no longer interfere with
-  custom error handlers.
+* `PhpParser\Parser` is now an interface, implemented by `Parser\Php5`, `Parser\Php7` and
+  `Parser\Multiple`. The `Multiple` parser will try multiple parsers, until one succeeds.
+* Token constants are now defined on `PhpParser\Parser\Tokens` rather than `PhpParser\Parser`.
+* The `Name->set()`, `Name->append()`, `Name->prepend()` and `Name->setFirst()` methods are
+  deprecated in favor of `Name::concat()` and `Name->slice()`.
+* The `NodeTraverser` no longer clones nodes by default. The old behavior can be restored by
+  passing `true` to the constructor.
+* The constructor for `Scalar` nodes no longer has a default value. E.g. `new LNumber()` should now
+  be written as `new LNumber(0)`.
 
 ---
 
-**This changelog only includes changes from the 1.0 series. For older changes see the [0.9 series changelog][1].**
-
- [1]: https://github.com/nikic/PHP-Parser/blob/0.9/CHANGELOG.md
+**This changelog only includes changes from the 2.0 series. For older changes see the
+[1.x series changelog](https://github.com/nikic/PHP-Parser/blob/1.x/CHANGELOG.md) and the
+[0.9 series changelog](https://github.com/nikic/PHP-Parser/blob/0.9/CHANGELOG.md).**

README.md

@@ -1,12 +1,16 @@
 PHP Parser
 ==========
 
-This is a PHP 5.2 to PHP 5.6 parser written in PHP. Its purpose is to simplify static code analysis and
+[![Build Status](https://travis-ci.org/nikic/PHP-Parser.svg?branch=master)](https://travis-ci.org/nikic/PHP-Parser) [![Coverage Status](https://coveralls.io/repos/github/nikic/PHP-Parser/badge.svg?branch=master)](https://coveralls.io/github/nikic/PHP-Parser?branch=master)
+
+This is a PHP 5.2 to PHP 7.1 parser written in PHP. Its purpose is to simplify static code analysis and
 manipulation.
 
-[**Documentation for version 1.x**][doc_master] (stable; for running on PHP >= 5.3).
+[**Documentation for version 3.x**][doc_master] (stable; for running on PHP >= 5.5; for parsing PHP 5.2 to PHP 7.1).
 
-[Documentation for version 0.9.x][doc_0_9] (unsupported; for running on PHP 5.2).
+[Documentation for version 2.x][doc_2_x] (stable; for running on PHP >= 5.4; for parsing PHP 5.2 to PHP 7.0).
+
+[Documentation for version 1.x][doc_1_x] (unsupported; for running on PHP >= 5.3; for parsing PHP 5.2 to PHP 5.6).
 
 In a Nutshell
 -------------
@@ -22,7 +26,7 @@ hello\world('foo', 'bar' . 'baz');
 
 You'll get a syntax tree looking roughly like this:
 
-```
+```php
 array(
     0: Stmt_Echo(
         exprs: array(
@@ -70,18 +74,26 @@ programming errors or security issues).
 Additionally, you can convert a syntax tree back to PHP code. This allows you to do code preprocessing
 (like automatedly porting code to older PHP versions).
 
+Installation
+------------
+
+The preferred installation method is [composer](https://getcomposer.org):
+
+    php composer.phar require nikic/php-parser
+
 Documentation
 -------------
 
  1. [Introduction](doc/0_Introduction.markdown)
- 2. [Installation](doc/1_Installation.markdown)
- 3. [Usage of basic components](doc/2_Usage_of_basic_components.markdown)
- 4. [Other node tree representations](doc/3_Other_node_tree_representations.markdown)
- 5. [Code generation](doc/4_Code_generation.markdown)
+ 2. [Usage of basic components](doc/2_Usage_of_basic_components.markdown)
+ 3. [Other node tree representations](doc/3_Other_node_tree_representations.markdown)
+ 4. [Code generation](doc/4_Code_generation.markdown)
 
 Component documentation:
 
- 1. [Lexer](doc/component/Lexer.markdown)
+ 1. [Error handling](doc/component/Error_handling.markdown)
+ 2. [Lexer](doc/component/Lexer.markdown)
 
- [doc_0_9]: https://github.com/nikic/PHP-Parser/tree/0.9/doc
+ [doc_1_x]: https://github.com/nikic/PHP-Parser/tree/1.x/doc
+ [doc_2_x]: https://github.com/nikic/PHP-Parser/tree/2.x/doc
  [doc_master]: https://github.com/nikic/PHP-Parser/tree/master/doc

UPGRADE-2.0.md

@@ -0,0 +1,74 @@
+Upgrading from PHP-Parser 1.x to 2.0
+====================================
+
+### PHP version requirements
+
+PHP-Parser now requires PHP 5.4 or newer to run. It is however still possible to *parse* PHP 5.2 and
+PHP 5.3 source code, while running on a newer version.
+
+### Creating a parser instance
+
+Parser instances should now be created through the `ParserFactory`. Old direct instantiation code
+will not work, because the parser class was renamed.
+
+Old:
+
+```php
+use PhpParser\Parser, PhpParser\Lexer;
+$parser = new Parser(new Lexer\Emulative);
+```
+
+New:
+
+```php
+use PhpParser\ParserFactory;
+$parser = (new ParserFactory)->create(ParserFactory::PREFER_PHP7);
+```
+
+The first argument to `ParserFactory` determines how different PHP versions are handled. The
+possible values are:
+
+ * `ParserFactory::PREFER_PHP7`: Try to parse code as PHP 7. If this fails, try to parse it as PHP 5.
+ * `ParserFactory::PREFER_PHP5`: Try to parse code as PHP 5. If this fails, try to parse it as PHP 7.
+ * `ParserFactory::ONLY_PHP7`: Parse code as PHP 7.
+ * `ParserFactory::ONLY_PHP5`: Parse code as PHP 5.
+
+For most practical purposes the difference between `PREFER_PHP7` and `PREFER_PHP5` is mainly whether
+a scalar type hint like `string` will be stored as `'string'` (PHP 7) or as `new Name('string')`
+(PHP 5).
+
+To use a custom lexer, pass it as the second argument to the `create()` method:
+
+```php
+use PhpParser\ParserFactory;
+$lexer = new MyLexer;
+$parser = (new ParserFactory)->create(ParserFactory::PREFER_PHP7, $lexer);
+```
+
+### Rename of the `PhpParser\Parser` class
+
+`PhpParser\Parser` is now an interface, which is implemented by `Parser\Php5`, `Parser\Php7` and
+`Parser\Multiple`. Parser tokens are now defined in `Parser\Tokens`. If you use the `ParserFactory`
+described above to create your parser instance, these changes should have no further impact on you.
+
+### Removal of legacy aliases
+
+All legacy aliases for classes have been removed. This includes the old non-namespaced `PHPParser_`
+classes, as well as the classes that had to be renamed for PHP 7 support.
+
+### Deprecations
+
+The `set()`, `setFirst()`, `append()` and `prepend()` methods of the `Node\Name` class have been
+deprecated. Instead `Name::concat()` and `Name->slice()` should be used.
+
+### Miscellaneous
+
+* The `NodeTraverser` no longer clones nodes by default. If you want to restore the old behavior,
+  pass `true` to the constructor.
+* The legacy node format has been removed. If you use custom nodes, they are now expected to
+  implement a `getSubNodeNames()` method.
+* The default value for `Scalar` node constructors was removed. This means that something like
+  `new LNumber()` should be replaced by `new LNumber(0)`.
+* String parts of encapsed strings are now represented using `Scalar\EncapsStringPart` nodes, while
+  previously raw strings were used. This affects the `parts` child of `Scalar\Encaps` and
+  `Expr\ShellExec`.

UPGRADE-3.0.md

@@ -0,0 +1,160 @@
+Upgrading from PHP-Parser 2.x to 3.0
+====================================
+
+The backwards-incompatible changes in this release may be summarized as follows:
+
+ * The specific details of the node representation have changed in some cases, primarily to
+   accomodate new PHP 7.1 features.
+ * There have been significant changes to the error recovery implementation. This may affect you,
+   if you used the error recovery mode or have a custom lexer implementation.
+ * A number of deprecated methods were removed.
+
+### PHP version requirements
+
+PHP-Parser now requires PHP 5.5 or newer to run. It is however still possible to *parse* PHP 5.2,
+5.3 and 5.4 source code, while running on a newer version.
+
+### Changes to the node structure
+
+The following changes are likely to require code changes if the respective nodes are used:
+
+ * The `List` subnode `vars` has been renamed to `items` and now contains `ArrayItem`s instead of
+   plain variables.
+ * The `Catch` subnode `type` has been renamed to `types` and is now an array of `Name`s.
+ * The `TryCatch` subnode `finallyStmts` has been replaced with a `finally` subnode that holds an
+   explicit `Finally` node.
+ * The `type` subnode on `Class`, `ClassMethod` and `Property` has been renamed to `flags`. The
+   `type` subnode has retained for backwards compatibility and is populated to the same value as
+   `flags`. However, writes to `type` will not update `flags` and use of `type` is discouraged.
+
+The following changes are unlikely to require code changes:
+
+ * The `ClassConst` constructor changed to accept an additional `flags` subnode.
+ * The `Trait` constructor now has the same form as the `Class` and `Interface` constructors: It
+   takes an array of subnodes. Unlike classes/interfaces, traits can only have a `stmts` subnode.
+ * The `Array` subnode `items` may now contain `null` elements (due to destructuring).
+ * `void` and `iterable` types are now stored as strings if the PHP 7 parser is used. Previously
+   these would have been represented as `Name` instances.
+
+### Changes to error recovery mode
+
+Previously, error recovery mode was enabled by setting the `throwOnError` option to `false` when
+creating the parser, while collected errors were retrieved using the `getErrors()` method:
+
+```php
+$lexer = ...;
+$parser = (new ParserFactory)->create(ParserFactor::ONLY_PHP7, $lexer, [
+    'throwOnError' => true,
+]);
+
+$stmts = $parser->parse($code);
+$errors = $parser->getErrors();
+if ($errors) {
+    handleErrors($errors);
+}
+processAst($stmts);
+```
+
+Both the `throwOnError` option and the `getErrors()` method have been removed in PHP-Parser 3.0.
+Instead an instance of `ErrorHandler\Collecting` should be passed to the `parse()` method:
+
+```php
+$lexer = ...;
+$parser = (new ParserFactory)->create(ParserFactor::ONLY_PHP7, $lexer);
+
+$errorHandler = new ErrorHandler\Collecting;
+$stmts = $parser->parse($code, $errorHandler);
+if ($errorHandler->hasErrors()) {
+    handleErrors($errorHandler->getErrors());
+}
+processAst($stmts);
+```
+
+#### Multiple parser fallback in error recovery mode
+
+As a result of this change, if a `Multiple` parser is used (e.g. through the `ParserFactory` using
+`PREFER_PHP7` or `PREFER_PHP5`), it will now return the result of the first *non-throwing* parse. As
+parsing never throws in error recovery mode, the result from the first parser will always be
+returned.
+
+The PHP 7 parser is a superset of the PHP 5 parser, with the exceptions that `=& new` and
+`global $$foo->bar` are not supported (other differences are in representation only). The PHP 7
+parser will be able to recover from the error in both cases. For this reason, this change will
+likely pass unnoticed if you do not specifically test for this syntax.
+
+It is possible to restore the precise previous behavior with the following code:
+
+```php
+$lexer = ...;
+$parser7 = new Parser\Php7($lexer);
+$parser5 = new Parser\Php5($lexer);
+
+$errors7 = new ErrorHandler\Collecting();
+$stmts7 = $parser7->parse($code, $errors7);
+if ($errors7->hasErrors()) {
+    $errors5 = new ErrorHandler\Collecting();
+    $stmts5 = $parser5->parse($code, $errors5);
+    if (!$errors5->hasErrors()) {
+        // If PHP 7 parse has errors but PHP 5 parse has no errors, use PHP 5 result
+        return [$stmts5, $errors5];
+    }
+}
+// If PHP 7 succeeds or both fail use PHP 7 result
+return [$stmts7, $errors7];
+```
+
+#### Error handling in the lexer
+
+In order to support recovery from lexer errors, the signature of the `startLexing()` method changed
+to optionally accept an `ErrorHandler`:
+
+```php
+// OLD
+public function startLexing($code);
+// NEW
+public function startLexing($code, ErrorHandler $errorHandler = null);
+```
+
+If you use a custom lexer with overriden `startLexing()` method, it needs to be changed to accept
+the extra parameter. The value should be passed on to the parent method.
+
+#### Error checks in node constructors
+
+The constructors of certain nodes used to contain additional checks for semantic errors, such as
+creating a try block without either catch or finally. These checks have been moved from the node
+constructors into the parser. This allows recovery from such errors, as well as representing the
+resulting (invalid) AST.
+
+This means that certain error conditions are no longer checked for manually constructed nodes.
+
+### Removed methods, arguments, options
+
+The following methods, arguments or options have been removed:
+
+ * `Comment::setLine()`, `Comment::setText()`: Create new `Comment` instances instead.
+ * `Name::set()`, `Name::setFirst()`, `Name::setLast()`, `Name::append()`, `Name::prepend()`:
+    Use `Name::concat()` in combination with `Name::slice()` instead.
+ * `Error::getRawLine()`, `Error::setRawLine()`. Use `Error::getStartLine()` and
+   `Error::setStartLine()` instead.
+ * `Parser::getErrors()`. Use `ErrorHandler\Collecting` instead.
+ * `$separator` argument of `Name::toString()`. Use `strtr()` instead, if you really need it.
+ * `$cloneNodes` argument of `NodeTraverser::__construct()`. Explicitly clone nodes in the visitor
+   instead.
+ * `throwOnError` parser option. Use `ErrorHandler\Collecting` instead.
+
+### Miscellaneous
+
+ * The `NameResolver` will now resolve unqualified function and constant names in the global
+   namespace into fully qualified names. For example `foo()` in the global namespace resolves to
+   `\foo()`. For names where no static resolution is possible, a `namespacedName` attribute is
+   added now, containing the namespaced variant of the name.
+ * All methods on `PrettyPrinter\Standard` are now protected. Previoulsy most of them were public.
+   The pretty printer should only be invoked using the `prettyPrint()`, `prettyPrintFile()` and
+   `prettyPrintExpr()` methods.
+ * The node dumper now prints numeric values that act as enums/flags in a string representation.
+   If node dumper results are used in tests, updates may be needed to account for this.
+ * The constants on `NameTraverserInterface` have been moved into the `NameTraverser` class.
+ * The emulative lexer now directly postprocesses tokens, instead of using `~__EMU__~` sequences.
+   This changes the protected API of the emulative lexer.
+ * The `Name::slice()` method now returns `null` for empty slices, previously `new Name([])` was
+   used. `Name::concat()` now also supports concatenation with `null`.

bin/php-parse

@@ -0,0 +1,202 @@
+#!/usr/bin/env php
+<?php
+
+foreach ([__DIR__ . '/../../../autoload.php', __DIR__ . '/../vendor/autoload.php'] as $file) {
+    if (file_exists($file)) {
+        require $file;
+        break;
+    }
+}
+
+ini_set('xdebug.max_nesting_level', 3000);
+
+// Disable XDebug var_dump() output truncation
+ini_set('xdebug.var_display_max_children', -1);
+ini_set('xdebug.var_display_max_data', -1);
+ini_set('xdebug.var_display_max_depth', -1);
+
+list($operations, $files, $attributes) = parseArgs($argv);
+
+/* Dump nodes by default */
+if (empty($operations)) {
+    $operations[] = 'dump';
+}
+
+if (empty($files)) {
+    showHelp("Must specify at least one file.");
+}
+
+$lexer = new PhpParser\Lexer\Emulative(array('usedAttributes' => array(
+    'startLine', 'endLine', 'startFilePos', 'endFilePos', 'comments'
+)));
+$parser = (new PhpParser\ParserFactory)->create(
+    PhpParser\ParserFactory::PREFER_PHP7,
+    $lexer
+);
+$dumper = new PhpParser\NodeDumper([
+    'dumpComments' => true,
+    'dumpPositions' => $attributes['with-positions'],
+]);
+$prettyPrinter = new PhpParser\PrettyPrinter\Standard;
+$serializer = new PhpParser\Serializer\XML;
+
+$traverser = new PhpParser\NodeTraverser();
+$traverser->addVisitor(new PhpParser\NodeVisitor\NameResolver);
+
+foreach ($files as $file) {
+    if (strpos($file, '<?php') === 0) {
+        $code = $file;
+        echo "====> Code $code\n";
+    } else {
+        if (!file_exists($file)) {
+            die("File $file does not exist.\n");
+        }
+
+        $code = file_get_contents($file);
+        echo "====> File $file:\n";
+    }
+
+    if ($attributes['with-recovery']) {
+        $errorHandler = new PhpParser\ErrorHandler\Collecting;
+        $stmts = $parser->parse($code, $errorHandler);
+        foreach ($errorHandler->getErrors() as $error) {
+            $message = formatErrorMessage($error, $code, $attributes['with-column-info']);
+            echo $message . "\n";
+        }
+        if (null === $stmts) {
+            continue;
+        }
+    } else {
+        try {
+            $stmts = $parser->parse($code);
+        } catch (PhpParser\Error $error) {
+            $message = formatErrorMessage($error, $code, $attributes['with-column-info']);
+            die($message . "\n");
+        }
+    }
+
+    foreach ($operations as $operation) {
+        if ('dump' === $operation) {
+            echo "==> Node dump:\n";
+            echo $dumper->dump($stmts, $code), "\n";
+        } elseif ('pretty-print' === $operation) {
+            echo "==> Pretty print:\n";
+            echo $prettyPrinter->prettyPrintFile($stmts), "\n";
+        } elseif ('serialize-xml' === $operation) {
+            echo "==> Serialized XML:\n";
+            echo $serializer->serialize($stmts), "\n";
+        } elseif ('var-dump' === $operation) {
+            echo "==> var_dump():\n";
+            var_dump($stmts);
+        } elseif ('resolve-names' === $operation) {
+            echo "==> Resolved names.\n";
+            $stmts = $traverser->traverse($stmts);
+        }
+    }
+}
+
+function formatErrorMessage(PhpParser\Error $e, $code, $withColumnInfo) {
+    if ($withColumnInfo && $e->hasColumnInfo()) {
+        return $e->getMessageWithColumnInfo($code);
+    } else {
+        return $e->getMessage();
+    }
+}
+
+function showHelp($error = '') {
+    if ($error) {
+        echo $error . "\n\n";
+    }
+    die(<<<OUTPUT
+Usage: php-parse [operations] file1.php [file2.php ...]
+   or: php-parse [operations] "<?php code"
+Turn PHP source code into an abstract syntax tree.
+
+Operations is a list of the following options (--dump by default):
+
+    -d, --dump              Dump nodes using NodeDumper
+    -p, --pretty-print      Pretty print file using PrettyPrinter\Standard
+        --serialize-xml     Serialize nodes using Serializer\XML
+        --var-dump          var_dump() nodes (for exact structure)
+    -N, --resolve-names     Resolve names using NodeVisitor\NameResolver
+    -c, --with-column-info  Show column-numbers for errors (if available)
+    -P, --with-positions    Show positions in node dumps
+    -r, --with-recovery     Use parsing with error recovery
+    -h, --help              Display this page
+
+Example:
+    php-parse -d -p -N -d file.php
+
+    Dumps nodes, pretty prints them, then resolves names and dumps them again.
+
+
+OUTPUT
+    );
+}
+
+function parseArgs($args) {
+    $operations = array();
+    $files = array();
+    $attributes = array(
+        'with-column-info' => false,
+        'with-posititions' => false,
+        'with-recovery' => false,
+    );
+
+    array_shift($args);
+    $parseOptions = true;
+    foreach ($args as $arg) {
+        if (!$parseOptions) {
+            $files[] = $arg;
+            continue;
+        }
+
+        switch ($arg) {
+            case '--dump':
+            case '-d':
+                $operations[] = 'dump';
+                break;
+            case '--pretty-print':
+            case '-p':
+                $operations[] = 'pretty-print';
+                break;
+            case '--serialize-xml':
+                $operations[] = 'serialize-xml';
+                break;
+            case '--var-dump':
+                $operations[] = 'var-dump';
+                break;
+            case '--resolve-names':
+            case '-N';
+                $operations[] = 'resolve-names';
+                break;
+            case '--with-column-info':
+            case '-c';
+                $attributes['with-column-info'] = true;
+                break;
+            case '--with-positions':
+            case '-P':
+                $attributes['with-positions'] = true;
+                break;
+            case '--with-recovery':
+            case '-r':
+                $attributes['with-recovery'] = true;
+                break;
+            case '--help':
+            case '-h';
+                showHelp();
+                break;
+            case '--':
+                $parseOptions = false;
+                break;
+            default:
+                if ($arg[0] === '-') {
+                    showHelp("Invalid operation $arg.");
+                } else {
+                    $files[] = $arg;
+                }
+        }
+    }
+
+    return array($operations, $files, $attributes);
+}

bin/php-parse.php

@@ -1,145 +0,0 @@
-#!/usr/bin/env php
-<?php
-
-require __DIR__ . '/../lib/bootstrap.php';
-
-ini_set('xdebug.max_nesting_level', 3000);
-
-// Disable XDebug var_dump() output truncation
-ini_set('xdebug.var_display_max_children', -1);
-ini_set('xdebug.var_display_max_data', -1);
-ini_set('xdebug.var_display_max_depth', -1);
-
-list($operations, $files) = parseArgs($argv);
-
-/* Dump nodes by default */
-if (empty($operations)) {
-    $operations[] = 'dump';
-}
-
-if (empty($files)) {
-    showHelp("Must specify at least one file.");
-}
-
-$parser = new PhpParser\Parser(new PhpParser\Lexer\Emulative);
-$dumper = new PhpParser\NodeDumper;
-$prettyPrinter = new PhpParser\PrettyPrinter\Standard;
-$serializer = new PhpParser\Serializer\XML;
-
-$traverser = new PhpParser\NodeTraverser();
-$traverser->addVisitor(new PhpParser\NodeVisitor\NameResolver);
-
-foreach ($files as $file) {
-    if (strpos($file, '<?php') === 0) {
-        $code = $file;
-        echo "====> Code $code\n";
-    } else {
-        if (!file_exists($file)) {
-            die("File $file does not exist.\n");
-        }
-
-        $code = file_get_contents($file);
-        echo "====> File $file:\n";
-    }
-
-    try {
-        $stmts = $parser->parse($code);
-    } catch (PhpParser\Error $e) {
-        die("==> Parse Error: {$e->getMessage()}\n");
-    }
-
-    foreach ($operations as $operation) {
-        if ('dump' === $operation) {
-            echo "==> Node dump:\n";
-            echo $dumper->dump($stmts), "\n";
-        } elseif ('pretty-print' === $operation) {
-            echo "==> Pretty print:\n";
-            echo $prettyPrinter->prettyPrintFile($stmts), "\n";
-        } elseif ('serialize-xml' === $operation) {
-            echo "==> Serialized XML:\n";
-            echo $serializer->serialize($stmts), "\n";
-        } elseif ('var-dump' === $operation) {
-            echo "==> var_dump():\n";
-            var_dump($stmts);
-        } elseif ('resolve-names' === $operation) {
-            echo "==> Resolved names.\n";
-            $stmts = $traverser->traverse($stmts);
-        }
-    }
-}
-
-function showHelp($error) {
-    die($error . "\n\n" .
-        <<<OUTPUT
-Usage:
-
-    php php-parse.php [operations] file1.php [file2.php ...]
-
-The file arguments can also be replaced with a code string:
-
-    php php-parse.php [operations] "<?php code"
-
-Operations is a list of the following options (--dump by default):
-
-    --dump           -d  Dump nodes using NodeDumper
-    --pretty-print   -p  Pretty print file using PrettyPrinter\Standard
-    --serialize-xml      Serialize nodes using Serializer\XML
-    --var-dump           var_dump() nodes (for exact structure)
-    --resolve-names  -N  Resolve names using NodeVisitor\NameResolver
-
-Example:
-
-    php php-parse.php -d -p -N -d file.php
-
-    Dumps nodes, pretty prints them, then resolves names and dumps them again.
-
-
-OUTPUT
-    );
-}
-
-function parseArgs($args) {
-    $operations = array();
-    $files = array();
-
-    array_shift($args);
-    $parseOptions = true;
-    foreach ($args as $arg) {
-        if (!$parseOptions) {
-            $files[] = $arg;
-            continue;
-        }
-
-        switch ($arg) {
-            case '--dump':
-            case '-d':
-                $operations[] = 'dump';
-                break;
-            case '--pretty-print':
-            case '-p':
-                $operations[] = 'pretty-print';
-                break;
-            case '--serialize-xml':
-                $operations[] = 'serialize-xml';
-                break;
-            case '--var-dump':
-                $operations[] = 'var-dump';
-                break;
-            case '--resolve-names':
-            case '-N';
-                $operations[] = 'resolve-names';
-                break;
-            case '--':
-                $parseOptions = false;
-                break;
-            default:
-                if ($arg[0] === '-') {
-                    showHelp("Invalid operation $arg.");
-                } else {
-                    $files[] = $arg;
-                }
-        }
-    }
-
-    return array($operations, $files);
-}

composer.json

@@ -10,15 +10,21 @@
         }
     ],
     "require": {
-        "php": ">=5.3",
+        "php": ">=5.5",
         "ext-tokenizer": "*"
     },
-    "autoload": {
-        "files": ["lib/bootstrap.php"]
+    "require-dev": {
+        "phpunit/phpunit": "~4.0|~5.0"
     },
+    "autoload": {
+        "psr-4": {
+            "PhpParser\\": "lib/PhpParser"
+        }
+    },
+    "bin": ["bin/php-parse"],
     "extra": {
         "branch-alias": {
-            "dev-master": "1.2-dev"
+            "dev-master": "3.0-dev"
         }
     }
 }

doc/0_Introduction.markdown

@@ -1,7 +1,7 @@
 Introduction
 ============
 
-This project is a PHP 5.2 to PHP 5.6 parser **written in PHP itself**.
+This project is a PHP 5.2 to PHP 7.1 parser **written in PHP itself**.
 
 What is this for?
 -----------------
@@ -18,7 +18,7 @@ For example an AST abstracts away the fact that in PHP variables can be written
 as `$$bar`, `${'foobar'}` or even `${!${''}=barfoo()}`. You don't have to worry about recognizing
 all the different syntaxes from a stream of tokens.
 
-Another questions is: Why would I want to have a PHP parser *written in PHP*? Well, PHP might not be
+Another question is: Why would I want to have a PHP parser *written in PHP*? Well, PHP might not be
 a language especially suited for fast parsing, but processing the AST is much easier in PHP than it
 would be in other, faster languages like C. Furthermore the people most probably wanting to do
 programmatic PHP code analysis are incidentally PHP developers, not C developers.
@@ -26,13 +26,12 @@ programmatic PHP code analysis are incidentally PHP developers, not C developers
 What can it parse?
 ------------------
 
-The parser uses a PHP 5.6 compliant grammar, which is backwards compatible with all PHP version from PHP 5.2
-upwards (and maybe older).
+The parser supports parsing PHP 5.2-5.6 and PHP 7.
 
 As the parser is based on the tokens returned by `token_get_all` (which is only able to lex the PHP
-version it runs on), additionally a wrapper for emulating new tokens from 5.3, 5.4, 5.5 and 5.6 is provided.
-This allows to parse PHP 5.6 source code running on PHP 5.3, for example. This emulation is very hacky and not
-perfect, but it should work well on any sane code.
+version it runs on), additionally a wrapper for emulating tokens from newer versions is provided.
+This allows to parse PHP 7.1 source code running on PHP 5.5, for example. This emulation is somewhat
+hacky and not perfect, but it should work well on any sane code.
 
 What output does it produce?
 ----------------------------

doc/1_Installation.markdown

@@ -1,31 +0,0 @@
-Installation
-============
-
-There are multiple ways to include the PHP parser into your project:
-
-Installing via Composer
------------------------
-
-Run the following command inside your project:
-
-    php composer.phar require nikic/php-parser
-
-If you haven't installed [Composer][1] yet, you can do so using:
-
-    curl -s http://getcomposer.org/installer | php
-
-Installing as a Git Submodule
------------------------------
-
-Run the following command to install the parser into the `vendor/PHP-Parser` folder:
-
-    git submodule add git://github.com/nikic/PHP-Parser.git vendor/PHP-Parser
-
-Installing from the Zip- or Tarball
------------------------------------
-
-Download the latest version from [the download page][2], unpack it and move the files somewhere into your project.
-
-
- [1]: https://getcomposer.org/
- [2]: https://github.com/nikic/PHP-Parser/tags

doc/2_Usage_of_basic_components.markdown

@@ -6,55 +6,60 @@ This document explains how to use the parser, the pretty printer and the node tr
 Bootstrapping
 -------------
 
-The library needs to register a class autoloader. You can either use the `vendor/autoload.php` file generated by
-Composer or by including the bundled `lib/bootstrap.php` file:
+To bootstrap the library, include the autoloader generated by composer:
 
 ```php
-<?php
-require 'path/to/PHP-Parser/lib/bootstrap.php';
-// Or, if you're using Composer:
 require 'path/to/vendor/autoload.php';
 ```
 
 Additionally you may want to set the `xdebug.max_nesting_level` ini option to a higher value:
 
 ```php
-<?php
 ini_set('xdebug.max_nesting_level', 3000);
 ```
 
-This ensures that there will be no errors when traversing highly nested node trees.
+This ensures that there will be no errors when traversing highly nested node trees. However, it is
+preferable to disable XDebug completely, as it can easily make this library more than five times
+slower.
 
 Parsing
 -------
 
-In order to parse some source code you first have to create a `PhpParser\Parser` object, which
-needs to be passed a `PhpParser\Lexer` instance:
+In order to parse code, you first have to create a parser instance:
 
 ```php
-<?php
-
-$parser = new PhpParser\Parser(new PhpParser\Lexer);
-// or
-$parser = new PhpParser\Parser(new PhpParser\Lexer\Emulative);
+use PhpParser\ParserFactory;
+$parser = (new ParserFactory)->create(ParserFactory::PREFER_PHP7);
 ```
 
-Use of the emulative lexer is required if you want to parse PHP code from newer versions than the one
-you're running on. For example it will allow you to parse PHP 5.6 code while running on PHP 5.3.
+The factory accepts a kind argument, that determines how different PHP versions are treated:
+
+Kind | Behavior
+-----|---------
+`ParserFactory::PREFER_PHP7` | Try to parse code as PHP 7. If this fails, try to parse it as PHP 5.
+`ParserFactory::PREFER_PHP5` | Try to parse code as PHP 5. If this fails, try to parse it as PHP 7.
+`ParserFactory::ONLY_PHP7` | Parse code as PHP 7.
+`ParserFactory::ONLY_PHP5` | Parse code as PHP 5.
+
+Unless you have strong reason to use something else, `PREFER_PHP7` is a reasonable default.
+
+The `create()` method optionally accepts a `Lexer` instance as the second argument. Some use cases
+that require customized lexers are discussed in the [lexer documentation](component/Lexer.markdown).
 
 Subsequently you can pass PHP code (including the opening `<?php` tag) to the `parse` method in order to
 create a syntax tree. If a syntax error is encountered, an `PhpParser\Error` exception will be thrown:
 
 ```php
-<?php
-$code = '<?php // some code';
+use PhpParser\Error;
+use PhpParser\ParserFactory;
 
-$parser = new PhpParser\Parser(new PhpParser\Lexer\Emulative);
+$code = '<?php // some code';
+$parser = (new ParserFactory)->create(ParserFactory::PREFER_PHP7);
 
 try {
     $stmts = $parser->parse($code);
     // $stmts is an array of statement nodes
-} catch (PhpParser\Error $e) {
+} catch (Error $e) {
     echo 'Parse Error: ', $e->getMessage();
 }
 ```
@@ -137,11 +142,14 @@ information the formatting is done using a specified scheme. Currently there is
 namely `PhpParser\PrettyPrinter\Standard`.
 
 ```php
-<?php
+use PhpParser\Error;
+use PhpParser\ParserFactory;
+use PhpParser\PrettyPrinter;
+
 $code = "<?php echo 'Hi ', hi\\getTarget();";
 
-$parser = new PhpParser\Parser(new PhpParser\Lexer);
-$prettyPrinter = new PhpParser\PrettyPrinter\Standard;
+$parser = (new ParserFactory)->create(ParserFactory::PREFER_PHP7);
+$prettyPrinter = new PrettyPrinter\Standard;
 
 try {
     // parse
@@ -158,7 +166,7 @@ try {
     $code = $prettyPrinter->prettyPrint($stmts);
 
     echo $code;
-} catch (PhpParser\Error $e) {
+} catch (Error $e) {
     echo 'Parse Error: ', $e->getMessage();
 }
 ```
@@ -188,11 +196,13 @@ For this purpose the parser provides a component for traversing and visiting the
 structure of a program using this `PhpParser\NodeTraverser` looks like this:
 
 ```php
-<?php
+use PhpParser\NodeTraverser;
+use PhpParser\ParserFactory;
+use PhpParser\PrettyPrinter;
 
-$parser        = new PhpParser\Parser(new PhpParser\Lexer\Emulative);
-$traverser     = new PhpParser\NodeTraverser;
-$prettyPrinter = new PhpParser\PrettyPrinter\Standard;
+$parser        = (new ParserFactory)->create(ParserFactory::PREFER_PHP7);
+$traverser     = new NodeTraverser;
+$prettyPrinter = new PrettyPrinter\Standard;
 
 // add your visitor
 $traverser->addVisitor(new MyNodeVisitor);
@@ -218,10 +228,10 @@ try {
 The corresponding node visitor might look like this:
 
 ```php
-<?php
 use PhpParser\Node;
+use PhpParser\NodeVisitorAbstract;
 
-class MyNodeVisitor extends PhpParser\NodeVisitorAbstract
+class MyNodeVisitor extends NodeVisitorAbstract
 {
     public function leaveNode(Node $node) {
         if ($node instanceof Node\Scalar\String_) {
@@ -236,10 +246,12 @@ The above node visitor would change all string literals in the program to `'foo'
 All visitors must implement the `PhpParser\NodeVisitor` interface, which defines the following four
 methods:
 
-    public function beforeTraverse(array $nodes);
-    public function enterNode(PhpParser\Node $node);
-    public function leaveNode(PhpParser\Node $node);
-    public function afterTraverse(array $nodes);
+```php
+public function beforeTraverse(array $nodes);
+public function enterNode(\PhpParser\Node $node);
+public function leaveNode(\PhpParser\Node $node);
+public function afterTraverse(array $nodes);
+```
 
 The `beforeTraverse()` method is called once before the traversal begins and is passed the nodes the
 traverser was called with. This method can be used for resetting values before traversation or
@@ -258,7 +270,7 @@ The `enterNode()` method can additionally return the value `NodeTraverser::DONT_
 which instructs the traverser to skip all children of the current node.
 
 The `leaveNode()` method can additionally return the value `NodeTraverser::REMOVE_NODE`, in which
-case the current node will be removed from the parent array. Furthermove it is possible to return
+case the current node will be removed from the parent array. Furthermore it is possible to return
 an array of nodes, which will be merged into the parent array at the offset of the current node.
 I.e. if in `array(A, B, C)` the node `B` should be replaced with `array(X, Y, Z)` the result will
 be `array(A, X, Y, Z, C)`.
@@ -299,20 +311,24 @@ assume that no dynamic features are used.
 We start off with the following base code:
 
 ```php
-<?php
+use PhpParser\ParserFactory;
+use PhpParser\PrettyPrinter;
+use PhpParser\NodeTraverser;
+use PhpParser\NodeVisitor\NameResolver;
+
 $inDir  = '/some/path';
 $outDir = '/some/other/path';
 
-$parser        = new PhpParser\Parser(new PhpParser\Lexer\Emulative);
-$traverser     = new PhpParser\NodeTraverser;
-$prettyPrinter = new PhpParser\PrettyPrinter\Standard;
+$parser        = (new ParserFactory)->create(ParserFactory::PREFER_PHP7);
+$traverser     = new NodeTraverser;
+$prettyPrinter = new PrettyPrinter\Standard;
 
-$traverser->addVisitor(new PhpParser\NodeVisitor\NameResolver); // we will need resolved names
-$traverser->addVisitor(new NodeVisitor\NamespaceConverter);     // our own node visitor
+$traverser->addVisitor(new NameResolver); // we will need resolved names
+$traverser->addVisitor(new NamespaceConverter); // our own node visitor
 
 // iterate over all .php files in the directory
-$files = new RecursiveIteratorIterator(new RecursiveDirectoryIterator($inDir));
-$files = new RegexIterator($files, '/\.php$/');
+$files = new \RecursiveIteratorIterator(new \RecursiveDirectoryIterator($inDir));
+$files = new \RegexIterator($files, '/\.php$/');
 
 foreach ($files as $file) {
     try {
@@ -343,9 +359,9 @@ Now lets start with the main code, the `NodeVisitor\NamespaceConverter`. One thi
 is convert `A\\B` style names to `A_B` style ones.
 
 ```php
-<?php
 use PhpParser\Node;
-class NodeVisitor_NamespaceConverter extends PhpParser\NodeVisitorAbstract
+
+class NamespaceConverter extends \PhpParser\NodeVisitorAbstract
 {
     public function leaveNode(Node $node) {
         if ($node instanceof Node\Name) {
@@ -362,14 +378,14 @@ create a name with backslashes either write `$node->toString()` or `(string) $no
 a new name from the string and return it. Returning a new node replaces the old node.
 
 Another thing we need to do is change the class/function/const declarations. Currently they contain
-only the shortname (i.e. the last part of the name), but they need to contain the complete name inclduing
+only the shortname (i.e. the last part of the name), but they need to contain the complete name including
 the namespace prefix:
 
 ```php
-<?php
 use PhpParser\Node;
 use PhpParser\Node\Stmt;
-class NodeVisitor_NamespaceConverter extends PhpParser\NodeVisitorAbstract
+
+class NodeVisitor_NamespaceConverter extends \PhpParser\NodeVisitorAbstract
 {
     public function leaveNode(Node $node) {
         if ($node instanceof Node\Name) {
@@ -392,10 +408,10 @@ There is not much more to it than converting the namespaced name to string with
 The last thing we need to do is remove the `namespace` and `use` statements:
 
 ```php
-<?php
 use PhpParser\Node;
 use PhpParser\Node\Stmt;
-class NodeVisitor_NamespaceConverter extends PhpParser\NodeVisitorAbstract
+
+class NodeVisitor_NamespaceConverter extends \PhpParser\NodeVisitorAbstract
 {
     public function leaveNode(Node $node) {
         if ($node instanceof Node\Name) {

doc/3_Other_node_tree_representations.markdown

@@ -8,7 +8,7 @@ Simple serialization
 
 It is possible to serialize the node tree using `serialize()` and also unserialize it using
 `unserialize()`. The output is not human readable and not easily processable from anything
-but PHP, but it is compact and generates fast. The main application thus is in caching.
+but PHP, but it is compact and generates quickly. The main application thus is in caching.
 
 Human readable dumping
 ----------------------
@@ -17,7 +17,6 @@ Furthermore it is possible to dump nodes into a human readable format using the
 `PhpParser\NodeDumper`. This can be used for debugging.
 
 ```php
-<?php
 $code = <<<'CODE'
 <?php
 
@@ -28,7 +27,7 @@ function printLine($msg) {
 printLine('Hello World!!!');
 CODE;
 
-$parser = new PhpParser\Parser(new PhpParser\Lexer);
+$parser = (new PhpParser\ParserFactory)->create(PhpParser\ParserFactory::PREFER_PHP7);
 $nodeDumper = new PhpParser\NodeDumper;
 
 try {
@@ -87,6 +86,134 @@ array(
 )
 ```
 
+JSON encoding
+-------------
+
+Nodes (and comments) implement the `JsonSerializable` interface. As such, it is possible to JSON
+encode the AST directly using `json_encode()`:
+
+```php
+$code = <<<'CODE'
+<?php
+
+function printLine($msg) {
+    echo $msg, "\n";
+}
+
+printLine('Hello World!!!');
+CODE;
+
+$parser = (new PhpParser\ParserFactory)->create(PhpParser\ParserFactory::PREFER_PHP7);
+$nodeDumper = new PhpParser\NodeDumper;
+
+try {
+    $stmts = $parser->parse($code);
+
+    echo json_encode($stmts, JSON_PRETTY_PRINT), "\n";
+} catch (PhpParser\Error $e) {
+    echo 'Parse Error: ', $e->getMessage();
+}
+```
+
+This will result in the following output (which includes attributes):
+
+```json
+[
+    {
+        "nodeType": "Stmt_Function",
+        "byRef": false,
+        "name": "printLine",
+        "params": [
+            {
+                "nodeType": "Param",
+                "type": null,
+                "byRef": false,
+                "variadic": false,
+                "name": "msg",
+                "default": null,
+                "attributes": {
+                    "startLine": 3,
+                    "endLine": 3
+                }
+            }
+        ],
+        "returnType": null,
+        "stmts": [
+            {
+                "nodeType": "Stmt_Echo",
+                "exprs": [
+                    {
+                        "nodeType": "Expr_Variable",
+                        "name": "msg",
+                        "attributes": {
+                            "startLine": 4,
+                            "endLine": 4
+                        }
+                    },
+                    {
+                        "nodeType": "Scalar_String",
+                        "value": "\n",
+                        "attributes": {
+                            "startLine": 4,
+                            "endLine": 4,
+                            "kind": 2
+                        }
+                    }
+                ],
+                "attributes": {
+                    "startLine": 4,
+                    "endLine": 4
+                }
+            }
+        ],
+        "attributes": {
+            "startLine": 3,
+            "endLine": 5
+        }
+    },
+    {
+        "nodeType": "Expr_FuncCall",
+        "name": {
+            "nodeType": "Name",
+            "parts": [
+                "printLine"
+            ],
+            "attributes": {
+                "startLine": 7,
+                "endLine": 7
+            }
+        },
+        "args": [
+            {
+                "nodeType": "Arg",
+                "value": {
+                    "nodeType": "Scalar_String",
+                    "value": "Hello World!!!",
+                    "attributes": {
+                        "startLine": 7,
+                        "endLine": 7,
+                        "kind": 1
+                    }
+                },
+                "byRef": false,
+                "unpack": false,
+                "attributes": {
+                    "startLine": 7,
+                    "endLine": 7
+                }
+            }
+        ],
+        "attributes": {
+            "startLine": 7,
+            "endLine": 7
+        }
+    }
+]
+```
+
+There is currently no mechanism to convert JSON back into a node tree. Furthermore, not all ASTs
+can be JSON encoded. In particular, JSON only supports UTF-8 strings.
+
 Serialization to XML
 --------------------
 
@@ -106,7 +233,7 @@ function printLine($msg) {
 printLine('Hello World!!!');
 CODE;
 
-$parser = new PhpParser\Parser(new PhpParser\Lexer);
+$parser = (new PhpParser\ParserFactory)->create(PhpParser\ParserFactory::PREFER_PHP7);
 $serializer = new PhpParser\Serializer\XML;
 
 try {

doc/4_Code_generation.markdown

@@ -14,8 +14,11 @@ the following syntactic elements:
 Here is an example:
 
 ```php
-<?php
-$factory = new PhpParser\BuilderFactory;
+use PhpParser\BuilderFactory;
+use PhpParser\PrettyPrinter;
+use PhpParser\Node;
+
+$factory = new BuilderFactory;
 $node = $factory->namespace('Name\Space')
     ->addStmt($factory->use('Some\Other\Thingy')->as('SomeOtherClass'))
     ->addStmt($factory->class('SomeClass')
@@ -26,6 +29,7 @@ $node = $factory->namespace('Name\Space')
         ->addStmt($factory->method('someMethod')
             ->makePublic()
             ->makeAbstract() // ->makeFinal()
+            ->setReturnType('bool')
             ->addParam($factory->param('someParam')->setTypeHint('SomeClass'))
             ->setDocComment('/**
                               * This method does something.
@@ -38,7 +42,7 @@ $node = $factory->namespace('Name\Space')
             ->makeProtected() // ->makePublic() [default], ->makePrivate()
             ->addParam($factory->param('someParam')->setDefault('test'))
             // it is possible to add manually created nodes
-            ->addStmt(new PhpParser\Node\Expr\Print_(new PhpParser\Node\Expr\Variable('someParam')))
+            ->addStmt(new Node\Expr\Print_(new Node\Expr\Variable('someParam')))
         )
 
         // properties will be correctly reordered above the methods
@@ -50,7 +54,7 @@ $node = $factory->namespace('Name\Space')
 ;
 
 $stmts = array($node);
-$prettyPrinter = new PhpParser\PrettyPrinter\Standard();
+$prettyPrinter = new PrettyPrinter\Standard();
 echo $prettyPrinter->prettyPrintFile($stmts);
 ```
 
@@ -71,7 +75,7 @@ abstract class SomeClass extends SomeOtherClass implements A\Few, \Interfaces
      *
      * @param SomeClass And takes a parameter
      */
-    public abstract function someMethod(SomeClass $someParam);
+    public abstract function someMethod(SomeClass $someParam) : bool;
     protected function anotherMethod($someParam = 'test')
     {
         print $someParam;

doc/component/Error_handling.markdown

@@ -0,0 +1,75 @@
+Error handling
+==============
+
+Errors during parsing or analysis are represented using the `PhpParser\Error` exception class. In addition to an error
+message, an error can also store additional information about the location the error occurred at.
+
+How much location information is available depends on the origin of the error and how many lexer attributes have been
+enabled. At a minimum the start line of the error is usually available.
+
+Column information
+------------------
+
+In order to receive information about not only the line, but also the column span an error occurred at, the file
+position attributes in the lexer need to be enabled:
+
+```php
+$lexer = new PhpParser\Lexer(array(
+    'usedAttributes' => array('comments', 'startLine', 'endLine', 'startFilePos', 'endFilePos'),
+));
+$parser = (new PhpParser\ParserFactory)->create(PhpParser\ParserFactory::PREFER_PHP7, $lexer);
+
+try {
+    $stmts = $parser->parse($code);
+    // ...
+} catch (PhpParser\Error $e) {
+    // ...
+}
+```
+
+Before using column information its availability needs to be checked with `$e->hasColumnInfo()`, as the precise
+location of an error cannot always be determined. The methods for retrieving column information also have to be passed
+the source code of the parsed file. An example for printing an error:
+
+```php
+if ($e->hasColumnInfo()) {
+    echo $e->getRawMessage() . ' from ' . $e->getStartLine() . ':' . $e->getStartColumn($code)
+        . ' to ' . $e->getEndLine() . ':' . $e->getEndColumn($code);
+    // or:
+    echo $e->getMessageWithColumnInfo();
+} else {
+    echo $e->getMessage();
+}
+```
+
+Both line numbers and column numbers are 1-based. EOF errors will be located at the position one past the end of the
+file.
+
+Error recovery
+--------------
+
+The error behavior of the parser (and other components) is controlled by an `ErrorHandler`. Whenever an error is
+encountered, `ErrorHandler::handleError()` is invoked. The default error handling strategy is `ErrorHandler\Throwing`,
+which will immediately throw when an error is encountered.
+
+To instead collect all encountered errors into an array, while trying to continue parsing the rest of the source code,
+an instance of `ErrorHandler\Collecting` can be passed to the `Parser::parse()` method. A usage example:
+
+```php
+$parser = (new PhpParser\ParserFactory)->create(PhpParser\ParserFactory::ONLY_PHP7);
+$errorHandler = new PhpParser\ErrorHandler\Collecting;
+
+$stmts = $parser->parse($code, $errorHandler);
+
+if ($errorHandler->hasErrors()) {
+    foreach ($errorHandler->getErrors() as $error) {
+        // $error is an ordinary PhpParser\Error
+    }
+}
+
+if (null !== $stmts) {
+    // $stmts is a best-effort partial AST
+}
+```
+
+The `NameResolver` visitor also accepts an `ErrorHandler` as a constructor argument.

doc/component/Lexer.markdown

@@ -72,7 +72,7 @@ $lexer = new PhpParser\Lexer(array(
         'comments', 'startLine', 'endLine', 'startTokenPos', 'endTokenPos'
     )
 ));
-$parser = new PhpParser\Parser($lexer);
+$parser = (new PhpParser\ParserFactory)->create(PhpParser\ParserFactory::PREFER_PHP7, $lexer);
 
 $visitor = new MyNodeVisitor();
 $traverser = new PhpParser\NodeTraverser();
@@ -95,13 +95,14 @@ Lexer extension
 
 A lexer has to define the following public interface:
 
-    void startLexing(string $code);
+    void startLexing(string $code, ErrorHandler $errorHandler = null);
     array getTokens();
     string handleHaltCompiler();
     int getNextToken(string &$value = null, array &$startAttributes = null, array &$endAttributes = null);
 
 The `startLexing()` method is invoked with the source code that is to be lexed (including the opening tag) whenever the
-`parse()` method of the parser is called. It can be used to reset state or preprocess the source code or tokens.
+`parse()` method of the parser is called. It can be used to reset state or preprocess the source code or tokens. The
+passes `ErrorHandler` should be used to report lexing errors.
 
 The `getTokens()` method returns the current token array, in the usual `token_get_all()` format. This method is not
 used by the parser (which uses `getNextToken()`), but is useful in combination with the token position attributes.
@@ -122,19 +123,24 @@ node and the `$endAttributes` from the last token that is part of the node.
 E.g. if the tokens `T_FUNCTION T_STRING ... '{' ... '}'` constitute a node, then the `$startAttributes` from the
 `T_FUNCTION` token will be taken and the `$endAttributes` from the `'}'` token.
 
-An application of custom attributes is storing the original formatting of literals: The parser does not retain
-information about the formatting of integers (like decimal vs. hexadecimal) or strings (like used quote type or used
-escape sequences). This can be remedied by storing the original value in an attribute:
+An application of custom attributes is storing the exact original formatting of literals: While the parser does retain
+some information about the formatting of integers (like decimal vs. hexadecimal) or strings (like used quote type), it
+does not preserve the exact original formatting (e.g. leading zeros for integers or escape sequences in strings). This
+can be remedied by storing the original value in an attribute:
 
 ```php
-class KeepOriginalValueLexer extends PHPParser\Lexer // or PHPParser\Lexer\Emulative
+use PhpParser\Lexer;
+use PhpParser\Parser\Tokens;
+
+class KeepOriginalValueLexer extends Lexer // or Lexer\Emulative
 {
     public function getNextToken(&$value = null, &$startAttributes = null, &$endAttributes = null) {
         $tokenId = parent::getNextToken($value, $startAttributes, $endAttributes);
 
-        if ($tokenId == PHPParser\Parser::T_CONSTANT_ENCAPSED_STRING // non-interpolated string
-            || $tokenId == PHPParser\Parser::T_LNUMBER               // integer
-            || $tokenId == PHPParser\Parser::T_DNUMBER               // floating point number
+        if ($tokenId == Tokens::T_CONSTANT_ENCAPSED_STRING   // non-interpolated string
+            || $tokenId == Tokens::T_ENCAPSED_AND_WHITESPACE // interpolated string
+            || $tokenId == Tokens::T_LNUMBER                 // integer
+            || $tokenId == Tokens::T_DNUMBER                 // floating point number
         ) {
             // could also use $startAttributes, doesn't really matter here
             $endAttributes['originalValue'] = $value;

grammar/README.md

@@ -1,29 +1,28 @@
 What do all those files mean?
 =============================
 
- * `zend_language_parser.phpy`: PHP grammer written in a pseudo language
- * `analyze.php`:               Analyzes the `.phpy`-grammer and outputs some info about it
- * `rebuildParser.php`:         Preprocesses the `.phpy`-grammar and builds the parser using `kmyacc`
- * `kmyacc.php.parser`:         A `kmyacc` parser prototype file for PHP
+ * `php5.y`:             PHP 5 grammar written in a pseudo language
+ * `php7.y`:             PHP 7 grammar written in a pseudo language
+ * `tokens.y`:           Tokens definition shared between PHP 5 and PHP 7 grammars
+ * `parser.template`:    A `kmyacc` parser prototype file for PHP
+ * `tokens.template`:    A `kmyacc` prototype file for the `Tokens` class
+ * `rebuildParsers.php`: Preprocesses the grammar and builds the parser using `kmyacc`
 
 .phpy pseudo language
 =====================
 
-The `.phpy` file is a normal grammer in `kmyacc` (`yacc`) style, with some transformations
+The `.y` file is a normal grammer in `kmyacc` (`yacc`) style, with some transformations
 applied to it:
 
  * Nodes are created using the syntax `Name[..., ...]`. This is transformed into
-   `new Node\Name(..., ..., $attributes)`
- * `Name::abc` is transformed to `Node\Name::abc`
- * Some function-like constructs are resolved (see `rebuildParser.php` for a list)
- * Associative arrays are written as `[key: value, ...]`, which is transformed to
-   `array('key' => value, ...)`
+   `new Name(..., ..., attributes())`
+ * Some function-like constructs are resolved (see `rebuildParsers.php` for a list)
 
 Building the parser
 ===================
 
 In order to rebuild the parser, you need [moriyoshi's fork of kmyacc](https://github.com/moriyoshi/kmyacc-forked).
-After you compiled/installed it, run the `rebuildParser.php` script.
+After you compiled/installed it, run the `rebuildParsers.php` script.
 
-By default only the `Parser.php` is built. If you want to additionally build `Parser/Debug.php` and `y.output` run the
-script with `--debug`. If you want to retain the preprocessed grammar pass `--keep-tmp-grammar`.
+By default only the `Parser.php` is built. If you want to additionally emit debug symbols and create `y.output`, run the
+script with `--debug`. If you want to retain the preprocessed grammar pass `--keep-tmp-grammar`.

grammar/analyze.php

@@ -1,96 +0,0 @@
-<?php
-
-const GRAMMAR_FILE = './zend_language_parser.phpy';
-
-const LIB = '(?(DEFINE)
-    (?<singleQuotedString>\'[^\\\\\']*+(?:\\\\.[^\\\\\']*+)*+\')
-    (?<doubleQuotedString>"[^\\\\"]*+(?:\\\\.[^\\\\"]*+)*+")
-    (?<string>(?&singleQuotedString)|(?&doubleQuotedString))
-    (?<comment>/\*[^*]*+(?:\*(?!/)[^*]*+)*+\*/)
-    (?<code>\{[^\'"/{}]*+(?:(?:(?&string)|(?&comment)|(?&code)|/)[^\'"/{}]*+)*+})
-)';
-
-const RULE_BLOCK = '(?<name>[a-z_]++):(?<rules>[^\'"/{};]*+(?:(?:(?&string)|(?&comment)|(?&code)|/|})[^\'"/{};]*+)*+);';
-
-$usedTerminals = array_flip(array(
-    'T_VARIABLE', 'T_STRING', 'T_INLINE_HTML', 'T_ENCAPSED_AND_WHITESPACE',
-    'T_LNUMBER', 'T_DNUMBER', 'T_CONSTANT_ENCAPSED_STRING', 'T_STRING_VARNAME', 'T_NUM_STRING'
-));
-$unusedNonterminals = array_flip(array(
-    'case_separator', 'optional_comma'
-));
-
-function regex($regex) {
-    return '~' . LIB . '(?:' . str_replace('~', '\~', $regex) . ')~';
-}
-
-function magicSplit($regex, $string) {
-    $pieces = preg_split(regex('(?:(?&string)|(?&comment)|(?&code))(*SKIP)(*FAIL)|' . $regex), $string);
-
-    foreach ($pieces as &$piece) {
-        $piece = trim($piece);
-    }
-
-    return array_filter($pieces);
-}
-
-echo '<pre>';
-
-////////////////////
-////////////////////
-////////////////////
-
-list($defs, $ruleBlocks) = magicSplit('%%', file_get_contents(GRAMMAR_FILE));
-
-if ('' !== trim(preg_replace(regex(RULE_BLOCK), '', $ruleBlocks))) {
-    die('Not all rule blocks were properly recognized!');
-}
-
-preg_match_all(regex(RULE_BLOCK), $ruleBlocks, $ruleBlocksMatches, PREG_SET_ORDER);
-foreach ($ruleBlocksMatches as $match) {
-    $ruleBlockName = $match['name'];
-    $rules = magicSplit('\|', $match['rules']);
-
-    foreach ($rules as &$rule) {
-        $parts = magicSplit('\s+', $rule);
-        $usedParts = array();
-
-        foreach ($parts as $part) {
-            if ('{' === $part[0]) {
-                preg_match_all('~\$([0-9]+)~', $part, $backReferencesMatches, PREG_SET_ORDER);
-                foreach ($backReferencesMatches as $match) {
-                    $usedParts[$match[1]] = true;
-                }
-            }
-        }
-
-        $i = 1;
-        foreach ($parts as &$part) {
-            if ('/' === $part[0]) {
-                continue;
-            }
-
-            if (isset($usedParts[$i])) {
-                if ('\'' === $part[0] || '{' === $part[0]
-                    || (ctype_upper($part[0]) && !isset($usedTerminals[$part]))
-                    || (ctype_lower($part[0]) && isset($unusedNonterminals[$part]))
-                ) {
-                    $part = '<span style="background-color: red; color: white;">' . $part . '</span>';
-                } else {
-                    $part = '<strong><em>' . $part . '</em></strong>';
-                }
-            } elseif ((ctype_upper($part[0]) && isset($usedTerminals[$part]))
-                      || (ctype_lower($part[0]) && !isset($unusedNonterminals[$part]))
-
-            ) {
-                $part = '<span style="background-color: blue; color: white;">' . $part . '</span>';
-            }
-
-            ++$i;
-        }
-
-        $rule = implode(' ', $parts);
-    }
-
-    echo $ruleBlockName, ':', "\n", '      ', implode("\n" . '    | ', $rules), "\n", ';', "\n\n";
-}

grammar/parser.template

@@ -2,35 +2,39 @@
 $meta #
 #semval($) $this->semValue
 #semval($,%t) $this->semValue
-#semval(%n) $this->semStack[$this->stackPos-(%l-%n)]
-#semval(%n,%t) $this->semStack[$this->stackPos-(%l-%n)]
-#include;
+#semval(%n) $this->stackPos-(%l-%n)
+#semval(%n,%t) $this->stackPos-(%l-%n)
 
-namespace PhpParser;
+namespace PhpParser\Parser;
+
+use PhpParser\Error;
+use PhpParser\Node;
+use PhpParser\Node\Expr;
+use PhpParser\Node\Name;
+use PhpParser\Node\Scalar;
+use PhpParser\Node\Stmt;
+#include;
 
 /* This is an automatically GENERATED file, which should not be manually edited.
  * Instead edit one of the following:
- *  * the grammar file grammar/zend_language_parser.phpy
- *  * the skeleton file grammar/kymacc.php.parser
- *  * the preprocessing script grammar/rebuildParser.php
+ *  * the grammar files grammar/php5.y or grammar/php7.y
+ *  * the skeleton file grammar/parser.template
+ *  * the preprocessing script grammar/rebuildParsers.php
  */
-class Parser extends ParserAbstract
+class #(-p) extends \PhpParser\ParserAbstract
 {
     protected $tokenToSymbolMapSize = #(YYMAXLEX);
     protected $actionTableSize = #(YYLAST);
     protected $gotoTableSize = #(YYGLAST);
 
-    protected $invalidToken = #(YYBADCH);
+    protected $invalidSymbol = #(YYBADCH);
+    protected $errorSymbol = #(YYINTERRTOK);
     protected $defaultAction = #(YYDEFAULT);
     protected $unexpectedTokenRule = #(YYUNEXPECTED);
 
     protected $YY2TBLSTATE  = #(YY2TBLSTATE);
     protected $YYNLSTATES   = #(YYNLSTATES);
 
-#tokenval
-    const %s = %n;
-#endtokenval
-
     protected $symbolToName = array(
         #listvar terminals
     );
@@ -86,7 +90,7 @@ class Parser extends ParserAbstract
 #endif
 #reduce
 
-    protected function reduceRule%n($attributes) {
+    protected function reduceRule%n() {
         %b
     }
 #noact

grammar/php5.y

@@ -1,114 +1,7 @@
 %pure_parser
-%expect 2
+%expect 6
 
-%left T_INCLUDE T_INCLUDE_ONCE T_EVAL T_REQUIRE T_REQUIRE_ONCE
-%left ','
-%left T_LOGICAL_OR
-%left T_LOGICAL_XOR
-%left T_LOGICAL_AND
-%right T_PRINT
-%right T_YIELD
-%left '=' T_PLUS_EQUAL T_MINUS_EQUAL T_MUL_EQUAL T_DIV_EQUAL T_CONCAT_EQUAL T_MOD_EQUAL T_AND_EQUAL T_OR_EQUAL T_XOR_EQUAL T_SL_EQUAL T_SR_EQUAL T_POW_EQUAL
-%left '?' ':'
-%right T_COALESCE
-%left T_BOOLEAN_OR
-%left T_BOOLEAN_AND
-%left '|'
-%left '^'
-%left '&'
-%nonassoc T_IS_EQUAL T_IS_NOT_EQUAL T_IS_IDENTICAL T_IS_NOT_IDENTICAL T_SPACESHIP
-%nonassoc '<' T_IS_SMALLER_OR_EQUAL '>' T_IS_GREATER_OR_EQUAL
-%left T_SL T_SR
-%left '+' '-' '.'
-%left '*' '/' '%'
-%right '!'
-%nonassoc T_INSTANCEOF
-%right '~' T_INC T_DEC T_INT_CAST T_DOUBLE_CAST T_STRING_CAST T_ARRAY_CAST T_OBJECT_CAST T_BOOL_CAST T_UNSET_CAST '@'
-%right T_POW
-%right '['
-%nonassoc T_NEW T_CLONE
-%token T_EXIT
-%token T_IF
-%left T_ELSEIF
-%left T_ELSE
-%left T_ENDIF
-%token T_LNUMBER
-%token T_DNUMBER
-%token T_STRING
-%token T_STRING_VARNAME
-%token T_VARIABLE
-%token T_NUM_STRING
-%token T_INLINE_HTML
-%token T_CHARACTER
-%token T_BAD_CHARACTER
-%token T_ENCAPSED_AND_WHITESPACE
-%token T_CONSTANT_ENCAPSED_STRING
-%token T_ECHO
-%token T_DO
-%token T_WHILE
-%token T_ENDWHILE
-%token T_FOR
-%token T_ENDFOR
-%token T_FOREACH
-%token T_ENDFOREACH
-%token T_DECLARE
-%token T_ENDDECLARE
-%token T_AS
-%token T_SWITCH
-%token T_ENDSWITCH
-%token T_CASE
-%token T_DEFAULT
-%token T_BREAK
-%token T_CONTINUE
-%token T_GOTO
-%token T_FUNCTION
-%token T_CONST
-%token T_RETURN
-%token T_TRY
-%token T_CATCH
-%token T_FINALLY
-%token T_THROW
-%token T_USE
-%token T_INSTEADOF
-%token T_GLOBAL
-%right T_STATIC T_ABSTRACT T_FINAL T_PRIVATE T_PROTECTED T_PUBLIC
-%token T_VAR
-%token T_UNSET
-%token T_ISSET
-%token T_EMPTY
-%token T_HALT_COMPILER
-%token T_CLASS
-%token T_TRAIT
-%token T_INTERFACE
-%token T_EXTENDS
-%token T_IMPLEMENTS
-%token T_OBJECT_OPERATOR
-%token T_DOUBLE_ARROW
-%token T_LIST
-%token T_ARRAY
-%token T_CALLABLE
-%token T_CLASS_C
-%token T_TRAIT_C
-%token T_METHOD_C
-%token T_FUNC_C
-%token T_LINE
-%token T_FILE
-%token T_COMMENT
-%token T_DOC_COMMENT
-%token T_OPEN_TAG
-%token T_OPEN_TAG_WITH_ECHO
-%token T_CLOSE_TAG
-%token T_WHITESPACE
-%token T_START_HEREDOC
-%token T_END_HEREDOC
-%token T_DOLLAR_OPEN_CURLY_BRACES
-%token T_CURLY_OPEN
-%token T_PAAMAYIM_NEKUDOTAYIM
-%token T_NAMESPACE
-%token T_NS_C
-%token T_DIR
-%token T_NS_SEPARATOR
-%token T_ELLIPSIS
+%tokens
 
 %%
 
@@ -116,11 +9,37 @@ start:
     top_statement_list                                      { $$ = $this->handleNamespaces($1); }
 ;
 
-top_statement_list:
-      top_statement_list top_statement                      { pushNormalizing($1, $2); }
+top_statement_list_ex:
+      top_statement_list_ex top_statement                   { pushNormalizing($1, $2); }
     | /* empty */                                           { init(); }
 ;
 
+top_statement_list:
+      top_statement_list_ex
+          { makeNop($nop, $this->lookaheadStartAttributes);
+            if ($nop !== null) { $1[] = $nop; } $$ = $1; }
+;
+
+reserved_non_modifiers:
+      T_INCLUDE | T_INCLUDE_ONCE | T_EVAL | T_REQUIRE | T_REQUIRE_ONCE | T_LOGICAL_OR | T_LOGICAL_XOR | T_LOGICAL_AND
+    | T_INSTANCEOF | T_NEW | T_CLONE | T_EXIT | T_IF | T_ELSEIF | T_ELSE | T_ENDIF | T_ECHO | T_DO | T_WHILE
+    | T_ENDWHILE | T_FOR | T_ENDFOR | T_FOREACH | T_ENDFOREACH | T_DECLARE | T_ENDDECLARE | T_AS | T_TRY | T_CATCH
+    | T_FINALLY | T_THROW | T_USE | T_INSTEADOF | T_GLOBAL | T_VAR | T_UNSET | T_ISSET | T_EMPTY | T_CONTINUE | T_GOTO
+    | T_FUNCTION | T_CONST | T_RETURN | T_PRINT | T_YIELD | T_LIST | T_SWITCH | T_ENDSWITCH | T_CASE | T_DEFAULT
+    | T_BREAK | T_ARRAY | T_CALLABLE | T_EXTENDS | T_IMPLEMENTS | T_NAMESPACE | T_TRAIT | T_INTERFACE | T_CLASS
+    | T_CLASS_C | T_TRAIT_C | T_FUNC_C | T_METHOD_C | T_LINE | T_FILE | T_DIR | T_NS_C | T_HALT_COMPILER
+;
+
+semi_reserved:
+      reserved_non_modifiers
+    | T_STATIC | T_ABSTRACT | T_FINAL | T_PRIVATE | T_PROTECTED | T_PUBLIC
+;
+
+identifier:
+      T_STRING                                              { $$ = $1; }
+    | semi_reserved                                         { $$ = $1; }
+;
+
 namespace_name_parts:
       T_STRING                                              { init($1); }
     | namespace_name_parts T_NS_SEPARATOR T_STRING          { push($1, $3); }
@@ -136,25 +55,66 @@ top_statement:
     | class_declaration_statement                           { $$ = $1; }
     | T_HALT_COMPILER
           { $$ = Stmt\HaltCompiler[$this->lexer->handleHaltCompiler()]; }
-    | T_NAMESPACE namespace_name ';'                        { $$ = Stmt\Namespace_[$2, null]; }
-    | T_NAMESPACE namespace_name '{' top_statement_list '}' { $$ = Stmt\Namespace_[$2, $4]; }
-    | T_NAMESPACE '{' top_statement_list '}'                { $$ = Stmt\Namespace_[null,     $3]; }
+    | T_NAMESPACE namespace_name ';'
+          { $$ = Stmt\Namespace_[$2, null]; $this->checkNamespace($$); }
+    | T_NAMESPACE namespace_name '{' top_statement_list '}'
+          { $$ = Stmt\Namespace_[$2, $4]; $this->checkNamespace($$); }
+    | T_NAMESPACE '{' top_statement_list '}'
+          { $$ = Stmt\Namespace_[null, $3]; $this->checkNamespace($$); }
     | T_USE use_declarations ';'                            { $$ = Stmt\Use_[$2, Stmt\Use_::TYPE_NORMAL]; }
-    | T_USE T_FUNCTION use_declarations ';'                 { $$ = Stmt\Use_[$3, Stmt\Use_::TYPE_FUNCTION]; }
-    | T_USE T_CONST use_declarations ';'                    { $$ = Stmt\Use_[$3, Stmt\Use_::TYPE_CONSTANT]; }
+    | T_USE use_type use_declarations ';'                   { $$ = Stmt\Use_[$3, $2]; }
+    | group_use_declaration ';'                             { $$ = $1; }
     | T_CONST constant_declaration_list ';'                 { $$ = Stmt\Const_[$2]; }
 ;
 
+use_type:
+      T_FUNCTION                                            { $$ = Stmt\Use_::TYPE_FUNCTION; }
+    | T_CONST                                               { $$ = Stmt\Use_::TYPE_CONSTANT; }
+;
+
+/* Using namespace_name_parts here to avoid s/r conflict on T_NS_SEPARATOR */
+group_use_declaration:
+      T_USE use_type namespace_name_parts T_NS_SEPARATOR '{' unprefixed_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($3, stackAttributes(#3)), $6, $2]; }
+    | T_USE use_type T_NS_SEPARATOR namespace_name_parts T_NS_SEPARATOR '{' unprefixed_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($4, stackAttributes(#4)), $7, $2]; }
+    | T_USE namespace_name_parts T_NS_SEPARATOR '{' inline_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($2, stackAttributes(#2)), $5, Stmt\Use_::TYPE_UNKNOWN]; }
+    | T_USE T_NS_SEPARATOR namespace_name_parts T_NS_SEPARATOR '{' inline_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($3, stackAttributes(#3)), $6, Stmt\Use_::TYPE_UNKNOWN]; }
+;
+
+unprefixed_use_declarations:
+      unprefixed_use_declarations ',' unprefixed_use_declaration
+          { push($1, $3); }
+    | unprefixed_use_declaration                            { init($1); }
+;
+
 use_declarations:
       use_declarations ',' use_declaration                  { push($1, $3); }
     | use_declaration                                       { init($1); }
 ;
 
+inline_use_declarations:
+      inline_use_declarations ',' inline_use_declaration    { push($1, $3); }
+    | inline_use_declaration                                { init($1); }
+;
+
+unprefixed_use_declaration:
+      namespace_name
+          { $$ = Stmt\UseUse[$1, null, Stmt\Use_::TYPE_UNKNOWN]; $this->checkUseUse($$, #1); }
+    | namespace_name T_AS T_STRING
+          { $$ = Stmt\UseUse[$1, $3, Stmt\Use_::TYPE_UNKNOWN]; $this->checkUseUse($$, #3); }
+;
+
 use_declaration:
-      namespace_name                                        { $$ = Stmt\UseUse[$1, null]; }
-    | namespace_name T_AS T_STRING                          { $$ = Stmt\UseUse[$1, $3]; }
-    | T_NS_SEPARATOR namespace_name                         { $$ = Stmt\UseUse[$2, null]; }
-    | T_NS_SEPARATOR namespace_name T_AS T_STRING           { $$ = Stmt\UseUse[$2, $4]; }
+      unprefixed_use_declaration                            { $$ = $1; }
+    | T_NS_SEPARATOR unprefixed_use_declaration             { $$ = $2; }
+;
+
+inline_use_declaration:
+      unprefixed_use_declaration                            { $$ = $1; $$->type = Stmt\Use_::TYPE_NORMAL; }
+    | use_type unprefixed_use_declaration                   { $$ = $2; $$->type = $1; }
 ;
 
 constant_declaration_list:
@@ -163,31 +123,47 @@ constant_declaration_list:
 ;
 
 constant_declaration:
-    T_STRING '=' static_scalar                              { $$ = Const_[$1, $3]; }
+    T_STRING '=' static_scalar                              { $$ = Node\Const_[$1, $3]; }
+;
+
+class_const_list:
+      class_const_list ',' class_const                      { push($1, $3); }
+    | class_const                                           { init($1); }
+;
+
+class_const:
+    identifier '=' static_scalar                            { $$ = Node\Const_[$1, $3]; }
+;
+
+inner_statement_list_ex:
+      inner_statement_list_ex inner_statement               { pushNormalizing($1, $2); }
+    | /* empty */                                           { init(); }
 ;
 
 inner_statement_list:
-      inner_statement_list inner_statement                  { pushNormalizing($1, $2); }
-    | /* empty */                                           { init(); }
+      inner_statement_list_ex
+          { makeNop($nop, $this->lookaheadStartAttributes);
+            if ($nop !== null) { $1[] = $nop; } $$ = $1; }
 ;
 
 inner_statement:
       statement                                             { $$ = $1; }
     | function_declaration_statement                        { $$ = $1; }
     | class_declaration_statement                           { $$ = $1; }
-    | T_HALT_COMPILER                                       { error('__HALT_COMPILER() can only be used from the outermost scope'); }
+    | T_HALT_COMPILER
+          { throw new Error('__HALT_COMPILER() can only be used from the outermost scope', attributes()); }
 ;
 
-statement:
-      '{' inner_statement_list '}'                          { $$ = $2; }
+non_empty_statement:
+      '{' inner_statement_list '}'                          { $$ = $2; prependLeadingComments($$); }
     | T_IF parentheses_expr statement elseif_list else_single
-          { $$ = Stmt\If_[$2, [stmts: toArray($3), elseifs: $4, else: $5]]; }
+          { $$ = Stmt\If_[$2, ['stmts' => toArray($3), 'elseifs' => $4, 'else' => $5]]; }
     | T_IF parentheses_expr ':' inner_statement_list new_elseif_list new_else_single T_ENDIF ';'
-          { $$ = Stmt\If_[$2, [stmts: $4, elseifs: $5, else: $6]]; }
+          { $$ = Stmt\If_[$2, ['stmts' => $4, 'elseifs' => $5, 'else' => $6]]; }
     | T_WHILE parentheses_expr while_statement              { $$ = Stmt\While_[$2, $3]; }
     | T_DO statement T_WHILE parentheses_expr ';'           { $$ = Stmt\Do_   [$4, toArray($2)]; }
     | T_FOR '(' for_expr ';'  for_expr ';' for_expr ')' for_statement
-          { $$ = Stmt\For_[[init: $3, cond: $5, loop: $7, stmts: $9]]; }
+          { $$ = Stmt\For_[['init' => $3, 'cond' => $5, 'loop' => $7, 'stmts' => $9]]; }
     | T_SWITCH parentheses_expr switch_case_list            { $$ = Stmt\Switch_[$2, $3]; }
     | T_BREAK ';'                                           { $$ = Stmt\Break_[null]; }
     | T_BREAK expr ';'                                      { $$ = Stmt\Break_[$2]; }
@@ -203,16 +179,24 @@ statement:
     | expr ';'                                              { $$ = $1; }
     | T_UNSET '(' variables_list ')' ';'                    { $$ = Stmt\Unset_[$3]; }
     | T_FOREACH '(' expr T_AS foreach_variable ')' foreach_statement
-          { $$ = Stmt\Foreach_[$3, $5[0], [keyVar: null, byRef: $5[1], stmts: $7]]; }
+          { $$ = Stmt\Foreach_[$3, $5[0], ['keyVar' => null, 'byRef' => $5[1], 'stmts' => $7]]; }
     | T_FOREACH '(' expr T_AS variable T_DOUBLE_ARROW foreach_variable ')' foreach_statement
-          { $$ = Stmt\Foreach_[$3, $7[0], [keyVar: $5, byRef: $7[1], stmts: $9]]; }
+          { $$ = Stmt\Foreach_[$3, $7[0], ['keyVar' => $5, 'byRef' => $7[1], 'stmts' => $9]]; }
     | T_DECLARE '(' declare_list ')' declare_statement      { $$ = Stmt\Declare_[$3, $5]; }
-    | ';'                                                   { $$ = array(); /* means: no statement */ }
     | T_TRY '{' inner_statement_list '}' catches optional_finally
-          { $$ = Stmt\TryCatch[$3, $5, $6]; }
+          { $$ = Stmt\TryCatch[$3, $5, $6]; $this->checkTryCatch($$); }
     | T_THROW expr ';'                                      { $$ = Stmt\Throw_[$2]; }
     | T_GOTO T_STRING ';'                                   { $$ = Stmt\Goto_[$2]; }
     | T_STRING ':'                                          { $$ = Stmt\Label[$1]; }
+    | expr error                                            { $$ = $1; }
+    | error                                                 { $$ = array(); /* means: no statement */ }
+;
+
+statement:
+      non_empty_statement                                   { $$ = $1; }
+    | ';'
+          { makeNop($$, $this->startAttributeStack[#1]);
+            if ($$ === null) $$ = array(); /* means: no statement */ }
 ;
 
 catches:
@@ -222,12 +206,12 @@ catches:
 
 catch:
     T_CATCH '(' name T_VARIABLE ')' '{' inner_statement_list '}'
-        { $$ = Stmt\Catch_[$3, parseVar($4), $7]; }
+        { $$ = Stmt\Catch_[array($3), parseVar($4), $7]; }
 ;
 
 optional_finally:
       /* empty */                                           { $$ = null; }
-    | T_FINALLY '{' inner_statement_list '}'                { $$ = $3; }
+    | T_FINALLY '{' inner_statement_list '}'                { $$ = Stmt\Finally_[$3]; }
 ;
 
 variables_list:
@@ -247,16 +231,18 @@ optional_ellipsis:
 
 function_declaration_statement:
     T_FUNCTION optional_ref T_STRING '(' parameter_list ')' optional_return_type '{' inner_statement_list '}'
-        { $$ = Stmt\Function_[$3, [byRef: $2, params: $5, returnType: $7, stmts: $9]]; }
+        { $$ = Stmt\Function_[$3, ['byRef' => $2, 'params' => $5, 'returnType' => $7, 'stmts' => $9]]; }
 ;
 
 class_declaration_statement:
       class_entry_type T_STRING extends_from implements_list '{' class_statement_list '}'
-          { $$ = Stmt\Class_[$2, [type: $1, extends: $3, implements: $4, stmts: $6]]; }
+          { $$ = Stmt\Class_[$2, ['type' => $1, 'extends' => $3, 'implements' => $4, 'stmts' => $6]];
+            $this->checkClass($$, #2); }
     | T_INTERFACE T_STRING interface_extends_list '{' class_statement_list '}'
-          { $$ = Stmt\Interface_[$2, [extends: $3, stmts: $5]]; }
+          { $$ = Stmt\Interface_[$2, ['extends' => $3, 'stmts' => $5]];
+            $this->checkInterface($$, #2); }
     | T_TRAIT T_STRING '{' class_statement_list '}'
-          { $$ = Stmt\Trait_[$2, $4]; }
+          { $$ = Stmt\Trait_[$2, ['stmts' => $4]]; }
 ;
 
 class_entry_type:
@@ -296,7 +282,8 @@ foreach_statement:
 ;
 
 declare_statement:
-      statement                                             { $$ = toArray($1); }
+      non_empty_statement                                   { $$ = toArray($1); }
+    | ';'                                                   { $$ = null; }
     | ':' inner_statement_list T_ENDDECLARE ';'             { $$ = $2; }
 ;
 
@@ -382,9 +369,9 @@ non_empty_parameter_list:
 
 parameter:
       optional_param_type optional_ref optional_ellipsis T_VARIABLE
-          { $$ = Param[parseVar($4), null, $1, $2, $3]; }
+          { $$ = Node\Param[parseVar($4), null, $1, $2, $3]; $this->checkParam($$); }
     | optional_param_type optional_ref optional_ellipsis T_VARIABLE '=' static_scalar
-          { $$ = Param[parseVar($4), $6, $1, $2, $3]; }
+          { $$ = Node\Param[parseVar($4), $6, $1, $2, $3]; $this->checkParam($$); }
 ;
 
 type:
@@ -406,7 +393,7 @@ optional_return_type:
 argument_list:
       '(' ')'                                               { $$ = array(); }
     | '(' non_empty_argument_list ')'                       { $$ = $2; }
-    | '(' yield_expr ')'                                    { $$ = array(Arg[$2, false, false]); }
+    | '(' yield_expr ')'                                    { $$ = array(Node\Arg[$2, false, false]); }
 ;
 
 non_empty_argument_list:
@@ -415,9 +402,9 @@ non_empty_argument_list:
 ;
 
 argument:
-      expr                                                  { $$ = Arg[$1, false, false]; }
-    | '&' variable                                          { $$ = Arg[$2, true, false]; }
-    | T_ELLIPSIS expr                                       { $$ = Arg[$2, false, true]; }
+      expr                                                  { $$ = Node\Arg[$1, false, false]; }
+    | '&' variable                                          { $$ = Node\Arg[$2, true, false]; }
+    | T_ELLIPSIS expr                                       { $$ = Node\Arg[$2, false, true]; }
 ;
 
 global_var_list:
@@ -447,10 +434,12 @@ class_statement_list:
 ;
 
 class_statement:
-      variable_modifiers property_declaration_list ';'      { $$ = Stmt\Property[$1, $2]; }
-    | T_CONST constant_declaration_list ';'                 { $$ = Stmt\ClassConst[$2]; }
-    | method_modifiers T_FUNCTION optional_ref T_STRING '(' parameter_list ')' optional_return_type method_body
-          { $$ = Stmt\ClassMethod[$4, [type: $1, byRef: $3, params: $6, returnType: $8, stmts: $9]]; }
+      variable_modifiers property_declaration_list ';'
+          { $$ = Stmt\Property[$1, $2]; $this->checkProperty($$, #1); }
+    | T_CONST class_const_list ';'                          { $$ = Stmt\ClassConst[$2, 0]; }
+    | method_modifiers T_FUNCTION optional_ref identifier '(' parameter_list ')' optional_return_type method_body
+          { $$ = Stmt\ClassMethod[$4, ['type' => $1, 'byRef' => $3, 'params' => $6, 'returnType' => $8, 'stmts' => $9]];
+            $this->checkClassMethod($$, #1); }
     | T_USE name_list trait_adaptations                     { $$ = Stmt\TraitUse[$2, $3]; }
 ;
 
@@ -467,20 +456,22 @@ trait_adaptation_list:
 trait_adaptation:
       trait_method_reference_fully_qualified T_INSTEADOF name_list ';'
           { $$ = Stmt\TraitUseAdaptation\Precedence[$1[0], $1[1], $3]; }
-    | trait_method_reference T_AS member_modifier T_STRING ';'
+    | trait_method_reference T_AS member_modifier identifier ';'
           { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], $3, $4]; }
     | trait_method_reference T_AS member_modifier ';'
           { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], $3, null]; }
     | trait_method_reference T_AS T_STRING ';'
           { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], null, $3]; }
+    | trait_method_reference T_AS reserved_non_modifiers ';'
+          { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], null, $3]; }
 ;
 
 trait_method_reference_fully_qualified:
-      name T_PAAMAYIM_NEKUDOTAYIM T_STRING                  { $$ = array($1, $3); }
+      name T_PAAMAYIM_NEKUDOTAYIM identifier                { $$ = array($1, $3); }
 ;
 trait_method_reference:
       trait_method_reference_fully_qualified                { $$ = $1; }
-    | T_STRING                                              { $$ = array(null, $1); }
+    | identifier                                            { $$ = array(null, $1); }
 ;
 
 method_body:
@@ -500,7 +491,7 @@ method_modifiers:
 
 non_empty_member_modifiers:
       member_modifier                                       { $$ = $1; }
-    | non_empty_member_modifiers member_modifier            { Stmt\Class_::verifyModifier($1, $2); $$ = $1 | $2; }
+    | non_empty_member_modifiers member_modifier            { $this->checkModifier($1, $2, #2); $$ = $1 | $2; }
 ;
 
 member_modifier:
@@ -607,7 +598,10 @@ expr:
     | T_OBJECT_CAST expr                                    { $$ = Expr\Cast\Object_ [$2]; }
     | T_BOOL_CAST expr                                      { $$ = Expr\Cast\Bool_   [$2]; }
     | T_UNSET_CAST expr                                     { $$ = Expr\Cast\Unset_  [$2]; }
-    | T_EXIT exit_expr                                      { $$ = Expr\Exit_        [$2]; }
+    | T_EXIT exit_expr
+          { $attrs = attributes();
+            $attrs['kind'] = strtolower($1) === 'exit' ? Expr\Exit_::KIND_EXIT : Expr\Exit_::KIND_DIE;
+            $$ = new Expr\Exit_($2, $attrs); }
     | '@' expr                                              { $$ = Expr\ErrorSuppress[$2]; }
     | scalar                                                { $$ = $1; }
     | array_expr                                            { $$ = $1; }
@@ -615,12 +609,13 @@ expr:
     | '`' backticks_expr '`'                                { $$ = Expr\ShellExec[$2]; }
     | T_PRINT expr                                          { $$ = Expr\Print_[$2]; }
     | T_YIELD                                               { $$ = Expr\Yield_[null, null]; }
+    | T_YIELD_FROM expr                                     { $$ = Expr\YieldFrom[$2]; }
     | T_FUNCTION optional_ref '(' parameter_list ')' lexical_vars optional_return_type
       '{' inner_statement_list '}'
-          { $$ = Expr\Closure[[static: false, byRef: $2, params: $4, uses: $6, returnType: $7, stmts: $9]]; }
+          { $$ = Expr\Closure[['static' => false, 'byRef' => $2, 'params' => $4, 'uses' => $6, 'returnType' => $7, 'stmts' => $9]]; }
     | T_STATIC T_FUNCTION optional_ref '(' parameter_list ')' lexical_vars optional_return_type
       '{' inner_statement_list '}'
-          { $$ = Expr\Closure[[static: true, byRef: $3, params: $5, uses: $7, returnType: $8, stmts: $10]]; }
+          { $$ = Expr\Closure[['static' => true, 'byRef' => $3, 'params' => $5, 'uses' => $7, 'returnType' => $8, 'stmts' => $10]]; }
 ;
 
 parentheses_expr:
@@ -634,21 +629,33 @@ yield_expr:
 ;
 
 array_expr:
-      T_ARRAY '(' array_pair_list ')'                       { $$ = Expr\Array_[$3]; }
-    | '[' array_pair_list ']'                               { $$ = Expr\Array_[$2]; }
+      T_ARRAY '(' array_pair_list ')'
+          { $attrs = attributes(); $attrs['kind'] = Expr\Array_::KIND_LONG;
+            $$ = new Expr\Array_($3, $attrs); }
+    | '[' array_pair_list ']'
+          { $attrs = attributes(); $attrs['kind'] = Expr\Array_::KIND_SHORT;
+            $$ = new Expr\Array_($2, $attrs); }
 ;
 
 scalar_dereference:
       array_expr '[' dim_offset ']'                         { $$ = Expr\ArrayDimFetch[$1, $3]; }
     | T_CONSTANT_ENCAPSED_STRING '[' dim_offset ']'
-          { $$ = Expr\ArrayDimFetch[Scalar\String_[Scalar\String_::parse($1)], $3]; }
+          { $attrs = attributes(); $attrs['kind'] = strKind($1);
+            $$ = Expr\ArrayDimFetch[new Scalar\String_(Scalar\String_::parse($1), $attrs), $3]; }
     | constant '[' dim_offset ']'                           { $$ = Expr\ArrayDimFetch[$1, $3]; }
     | scalar_dereference '[' dim_offset ']'                 { $$ = Expr\ArrayDimFetch[$1, $3]; }
     /* alternative array syntax missing intentionally */
 ;
 
+anonymous_class:
+      T_CLASS ctor_arguments extends_from implements_list '{' class_statement_list '}'
+          { $$ = array(Stmt\Class_[null, ['type' => 0, 'extends' => $3, 'implements' => $4, 'stmts' => $6]], $2);
+            $this->checkClass($$[0], -1); }
+
 new_expr:
       T_NEW class_name_reference ctor_arguments             { $$ = Expr\New_[$2, $3]; }
+    | T_NEW anonymous_class
+          { list($class, $ctorArgs) = $2; $$ = Expr\New_[$class, $ctorArgs]; }
 ;
 
 lexical_vars:
@@ -667,7 +674,7 @@ lexical_var:
 
 function_call:
       name argument_list                                    { $$ = Expr\FuncCall[$1, $2]; }
-    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM T_STRING argument_list
+    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM identifier argument_list
           { $$ = Expr\StaticCall[$1, $3, $4]; }
     | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM '{' expr '}' argument_list
           { $$ = Expr\StaticCall[$1, $4, $6]; }
@@ -735,8 +742,9 @@ exit_expr:
 
 backticks_expr:
       /* empty */                                           { $$ = array(); }
-    | T_ENCAPSED_AND_WHITESPACE                             { $$ = array(Scalar\String_::parseEscapeSequences($1, '`')); }
-    | encaps_list                                           { parseEncapsed($1, '`'); $$ = $1; }
+    | T_ENCAPSED_AND_WHITESPACE
+          { $$ = array(Scalar\EncapsedStringPart[Scalar\String_::parseEscapeSequences($1, '`', false)]); }
+    | encaps_list                                           { parseEncapsed($1, '`', false); $$ = $1; }
 ;
 
 ctor_arguments:
@@ -745,9 +753,11 @@ ctor_arguments:
 ;
 
 common_scalar:
-      T_LNUMBER                                             { $$ = Scalar\LNumber[Scalar\LNumber::parse($1)]; }
+      T_LNUMBER                                             { $$ = $this->parseLNumber($1, attributes(), true); }
     | T_DNUMBER                                             { $$ = Scalar\DNumber[Scalar\DNumber::parse($1)]; }
-    | T_CONSTANT_ENCAPSED_STRING                            { $$ = Scalar\String_[Scalar\String_::parse($1)]; }
+    | T_CONSTANT_ENCAPSED_STRING
+          { $attrs = attributes(); $attrs['kind'] = strKind($1);
+            $$ = new Scalar\String_(Scalar\String_::parse($1, false), $attrs); }
     | T_LINE                                                { $$ = Scalar\MagicConst\Line[]; }
     | T_FILE                                                { $$ = Scalar\MagicConst\File[]; }
     | T_DIR                                                 { $$ = Scalar\MagicConst\Dir[]; }
@@ -757,14 +767,16 @@ common_scalar:
     | T_FUNC_C                                              { $$ = Scalar\MagicConst\Function_[]; }
     | T_NS_C                                                { $$ = Scalar\MagicConst\Namespace_[]; }
     | T_START_HEREDOC T_ENCAPSED_AND_WHITESPACE T_END_HEREDOC
-          { $$ = Scalar\String_[Scalar\String_::parseDocString($1, $2)]; }
+          { $attrs = attributes(); setDocStringAttrs($attrs, $1);
+            $$ = new Scalar\String_(Scalar\String_::parseDocString($1, $2, false), $attrs); }
     | T_START_HEREDOC T_END_HEREDOC
-          { $$ = Scalar\String_['']; }
+          { $attrs = attributes(); setDocStringAttrs($attrs, $1);
+            $$ = new Scalar\String_('', $attrs); }
 ;
 
 static_scalar:
       common_scalar                                         { $$ = $1; }
-    | class_name T_PAAMAYIM_NEKUDOTAYIM class_const_name    { $$ = Expr\ClassConstFetch[$1, $3]; }
+    | class_name T_PAAMAYIM_NEKUDOTAYIM identifier          { $$ = Expr\ClassConstFetch[$1, $3]; }
     | name                                                  { $$ = Expr\ConstFetch[$1]; }
     | T_ARRAY '(' static_array_pair_list ')'                { $$ = Expr\Array_[$3]; }
     | '[' static_array_pair_list ']'                        { $$ = Expr\Array_[$2]; }
@@ -809,7 +821,7 @@ static_operation:
 
 constant:
       name                                                  { $$ = Expr\ConstFetch[$1]; }
-    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM class_const_name
+    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM identifier
           { $$ = Expr\ClassConstFetch[$1, $3]; }
 ;
 
@@ -817,14 +829,11 @@ scalar:
       common_scalar                                         { $$ = $1; }
     | constant                                              { $$ = $1; }
     | '"' encaps_list '"'
-          { parseEncapsed($2, '"'); $$ = Scalar\Encapsed[$2]; }
+          { $attrs = attributes(); $attrs['kind'] = Scalar\String_::KIND_DOUBLE_QUOTED;
+            parseEncapsed($2, '"', true); $$ = new Scalar\Encapsed($2, $attrs); }
     | T_START_HEREDOC encaps_list T_END_HEREDOC
-          { parseEncapsedDoc($2); $$ = Scalar\Encapsed[$2]; }
-;
-
-class_const_name:
-      T_STRING                                              { $$ = $1; }
-    | T_CLASS                                               { $$ = 'class'; }
+          { $attrs = attributes(); setDocStringAttrs($attrs, $1);
+            parseEncapsedDoc($2, true); $$ = new Scalar\Encapsed($2, $attrs); }
 ;
 
 static_array_pair_list:
@@ -916,6 +925,7 @@ object_property:
       T_STRING                                              { $$ = $1; }
     | '{' expr '}'                                          { $$ = $2; }
     | variable_without_objects                              { $$ = $1; }
+    | error                                                 { $$ = Expr\Error[]; $this->errorState = 2; }
 ;
 
 list_expr:
@@ -928,8 +938,8 @@ list_expr_elements:
 ;
 
 list_expr_element:
-      variable                                              { $$ = $1; }
-    | list_expr                                             { $$ = $1; }
+      variable                                              { $$ = Expr\ArrayItem[$1, null, false]; }
+    | list_expr                                             { $$ = Expr\ArrayItem[$1, null, false]; }
     | /* empty */                                           { $$ = null; }
 ;
 
@@ -952,15 +962,23 @@ array_pair:
 
 encaps_list:
       encaps_list encaps_var                                { push($1, $2); }
-    | encaps_list T_ENCAPSED_AND_WHITESPACE                 { push($1, $2); }
+    | encaps_list encaps_string_part                        { push($1, $2); }
     | encaps_var                                            { init($1); }
-    | T_ENCAPSED_AND_WHITESPACE encaps_var                  { init($1, $2); }
+    | encaps_string_part encaps_var                         { init($1, $2); }
+;
+
+encaps_string_part:
+      T_ENCAPSED_AND_WHITESPACE                             { $$ = Scalar\EncapsedStringPart[$1]; }
+;
+
+encaps_base_var:
+      T_VARIABLE                                            { $$ = Expr\Variable[parseVar($1)]; }
 ;
 
 encaps_var:
-      T_VARIABLE                                            { $$ = Expr\Variable[parseVar($1)]; }
-    | T_VARIABLE '[' encaps_var_offset ']'                  { $$ = Expr\ArrayDimFetch[Expr\Variable[parseVar($1)], $3]; }
-    | T_VARIABLE T_OBJECT_OPERATOR T_STRING                 { $$ = Expr\PropertyFetch[Expr\Variable[parseVar($1)], $3]; }
+      encaps_base_var                                       { $$ = $1; }
+    | encaps_base_var '[' encaps_var_offset ']'             { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | encaps_base_var T_OBJECT_OPERATOR T_STRING            { $$ = Expr\PropertyFetch[$1, $3]; }
     | T_DOLLAR_OPEN_CURLY_BRACES expr '}'                   { $$ = Expr\Variable[$2]; }
     | T_DOLLAR_OPEN_CURLY_BRACES T_STRING_VARNAME '}'       { $$ = Expr\Variable[$2]; }
     | T_DOLLAR_OPEN_CURLY_BRACES T_STRING_VARNAME '[' expr ']' '}'
@@ -970,7 +988,7 @@ encaps_var:
 
 encaps_var_offset:
       T_STRING                                              { $$ = Scalar\String_[$1]; }
-    | T_NUM_STRING                                          { $$ = Scalar\String_[$1]; }
+    | T_NUM_STRING                                          { $$ = $this->parseNumString($1, attributes()); }
     | T_VARIABLE                                            { $$ = Expr\Variable[parseVar($1)]; }
 ;
 

grammar/php7.y

@@ -0,0 +1,883 @@
+%pure_parser
+%expect 2
+
+%tokens
+
+%%
+
+start:
+    top_statement_list                                      { $$ = $this->handleNamespaces($1); }
+;
+
+top_statement_list_ex:
+      top_statement_list_ex top_statement                   { pushNormalizing($1, $2); }
+    | /* empty */                                           { init(); }
+;
+
+top_statement_list:
+      top_statement_list_ex
+          { makeNop($nop, $this->lookaheadStartAttributes);
+            if ($nop !== null) { $1[] = $nop; } $$ = $1; }
+;
+
+reserved_non_modifiers:
+      T_INCLUDE | T_INCLUDE_ONCE | T_EVAL | T_REQUIRE | T_REQUIRE_ONCE | T_LOGICAL_OR | T_LOGICAL_XOR | T_LOGICAL_AND
+    | T_INSTANCEOF | T_NEW | T_CLONE | T_EXIT | T_IF | T_ELSEIF | T_ELSE | T_ENDIF | T_ECHO | T_DO | T_WHILE
+    | T_ENDWHILE | T_FOR | T_ENDFOR | T_FOREACH | T_ENDFOREACH | T_DECLARE | T_ENDDECLARE | T_AS | T_TRY | T_CATCH
+    | T_FINALLY | T_THROW | T_USE | T_INSTEADOF | T_GLOBAL | T_VAR | T_UNSET | T_ISSET | T_EMPTY | T_CONTINUE | T_GOTO
+    | T_FUNCTION | T_CONST | T_RETURN | T_PRINT | T_YIELD | T_LIST | T_SWITCH | T_ENDSWITCH | T_CASE | T_DEFAULT
+    | T_BREAK | T_ARRAY | T_CALLABLE | T_EXTENDS | T_IMPLEMENTS | T_NAMESPACE | T_TRAIT | T_INTERFACE | T_CLASS
+    | T_CLASS_C | T_TRAIT_C | T_FUNC_C | T_METHOD_C | T_LINE | T_FILE | T_DIR | T_NS_C | T_HALT_COMPILER
+;
+
+semi_reserved:
+      reserved_non_modifiers
+    | T_STATIC | T_ABSTRACT | T_FINAL | T_PRIVATE | T_PROTECTED | T_PUBLIC
+;
+
+identifier:
+      T_STRING                                              { $$ = $1; }
+    | semi_reserved                                         { $$ = $1; }
+;
+
+namespace_name_parts:
+      T_STRING                                              { init($1); }
+    | namespace_name_parts T_NS_SEPARATOR T_STRING          { push($1, $3); }
+;
+
+namespace_name:
+      namespace_name_parts                                  { $$ = Name[$1]; }
+;
+
+top_statement:
+      statement                                             { $$ = $1; }
+    | function_declaration_statement                        { $$ = $1; }
+    | class_declaration_statement                           { $$ = $1; }
+    | T_HALT_COMPILER
+          { $$ = Stmt\HaltCompiler[$this->lexer->handleHaltCompiler()]; }
+    | T_NAMESPACE namespace_name ';'
+          { $$ = Stmt\Namespace_[$2, null]; $this->checkNamespace($$); }
+    | T_NAMESPACE namespace_name '{' top_statement_list '}'
+          { $$ = Stmt\Namespace_[$2, $4]; $this->checkNamespace($$); }
+    | T_NAMESPACE '{' top_statement_list '}'
+          { $$ = Stmt\Namespace_[null, $3]; $this->checkNamespace($$); }
+    | T_USE use_declarations ';'                            { $$ = Stmt\Use_[$2, Stmt\Use_::TYPE_NORMAL]; }
+    | T_USE use_type use_declarations ';'                   { $$ = Stmt\Use_[$3, $2]; }
+    | group_use_declaration ';'                             { $$ = $1; }
+    | T_CONST constant_declaration_list ';'                 { $$ = Stmt\Const_[$2]; }
+;
+
+use_type:
+      T_FUNCTION                                            { $$ = Stmt\Use_::TYPE_FUNCTION; }
+    | T_CONST                                               { $$ = Stmt\Use_::TYPE_CONSTANT; }
+;
+
+/* Using namespace_name_parts here to avoid s/r conflict on T_NS_SEPARATOR */
+group_use_declaration:
+      T_USE use_type namespace_name_parts T_NS_SEPARATOR '{' unprefixed_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($3, stackAttributes(#3)), $6, $2]; }
+    | T_USE use_type T_NS_SEPARATOR namespace_name_parts T_NS_SEPARATOR '{' unprefixed_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($4, stackAttributes(#4)), $7, $2]; }
+    | T_USE namespace_name_parts T_NS_SEPARATOR '{' inline_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($2, stackAttributes(#2)), $5, Stmt\Use_::TYPE_UNKNOWN]; }
+    | T_USE T_NS_SEPARATOR namespace_name_parts T_NS_SEPARATOR '{' inline_use_declarations '}'
+          { $$ = Stmt\GroupUse[new Name($3, stackAttributes(#3)), $6, Stmt\Use_::TYPE_UNKNOWN]; }
+;
+
+unprefixed_use_declarations:
+      unprefixed_use_declarations ',' unprefixed_use_declaration
+          { push($1, $3); }
+    | unprefixed_use_declaration                            { init($1); }
+;
+
+use_declarations:
+      use_declarations ',' use_declaration                  { push($1, $3); }
+    | use_declaration                                       { init($1); }
+;
+
+inline_use_declarations:
+      inline_use_declarations ',' inline_use_declaration    { push($1, $3); }
+    | inline_use_declaration                                { init($1); }
+;
+
+unprefixed_use_declaration:
+      namespace_name
+          { $$ = Stmt\UseUse[$1, null, Stmt\Use_::TYPE_UNKNOWN]; $this->checkUseUse($$, #1); }
+    | namespace_name T_AS T_STRING
+          { $$ = Stmt\UseUse[$1, $3, Stmt\Use_::TYPE_UNKNOWN]; $this->checkUseUse($$, #3); }
+;
+
+use_declaration:
+      unprefixed_use_declaration                            { $$ = $1; }
+    | T_NS_SEPARATOR unprefixed_use_declaration             { $$ = $2; }
+;
+
+inline_use_declaration:
+      unprefixed_use_declaration                            { $$ = $1; $$->type = Stmt\Use_::TYPE_NORMAL; }
+    | use_type unprefixed_use_declaration                   { $$ = $2; $$->type = $1; }
+;
+
+constant_declaration_list:
+      constant_declaration_list ',' constant_declaration    { push($1, $3); }
+    | constant_declaration                                  { init($1); }
+;
+
+constant_declaration:
+    T_STRING '=' expr                                       { $$ = Node\Const_[$1, $3]; }
+;
+
+class_const_list:
+      class_const_list ',' class_const                      { push($1, $3); }
+    | class_const                                           { init($1); }
+;
+
+class_const:
+    identifier '=' expr                                     { $$ = Node\Const_[$1, $3]; }
+;
+
+inner_statement_list_ex:
+      inner_statement_list_ex inner_statement               { pushNormalizing($1, $2); }
+    | /* empty */                                           { init(); }
+;
+
+inner_statement_list:
+      inner_statement_list_ex
+          { makeNop($nop, $this->lookaheadStartAttributes);
+            if ($nop !== null) { $1[] = $nop; } $$ = $1; }
+;
+
+inner_statement:
+      statement                                             { $$ = $1; }
+    | function_declaration_statement                        { $$ = $1; }
+    | class_declaration_statement                           { $$ = $1; }
+    | T_HALT_COMPILER
+          { throw new Error('__HALT_COMPILER() can only be used from the outermost scope', attributes()); }
+;
+
+non_empty_statement:
+      '{' inner_statement_list '}'                          { $$ = $2; prependLeadingComments($$); }
+    | T_IF '(' expr ')' statement elseif_list else_single
+          { $$ = Stmt\If_[$3, ['stmts' => toArray($5), 'elseifs' => $6, 'else' => $7]]; }
+    | T_IF '(' expr ')' ':' inner_statement_list new_elseif_list new_else_single T_ENDIF ';'
+          { $$ = Stmt\If_[$3, ['stmts' => $6, 'elseifs' => $7, 'else' => $8]]; }
+    | T_WHILE '(' expr ')' while_statement                  { $$ = Stmt\While_[$3, $5]; }
+    | T_DO statement T_WHILE '(' expr ')' ';'               { $$ = Stmt\Do_   [$5, toArray($2)]; }
+    | T_FOR '(' for_expr ';'  for_expr ';' for_expr ')' for_statement
+          { $$ = Stmt\For_[['init' => $3, 'cond' => $5, 'loop' => $7, 'stmts' => $9]]; }
+    | T_SWITCH '(' expr ')' switch_case_list                { $$ = Stmt\Switch_[$3, $5]; }
+    | T_BREAK optional_expr ';'                             { $$ = Stmt\Break_[$2]; }
+    | T_CONTINUE optional_expr ';'                          { $$ = Stmt\Continue_[$2]; }
+    | T_RETURN optional_expr ';'                            { $$ = Stmt\Return_[$2]; }
+    | T_GLOBAL global_var_list ';'                          { $$ = Stmt\Global_[$2]; }
+    | T_STATIC static_var_list ';'                          { $$ = Stmt\Static_[$2]; }
+    | T_ECHO expr_list ';'                                  { $$ = Stmt\Echo_[$2]; }
+    | T_INLINE_HTML                                         { $$ = Stmt\InlineHTML[$1]; }
+    | expr ';'                                              { $$ = $1; }
+    | T_UNSET '(' variables_list ')' ';'                    { $$ = Stmt\Unset_[$3]; }
+    | T_FOREACH '(' expr T_AS foreach_variable ')' foreach_statement
+          { $$ = Stmt\Foreach_[$3, $5[0], ['keyVar' => null, 'byRef' => $5[1], 'stmts' => $7]]; }
+    | T_FOREACH '(' expr T_AS variable T_DOUBLE_ARROW foreach_variable ')' foreach_statement
+          { $$ = Stmt\Foreach_[$3, $7[0], ['keyVar' => $5, 'byRef' => $7[1], 'stmts' => $9]]; }
+    | T_DECLARE '(' declare_list ')' declare_statement      { $$ = Stmt\Declare_[$3, $5]; }
+    | T_TRY '{' inner_statement_list '}' catches optional_finally
+          { $$ = Stmt\TryCatch[$3, $5, $6]; $this->checkTryCatch($$); }
+    | T_THROW expr ';'                                      { $$ = Stmt\Throw_[$2]; }
+    | T_GOTO T_STRING ';'                                   { $$ = Stmt\Goto_[$2]; }
+    | T_STRING ':'                                          { $$ = Stmt\Label[$1]; }
+    | expr error                                            { $$ = $1; }
+    | error                                                 { $$ = array(); /* means: no statement */ }
+;
+
+statement:
+      non_empty_statement                                   { $$ = $1; }
+    | ';'
+          { makeNop($$, $this->startAttributeStack[#1]);
+            if ($$ === null) $$ = array(); /* means: no statement */ }
+;
+
+catches:
+      /* empty */                                           { init(); }
+    | catches catch                                         { push($1, $2); }
+;
+
+name_union:
+      name                                                  { init($1); }
+    | name_union '|' name                                   { push($1, $3); }
+;
+
+catch:
+    T_CATCH '(' name_union T_VARIABLE ')' '{' inner_statement_list '}'
+        { $$ = Stmt\Catch_[$3, parseVar($4), $7]; }
+;
+
+optional_finally:
+      /* empty */                                           { $$ = null; }
+    | T_FINALLY '{' inner_statement_list '}'                { $$ = Stmt\Finally_[$3]; }
+;
+
+variables_list:
+      variable                                              { init($1); }
+    | variables_list ',' variable                           { push($1, $3); }
+;
+
+optional_ref:
+      /* empty */                                           { $$ = false; }
+    | '&'                                                   { $$ = true; }
+;
+
+optional_ellipsis:
+      /* empty */                                           { $$ = false; }
+    | T_ELLIPSIS                                            { $$ = true; }
+;
+
+function_declaration_statement:
+    T_FUNCTION optional_ref T_STRING '(' parameter_list ')' optional_return_type '{' inner_statement_list '}'
+        { $$ = Stmt\Function_[$3, ['byRef' => $2, 'params' => $5, 'returnType' => $7, 'stmts' => $9]]; }
+;
+
+class_declaration_statement:
+      class_entry_type T_STRING extends_from implements_list '{' class_statement_list '}'
+          { $$ = Stmt\Class_[$2, ['type' => $1, 'extends' => $3, 'implements' => $4, 'stmts' => $6]];
+            $this->checkClass($$, #2); }
+    | T_INTERFACE T_STRING interface_extends_list '{' class_statement_list '}'
+          { $$ = Stmt\Interface_[$2, ['extends' => $3, 'stmts' => $5]];
+            $this->checkInterface($$, #2); }
+    | T_TRAIT T_STRING '{' class_statement_list '}'
+          { $$ = Stmt\Trait_[$2, ['stmts' => $4]]; }
+;
+
+class_entry_type:
+      T_CLASS                                               { $$ = 0; }
+    | T_ABSTRACT T_CLASS                                    { $$ = Stmt\Class_::MODIFIER_ABSTRACT; }
+    | T_FINAL T_CLASS                                       { $$ = Stmt\Class_::MODIFIER_FINAL; }
+;
+
+extends_from:
+      /* empty */                                           { $$ = null; }
+    | T_EXTENDS name                                        { $$ = $2; }
+;
+
+interface_extends_list:
+      /* empty */                                           { $$ = array(); }
+    | T_EXTENDS name_list                                   { $$ = $2; }
+;
+
+implements_list:
+      /* empty */                                           { $$ = array(); }
+    | T_IMPLEMENTS name_list                                { $$ = $2; }
+;
+
+name_list:
+      name                                                  { init($1); }
+    | name_list ',' name                                    { push($1, $3); }
+;
+
+for_statement:
+      statement                                             { $$ = toArray($1); }
+    | ':' inner_statement_list T_ENDFOR ';'                 { $$ = $2; }
+;
+
+foreach_statement:
+      statement                                             { $$ = toArray($1); }
+    | ':' inner_statement_list T_ENDFOREACH ';'             { $$ = $2; }
+;
+
+declare_statement:
+      non_empty_statement                                   { $$ = toArray($1); }
+    | ';'                                                   { $$ = null; }
+    | ':' inner_statement_list T_ENDDECLARE ';'             { $$ = $2; }
+;
+
+declare_list:
+      declare_list_element                                  { init($1); }
+    | declare_list ',' declare_list_element                 { push($1, $3); }
+;
+
+declare_list_element:
+      T_STRING '=' expr                                     { $$ = Stmt\DeclareDeclare[$1, $3]; }
+;
+
+switch_case_list:
+      '{' case_list '}'                                     { $$ = $2; }
+    | '{' ';' case_list '}'                                 { $$ = $3; }
+    | ':' case_list T_ENDSWITCH ';'                         { $$ = $2; }
+    | ':' ';' case_list T_ENDSWITCH ';'                     { $$ = $3; }
+;
+
+case_list:
+      /* empty */                                           { init(); }
+    | case_list case                                        { push($1, $2); }
+;
+
+case:
+      T_CASE expr case_separator inner_statement_list       { $$ = Stmt\Case_[$2, $4]; }
+    | T_DEFAULT case_separator inner_statement_list         { $$ = Stmt\Case_[null, $3]; }
+;
+
+case_separator:
+      ':'
+    | ';'
+;
+
+while_statement:
+      statement                                             { $$ = toArray($1); }
+    | ':' inner_statement_list T_ENDWHILE ';'               { $$ = $2; }
+;
+
+elseif_list:
+      /* empty */                                           { init(); }
+    | elseif_list elseif                                    { push($1, $2); }
+;
+
+elseif:
+      T_ELSEIF '(' expr ')' statement                       { $$ = Stmt\ElseIf_[$3, toArray($5)]; }
+;
+
+new_elseif_list:
+      /* empty */                                           { init(); }
+    | new_elseif_list new_elseif                            { push($1, $2); }
+;
+
+new_elseif:
+     T_ELSEIF '(' expr ')' ':' inner_statement_list         { $$ = Stmt\ElseIf_[$3, $6]; }
+;
+
+else_single:
+      /* empty */                                           { $$ = null; }
+    | T_ELSE statement                                      { $$ = Stmt\Else_[toArray($2)]; }
+;
+
+new_else_single:
+      /* empty */                                           { $$ = null; }
+    | T_ELSE ':' inner_statement_list                       { $$ = Stmt\Else_[$3]; }
+;
+
+foreach_variable:
+      variable                                              { $$ = array($1, false); }
+    | '&' variable                                          { $$ = array($2, true); }
+    | list_expr                                             { $$ = array($1, false); }
+    | array_short_syntax                                    { $$ = array($1, false); }
+;
+
+parameter_list:
+      non_empty_parameter_list                              { $$ = $1; }
+    | /* empty */                                           { $$ = array(); }
+;
+
+non_empty_parameter_list:
+      parameter                                             { init($1); }
+    | non_empty_parameter_list ',' parameter                { push($1, $3); }
+;
+
+parameter:
+      optional_param_type optional_ref optional_ellipsis T_VARIABLE
+          { $$ = Node\Param[parseVar($4), null, $1, $2, $3]; $this->checkParam($$); }
+    | optional_param_type optional_ref optional_ellipsis T_VARIABLE '=' expr
+          { $$ = Node\Param[parseVar($4), $6, $1, $2, $3]; $this->checkParam($$); }
+;
+
+type_expr:
+      type                                                  { $$ = $1; }
+    | '?' type                                              { $$ = Node\NullableType[$2]; }
+;
+
+type:
+      name                                                  { $$ = $this->handleBuiltinTypes($1); }
+    | T_ARRAY                                               { $$ = 'array'; }
+    | T_CALLABLE                                            { $$ = 'callable'; }
+;
+
+optional_param_type:
+      /* empty */                                           { $$ = null; }
+    | type_expr                                             { $$ = $1; }
+;
+
+optional_return_type:
+      /* empty */                                           { $$ = null; }
+    | ':' type_expr                                         { $$ = $2; }
+;
+
+argument_list:
+      '(' ')'                                               { $$ = array(); }
+    | '(' non_empty_argument_list ')'                       { $$ = $2; }
+;
+
+non_empty_argument_list:
+      argument                                              { init($1); }
+    | non_empty_argument_list ',' argument                  { push($1, $3); }
+;
+
+argument:
+      expr                                                  { $$ = Node\Arg[$1, false, false]; }
+    | '&' variable                                          { $$ = Node\Arg[$2, true, false]; }
+    | T_ELLIPSIS expr                                       { $$ = Node\Arg[$2, false, true]; }
+;
+
+global_var_list:
+      global_var_list ',' global_var                        { push($1, $3); }
+    | global_var                                            { init($1); }
+;
+
+global_var:
+      simple_variable                                       { $$ = Expr\Variable[$1]; }
+;
+
+static_var_list:
+      static_var_list ',' static_var                        { push($1, $3); }
+    | static_var                                            { init($1); }
+;
+
+static_var:
+      T_VARIABLE                                            { $$ = Stmt\StaticVar[parseVar($1), null]; }
+    | T_VARIABLE '=' expr                                   { $$ = Stmt\StaticVar[parseVar($1), $3]; }
+;
+
+class_statement_list:
+      class_statement_list class_statement                  { push($1, $2); }
+    | /* empty */                                           { init(); }
+;
+
+class_statement:
+      variable_modifiers property_declaration_list ';'
+          { $$ = Stmt\Property[$1, $2]; $this->checkProperty($$, #1); }
+    | method_modifiers T_CONST class_const_list ';'
+          { $$ = Stmt\ClassConst[$3, $1]; $this->checkClassConst($$, #1); }
+    | method_modifiers T_FUNCTION optional_ref identifier '(' parameter_list ')' optional_return_type method_body
+          { $$ = Stmt\ClassMethod[$4, ['type' => $1, 'byRef' => $3, 'params' => $6, 'returnType' => $8, 'stmts' => $9]];
+            $this->checkClassMethod($$, #1); }
+    | T_USE name_list trait_adaptations                     { $$ = Stmt\TraitUse[$2, $3]; }
+;
+
+trait_adaptations:
+      ';'                                                   { $$ = array(); }
+    | '{' trait_adaptation_list '}'                         { $$ = $2; }
+;
+
+trait_adaptation_list:
+      /* empty */                                           { init(); }
+    | trait_adaptation_list trait_adaptation                { push($1, $2); }
+;
+
+trait_adaptation:
+      trait_method_reference_fully_qualified T_INSTEADOF name_list ';'
+          { $$ = Stmt\TraitUseAdaptation\Precedence[$1[0], $1[1], $3]; }
+    | trait_method_reference T_AS member_modifier identifier ';'
+          { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], $3, $4]; }
+    | trait_method_reference T_AS member_modifier ';'
+          { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], $3, null]; }
+    | trait_method_reference T_AS T_STRING ';'
+          { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], null, $3]; }
+    | trait_method_reference T_AS reserved_non_modifiers ';'
+          { $$ = Stmt\TraitUseAdaptation\Alias[$1[0], $1[1], null, $3]; }
+;
+
+trait_method_reference_fully_qualified:
+      name T_PAAMAYIM_NEKUDOTAYIM identifier                { $$ = array($1, $3); }
+;
+trait_method_reference:
+      trait_method_reference_fully_qualified                { $$ = $1; }
+    | identifier                                            { $$ = array(null, $1); }
+;
+
+method_body:
+      ';' /* abstract method */                             { $$ = null; }
+    | '{' inner_statement_list '}'                          { $$ = $2; }
+;
+
+variable_modifiers:
+      non_empty_member_modifiers                            { $$ = $1; }
+    | T_VAR                                                 { $$ = 0; }
+;
+
+method_modifiers:
+      /* empty */                                           { $$ = 0; }
+    | non_empty_member_modifiers                            { $$ = $1; }
+;
+
+non_empty_member_modifiers:
+      member_modifier                                       { $$ = $1; }
+    | non_empty_member_modifiers member_modifier            { $this->checkModifier($1, $2, #2); $$ = $1 | $2; }
+;
+
+member_modifier:
+      T_PUBLIC                                              { $$ = Stmt\Class_::MODIFIER_PUBLIC; }
+    | T_PROTECTED                                           { $$ = Stmt\Class_::MODIFIER_PROTECTED; }
+    | T_PRIVATE                                             { $$ = Stmt\Class_::MODIFIER_PRIVATE; }
+    | T_STATIC                                              { $$ = Stmt\Class_::MODIFIER_STATIC; }
+    | T_ABSTRACT                                            { $$ = Stmt\Class_::MODIFIER_ABSTRACT; }
+    | T_FINAL                                               { $$ = Stmt\Class_::MODIFIER_FINAL; }
+;
+
+property_declaration_list:
+      property_declaration                                  { init($1); }
+    | property_declaration_list ',' property_declaration    { push($1, $3); }
+;
+
+property_declaration:
+      T_VARIABLE                                            { $$ = Stmt\PropertyProperty[parseVar($1), null]; }
+    | T_VARIABLE '=' expr                                   { $$ = Stmt\PropertyProperty[parseVar($1), $3]; }
+;
+
+expr_list:
+      expr_list ',' expr                                    { push($1, $3); }
+    | expr                                                  { init($1); }
+;
+
+for_expr:
+      /* empty */                                           { $$ = array(); }
+    | expr_list                                             { $$ = $1; }
+;
+
+expr:
+      variable                                              { $$ = $1; }
+    | list_expr '=' expr                                    { $$ = Expr\Assign[$1, $3]; }
+    | array_short_syntax '=' expr                           { $$ = Expr\Assign[$1, $3]; }
+    | variable '=' expr                                     { $$ = Expr\Assign[$1, $3]; }
+    | variable '=' '&' variable                             { $$ = Expr\AssignRef[$1, $4]; }
+    | new_expr                                              { $$ = $1; }
+    | T_CLONE expr                                          { $$ = Expr\Clone_[$2]; }
+    | variable T_PLUS_EQUAL expr                            { $$ = Expr\AssignOp\Plus      [$1, $3]; }
+    | variable T_MINUS_EQUAL expr                           { $$ = Expr\AssignOp\Minus     [$1, $3]; }
+    | variable T_MUL_EQUAL expr                             { $$ = Expr\AssignOp\Mul       [$1, $3]; }
+    | variable T_DIV_EQUAL expr                             { $$ = Expr\AssignOp\Div       [$1, $3]; }
+    | variable T_CONCAT_EQUAL expr                          { $$ = Expr\AssignOp\Concat    [$1, $3]; }
+    | variable T_MOD_EQUAL expr                             { $$ = Expr\AssignOp\Mod       [$1, $3]; }
+    | variable T_AND_EQUAL expr                             { $$ = Expr\AssignOp\BitwiseAnd[$1, $3]; }
+    | variable T_OR_EQUAL expr                              { $$ = Expr\AssignOp\BitwiseOr [$1, $3]; }
+    | variable T_XOR_EQUAL expr                             { $$ = Expr\AssignOp\BitwiseXor[$1, $3]; }
+    | variable T_SL_EQUAL expr                              { $$ = Expr\AssignOp\ShiftLeft [$1, $3]; }
+    | variable T_SR_EQUAL expr                              { $$ = Expr\AssignOp\ShiftRight[$1, $3]; }
+    | variable T_POW_EQUAL expr                             { $$ = Expr\AssignOp\Pow       [$1, $3]; }
+    | variable T_INC                                        { $$ = Expr\PostInc[$1]; }
+    | T_INC variable                                        { $$ = Expr\PreInc [$2]; }
+    | variable T_DEC                                        { $$ = Expr\PostDec[$1]; }
+    | T_DEC variable                                        { $$ = Expr\PreDec [$2]; }
+    | expr T_BOOLEAN_OR expr                                { $$ = Expr\BinaryOp\BooleanOr [$1, $3]; }
+    | expr T_BOOLEAN_AND expr                               { $$ = Expr\BinaryOp\BooleanAnd[$1, $3]; }
+    | expr T_LOGICAL_OR expr                                { $$ = Expr\BinaryOp\LogicalOr [$1, $3]; }
+    | expr T_LOGICAL_AND expr                               { $$ = Expr\BinaryOp\LogicalAnd[$1, $3]; }
+    | expr T_LOGICAL_XOR expr                               { $$ = Expr\BinaryOp\LogicalXor[$1, $3]; }
+    | expr '|' expr                                         { $$ = Expr\BinaryOp\BitwiseOr [$1, $3]; }
+    | expr '&' expr                                         { $$ = Expr\BinaryOp\BitwiseAnd[$1, $3]; }
+    | expr '^' expr                                         { $$ = Expr\BinaryOp\BitwiseXor[$1, $3]; }
+    | expr '.' expr                                         { $$ = Expr\BinaryOp\Concat    [$1, $3]; }
+    | expr '+' expr                                         { $$ = Expr\BinaryOp\Plus      [$1, $3]; }
+    | expr '-' expr                                         { $$ = Expr\BinaryOp\Minus     [$1, $3]; }
+    | expr '*' expr                                         { $$ = Expr\BinaryOp\Mul       [$1, $3]; }
+    | expr '/' expr                                         { $$ = Expr\BinaryOp\Div       [$1, $3]; }
+    | expr '%' expr                                         { $$ = Expr\BinaryOp\Mod       [$1, $3]; }
+    | expr T_SL expr                                        { $$ = Expr\BinaryOp\ShiftLeft [$1, $3]; }
+    | expr T_SR expr                                        { $$ = Expr\BinaryOp\ShiftRight[$1, $3]; }
+    | expr T_POW expr                                       { $$ = Expr\BinaryOp\Pow       [$1, $3]; }
+    | '+' expr %prec T_INC                                  { $$ = Expr\UnaryPlus [$2]; }
+    | '-' expr %prec T_INC                                  { $$ = Expr\UnaryMinus[$2]; }
+    | '!' expr                                              { $$ = Expr\BooleanNot[$2]; }
+    | '~' expr                                              { $$ = Expr\BitwiseNot[$2]; }
+    | expr T_IS_IDENTICAL expr                              { $$ = Expr\BinaryOp\Identical     [$1, $3]; }
+    | expr T_IS_NOT_IDENTICAL expr                          { $$ = Expr\BinaryOp\NotIdentical  [$1, $3]; }
+    | expr T_IS_EQUAL expr                                  { $$ = Expr\BinaryOp\Equal         [$1, $3]; }
+    | expr T_IS_NOT_EQUAL expr                              { $$ = Expr\BinaryOp\NotEqual      [$1, $3]; }
+    | expr T_SPACESHIP expr                                 { $$ = Expr\BinaryOp\Spaceship     [$1, $3]; }
+    | expr '<' expr                                         { $$ = Expr\BinaryOp\Smaller       [$1, $3]; }
+    | expr T_IS_SMALLER_OR_EQUAL expr                       { $$ = Expr\BinaryOp\SmallerOrEqual[$1, $3]; }
+    | expr '>' expr                                         { $$ = Expr\BinaryOp\Greater       [$1, $3]; }
+    | expr T_IS_GREATER_OR_EQUAL expr                       { $$ = Expr\BinaryOp\GreaterOrEqual[$1, $3]; }
+    | expr T_INSTANCEOF class_name_reference                { $$ = Expr\Instanceof_[$1, $3]; }
+    | '(' expr ')'                                          { $$ = $2; }
+    | expr '?' expr ':' expr                                { $$ = Expr\Ternary[$1, $3,   $5]; }
+    | expr '?' ':' expr                                     { $$ = Expr\Ternary[$1, null, $4]; }
+    | expr T_COALESCE expr                                  { $$ = Expr\BinaryOp\Coalesce[$1, $3]; }
+    | T_ISSET '(' variables_list ')'                        { $$ = Expr\Isset_[$3]; }
+    | T_EMPTY '(' expr ')'                                  { $$ = Expr\Empty_[$3]; }
+    | T_INCLUDE expr                                        { $$ = Expr\Include_[$2, Expr\Include_::TYPE_INCLUDE]; }
+    | T_INCLUDE_ONCE expr                                   { $$ = Expr\Include_[$2, Expr\Include_::TYPE_INCLUDE_ONCE]; }
+    | T_EVAL '(' expr ')'                                   { $$ = Expr\Eval_[$3]; }
+    | T_REQUIRE expr                                        { $$ = Expr\Include_[$2, Expr\Include_::TYPE_REQUIRE]; }
+    | T_REQUIRE_ONCE expr                                   { $$ = Expr\Include_[$2, Expr\Include_::TYPE_REQUIRE_ONCE]; }
+    | T_INT_CAST expr                                       { $$ = Expr\Cast\Int_    [$2]; }
+    | T_DOUBLE_CAST expr                                    { $$ = Expr\Cast\Double  [$2]; }
+    | T_STRING_CAST expr                                    { $$ = Expr\Cast\String_ [$2]; }
+    | T_ARRAY_CAST expr                                     { $$ = Expr\Cast\Array_  [$2]; }
+    | T_OBJECT_CAST expr                                    { $$ = Expr\Cast\Object_ [$2]; }
+    | T_BOOL_CAST expr                                      { $$ = Expr\Cast\Bool_   [$2]; }
+    | T_UNSET_CAST expr                                     { $$ = Expr\Cast\Unset_  [$2]; }
+    | T_EXIT exit_expr
+          { $attrs = attributes();
+            $attrs['kind'] = strtolower($1) === 'exit' ? Expr\Exit_::KIND_EXIT : Expr\Exit_::KIND_DIE;
+            $$ = new Expr\Exit_($2, $attrs); }
+    | '@' expr                                              { $$ = Expr\ErrorSuppress[$2]; }
+    | scalar                                                { $$ = $1; }
+    | '`' backticks_expr '`'                                { $$ = Expr\ShellExec[$2]; }
+    | T_PRINT expr                                          { $$ = Expr\Print_[$2]; }
+    | T_YIELD                                               { $$ = Expr\Yield_[null, null]; }
+    | T_YIELD expr                                          { $$ = Expr\Yield_[$2, null]; }
+    | T_YIELD expr T_DOUBLE_ARROW expr                      { $$ = Expr\Yield_[$4, $2]; }
+    | T_YIELD_FROM expr                                     { $$ = Expr\YieldFrom[$2]; }
+    | T_FUNCTION optional_ref '(' parameter_list ')' lexical_vars optional_return_type
+      '{' inner_statement_list '}'
+          { $$ = Expr\Closure[['static' => false, 'byRef' => $2, 'params' => $4, 'uses' => $6, 'returnType' => $7, 'stmts' => $9]]; }
+    | T_STATIC T_FUNCTION optional_ref '(' parameter_list ')' lexical_vars optional_return_type
+      '{' inner_statement_list '}'
+          { $$ = Expr\Closure[['static' => true, 'byRef' => $3, 'params' => $5, 'uses' => $7, 'returnType' => $8, 'stmts' => $10]]; }
+;
+
+anonymous_class:
+      T_CLASS ctor_arguments extends_from implements_list '{' class_statement_list '}'
+          { $$ = array(Stmt\Class_[null, ['type' => 0, 'extends' => $3, 'implements' => $4, 'stmts' => $6]], $2);
+            $this->checkClass($$[0], -1); }
+
+new_expr:
+      T_NEW class_name_reference ctor_arguments             { $$ = Expr\New_[$2, $3]; }
+    | T_NEW anonymous_class
+          { list($class, $ctorArgs) = $2; $$ = Expr\New_[$class, $ctorArgs]; }
+;
+
+lexical_vars:
+      /* empty */                                           { $$ = array(); }
+    | T_USE '(' lexical_var_list ')'                        { $$ = $3; }
+;
+
+lexical_var_list:
+      lexical_var                                           { init($1); }
+    | lexical_var_list ',' lexical_var                      { push($1, $3); }
+;
+
+lexical_var:
+      optional_ref T_VARIABLE                               { $$ = Expr\ClosureUse[parseVar($2), $1]; }
+;
+
+function_call:
+      name argument_list                                    { $$ = Expr\FuncCall[$1, $2]; }
+    | callable_expr argument_list                           { $$ = Expr\FuncCall[$1, $2]; }
+    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM member_name argument_list
+          { $$ = Expr\StaticCall[$1, $3, $4]; }
+;
+
+class_name:
+      T_STATIC                                              { $$ = Name[$1]; }
+    | name                                                  { $$ = $1; }
+;
+
+name:
+      namespace_name_parts                                  { $$ = Name[$1]; }
+    | T_NS_SEPARATOR namespace_name_parts                   { $$ = Name\FullyQualified[$2]; }
+    | T_NAMESPACE T_NS_SEPARATOR namespace_name_parts       { $$ = Name\Relative[$3]; }
+;
+
+class_name_reference:
+      class_name                                            { $$ = $1; }
+    | new_variable                                          { $$ = $1; }
+    | error                                                 { $$ = Expr\Error[]; $this->errorState = 2; }
+;
+
+class_name_or_var:
+      class_name                                            { $$ = $1; }
+    | dereferencable                                        { $$ = $1; }
+;
+
+exit_expr:
+      /* empty */                                           { $$ = null; }
+    | '(' optional_expr ')'                                 { $$ = $2; }
+;
+
+backticks_expr:
+      /* empty */                                           { $$ = array(); }
+    | T_ENCAPSED_AND_WHITESPACE
+          { $$ = array(Scalar\EncapsedStringPart[Scalar\String_::parseEscapeSequences($1, '`')]); }
+    | encaps_list                                           { parseEncapsed($1, '`', true); $$ = $1; }
+;
+
+ctor_arguments:
+      /* empty */                                           { $$ = array(); }
+    | argument_list                                         { $$ = $1; }
+;
+
+constant:
+      name                                                  { $$ = Expr\ConstFetch[$1]; }
+    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM identifier
+          { $$ = Expr\ClassConstFetch[$1, $3]; }
+    /* We interpret and isolated FOO:: as an unfinished class constant fetch. It could also be
+       an unfinished static property fetch or unfinished scoped call. */
+    | class_name_or_var T_PAAMAYIM_NEKUDOTAYIM error
+          { $$ = Expr\ClassConstFetch[$1, Expr\Error[]]; $this->errorState = 2; }
+;
+
+array_short_syntax:
+      '[' array_pair_list ']'
+          { $attrs = attributes(); $attrs['kind'] = Expr\Array_::KIND_SHORT;
+            $$ = new Expr\Array_($2, $attrs); }
+;
+
+dereferencable_scalar:
+      T_ARRAY '(' array_pair_list ')'
+          { $attrs = attributes(); $attrs['kind'] = Expr\Array_::KIND_LONG;
+            $$ = new Expr\Array_($3, $attrs); }
+    | array_short_syntax                                    { $$ = $1; }
+    | T_CONSTANT_ENCAPSED_STRING
+          { $attrs = attributes(); $attrs['kind'] = strKind($1);
+            $$ = new Scalar\String_(Scalar\String_::parse($1), $attrs); }
+;
+
+scalar:
+      T_LNUMBER                                             { $$ = $this->parseLNumber($1, attributes()); }
+    | T_DNUMBER                                             { $$ = Scalar\DNumber[Scalar\DNumber::parse($1)]; }
+    | T_LINE                                                { $$ = Scalar\MagicConst\Line[]; }
+    | T_FILE                                                { $$ = Scalar\MagicConst\File[]; }
+    | T_DIR                                                 { $$ = Scalar\MagicConst\Dir[]; }
+    | T_CLASS_C                                             { $$ = Scalar\MagicConst\Class_[]; }
+    | T_TRAIT_C                                             { $$ = Scalar\MagicConst\Trait_[]; }
+    | T_METHOD_C                                            { $$ = Scalar\MagicConst\Method[]; }
+    | T_FUNC_C                                              { $$ = Scalar\MagicConst\Function_[]; }
+    | T_NS_C                                                { $$ = Scalar\MagicConst\Namespace_[]; }
+    | dereferencable_scalar                                 { $$ = $1; }
+    | constant                                              { $$ = $1; }
+    | T_START_HEREDOC T_ENCAPSED_AND_WHITESPACE T_END_HEREDOC
+          { $attrs = attributes(); setDocStringAttrs($attrs, $1);
+            $$ = new Scalar\String_(Scalar\String_::parseDocString($1, $2), $attrs); }
+    | T_START_HEREDOC T_END_HEREDOC
+          { $attrs = attributes(); setDocStringAttrs($attrs, $1);
+            $$ = new Scalar\String_('', $attrs); }
+    | '"' encaps_list '"'
+          { $attrs = attributes(); $attrs['kind'] = Scalar\String_::KIND_DOUBLE_QUOTED;
+            parseEncapsed($2, '"', true); $$ = new Scalar\Encapsed($2, $attrs); }
+    | T_START_HEREDOC encaps_list T_END_HEREDOC
+          { $attrs = attributes(); setDocStringAttrs($attrs, $1);
+            parseEncapsedDoc($2, true); $$ = new Scalar\Encapsed($2, $attrs); }
+;
+
+optional_expr:
+      /* empty */                                           { $$ = null; }
+    | expr                                                  { $$ = $1; }
+;
+
+dereferencable:
+      variable                                              { $$ = $1; }
+    | '(' expr ')'                                          { $$ = $2; }
+    | dereferencable_scalar                                 { $$ = $1; }
+;
+
+callable_expr:
+      callable_variable                                     { $$ = $1; }
+    | '(' expr ')'                                          { $$ = $2; }
+    | dereferencable_scalar                                 { $$ = $1; }
+;
+
+callable_variable:
+      simple_variable                                       { $$ = Expr\Variable[$1]; }
+    | dereferencable '[' optional_expr ']'                  { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | constant '[' optional_expr ']'                        { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | dereferencable '{' expr '}'                           { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | function_call                                         { $$ = $1; }
+    | dereferencable T_OBJECT_OPERATOR property_name argument_list
+          { $$ = Expr\MethodCall[$1, $3, $4]; }
+;
+
+variable:
+      callable_variable                                     { $$ = $1; }
+    | static_member                                         { $$ = $1; }
+    | dereferencable T_OBJECT_OPERATOR property_name        { $$ = Expr\PropertyFetch[$1, $3]; }
+;
+
+simple_variable:
+      T_VARIABLE                                            { $$ = parseVar($1); }
+    | '$' '{' expr '}'                                      { $$ = $3; }
+    | '$' simple_variable                                   { $$ = Expr\Variable[$2]; }
+    | '$' error                                             { $$ = Expr\Error[]; $this->errorState = 2; }
+;
+
+static_member:
+      class_name_or_var T_PAAMAYIM_NEKUDOTAYIM simple_variable
+          { $$ = Expr\StaticPropertyFetch[$1, $3]; }
+;
+
+new_variable:
+      simple_variable                                       { $$ = Expr\Variable[$1]; }
+    | new_variable '[' optional_expr ']'                    { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | new_variable '{' expr '}'                             { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | new_variable T_OBJECT_OPERATOR property_name          { $$ = Expr\PropertyFetch[$1, $3]; }
+    | class_name T_PAAMAYIM_NEKUDOTAYIM simple_variable     { $$ = Expr\StaticPropertyFetch[$1, $3]; }
+    | new_variable T_PAAMAYIM_NEKUDOTAYIM simple_variable   { $$ = Expr\StaticPropertyFetch[$1, $3]; }
+;
+
+member_name:
+      identifier                                            { $$ = $1; }
+    | '{' expr '}'	                                        { $$ = $2; }
+    | simple_variable	                                    { $$ = Expr\Variable[$1]; }
+;
+
+property_name:
+      T_STRING                                              { $$ = $1; }
+    | '{' expr '}'	                                        { $$ = $2; }
+    | simple_variable	                                    { $$ = Expr\Variable[$1]; }
+    | error                                                 { $$ = Expr\Error[]; $this->errorState = 2; }
+;
+
+list_expr:
+      T_LIST '(' list_expr_elements ')'                     { $$ = Expr\List_[$3]; }
+;
+
+list_expr_elements:
+      list_expr_elements ',' list_expr_element              { push($1, $3); }
+    | list_expr_element                                     { init($1); }
+;
+
+list_expr_element:
+      variable                                              { $$ = Expr\ArrayItem[$1, null, false]; }
+    | list_expr                                             { $$ = Expr\ArrayItem[$1, null, false]; }
+    | expr T_DOUBLE_ARROW variable                          { $$ = Expr\ArrayItem[$3, $1, false]; }
+    | expr T_DOUBLE_ARROW list_expr                         { $$ = Expr\ArrayItem[$3, $1, false]; }
+    | /* empty */                                           { $$ = null; }
+;
+
+array_pair_list:
+      inner_array_pair_list
+          { $$ = $1; $end = count($$)-1; if ($$[$end] === null) unset($$[$end]); }
+;
+
+inner_array_pair_list:
+      inner_array_pair_list ',' array_pair                  { push($1, $3); }
+    | array_pair                                            { init($1); }
+;
+
+array_pair:
+      expr T_DOUBLE_ARROW expr                              { $$ = Expr\ArrayItem[$3, $1,   false]; }
+    | expr                                                  { $$ = Expr\ArrayItem[$1, null, false]; }
+    | expr T_DOUBLE_ARROW '&' variable                      { $$ = Expr\ArrayItem[$4, $1,   true]; }
+    | '&' variable                                          { $$ = Expr\ArrayItem[$2, null, true]; }
+    | /* empty */                                           { $$ = null; }
+;
+
+encaps_list:
+      encaps_list encaps_var                                { push($1, $2); }
+    | encaps_list encaps_string_part                        { push($1, $2); }
+    | encaps_var                                            { init($1); }
+    | encaps_string_part encaps_var                         { init($1, $2); }
+;
+
+encaps_string_part:
+      T_ENCAPSED_AND_WHITESPACE                             { $$ = Scalar\EncapsedStringPart[$1]; }
+;
+
+encaps_base_var:
+      T_VARIABLE                                            { $$ = Expr\Variable[parseVar($1)]; }
+;
+
+encaps_var:
+      encaps_base_var                                       { $$ = $1; }
+    | encaps_base_var '[' encaps_var_offset ']'             { $$ = Expr\ArrayDimFetch[$1, $3]; }
+    | encaps_base_var T_OBJECT_OPERATOR T_STRING            { $$ = Expr\PropertyFetch[$1, $3]; }
+    | T_DOLLAR_OPEN_CURLY_BRACES expr '}'                   { $$ = Expr\Variable[$2]; }
+    | T_DOLLAR_OPEN_CURLY_BRACES T_STRING_VARNAME '}'       { $$ = Expr\Variable[$2]; }
+    | T_DOLLAR_OPEN_CURLY_BRACES T_STRING_VARNAME '[' expr ']' '}'
+          { $$ = Expr\ArrayDimFetch[Expr\Variable[$2], $4]; }
+    | T_CURLY_OPEN variable '}'                             { $$ = $2; }
+;
+
+encaps_var_offset:
+      T_STRING                                              { $$ = Scalar\String_[$1]; }
+    | T_NUM_STRING                                          { $$ = $this->parseNumString($1, attributes()); }
+    | '-' T_NUM_STRING                                      { $$ = $this->parseNumString('-' . $2, attributes()); }
+    | T_VARIABLE                                            { $$ = Expr\Variable[parseVar($1)]; }
+;
+
+%%

grammar/rebuildParser.php

@@ -1,218 +0,0 @@
-<?php
-
-$grammarFile           = __DIR__ . '/zend_language_parser.phpy';
-$skeletonFile          = __DIR__ . '/kmyacc.php.parser';
-$tmpGrammarFile        = __DIR__ . '/tmp_parser.phpy';
-$tmpResultFile         = __DIR__ . '/tmp_parser.php';
-$parserResultFile      = __DIR__ . '/../lib/PhpParser/Parser.php';
-
-// check for kmyacc.exe binary in this directory, otherwise fall back to global name
-$kmyacc = __DIR__ . '/kmyacc.exe';
-if (!file_exists($kmyacc)) {
-    $kmyacc = 'kmyacc';
-}
-
-$options = array_flip($argv);
-$optionDebug = isset($options['--debug']);
-$optionKeepTmpGrammar = isset($options['--keep-tmp-grammar']);
-
-///////////////////////////////
-/// Utility regex constants ///
-///////////////////////////////
-
-const LIB = '(?(DEFINE)
-    (?<singleQuotedString>\'[^\\\\\']*+(?:\\\\.[^\\\\\']*+)*+\')
-    (?<doubleQuotedString>"[^\\\\"]*+(?:\\\\.[^\\\\"]*+)*+")
-    (?<string>(?&singleQuotedString)|(?&doubleQuotedString))
-    (?<comment>/\*[^*]*+(?:\*(?!/)[^*]*+)*+\*/)
-    (?<code>\{[^\'"/{}]*+(?:(?:(?&string)|(?&comment)|(?&code)|/)[^\'"/{}]*+)*+})
-)';
-
-const PARAMS = '\[(?<params>[^[\]]*+(?:\[(?&params)\][^[\]]*+)*+)\]';
-const ARGS   = '\((?<args>[^()]*+(?:\((?&args)\)[^()]*+)*+)\)';
-
-///////////////////
-/// Main script ///
-///////////////////
-
-echo 'Building temporary preproprocessed grammar file.', "\n";
-
-$grammarCode = file_get_contents($grammarFile);
-
-$grammarCode = resolveConstants($grammarCode);
-$grammarCode = resolveNodes($grammarCode);
-$grammarCode = resolveMacros($grammarCode);
-$grammarCode = resolveArrays($grammarCode);
-
-file_put_contents($tmpGrammarFile, $grammarCode);
-
-$additionalArgs = $optionDebug ? '-t -v' : '';
-
-echo "Building parser.\n";
-$output = trim(shell_exec("$kmyacc $additionalArgs -l -m $skeletonFile $tmpGrammarFile 2>&1"));
-echo "Output: \"$output\"\n";
-
-moveFileWithDirCheck($tmpResultFile, $parserResultFile);
-
-if (!$optionKeepTmpGrammar) {
-    unlink($tmpGrammarFile);
-}
-
-///////////////////////////////
-/// Preprocessing functions ///
-///////////////////////////////
-
-function resolveConstants($code) {
-    return preg_replace('~[A-Z][a-zA-Z_\\\\]++::~', 'Node\\\\$0', $code);
-}
-
-function resolveNodes($code) {
-    return preg_replace_callback(
-        '~(?<name>[A-Z][a-zA-Z_\\\\]++)\s*' . PARAMS . '~',
-        function($matches) {
-            // recurse
-            $matches['params'] = resolveNodes($matches['params']);
-
-            $params = magicSplit(
-                '(?:' . PARAMS . '|' . ARGS . ')(*SKIP)(*FAIL)|,',
-                $matches['params']
-            );
-
-            $paramCode = '';
-            foreach ($params as $param) {
-                $paramCode .= $param . ', ';
-            }
-
-            return 'new Node\\' . $matches['name'] . '(' . $paramCode . '$attributes)';
-        },
-        $code
-    );
-}
-
-function resolveMacros($code) {
-    return preg_replace_callback(
-        '~\b(?<!::|->)(?!array\()(?<name>[a-z][A-Za-z]++)' . ARGS . '~',
-        function($matches) {
-            // recurse
-            $matches['args'] = resolveMacros($matches['args']);
-
-            $name = $matches['name'];
-            $args = magicSplit(
-                '(?:' . PARAMS . '|' . ARGS . ')(*SKIP)(*FAIL)|,',
-                $matches['args']
-            );
-
-            if ('error' == $name) {
-                assertArgs(1, $args, $name);
-
-                return 'throw new Error(' . $args[0] . ')';
-            }
-
-            if ('init' == $name) {
-                return '$$ = array(' . implode(', ', $args) . ')';
-            }
-
-            if ('push' == $name) {
-                assertArgs(2, $args, $name);
-
-                return $args[0] . '[] = ' . $args[1] . '; $$ = ' . $args[0];
-            }
-
-            if ('pushNormalizing' == $name) {
-                assertArgs(2, $args, $name);
-
-                return 'if (is_array(' . $args[1] . ')) { $$ = array_merge(' . $args[0] . ', ' . $args[1] . '); } else { ' . $args[0] . '[] = ' . $args[1] . '; $$ = ' . $args[0] . '; }';
-            }
-
-            if ('toArray' == $name) {
-                assertArgs(1, $args, $name);
-
-                return 'is_array(' . $args[0] . ') ? ' . $args[0] . ' : array(' . $args[0] . ')';
-            }
-
-            if ('parseVar' == $name) {
-                assertArgs(1, $args, $name);
-
-                return 'substr(' . $args[0] . ', 1)';
-            }
-
-            if ('parseEncapsed' == $name) {
-                assertArgs(2, $args, $name);
-
-                return 'foreach (' . $args[0] . ' as &$s) { if (is_string($s)) { $s = Node\Scalar\String_::parseEscapeSequences($s, ' . $args[1] . '); } }';
-            }
-
-            if ('parseEncapsedDoc' == $name) {
-                assertArgs(1, $args, $name);
-
-                return 'foreach (' . $args[0] . ' as &$s) { if (is_string($s)) { $s = Node\Scalar\String_::parseEscapeSequences($s, null); } } $s = preg_replace(\'~(\r\n|\n|\r)$~\', \'\', $s); if (\'\' === $s) array_pop(' . $args[0] . ');';
-            }
-
-            throw new Exception(sprintf('Unknown macro "%s"', $name));
-        },
-        $code
-    );
-}
-
-function assertArgs($num, $args, $name) {
-    if ($num != count($args)) {
-        die('Wrong argument count for ' . $name . '().');
-    }
-}
-
-function resolveArrays($code) {
-    return preg_replace_callback(
-        '~' . PARAMS . '~',
-        function ($matches) {
-            $elements = magicSplit(
-                '(?:' . PARAMS . '|' . ARGS . ')(*SKIP)(*FAIL)|,',
-                $matches['params']
-            );
-
-            // don't convert [] to array, it might have different meaning
-            if (empty($elements)) {
-                return $matches[0];
-            }
-
-            $elementCodes = array();
-            foreach ($elements as $element) {
-                // convert only arrays where all elements have keys
-                if (false === strpos($element, ':')) {
-                    return $matches[0];
-                }
-
-                list($key, $value) = explode(':', $element, 2);
-                $elementCodes[] = "'" . $key . "' =>" . $value;
-            }
-
-            return 'array(' . implode(', ', $elementCodes) . ')';
-        },
-        $code
-    );
-}
-
-function moveFileWithDirCheck($fromPath, $toPath) {
-    $dir = dirname($toPath);
-    if (!is_dir($dir)) {
-        mkdir($dir, 0777, true);
-    }
-    rename($fromPath, $toPath);
-}
-
-//////////////////////////////
-/// Regex helper functions ///
-//////////////////////////////
-
-function regex($regex) {
-    return '~' . LIB . '(?:' . str_replace('~', '\~', $regex) . ')~';
-}
-
-function magicSplit($regex, $string) {
-    $pieces = preg_split(regex('(?:(?&string)|(?&comment)|(?&code))(*SKIP)(*FAIL)|' . $regex), $string);
-
-    foreach ($pieces as &$piece) {
-        $piece = trim($piece);
-    }
-
-    return array_filter($pieces);
-}

grammar/rebuildParsers.php

@@ -0,0 +1,263 @@
+<?php
+
+$grammarFileToName = [
+    __DIR__ . '/php5.y' => 'Php5',
+    __DIR__ . '/php7.y' => 'Php7',
+];
+
+$tokensFile     = __DIR__ . '/tokens.y';
+$tokensTemplate = __DIR__ . '/tokens.template';
+$skeletonFile   = __DIR__ . '/parser.template';
+$tmpGrammarFile = __DIR__ . '/tmp_parser.phpy';
+$tmpResultFile  = __DIR__ . '/tmp_parser.php';
+$resultDir = __DIR__ . '/../lib/PhpParser/Parser';
+$tokensResultsFile = $resultDir . '/Tokens.php';
+
+// check for kmyacc.exe binary in this directory, otherwise fall back to global name
+$kmyacc = __DIR__ . '/kmyacc.exe';
+if (!file_exists($kmyacc)) {
+    $kmyacc = 'kmyacc';
+}
+
+$options = array_flip($argv);
+$optionDebug = isset($options['--debug']);
+$optionKeepTmpGrammar = isset($options['--keep-tmp-grammar']);
+
+///////////////////////////////
+/// Utility regex constants ///
+///////////////////////////////
+
+const LIB = '(?(DEFINE)
+    (?<singleQuotedString>\'[^\\\\\']*+(?:\\\\.[^\\\\\']*+)*+\')
+    (?<doubleQuotedString>"[^\\\\"]*+(?:\\\\.[^\\\\"]*+)*+")
+    (?<string>(?&singleQuotedString)|(?&doubleQuotedString))
+    (?<comment>/\*[^*]*+(?:\*(?!/)[^*]*+)*+\*/)
+    (?<code>\{[^\'"/{}]*+(?:(?:(?&string)|(?&comment)|(?&code)|/)[^\'"/{}]*+)*+})
+)';
+
+const PARAMS = '\[(?<params>[^[\]]*+(?:\[(?&params)\][^[\]]*+)*+)\]';
+const ARGS   = '\((?<args>[^()]*+(?:\((?&args)\)[^()]*+)*+)\)';
+
+///////////////////
+/// Main script ///
+///////////////////
+
+$tokens = file_get_contents($tokensFile);
+
+foreach ($grammarFileToName as $grammarFile => $name) {
+    echo "Building temporary $name grammar file.\n";
+
+    $grammarCode = file_get_contents($grammarFile);
+    $grammarCode = str_replace('%tokens', $tokens, $grammarCode);
+
+    $grammarCode = resolveNodes($grammarCode);
+    $grammarCode = resolveMacros($grammarCode);
+    $grammarCode = resolveStackAccess($grammarCode);
+
+    file_put_contents($tmpGrammarFile, $grammarCode);
+
+    $additionalArgs = $optionDebug ? '-t -v' : '';
+
+    echo "Building $name parser.\n";
+    $output = trim(shell_exec("$kmyacc $additionalArgs -l -m $skeletonFile -p $name $tmpGrammarFile 2>&1"));
+    echo "Output: \"$output\"\n";
+
+    $resultCode = file_get_contents($tmpResultFile);
+    $resultCode = removeTrailingWhitespace($resultCode);
+
+    ensureDirExists($resultDir);
+    file_put_contents("$resultDir/$name.php", $resultCode);
+    unlink($tmpResultFile);
+
+    echo "Building token definition.\n";
+    $output = trim(shell_exec("$kmyacc -l -m $tokensTemplate $tmpGrammarFile 2>&1"));
+    assert($output === '');
+    rename($tmpResultFile, $tokensResultsFile);
+
+    if (!$optionKeepTmpGrammar) {
+        unlink($tmpGrammarFile);
+    }
+}
+
+///////////////////////////////
+/// Preprocessing functions ///
+///////////////////////////////
+
+function resolveNodes($code) {
+    return preg_replace_callback(
+        '~\b(?<name>[A-Z][a-zA-Z_\\\\]++)\s*' . PARAMS . '~',
+        function($matches) {
+            // recurse
+            $matches['params'] = resolveNodes($matches['params']);
+
+            $params = magicSplit(
+                '(?:' . PARAMS . '|' . ARGS . ')(*SKIP)(*FAIL)|,',
+                $matches['params']
+            );
+
+            $paramCode = '';
+            foreach ($params as $param) {
+                $paramCode .= $param . ', ';
+            }
+
+            return 'new ' . $matches['name'] . '(' . $paramCode . 'attributes())';
+        },
+        $code
+    );
+}
+
+function resolveMacros($code) {
+    return preg_replace_callback(
+        '~\b(?<!::|->)(?!array\()(?<name>[a-z][A-Za-z]++)' . ARGS . '~',
+        function($matches) {
+            // recurse
+            $matches['args'] = resolveMacros($matches['args']);
+
+            $name = $matches['name'];
+            $args = magicSplit(
+                '(?:' . PARAMS . '|' . ARGS . ')(*SKIP)(*FAIL)|,',
+                $matches['args']
+            );
+
+            if ('attributes' == $name) {
+                assertArgs(0, $args, $name);
+                return '$this->startAttributeStack[#1] + $this->endAttributes';
+            }
+
+            if ('stackAttributes' == $name) {
+                assertArgs(1, $args, $name);
+                return '$this->startAttributeStack[' . $args[0] . ']'
+                     . ' + $this->endAttributeStack[' . $args[0] . ']';
+            }
+
+            if ('init' == $name) {
+                return '$$ = array(' . implode(', ', $args) . ')';
+            }
+
+            if ('push' == $name) {
+                assertArgs(2, $args, $name);
+
+                return $args[0] . '[] = ' . $args[1] . '; $$ = ' . $args[0];
+            }
+
+            if ('pushNormalizing' == $name) {
+                assertArgs(2, $args, $name);
+
+                return 'if (is_array(' . $args[1] . ')) { $$ = array_merge(' . $args[0] . ', ' . $args[1] . '); }'
+                     . ' else { ' . $args[0] . '[] = ' . $args[1] . '; $$ = ' . $args[0] . '; }';
+            }
+
+            if ('toArray' == $name) {
+                assertArgs(1, $args, $name);
+
+                return 'is_array(' . $args[0] . ') ? ' . $args[0] . ' : array(' . $args[0] . ')';
+            }
+
+            if ('parseVar' == $name) {
+                assertArgs(1, $args, $name);
+
+                return 'substr(' . $args[0] . ', 1)';
+            }
+
+            if ('parseEncapsed' == $name) {
+                assertArgs(3, $args, $name);
+
+                return 'foreach (' . $args[0] . ' as $s) { if ($s instanceof Node\Scalar\EncapsedStringPart) {'
+                     . ' $s->value = Node\Scalar\String_::parseEscapeSequences($s->value, ' . $args[1] . ', ' . $args[2] . '); } }';
+            }
+
+            if ('parseEncapsedDoc' == $name) {
+                assertArgs(2, $args, $name);
+
+                return 'foreach (' . $args[0] . ' as $s) { if ($s instanceof Node\Scalar\EncapsedStringPart) {'
+                     . ' $s->value = Node\Scalar\String_::parseEscapeSequences($s->value, null, ' . $args[1] . '); } }'
+                     . ' $s->value = preg_replace(\'~(\r\n|\n|\r)\z~\', \'\', $s->value);'
+                     . ' if (\'\' === $s->value) array_pop(' . $args[0] . ');';
+            }
+
+            if ('makeNop' == $name) {
+                assertArgs(2, $args, $name);
+
+                return '$startAttributes = ' . $args[1] . ';'
+                . ' if (isset($startAttributes[\'comments\']))'
+                . ' { ' . $args[0] . ' = new Stmt\Nop([\'comments\' => $startAttributes[\'comments\']]); }'
+                . ' else { ' . $args[0] . ' = null; }';
+            }
+
+            if ('strKind' == $name) {
+                assertArgs(1, $args, $name);
+
+                return '(' . $args[0] . '[0] === "\'" || (' . $args[0] . '[1] === "\'" && '
+                     . '(' . $args[0] . '[0] === \'b\' || ' . $args[0] . '[0] === \'B\')) '
+                     . '? Scalar\String_::KIND_SINGLE_QUOTED : Scalar\String_::KIND_DOUBLE_QUOTED)';
+            }
+
+            if ('setDocStringAttrs' == $name) {
+                assertArgs(2, $args, $name);
+
+                return $args[0] . '[\'kind\'] = strpos(' . $args[1] . ', "\'") === false '
+                     . '? Scalar\String_::KIND_HEREDOC : Scalar\String_::KIND_NOWDOC; '
+                     . 'preg_match(\'/\A[bB]?<<<[ \t]*[\\\'"]?([a-zA-Z_\x7f-\xff][a-zA-Z0-9_\x7f-\xff]*)[\\\'"]?(?:\r\n|\n|\r)\z/\', ' . $args[1] . ', $matches); '
+                     . $args[0] . '[\'docLabel\'] = $matches[1];';
+            }
+
+            if ('prependLeadingComments' == $name) {
+                assertArgs(1, $args, $name);
+
+                return '$attrs = $this->startAttributeStack[#1]; $stmts = ' . $args[0] . '; '
+                . 'if (!empty($attrs[\'comments\']) && isset($stmts[0])) {'
+                . '$stmts[0]->setAttribute(\'comments\', '
+                . 'array_merge($attrs[\'comments\'], $stmts[0]->getAttribute(\'comments\', []))); }';
+            }
+
+            return $matches[0];
+        },
+        $code
+    );
+}
+
+function assertArgs($num, $args, $name) {
+    if ($num != count($args)) {
+        die('Wrong argument count for ' . $name . '().');
+    }
+}
+
+function resolveStackAccess($code) {
+    $code = preg_replace('/\$\d+/', '$this->semStack[$0]', $code);
+    $code = preg_replace('/#(\d+)/', '$$1', $code);
+    return $code;
+}
+
+function removeTrailingWhitespace($code) {
+    $lines = explode("\n", $code);
+    $lines = array_map('rtrim', $lines);
+    return implode("\n", $lines);
+}
+
+function ensureDirExists($dir) {
+    if (!is_dir($dir)) {
+        mkdir($dir, 0777, true);
+    }
+}
+
+//////////////////////////////
+/// Regex helper functions ///
+//////////////////////////////
+
+function regex($regex) {
+    return '~' . LIB . '(?:' . str_replace('~', '\~', $regex) . ')~';
+}
+
+function magicSplit($regex, $string) {
+    $pieces = preg_split(regex('(?:(?&string)|(?&comment)|(?&code))(*SKIP)(*FAIL)|' . $regex), $string);
+
+    foreach ($pieces as &$piece) {
+        $piece = trim($piece);
+    }
+
+    if ($pieces === ['']) {
+        return [];
+    }
+
+    return $pieces;
+}

grammar/tokens.template

@@ -0,0 +1,17 @@
+<?php
+$meta #
+#semval($) $this->semValue
+#semval($,%t) $this->semValue
+#semval(%n) $this->stackPos-(%l-%n)
+#semval(%n,%t) $this->stackPos-(%l-%n)
+
+namespace PhpParser\Parser;
+#include;
+
+/* GENERATED file based on grammar/tokens.y */
+final class Tokens
+{
+#tokenval
+    const %s = %n;
+#endtokenval
+}

grammar/tokens.y

@@ -0,0 +1,113 @@
+/* We currently rely on the token ID mapping to be the same between PHP 5 and PHP 7 - so the same lexer can be used for
+ * both. This is enforced by sharing this token file. */
+
+%left T_INCLUDE T_INCLUDE_ONCE T_EVAL T_REQUIRE T_REQUIRE_ONCE
+%left ','
+%left T_LOGICAL_OR
+%left T_LOGICAL_XOR
+%left T_LOGICAL_AND
+%right T_PRINT
+%right T_YIELD
+%right T_DOUBLE_ARROW
+%right T_YIELD_FROM
+%left '=' T_PLUS_EQUAL T_MINUS_EQUAL T_MUL_EQUAL T_DIV_EQUAL T_CONCAT_EQUAL T_MOD_EQUAL T_AND_EQUAL T_OR_EQUAL T_XOR_EQUAL T_SL_EQUAL T_SR_EQUAL T_POW_EQUAL
+%left '?' ':'
+%right T_COALESCE
+%left T_BOOLEAN_OR
+%left T_BOOLEAN_AND
+%left '|'
+%left '^'
+%left '&'
+%nonassoc T_IS_EQUAL T_IS_NOT_EQUAL T_IS_IDENTICAL T_IS_NOT_IDENTICAL T_SPACESHIP
+%nonassoc '<' T_IS_SMALLER_OR_EQUAL '>' T_IS_GREATER_OR_EQUAL
+%left T_SL T_SR
+%left '+' '-' '.'
+%left '*' '/' '%'
+%right '!'
+%nonassoc T_INSTANCEOF
+%right '~' T_INC T_DEC T_INT_CAST T_DOUBLE_CAST T_STRING_CAST T_ARRAY_CAST T_OBJECT_CAST T_BOOL_CAST T_UNSET_CAST '@'
+%right T_POW
+%right '['
+%nonassoc T_NEW T_CLONE
+%token T_EXIT
+%token T_IF
+%left T_ELSEIF
+%left T_ELSE
+%left T_ENDIF
+%token T_LNUMBER
+%token T_DNUMBER
+%token T_STRING
+%token T_STRING_VARNAME
+%token T_VARIABLE
+%token T_NUM_STRING
+%token T_INLINE_HTML
+%token T_CHARACTER
+%token T_BAD_CHARACTER
+%token T_ENCAPSED_AND_WHITESPACE
+%token T_CONSTANT_ENCAPSED_STRING
+%token T_ECHO
+%token T_DO
+%token T_WHILE
+%token T_ENDWHILE
+%token T_FOR
+%token T_ENDFOR
+%token T_FOREACH
+%token T_ENDFOREACH
+%token T_DECLARE
+%token T_ENDDECLARE
+%token T_AS
+%token T_SWITCH
+%token T_ENDSWITCH
+%token T_CASE
+%token T_DEFAULT
+%token T_BREAK
+%token T_CONTINUE
+%token T_GOTO
+%token T_FUNCTION
+%token T_CONST
+%token T_RETURN
+%token T_TRY
+%token T_CATCH
+%token T_FINALLY
+%token T_THROW
+%token T_USE
+%token T_INSTEADOF
+%token T_GLOBAL
+%right T_STATIC T_ABSTRACT T_FINAL T_PRIVATE T_PROTECTED T_PUBLIC
+%token T_VAR
+%token T_UNSET
+%token T_ISSET
+%token T_EMPTY
+%token T_HALT_COMPILER
+%token T_CLASS
+%token T_TRAIT
+%token T_INTERFACE
+%token T_EXTENDS
+%token T_IMPLEMENTS
+%token T_OBJECT_OPERATOR
+%token T_DOUBLE_ARROW
+%token T_LIST
+%token T_ARRAY
+%token T_CALLABLE
+%token T_CLASS_C
+%token T_TRAIT_C
+%token T_METHOD_C
+%token T_FUNC_C
+%token T_LINE
+%token T_FILE
+%token T_COMMENT
+%token T_DOC_COMMENT
+%token T_OPEN_TAG
+%token T_OPEN_TAG_WITH_ECHO
+%token T_CLOSE_TAG
+%token T_WHITESPACE
+%token T_START_HEREDOC
+%token T_END_HEREDOC
+%token T_DOLLAR_OPEN_CURLY_BRACES
+%token T_CURLY_OPEN
+%token T_PAAMAYIM_NEKUDOTAYIM
+%token T_NAMESPACE
+%token T_NS_C
+%token T_DIR
+%token T_NS_SEPARATOR
+%token T_ELLIPSIS

lib/PhpParser/Autoloader.php

@@ -10,9 +10,6 @@ class Autoloader
     /** @var bool Whether the autoloader has been registered. */
     private static $registered = false;
 
-    /** @var bool Whether we're running on PHP 7. */
-    private static $runningOnPhp7;
-
     /**
      * Registers PhpParser\Autoloader as an SPL autoloader.
      *
@@ -23,10 +20,8 @@ class Autoloader
             return;
         }
 
-        ini_set('unserialize_callback_func', 'spl_autoload_call');
         spl_autoload_register(array(__CLASS__, 'autoload'), true, $prepend);
         self::$registered = true;
-        self::$runningOnPhp7 = version_compare(PHP_VERSION, '7.0-dev', '>=');
     }
 
     /**
@@ -36,245 +31,10 @@ class Autoloader
      */
     static public function autoload($class) {
         if (0 === strpos($class, 'PhpParser\\')) {
-            if (isset(self::$php7AliasesOldToNew[$class])) {
-                // Old class name was used, register alias to new one (which will
-                // be autoloaded, if it wasn't yet).
-                self::registerPhp7Alias(self::$php7AliasesOldToNew[$class], $class);
-                return;
-            }
-
-            $fileName = dirname(__DIR__) . '/' . strtr($class, '\\', '/') . '.php';
+            $fileName = __DIR__ . strtr(substr($class, 9), '\\', '/') . '.php';
             if (file_exists($fileName)) {
                 require $fileName;
             }
-
-            if (isset(self::$php7AliasesNewToOld[$class])) {
-                // New class name was used, register alias for old one, otherwise
-                // it won't be usable in "instanceof" and other non-autoloading places.
-                self::registerPhp7Alias($class, self::$php7AliasesNewToOld[$class]);
-            }
-        } else if (0 === strpos($class, 'PHPParser_')) {
-            if (isset(self::$nonNamespacedAliases[$class])) {
-                // Register all aliases at once to avoid dependency issues
-                self::registerNonNamespacedAliases();
-            }
         }
     }
-
-    private static function registerPhp7Alias($old, $new) {
-        // Registering these aliases would throw a fatal error on PHP 7,
-        // we want to avoid that.
-        if (!self::$runningOnPhp7) {
-            class_alias($old, $new);
-        }
-    }
-
-    private static function registerNonNamespacedAliases() {
-        foreach (self::$nonNamespacedAliases as $old => $new) {
-            class_alias($new, $old);
-        }
-    }
-
-    private static $php7AliasesOldToNew = array(
-        'PhpParser\Node\Expr\Cast\Bool' => 'PhpParser\Node\Expr\Cast\Bool_',
-        'PhpParser\Node\Expr\Cast\Int' => 'PhpParser\Node\Expr\Cast\Int_',
-        'PhpParser\Node\Expr\Cast\Object' => 'PhpParser\Node\Expr\Cast\Object_',
-        'PhpParser\Node\Expr\Cast\String' => 'PhpParser\Node\Expr\Cast\String_',
-        'PhpParser\Node\Scalar\String' => 'PhpParser\Node\Scalar\String_',
-    );
-
-    private static $php7AliasesNewToOld = array(
-        'PhpParser\Node\Expr\Cast\Bool_' => 'PhpParser\Node\Expr\Cast\Bool',
-        'PhpParser\Node\Expr\Cast\Int_' => 'PhpParser\Node\Expr\Cast\Int',
-        'PhpParser\Node\Expr\Cast\Object_' => 'PhpParser\Node\Expr\Cast\Object',
-        'PhpParser\Node\Expr\Cast\String_' => 'PhpParser\Node\Expr\Cast\String',
-        'PhpParser\Node\Scalar\String_' => 'PhpParser\Node\Scalar\String',
-    );
-
-    private static $nonNamespacedAliases = array(
-        'PHPParser_Builder' => 'PhpParser\Builder',
-        'PHPParser_BuilderAbstract' => 'PhpParser\BuilderAbstract',
-        'PHPParser_BuilderFactory' => 'PhpParser\BuilderFactory',
-        'PHPParser_Comment' => 'PhpParser\Comment',
-        'PHPParser_Comment_Doc' => 'PhpParser\Comment\Doc',
-        'PHPParser_Error' => 'PhpParser\Error',
-        'PHPParser_Lexer' => 'PhpParser\Lexer',
-        'PHPParser_Lexer_Emulative' => 'PhpParser\Lexer\Emulative',
-        'PHPParser_Node' => 'PhpParser\Node',
-        'PHPParser_NodeAbstract' => 'PhpParser\NodeAbstract',
-        'PHPParser_NodeDumper' => 'PhpParser\NodeDumper',
-        'PHPParser_NodeTraverser' => 'PhpParser\NodeTraverser',
-        'PHPParser_NodeTraverserInterface' => 'PhpParser\NodeTraverserInterface',
-        'PHPParser_NodeVisitor' => 'PhpParser\NodeVisitor',
-        'PHPParser_NodeVisitor_NameResolver' => 'PhpParser\NodeVisitor\NameResolver',
-        'PHPParser_NodeVisitorAbstract' => 'PhpParser\NodeVisitorAbstract',
-        'PHPParser_Parser' => 'PhpParser\Parser',
-        'PHPParser_PrettyPrinterAbstract' => 'PhpParser\PrettyPrinterAbstract',
-        'PHPParser_PrettyPrinter_Default' => 'PhpParser\PrettyPrinter\Standard',
-        'PHPParser_PrettyPrinter_Zend' => 'PhpParser\PrettyPrinter\Standard',
-        'PHPParser_Serializer' => 'PhpParser\Serializer',
-        'PHPParser_Serializer_XML' => 'PhpParser\Serializer\XML',
-        'PHPParser_Unserializer' => 'PhpParser\Unserializer',
-        'PHPParser_Unserializer_XML' => 'PhpParser\Unserializer\XML',
-
-        'PHPParser_Builder_Class' => 'PhpParser\Builder\Class_',
-        'PHPParser_Builder_Function' => 'PhpParser\Builder\Function_',
-        'PHPParser_Builder_Interface' => 'PhpParser\Builder\Interface_',
-        'PHPParser_Builder_Method' => 'PhpParser\Builder\Method',
-        'PHPParser_Builder_Param' => 'PhpParser\Builder\Param',
-        'PHPParser_Builder_Property' => 'PhpParser\Builder\Property',
-
-        'PHPParser_Node_Arg' => 'PhpParser\Node\Arg',
-        'PHPParser_Node_Const' => 'PhpParser\Node\Const_',
-        'PHPParser_Node_Expr' => 'PhpParser\Node\Expr',
-        'PHPParser_Node_Name' => 'PhpParser\Node\Name',
-        'PHPParser_Node_Name_FullyQualified' => 'PhpParser\Node\Name\FullyQualified',
-        'PHPParser_Node_Name_Relative' => 'PhpParser\Node\Name\Relative',
-        'PHPParser_Node_Param' => 'PhpParser\Node\Param',
-        'PHPParser_Node_Scalar' => 'PhpParser\Node\Scalar',
-        'PHPParser_Node_Stmt' => 'PhpParser\Node\Stmt',
-
-        'PHPParser_Node_Stmt_Break' => 'PhpParser\Node\Stmt\Break_',
-        'PHPParser_Node_Stmt_Case' => 'PhpParser\Node\Stmt\Case_',
-        'PHPParser_Node_Stmt_Catch' => 'PhpParser\Node\Stmt\Catch_',
-        'PHPParser_Node_Stmt_Class' => 'PhpParser\Node\Stmt\Class_',
-        'PHPParser_Node_Stmt_ClassConst' => 'PhpParser\Node\Stmt\ClassConst',
-        'PHPParser_Node_Stmt_ClassMethod' => 'PhpParser\Node\Stmt\ClassMethod',
-        'PHPParser_Node_Stmt_Const' => 'PhpParser\Node\Stmt\Const_',
-        'PHPParser_Node_Stmt_Continue' => 'PhpParser\Node\Stmt\Continue_',
-        'PHPParser_Node_Stmt_Declare' => 'PhpParser\Node\Stmt\Declare_',
-        'PHPParser_Node_Stmt_DeclareDeclare' => 'PhpParser\Node\Stmt\DeclareDeclare',
-        'PHPParser_Node_Stmt_Do' => 'PhpParser\Node\Stmt\Do_',
-        'PHPParser_Node_Stmt_Echo' => 'PhpParser\Node\Stmt\Echo_',
-        'PHPParser_Node_Stmt_Else' => 'PhpParser\Node\Stmt\Else_',
-        'PHPParser_Node_Stmt_ElseIf' => 'PhpParser\Node\Stmt\ElseIf_',
-        'PHPParser_Node_Stmt_For' => 'PhpParser\Node\Stmt\For_',
-        'PHPParser_Node_Stmt_Foreach' => 'PhpParser\Node\Stmt\Foreach_',
-        'PHPParser_Node_Stmt_Function' => 'PhpParser\Node\Stmt\Function_',
-        'PHPParser_Node_Stmt_Global' => 'PhpParser\Node\Stmt\Global_',
-        'PHPParser_Node_Stmt_Goto' => 'PhpParser\Node\Stmt\Goto_',
-        'PHPParser_Node_Stmt_HaltCompiler' => 'PhpParser\Node\Stmt\HaltCompiler',
-        'PHPParser_Node_Stmt_If' => 'PhpParser\Node\Stmt\If_',
-        'PHPParser_Node_Stmt_InlineHTML' => 'PhpParser\Node\Stmt\InlineHTML',
-        'PHPParser_Node_Stmt_Interface' => 'PhpParser\Node\Stmt\Interface_',
-        'PHPParser_Node_Stmt_Label' => 'PhpParser\Node\Stmt\Label',
-        'PHPParser_Node_Stmt_Namespace' => 'PhpParser\Node\Stmt\Namespace_',
-        'PHPParser_Node_Stmt_Property' => 'PhpParser\Node\Stmt\Property',
-        'PHPParser_Node_Stmt_PropertyProperty' => 'PhpParser\Node\Stmt\PropertyProperty',
-        'PHPParser_Node_Stmt_Return' => 'PhpParser\Node\Stmt\Return_',
-        'PHPParser_Node_Stmt_Static' => 'PhpParser\Node\Stmt\Static_',
-        'PHPParser_Node_Stmt_StaticVar' => 'PhpParser\Node\Stmt\StaticVar',
-        'PHPParser_Node_Stmt_Switch' => 'PhpParser\Node\Stmt\Switch_',
-        'PHPParser_Node_Stmt_Throw' => 'PhpParser\Node\Stmt\Throw_',
-        'PHPParser_Node_Stmt_Trait' => 'PhpParser\Node\Stmt\Trait_',
-        'PHPParser_Node_Stmt_TraitUse' => 'PhpParser\Node\Stmt\TraitUse',
-        'PHPParser_Node_Stmt_TraitUseAdaptation' => 'PhpParser\Node\Stmt\TraitUseAdaptation',
-        'PHPParser_Node_Stmt_TraitUseAdaptation_Alias' => 'PhpParser\Node\Stmt\TraitUseAdaptation\Alias',
-        'PHPParser_Node_Stmt_TraitUseAdaptation_Precedence' => 'PhpParser\Node\Stmt\TraitUseAdaptation\Precedence',
-        'PHPParser_Node_Stmt_TryCatch' => 'PhpParser\Node\Stmt\TryCatch',
-        'PHPParser_Node_Stmt_Unset' => 'PhpParser\Node\Stmt\Unset_',
-        'PHPParser_Node_Stmt_UseUse' => 'PhpParser\Node\Stmt\UseUse',
-        'PHPParser_Node_Stmt_Use' => 'PhpParser\Node\Stmt\Use_',
-        'PHPParser_Node_Stmt_While' => 'PhpParser\Node\Stmt\While_',
-
-        'PHPParser_Node_Expr_AssignBitwiseAnd' => 'PhpParser\Node\Expr\AssignOp\BitwiseAnd',
-        'PHPParser_Node_Expr_AssignBitwiseOr' => 'PhpParser\Node\Expr\AssignOp\BitwiseOr',
-        'PHPParser_Node_Expr_AssignBitwiseXor' => 'PhpParser\Node\Expr\AssignOp\BitwiseXor',
-        'PHPParser_Node_Expr_AssignConcat' => 'PhpParser\Node\Expr\AssignOp\Concat',
-        'PHPParser_Node_Expr_AssignDiv' => 'PhpParser\Node\Expr\AssignOp\Div',
-        'PHPParser_Node_Expr_AssignMinus' => 'PhpParser\Node\Expr\AssignOp\Minus',
-        'PHPParser_Node_Expr_AssignMod' => 'PhpParser\Node\Expr\AssignOp\Mod',
-        'PHPParser_Node_Expr_AssignMul' => 'PhpParser\Node\Expr\AssignOp\Mul',
-        'PHPParser_Node_Expr_AssignPlus' => 'PhpParser\Node\Expr\AssignOp\Plus',
-        'PHPParser_Node_Expr_AssignShiftLeft' => 'PhpParser\Node\Expr\AssignOp\ShiftLeft',
-        'PHPParser_Node_Expr_AssignShiftRight' => 'PhpParser\Node\Expr\AssignOp\ShiftRight',
-
-        'PHPParser_Node_Expr_Cast' => 'PhpParser\Node\Expr\Cast',
-        'PHPParser_Node_Expr_Cast_Array' => 'PhpParser\Node\Expr\Cast\Array_',
-        'PHPParser_Node_Expr_Cast_Bool' => 'PhpParser\Node\Expr\Cast\Bool_',
-        'PHPParser_Node_Expr_Cast_Double' => 'PhpParser\Node\Expr\Cast\Double',
-        'PHPParser_Node_Expr_Cast_Int' => 'PhpParser\Node\Expr\Cast\Int_',
-        'PHPParser_Node_Expr_Cast_Object' => 'PhpParser\Node\Expr\Cast\Object_',
-        'PHPParser_Node_Expr_Cast_String' => 'PhpParser\Node\Expr\Cast\String_',
-        'PHPParser_Node_Expr_Cast_Unset' => 'PhpParser\Node\Expr\Cast\Unset_',
-
-        'PHPParser_Node_Expr_BitwiseAnd' => 'PhpParser\Node\Expr\BinaryOp\BitwiseAnd',
-        'PHPParser_Node_Expr_BitwiseOr' => 'PhpParser\Node\Expr\BinaryOp\BitwiseOr',
-        'PHPParser_Node_Expr_BitwiseXor' => 'PhpParser\Node\Expr\BinaryOp\BitwiseXor',
-        'PHPParser_Node_Expr_BooleanAnd' => 'PhpParser\Node\Expr\BinaryOp\BooleanAnd',
-        'PHPParser_Node_Expr_BooleanOr' => 'PhpParser\Node\Expr\BinaryOp\BooleanOr',
-        'PHPParser_Node_Expr_Concat' => 'PhpParser\Node\Expr\BinaryOp\Concat',
-        'PHPParser_Node_Expr_Div' => 'PhpParser\Node\Expr\BinaryOp\Div',
-        'PHPParser_Node_Expr_Equal' => 'PhpParser\Node\Expr\BinaryOp\Equal',
-        'PHPParser_Node_Expr_Greater' => 'PhpParser\Node\Expr\BinaryOp\Greater',
-        'PHPParser_Node_Expr_GreaterOrEqual' => 'PhpParser\Node\Expr\BinaryOp\GreaterOrEqual',
-        'PHPParser_Node_Expr_Identical' => 'PhpParser\Node\Expr\BinaryOp\Identical',
-        'PHPParser_Node_Expr_LogicalAnd' => 'PhpParser\Node\Expr\BinaryOp\LogicalAnd',
-        'PHPParser_Node_Expr_LogicalOr' => 'PhpParser\Node\Expr\BinaryOp\LogicalOr',
-        'PHPParser_Node_Expr_LogicalXor' => 'PhpParser\Node\Expr\BinaryOp\LogicalXor',
-        'PHPParser_Node_Expr_Minus' => 'PhpParser\Node\Expr\BinaryOp\Minus',
-        'PHPParser_Node_Expr_Mod' => 'PhpParser\Node\Expr\BinaryOp\Mod',
-        'PHPParser_Node_Expr_Mul' => 'PhpParser\Node\Expr\BinaryOp\Mul',
-        'PHPParser_Node_Expr_NotEqual' => 'PhpParser\Node\Expr\BinaryOp\NotEqual',
-        'PHPParser_Node_Expr_NotIdentical' => 'PhpParser\Node\Expr\BinaryOp\NotIdentical',
-        'PHPParser_Node_Expr_Plus' => 'PhpParser\Node\Expr\BinaryOp\Plus',
-        'PHPParser_Node_Expr_ShiftLeft' => 'PhpParser\Node\Expr\BinaryOp\ShiftLeft',
-        'PHPParser_Node_Expr_ShiftRight' => 'PhpParser\Node\Expr\BinaryOp\ShiftRight',
-        'PHPParser_Node_Expr_Smaller' => 'PhpParser\Node\Expr\BinaryOp\Smaller',
-        'PHPParser_Node_Expr_SmallerOrEqual' => 'PhpParser\Node\Expr\BinaryOp\SmallerOrEqual',
-
-        'PHPParser_Node_Expr_Array' => 'PhpParser\Node\Expr\Array_',
-        'PHPParser_Node_Expr_ArrayDimFetch' => 'PhpParser\Node\Expr\ArrayDimFetch',
-        'PHPParser_Node_Expr_ArrayItem' => 'PhpParser\Node\Expr\ArrayItem',
-        'PHPParser_Node_Expr_Assign' => 'PhpParser\Node\Expr\Assign',
-        'PHPParser_Node_Expr_AssignRef' => 'PhpParser\Node\Expr\AssignRef',
-        'PHPParser_Node_Expr_BitwiseNot' => 'PhpParser\Node\Expr\BitwiseNot',
-        'PHPParser_Node_Expr_BooleanNot' => 'PhpParser\Node\Expr\BooleanNot',
-        'PHPParser_Node_Expr_ClassConstFetch' => 'PhpParser\Node\Expr\ClassConstFetch',
-        'PHPParser_Node_Expr_Clone' => 'PhpParser\Node\Expr\Clone_',
-        'PHPParser_Node_Expr_Closure' => 'PhpParser\Node\Expr\Closure',
-        'PHPParser_Node_Expr_ClosureUse' => 'PhpParser\Node\Expr\ClosureUse',
-        'PHPParser_Node_Expr_ConstFetch' => 'PhpParser\Node\Expr\ConstFetch',
-        'PHPParser_Node_Expr_Empty' => 'PhpParser\Node\Expr\Empty_',
-        'PHPParser_Node_Expr_ErrorSuppress' => 'PhpParser\Node\Expr\ErrorSuppress',
-        'PHPParser_Node_Expr_Eval' => 'PhpParser\Node\Expr\Eval_',
-        'PHPParser_Node_Expr_Exit' => 'PhpParser\Node\Expr\Exit_',
-        'PHPParser_Node_Expr_FuncCall' => 'PhpParser\Node\Expr\FuncCall',
-        'PHPParser_Node_Expr_Include' => 'PhpParser\Node\Expr\Include_',
-        'PHPParser_Node_Expr_Instanceof' => 'PhpParser\Node\Expr\Instanceof_',
-        'PHPParser_Node_Expr_Isset' => 'PhpParser\Node\Expr\Isset_',
-        'PHPParser_Node_Expr_List' => 'PhpParser\Node\Expr\List_',
-        'PHPParser_Node_Expr_MethodCall' => 'PhpParser\Node\Expr\MethodCall',
-        'PHPParser_Node_Expr_New' => 'PhpParser\Node\Expr\New_',
-        'PHPParser_Node_Expr_PostDec' => 'PhpParser\Node\Expr\PostDec',
-        'PHPParser_Node_Expr_PostInc' => 'PhpParser\Node\Expr\PostInc',
-        'PHPParser_Node_Expr_PreDec' => 'PhpParser\Node\Expr\PreDec',
-        'PHPParser_Node_Expr_PreInc' => 'PhpParser\Node\Expr\PreInc',
-        'PHPParser_Node_Expr_Print' => 'PhpParser\Node\Expr\Print_',
-        'PHPParser_Node_Expr_PropertyFetch' => 'PhpParser\Node\Expr\PropertyFetch',
-        'PHPParser_Node_Expr_ShellExec' => 'PhpParser\Node\Expr\ShellExec',
-        'PHPParser_Node_Expr_StaticCall' => 'PhpParser\Node\Expr\StaticCall',
-        'PHPParser_Node_Expr_StaticPropertyFetch' => 'PhpParser\Node\Expr\StaticPropertyFetch',
-        'PHPParser_Node_Expr_Ternary' => 'PhpParser\Node\Expr\Ternary',
-        'PHPParser_Node_Expr_UnaryMinus' => 'PhpParser\Node\Expr\UnaryMinus',
-        'PHPParser_Node_Expr_UnaryPlus' => 'PhpParser\Node\Expr\UnaryPlus',
-        'PHPParser_Node_Expr_Variable' => 'PhpParser\Node\Expr\Variable',
-        'PHPParser_Node_Expr_Yield' => 'PhpParser\Node\Expr\Yield_',
-
-        'PHPParser_Node_Scalar_ClassConst' => 'PhpParser\Node\Scalar\MagicConst\Class_',
-        'PHPParser_Node_Scalar_DirConst' => 'PhpParser\Node\Scalar\MagicConst\Dir',
-        'PHPParser_Node_Scalar_FileConst' => 'PhpParser\Node\Scalar\MagicConst\File',
-        'PHPParser_Node_Scalar_FuncConst' => 'PhpParser\Node\Scalar\MagicConst\Function_',
-        'PHPParser_Node_Scalar_LineConst' => 'PhpParser\Node\Scalar\MagicConst\Line',
-        'PHPParser_Node_Scalar_MethodConst' => 'PhpParser\Node\Scalar\MagicConst\Method',
-        'PHPParser_Node_Scalar_NSConst' => 'PhpParser\Node\Scalar\MagicConst\Namespace_',
-        'PHPParser_Node_Scalar_TraitConst' => 'PhpParser\Node\Scalar\MagicConst\Trait_',
-
-        'PHPParser_Node_Scalar_DNumber' => 'PhpParser\Node\Scalar\DNumber',
-        'PHPParser_Node_Scalar_Encapsed' => 'PhpParser\Node\Scalar\Encapsed',
-        'PHPParser_Node_Scalar_LNumber' => 'PhpParser\Node\Scalar\LNumber',
-        'PHPParser_Node_Scalar_String' => 'PhpParser\Node\Scalar\String_',
-    );
 }
-
-class_alias('PhpParser\Autoloader', 'PHPParser_Autoloader');

lib/PhpParser/Builder/Class_.php

@@ -12,7 +12,7 @@ class Class_ extends Declaration
 
     protected $extends = null;
     protected $implements = array();
-    protected $type = 0;
+    protected $flags = 0;
 
     protected $uses = array();
     protected $constants = array();
@@ -44,8 +44,7 @@ class Class_ extends Declaration
     /**
      * Implements one or more interfaces.
      *
-     * @param Name|string $interface Name of interface to implement
-     * @param Name|string $...       More interfaces to implement
+     * @param Name|string ...$interfaces Names of interfaces to implement
      *
      * @return $this The builder instance (for fluid interface)
      */
@@ -113,7 +112,7 @@ class Class_ extends Declaration
      */
     public function getNode() {
         return new Stmt\Class_($this->name, array(
-            'type' => $this->type,
+            'flags' => $this->flags,
             'extends' => $this->extends,
             'implements' => $this->implements,
             'stmts' => array_merge($this->uses, $this->constants, $this->properties, $this->methods),

lib/PhpParser/Builder/Declaration.php

@@ -3,8 +3,6 @@
 namespace PhpParser\Builder;
 
 use PhpParser;
-use PhpParser\Node;
-use PhpParser\Node\Stmt;
 
 abstract class Declaration extends PhpParser\BuilderAbstract
 {

lib/PhpParser/Builder/FunctionLike.php

@@ -4,13 +4,15 @@ namespace PhpParser\Builder;
 
 use PhpParser;
 use PhpParser\Node;
-use PhpParser\Node\Stmt;
 
 abstract class FunctionLike extends Declaration
 {
     protected $returnByRef = false;
     protected $params = array();
 
+    /** @var string|Node\Name|Node\NullableType|null */
+    protected $returnType = null;
+
     /**
      * Make the function return by reference.
      *
@@ -55,4 +57,19 @@ abstract class FunctionLike extends Declaration
 
         return $this;
     }
-}
+
+    /**
+     * Sets the return type for PHP 7.
+     *
+     * @param string|Node\Name|Node\NullableType $type One of array, callable, string, int, float, bool, iterable,
+     *                               or a class/interface name.
+     *
+     * @return $this The builder instance (for fluid interface)
+     */
+    public function setReturnType($type)
+    {
+        $this->returnType = $this->normalizeType($type);
+
+        return $this;
+    }
+}

lib/PhpParser/Builder/Function_.php

@@ -40,9 +40,10 @@ class Function_ extends FunctionLike
      */
     public function getNode() {
         return new Stmt\Function_($this->name, array(
-            'byRef'  => $this->returnByRef,
-            'params' => $this->params,
-            'stmts'  => $this->stmts,
+            'byRef'      => $this->returnByRef,
+            'params'     => $this->params,
+            'returnType' => $this->returnType,
+            'stmts'      => $this->stmts,
         ), $this->attributes);
     }
-}
+}

lib/PhpParser/Builder/Interface_.php

@@ -25,8 +25,7 @@ class Interface_ extends Declaration
     /**
      * Extends one or more interfaces.
      *
-     * @param Name|string $interface Name of interface to extend
-     * @param Name|string $...       More interfaces to extend
+     * @param Name|string ...$interfaces Names of interfaces to extend
      *
      * @return $this The builder instance (for fluid interface)
      */

lib/PhpParser/Builder/Method.php

@@ -9,7 +9,9 @@ use PhpParser\Node\Stmt;
 class Method extends FunctionLike
 {
     protected $name;
-    protected $type = 0;
+    protected $flags = 0;
+
+    /** @var array|null */
     protected $stmts = array();
 
     /**
@@ -116,10 +118,11 @@ class Method extends FunctionLike
      */
     public function getNode() {
         return new Stmt\ClassMethod($this->name, array(
-            'type'   => $this->type,
-            'byRef'  => $this->returnByRef,
-            'params' => $this->params,
-            'stmts'  => $this->stmts,
+            'flags'      => $this->flags,
+            'byRef'      => $this->returnByRef,
+            'params'     => $this->params,
+            'returnType' => $this->returnType,
+            'stmts'      => $this->stmts,
         ), $this->attributes);
     }
 }

lib/PhpParser/Builder/Param.php

@@ -10,7 +10,10 @@ class Param extends PhpParser\BuilderAbstract
     protected $name;
 
     protected $default = null;
+
+    /** @var string|Node\Name|Node\NullableType|null */
     protected $type = null;
+
     protected $byRef = false;
 
     /**
@@ -38,15 +41,14 @@ class Param extends PhpParser\BuilderAbstract
     /**
      * Sets type hint for the parameter.
      *
-     * @param string|Node\Name $type Type hint to use
+     * @param string|Node\Name|Node\NullableType $type Type hint to use
      *
      * @return $this The builder instance (for fluid interface)
      */
     public function setTypeHint($type) {
-        if ($type === 'array' || $type === 'callable') {
-            $this->type = $type;
-        } else {
-            $this->type = $this->normalizeName($type);
+        $this->type = $this->normalizeType($type);
+        if ($this->type === 'void') {
+            throw new \LogicException('Parameter type cannot be void');
         }
 
         return $this;
@@ -73,4 +75,4 @@ class Param extends PhpParser\BuilderAbstract
             $this->name, $this->default, $this->type, $this->byRef
         );
     }
-}
+}

lib/PhpParser/Builder/Property.php

@@ -9,7 +9,7 @@ class Property extends PhpParser\BuilderAbstract
 {
     protected $name;
 
-    protected $type = 0;
+    protected $flags = 0;
     protected $default = null;
     protected $attributes = array();
 
@@ -101,7 +101,7 @@ class Property extends PhpParser\BuilderAbstract
      */
     public function getNode() {
         return new Stmt\Property(
-            $this->type !== 0 ? $this->type : Stmt\Class_::MODIFIER_PUBLIC,
+            $this->flags !== 0 ? $this->flags : Stmt\Class_::MODIFIER_PUBLIC,
             array(
                 new Stmt\PropertyProperty($this->name, $this->default)
             ),

lib/PhpParser/Builder/Trait_.php

@@ -3,12 +3,12 @@
 namespace PhpParser\Builder;
 
 use PhpParser;
-use PhpParser\Node\Name;
 use PhpParser\Node\Stmt;
 
 class Trait_ extends Declaration
 {
     protected $name;
+    protected $properties = array();
     protected $methods = array();
 
     /**
@@ -29,12 +29,15 @@ class Trait_ extends Declaration
      */
     public function addStmt($stmt) {
         $stmt = $this->normalizeNode($stmt);
-        if (!$stmt instanceof Stmt\ClassMethod) {
+
+        if ($stmt instanceof Stmt\Property) {
+            $this->properties[] = $stmt;
+        } else if ($stmt instanceof Stmt\ClassMethod) {
+            $this->methods[] = $stmt;
+        } else {
             throw new \LogicException(sprintf('Unexpected node of type "%s"', $stmt->getType()));
         }
 
-        $this->methods[] = $stmt;
-
         return $this;
     }
 
@@ -44,6 +47,10 @@ class Trait_ extends Declaration
      * @return Stmt\Trait_ The built interface node
      */
     public function getNode() {
-        return new Stmt\Trait_($this->name, $this->methods, $this->attributes);
+        return new Stmt\Trait_(
+            $this->name, array(
+                'stmts' => array_merge($this->properties, $this->methods)
+            ), $this->attributes
+        );
     }
 }

lib/PhpParser/Builder/Use_.php

@@ -36,8 +36,12 @@ class Use_ extends BuilderAbstract {
         $this->alias = $alias;
         return $this;
     }
-    public function __call($method, $args) {
-        return call_user_func_array(array($this, $method . '_'), $args);
+    public function __call($name, $args) {
+        if (method_exists($this, $name . '_')) {
+            return call_user_func_array(array($this, $name . '_'), $args);
+        }
+
+        throw new \LogicException(sprintf('Method "%s" does not exist', $name));
     }
 
     /**

lib/PhpParser/BuilderAbstract.php

@@ -2,11 +2,12 @@
 
 namespace PhpParser;
 
-use PhpParser\Node\Name;
-use PhpParser\Node\Expr;
-use PhpParser\Node\Stmt;
-use PhpParser\Node\Scalar;
 use PhpParser\Comment;
+use PhpParser\Node\Expr;
+use PhpParser\Node\Name;
+use PhpParser\Node\NullableType;
+use PhpParser\Node\Scalar;
+use PhpParser\Node\Stmt;
 
 abstract class BuilderAbstract implements Builder {
     /**
@@ -53,6 +54,49 @@ abstract class BuilderAbstract implements Builder {
         throw new \LogicException('Name must be a string or an instance of PhpParser\Node\Name');
     }
 
+    /**
+     * Normalizes a type: Converts plain-text type names into proper AST representation.
+     *
+     * In particular, builtin types are left as strings, custom types become Names and nullables
+     * are wrapped in NullableType nodes.
+     *
+     * @param Name|string|NullableType $type The type to normalize
+     *
+     * @return Name|string|NullableType The normalized type
+     */
+    protected function normalizeType($type) {
+        if (!is_string($type)) {
+            if (!$type instanceof Name && !$type instanceof NullableType) {
+                throw new \LogicException(
+                    'Type must be a string, or an instance of Name or NullableType');
+            }
+            return $type;
+        }
+
+        $nullable = false;
+        if (strlen($type) > 0 && $type[0] === '?') {
+            $nullable = true;
+            $type = substr($type, 1);
+        }
+
+        $builtinTypes = array(
+            'array', 'callable', 'string', 'int', 'float', 'bool', 'iterable', 'void'
+        );
+
+        $lowerType = strtolower($type);
+        if (in_array($lowerType, $builtinTypes)) {
+            $type = $lowerType;
+        } else {
+            $type = $this->normalizeName($type);
+        }
+
+        if ($nullable && $type === 'void') {
+            throw new \LogicException('void type cannot be nullable');
+        }
+
+        return $nullable ? new Node\NullableType($type) : $type;
+    }
+
     /**
      * Normalizes a value: Converts nulls, booleans, integers,
      * floats, strings and arrays into their respective nodes
@@ -125,7 +169,7 @@ abstract class BuilderAbstract implements Builder {
      * @param int $modifier Modifier to set
      */
     protected function setModifier($modifier) {
-        Stmt\Class_::verifyModifier($this->type, $modifier);
-        $this->type |= $modifier;
+        Stmt\Class_::verifyModifier($this->flags, $modifier);
+        $this->flags |= $modifier;
     }
 }

lib/PhpParser/BuilderFactory.php

@@ -20,7 +20,7 @@ class BuilderFactory
 {
     /**
      * Creates a namespace builder.
-     * 
+     *
      * @param null|string|Node\Name $name Name of the namespace
      *
      * @return Builder\Namespace_ The created namespace builder

lib/PhpParser/Comment.php

@@ -2,20 +2,23 @@
 
 namespace PhpParser;
 
-class Comment
+class Comment implements \JsonSerializable
 {
     protected $text;
     protected $line;
+    protected $filePos;
 
     /**
      * Constructs a comment node.
      *
-     * @param string $text Comment text (including comment delimiters like /*)
-     * @param int    $line Line number the comment started on
+     * @param string $text         Comment text (including comment delimiters like /*)
+     * @param int    $startLine    Line number the comment started on
+     * @param int    $startFilePos File offset the comment started on
      */
-    public function __construct($text, $line = -1) {
+    public function __construct($text, $startLine = -1, $startFilePos = -1) {
         $this->text = $text;
-        $this->line = $line;
+        $this->line = $startLine;
+        $this->filePos = $startFilePos;
     }
 
     /**
@@ -27,15 +30,6 @@ class Comment
         return $this->text;
     }
 
-    /**
-     * Sets the comment text.
-     *
-     * @param string $text The comment text (including comment delimiters like /*)
-     */
-    public function setText($text) {
-        $this->text = $text;
-    }
-
     /**
      * Gets the line number the comment started on.
      *
@@ -46,12 +40,12 @@ class Comment
     }
 
     /**
-     * Sets the line number the comment started on.
+     * Gets the file offset the comment started on.
      *
-     * @param int $line Line number
+     * @return int File offset
      */
-    public function setLine($line) {
-        $this->line = $line;
+    public function getFilePos() {
+        return $this->filePos;
     }
 
     /**
@@ -75,7 +69,8 @@ class Comment
      */
     public function getReformattedText() {
         $text = trim($this->text);
-        if (false === strpos($text, "\n")) {
+        $newlinePos = strpos($text, "\n");
+        if (false === $newlinePos) {
             // Single line comments don't need further processing
             return $text;
         } elseif (preg_match('((*BSR_ANYCRLF)(*ANYCRLF)^.*(?:\R\s+\*.*)+$)', $text)) {
@@ -105,15 +100,41 @@ class Comment
             //
             //     /* Some text.
             //        Some more text.
+            //          Indented text.
             //        Even more text. */
             //
-            // is handled by taking the length of the "/* " segment and leaving only that
-            // many space characters before the lines. Thus in the above example only three
-            // space characters are left at the start of every line.
-            return preg_replace('(^\s*(?= {' . strlen($matches[0]) . '}(?!\s)))m', '', $text);
+            // is handled by removing the difference between the shortest whitespace prefix on all
+            // lines and the length of the "/* " opening sequence.
+            $prefixLen = $this->getShortestWhitespacePrefixLen(substr($text, $newlinePos + 1));
+            $removeLen = $prefixLen - strlen($matches[0]);
+            return preg_replace('(^\s{' . $removeLen . '})m', '', $text);
         }
 
         // No idea how to format this comment, so simply return as is
         return $text;
     }
+
+    private function getShortestWhitespacePrefixLen($str) {
+        $lines = explode("\n", $str);
+        $shortestPrefixLen = INF;
+        foreach ($lines as $line) {
+            preg_match('(^\s*)', $line, $matches);
+            $prefixLen = strlen($matches[0]);
+            if ($prefixLen < $shortestPrefixLen) {
+                $shortestPrefixLen = $prefixLen;
+            }
+        }
+        return $shortestPrefixLen;
+    }
+
+    public function jsonSerialize() {
+        // Technically not a node, but we make it look like one anyway
+        $type = $this instanceof Comment\Doc ? 'Comment_Doc' : 'Comment';
+        return [
+            'nodeType' => $type,
+            'text' => $this->text,
+            'line' => $this->line,
+            'filePos' => $this->filePos,
+        ];
+    }
 }

lib/PhpParser/Error.php

@@ -5,17 +5,22 @@ namespace PhpParser;
 class Error extends \RuntimeException
 {
     protected $rawMessage;
-    protected $rawLine;
+    protected $attributes;
 
     /**
      * Creates an Exception signifying a parse error.
      *
-     * @param string $message Error message
-     * @param int    $line    Error line in PHP file
+     * @param string    $message    Error message
+     * @param array|int $attributes Attributes of node/token where error occurred
+     *                              (or start line of error -- deprecated)
      */
-    public function __construct($message, $line = -1) {
+    public function __construct($message, $attributes = array()) {
         $this->rawMessage = (string) $message;
-        $this->rawLine    = (int) $line;
+        if (is_array($attributes)) {
+            $this->attributes = $attributes;
+        } else {
+            $this->attributes = array('startLine' => $attributes);
+        }
         $this->updateMessage();
     }
 
@@ -28,6 +33,44 @@ class Error extends \RuntimeException
         return $this->rawMessage;
     }
 
+    /**
+     * Gets the line the error starts in.
+     *
+     * @return int Error start line
+     */
+    public function getStartLine() {
+        return isset($this->attributes['startLine']) ? $this->attributes['startLine'] : -1;
+    }
+
+    /**
+     * Gets the line the error ends in.
+     *
+     * @return int Error end line
+     */
+    public function getEndLine() {
+        return isset($this->attributes['endLine']) ? $this->attributes['endLine'] : -1;
+    }
+
+
+    /**
+     * Gets the attributes of the node/token the error occurred at.
+     *
+     * @return array
+     */
+    public function getAttributes() {
+        return $this->attributes;
+    }
+
+    /**
+     * Sets the attributes of the node/token the error occured at.
+     *
+     * @param array $attributes
+     */
+    public function setAttributes(array $attributes) {
+        $this->attributes = $attributes;
+        $this->updateMessage();
+    }
+
     /**
      * Sets the line of the PHP file the error occurred in.
      *
@@ -39,22 +82,73 @@ class Error extends \RuntimeException
     }
 
     /**
-     * Gets the error line in the PHP file.
+     * Sets the line the error starts in.
      *
-     * @return int Error line in the PHP file
+     * @param int $line Error start line
      */
-    public function getRawLine() {
-        return $this->rawLine;
+    public function setStartLine($line) {
+        $this->attributes['startLine'] = (int) $line;
+        $this->updateMessage();
     }
 
     /**
-     * Sets the line of the PHP file the error occurred in.
+     * Returns whether the error has start and end column information.
      *
-     * @param int $line Error line in the PHP file
+     * For column information enable the startFilePos and endFilePos in the lexer options.
+     *
+     * @return bool
      */
-    public function setRawLine($line) {
-        $this->rawLine = (int) $line;
-        $this->updateMessage();
+    public function hasColumnInfo() {
+        return isset($this->attributes['startFilePos']) && isset($this->attributes['endFilePos']);
+    }
+
+    /**
+     * Gets the start column (1-based) into the line where the error started.
+     *
+     * @param string $code Source code of the file
+     * @return int
+     */
+    public function getStartColumn($code) {
+        if (!$this->hasColumnInfo()) {
+            throw new \RuntimeException('Error does not have column information');
+        }
+
+        return $this->toColumn($code, $this->attributes['startFilePos']);
+    }
+
+    /**
+     * Gets the end column (1-based) into the line where the error ended.
+     *
+     * @param string $code Source code of the file
+     * @return int
+     */
+    public function getEndColumn($code) {
+        if (!$this->hasColumnInfo()) {
+            throw new \RuntimeException('Error does not have column information');
+        }
+
+        return $this->toColumn($code, $this->attributes['endFilePos']);
+    }
+
+    public function getMessageWithColumnInfo($code) {
+        return sprintf(
+            '%s from %d:%d to %d:%d', $this->getRawMessage(),
+            $this->getStartLine(), $this->getStartColumn($code),
+            $this->getEndLine(), $this->getEndColumn($code)
+        );
+    }
+
+    private function toColumn($code, $pos) {
+        if ($pos > strlen($code)) {
+            throw new \RuntimeException('Invalid position information');
+        }
+
+        $lineStartPos = strrpos($code, "\n", $pos - strlen($code));
+        if (false === $lineStartPos) {
+            $lineStartPos = -1;
+        }
+
+        return $pos - $lineStartPos;
     }
 
     /**
@@ -63,10 +157,10 @@ class Error extends \RuntimeException
     protected function updateMessage() {
         $this->message = $this->rawMessage;
 
-        if (-1 === $this->rawLine) {
+        if (-1 === $this->getStartLine()) {
             $this->message .= ' on unknown line';
         } else {
-            $this->message .= ' on line ' . $this->rawLine;
+            $this->message .= ' on line ' . $this->getStartLine();
         }
     }
-}
+}

lib/PhpParser/ErrorHandler.php

@@ -0,0 +1,13 @@
+<?php
+
+namespace PhpParser;
+
+interface ErrorHandler
+{
+    /**
+     * Handle an error generated during lexing, parsing or some other operation.
+     *
+     * @param Error $error The error that needs to be handled
+     */
+    public function handleError(Error $error);
+}

lib/PhpParser/ErrorHandler/Collecting.php

@@ -0,0 +1,46 @@
+<?php
+
+namespace PhpParser\ErrorHandler;
+
+use PhpParser\Error;
+use PhpParser\ErrorHandler;
+
+/**
+ * Error handler that collects all errors into an array.
+ *
+ * This allows graceful handling of errors.
+ */
+class Collecting implements ErrorHandler
+{
+    /** @var Error[] Collected errors */
+    private $errors = [];
+
+    public function handleError(Error $error) {
+        $this->errors[] = $error;
+    }
+
+    /**
+     * Get collected errors.
+     *
+     * @return Error[]
+     */
+    public function getErrors() {
+        return $this->errors;
+    }
+
+    /**
+     * Check whether there are any errors.
+     *
+     * @return bool
+     */
+    public function hasErrors() {
+        return !empty($this->errors);
+    }
+
+    /**
+     * Reset/clear collected errors.
+     */
+    public function clearErrors() {
+        $this->errors = [];
+    }
+}

lib/PhpParser/ErrorHandler/Throwing.php

@@ -0,0 +1,18 @@
+<?php
+
+namespace PhpParser\ErrorHandler;
+
+use PhpParser\Error;
+use PhpParser\ErrorHandler;
+
+/**
+ * Error handler that handles all errors by throwing them.
+ *
+ * This is the default strategy used by all components.
+ */
+class Throwing implements ErrorHandler
+{
+    public function handleError(Error $error) {
+        throw $error;
+    }
+}

lib/PhpParser/Lexer.php

@@ -2,6 +2,8 @@
 
 namespace PhpParser;
 
+use PhpParser\Parser\Tokens;
+
 class Lexer
 {
     protected $code;
@@ -9,6 +11,7 @@ class Lexer
     protected $pos;
     protected $line;
     protected $filePos;
+    protected $prevCloseTagHasNewline;
 
     protected $tokenMap;
     protected $dropTokens;
@@ -19,10 +22,10 @@ class Lexer
      * Creates a Lexer.
      *
      * @param array $options Options array. Currently only the 'usedAttributes' option is supported,
-     *                       which is an array of attributes to add to the AST nodes. Possible attributes
-     *                       are: 'comments', 'startLine', 'endLine', 'startTokenPos', 'endTokenPos',
-     *                       'startFilePos', 'endFilePos'. The option defaults to the first three.
-     *                       For more info see getNextToken() docs.
+     *                       which is an array of attributes to add to the AST nodes. Possible
+     *                       attributes are: 'comments', 'startLine', 'endLine', 'startTokenPos',
+     *                       'endTokenPos', 'startFilePos', 'endFilePos'. The option defaults to the
+     *                       first three. For more info see getNextToken() docs.
      */
     public function __construct(array $options = array()) {
         // map from internal tokens to PhpParser tokens
@@ -30,7 +33,9 @@ class Lexer
 
         // map of tokens to drop while lexing (the map is only used for isset lookup,
         // that's why the value is simply set to 1; the value is never actually used.)
-        $this->dropTokens = array_fill_keys(array(T_WHITESPACE, T_OPEN_TAG), 1);
+        $this->dropTokens = array_fill_keys(
+            array(T_WHITESPACE, T_OPEN_TAG, T_COMMENT, T_DOC_COMMENT), 1
+        );
 
         // the usedAttributes member is a map of the used attribute names to a dummy
         // value (here "true")
@@ -43,57 +48,152 @@ class Lexer
     /**
      * Initializes the lexer for lexing the provided source code.
      *
-     * @param string $code The source code to lex
+     * This function does not throw if lexing errors occur. Instead, errors may be retrieved using
+     * the getErrors() method.
      *
-     * @throws Error on lexing errors (unterminated comment or unexpected character)
+     * @param string $code The source code to lex
+     * @param ErrorHandler|null $errorHandler Error handler to use for lexing errors. Defaults to
+     *                                        ErrorHandler\Throwing
      */
-    public function startLexing($code) {
-        $scream = ini_set('xdebug.scream', '0');
-
-        $this->resetErrors();
-        $this->tokens = @token_get_all($code);
-        $this->handleErrors();
-
-        if (false !== $scream) {
-            ini_set('xdebug.scream', $scream);
+    public function startLexing($code, ErrorHandler $errorHandler = null) {
+        if (null === $errorHandler) {
+            $errorHandler = new ErrorHandler\Throwing();
         }
 
         $this->code = $code; // keep the code around for __halt_compiler() handling
         $this->pos  = -1;
         $this->line =  1;
         $this->filePos = 0;
+
+        // If inline HTML occurs without preceding code, treat it as if it had a leading newline.
+        // This ensures proper composability, because having a newline is the "safe" assumption.
+        $this->prevCloseTagHasNewline = true;
+
+        $scream = ini_set('xdebug.scream', '0');
+
+        $this->resetErrors();
+        $this->tokens = @token_get_all($code);
+        $this->handleErrors($errorHandler);
+
+        if (false !== $scream) {
+            ini_set('xdebug.scream', $scream);
+        }
     }
 
     protected function resetErrors() {
-        // set error_get_last() to defined state by forcing an undefined variable error
-        set_error_handler(function() { return false; }, 0);
-        @$undefinedVariable;
-        restore_error_handler();
+        if (function_exists('error_clear_last')) {
+            error_clear_last();
+        } else {
+            // set error_get_last() to defined state by forcing an undefined variable error
+            set_error_handler(function() { return false; }, 0);
+            @$undefinedVariable;
+            restore_error_handler();
+        }
     }
 
-    protected function handleErrors() {
+    private function handleInvalidCharacterRange($start, $end, $line, ErrorHandler $errorHandler) {
+        for ($i = $start; $i < $end; $i++) {
+            $chr = $this->code[$i];
+            if ($chr === 'b' || $chr === 'B') {
+                // HHVM does not treat b" tokens correctly, so ignore these
+                continue;
+            }
+
+            if ($chr === "\0") {
+                // PHP cuts error message after null byte, so need special case
+                $errorMsg = 'Unexpected null byte';
+            } else {
+                $errorMsg = sprintf(
+                    'Unexpected character "%s" (ASCII %d)', $chr, ord($chr)
+                );
+            }
+
+            $errorHandler->handleError(new Error($errorMsg, [
+                'startLine' => $line,
+                'endLine' => $line,
+                'startFilePos' => $i,
+                'endFilePos' => $i,
+            ]));
+        }
+    }
+
+    private function isUnterminatedComment($token) {
+        return ($token[0] === T_COMMENT || $token[0] === T_DOC_COMMENT)
+            && substr($token[1], 0, 2) === '/*'
+            && substr($token[1], -2) !== '*/';
+    }
+
+    private function errorMayHaveOccurred() {
+        if (defined('HHVM_VERSION')) {
+            // In HHVM token_get_all() does not throw warnings, so we need to conservatively
+            // assume that an error occurred
+            return true;
+        }
+
         $error = error_get_last();
+        return null !== $error
+            && false === strpos($error['message'], 'Undefined variable');
+    }
 
-        if (preg_match(
-            '~^Unterminated comment starting line ([0-9]+)$~',
-            $error['message'], $matches
-        )) {
-            throw new Error('Unterminated comment', (int) $matches[1]);
+    protected function handleErrors(ErrorHandler $errorHandler) {
+        if (!$this->errorMayHaveOccurred()) {
+            return;
         }
 
-        if (preg_match(
-            '~^Unexpected character in input:  \'(.)\' \(ASCII=([0-9]+)\)~s',
-            $error['message'], $matches
-        )) {
-            throw new Error(sprintf(
-                'Unexpected character "%s" (ASCII %d)',
-                $matches[1], $matches[2]
-            ));
+        // PHP's error handling for token_get_all() is rather bad, so if we want detailed
+        // error information we need to compute it ourselves. Invalid character errors are
+        // detected by finding "gaps" in the token array. Unterminated comments are detected
+        // by checking if a trailing comment has a "*/" at the end.
+
+        $filePos = 0;
+        $line = 1;
+        foreach ($this->tokens as $i => $token) {
+            $tokenValue = \is_string($token) ? $token : $token[1];
+            $tokenLen = \strlen($tokenValue);
+
+            if (substr($this->code, $filePos, $tokenLen) !== $tokenValue) {
+                // Something is missing, must be an invalid character
+                $nextFilePos = strpos($this->code, $tokenValue, $filePos);
+                $this->handleInvalidCharacterRange(
+                    $filePos, $nextFilePos, $line, $errorHandler);
+                $filePos = $nextFilePos;
+            }
+
+            $filePos += $tokenLen;
+            $line += substr_count($tokenValue, "\n");
         }
 
-        // PHP cuts error message after null byte, so need special case
-        if (preg_match('~^Unexpected character in input:  \'$~', $error['message'])) {
-            throw new Error('Unexpected null byte');
+        if ($filePos !== \strlen($this->code)) {
+            if (substr($this->code, $filePos, 2) === '/*') {
+                // Unlike PHP, HHVM will drop unterminated comments entirely
+                $comment = substr($this->code, $filePos);
+                $errorHandler->handleError(new Error('Unterminated comment', [
+                    'startLine' => $line,
+                    'endLine' => $line + substr_count($comment, "\n"),
+                    'startFilePos' => $filePos,
+                    'endFilePos' => $filePos + \strlen($comment),
+                ]));
+
+                // Emulate the PHP behavior
+                $isDocComment = isset($comment[3]) && $comment[3] === '*';
+                $this->tokens[] = [$isDocComment ? T_DOC_COMMENT : T_COMMENT, $comment, $line];
+            } else {
+                // Invalid characters at the end of the input
+                $this->handleInvalidCharacterRange(
+                    $filePos, \strlen($this->code), $line, $errorHandler);
+            }
+            return;
+        }
+
+        // Check for unterminated comment
+        $lastToken = $this->tokens[count($this->tokens) - 1];
+        if ($this->isUnterminatedComment($lastToken)) {
+            $errorHandler->handleError(new Error('Unterminated comment', [
+                'startLine' => $line - substr_count($lastToken[1], "\n"),
+                'endLine' => $line,
+                'startFilePos' => $filePos - \strlen($lastToken[1]),
+                'endFilePos' => $filePos,
+            ]));
         }
     }
 
@@ -111,7 +211,7 @@ class Lexer
      *  * 'startTokenPos' => Offset into the token array of the first token in the node.
      *  * 'endTokenPos'   => Offset into the token array of the last token in the node.
      *  * 'startFilePos'  => Offset into the code string of the first character that is part of the node.
-     *  * 'endFilePos'    => Offset into the code string of the last character that is part of the node
+     *  * 'endFilePos'    => Offset into the code string of the last character that is part of the node.
      *
      * @param mixed $value           Variable to store token content in
      * @param mixed $startAttributes Variable to store start attributes in
@@ -123,9 +223,17 @@ class Lexer
         $startAttributes = array();
         $endAttributes   = array();
 
-        while (isset($this->tokens[++$this->pos])) {
-            $token = $this->tokens[$this->pos];
+        while (1) {
+            if (isset($this->tokens[++$this->pos])) {
+                $token = $this->tokens[$this->pos];
+            } else {
+                // EOF token with ID 0
+                $token = "\0";
+            }
 
+            if (isset($this->usedAttributes['startLine'])) {
+                $startAttributes['startLine'] = $this->line;
+            }
             if (isset($this->usedAttributes['startTokenPos'])) {
                 $startAttributes['startTokenPos'] = $this->pos;
             }
@@ -133,69 +241,56 @@ class Lexer
                 $startAttributes['startFilePos'] = $this->filePos;
             }
 
-            if (is_string($token)) {
-                // bug in token_get_all
-                if ('b"' === $token) {
-                    $value = 'b"';
+            if (\is_string($token)) {
+                $value = $token;
+                if (isset($token[1])) {
+                    // bug in token_get_all
                     $this->filePos += 2;
                     $id = ord('"');
                 } else {
-                    $value = $token;
                     $this->filePos += 1;
                     $id = ord($token);
                 }
-
-                if (isset($this->usedAttributes['startLine'])) {
-                    $startAttributes['startLine'] = $this->line;
-                }
-                if (isset($this->usedAttributes['endLine'])) {
-                    $endAttributes['endLine'] = $this->line;
-                }
-                if (isset($this->usedAttributes['endTokenPos'])) {
-                    $endAttributes['endTokenPos'] = $this->pos;
-                }
-                if (isset($this->usedAttributes['endFilePos'])) {
-                    $endAttributes['endFilePos'] = $this->filePos - 1;
+            } elseif (!isset($this->dropTokens[$token[0]])) {
+                $value = $token[1];
+                $id = $this->tokenMap[$token[0]];
+                if (T_CLOSE_TAG === $token[0]) {
+                    $this->prevCloseTagHasNewline = false !== strpos($token[1], "\n");
+                } else if (T_INLINE_HTML === $token[0]) {
+                    $startAttributes['hasLeadingNewline'] = $this->prevCloseTagHasNewline;
                 }
 
-                return $id;
+                $this->line += substr_count($value, "\n");
+                $this->filePos += \strlen($value);
             } else {
-                $this->line += substr_count($token[1], "\n");
-                $this->filePos += strlen($token[1]);
-
-                if (T_COMMENT === $token[0]) {
+                if (T_COMMENT === $token[0] || T_DOC_COMMENT === $token[0]) {
                     if (isset($this->usedAttributes['comments'])) {
-                        $startAttributes['comments'][] = new Comment($token[1], $token[2]);
+                        $comment = T_DOC_COMMENT === $token[0]
+                            ? new Comment\Doc($token[1], $this->line, $this->filePos)
+                            : new Comment($token[1], $this->line, $this->filePos);
+                        $startAttributes['comments'][] = $comment;
                     }
-                } elseif (T_DOC_COMMENT === $token[0]) {
-                    if (isset($this->usedAttributes['comments'])) {
-                        $startAttributes['comments'][] = new Comment\Doc($token[1], $token[2]);
-                    }
-                } elseif (!isset($this->dropTokens[$token[0]])) {
-                    $value = $token[1];
-
-                    if (isset($this->usedAttributes['startLine'])) {
-                        $startAttributes['startLine'] = $token[2];
-                    }
-                    if (isset($this->usedAttributes['endLine'])) {
-                        $endAttributes['endLine'] = $this->line;
-                    }
-                    if (isset($this->usedAttributes['endTokenPos'])) {
-                        $endAttributes['endTokenPos'] = $this->pos;
-                    }
-                    if (isset($this->usedAttributes['endFilePos'])) {
-                        $endAttributes['endFilePos'] = $this->filePos - 1;
-                    }
-
-                    return $this->tokenMap[$token[0]];
                 }
+
+                $this->line += substr_count($token[1], "\n");
+                $this->filePos += \strlen($token[1]);
+                continue;
             }
+
+            if (isset($this->usedAttributes['endLine'])) {
+                $endAttributes['endLine'] = $this->line;
+            }
+            if (isset($this->usedAttributes['endTokenPos'])) {
+                $endAttributes['endTokenPos'] = $this->pos;
+            }
+            if (isset($this->usedAttributes['endFilePos'])) {
+                $endAttributes['endFilePos'] = $this->filePos - 1;
+            }
+
+            return $id;
         }
 
-        $startAttributes['startLine'] = $this->line;
-
-        // 0 is the EOF token
-        return 0;
+        throw new \RuntimeException('Reached end of lexer loop');
     }
 
     /**
@@ -218,23 +313,13 @@ class Lexer
      * @return string Remaining text
      */
     public function handleHaltCompiler() {
-        // get the length of the text before the T_HALT_COMPILER token
-        $textBefore = '';
-        for ($i = 0; $i <= $this->pos; ++$i) {
-            if (is_string($this->tokens[$i])) {
-                $textBefore .= $this->tokens[$i];
-            } else {
-                $textBefore .= $this->tokens[$i][1];
-            }
-        }
-
         // text after T_HALT_COMPILER, still including ();
-        $textAfter = substr($this->code, strlen($textBefore));
+        $textAfter = substr($this->code, $this->filePos);
 
         // ensure that it is followed by ();
         // this simplifies the situation, by not allowing any comments
         // in between of the tokens.
-        if (!preg_match('~\s*\(\s*\)\s*(?:;|\?>\r?\n?)~', $textAfter, $matches)) {
+        if (!preg_match('~^\s*\(\s*\)\s*(?:;|\?>\r?\n?)~', $textAfter, $matches)) {
             throw new Error('__HALT_COMPILER must be followed by "();"');
         }
 
@@ -262,18 +347,18 @@ class Lexer
         for ($i = 256; $i < 1000; ++$i) {
             if (T_DOUBLE_COLON === $i) {
                 // T_DOUBLE_COLON is equivalent to T_PAAMAYIM_NEKUDOTAYIM
-                $tokenMap[$i] = Parser::T_PAAMAYIM_NEKUDOTAYIM;
+                $tokenMap[$i] = Tokens::T_PAAMAYIM_NEKUDOTAYIM;
             } elseif(T_OPEN_TAG_WITH_ECHO === $i) {
                 // T_OPEN_TAG_WITH_ECHO with dropped T_OPEN_TAG results in T_ECHO
-                $tokenMap[$i] = Parser::T_ECHO;
+                $tokenMap[$i] = Tokens::T_ECHO;
             } elseif(T_CLOSE_TAG === $i) {
                 // T_CLOSE_TAG is equivalent to ';'
                 $tokenMap[$i] = ord(';');
             } elseif ('UNKNOWN' !== $name = token_name($i)) {
                 if ('T_HASHBANG' === $name) {
                     // HHVM uses a special token for #! hashbang lines
-                    $tokenMap[$i] = Parser::T_INLINE_HTML;
-                } else if (defined($name = 'PhpParser\Parser::' . $name)) {
+                    $tokenMap[$i] = Tokens::T_INLINE_HTML;
+                } else if (defined($name = 'PhpParser\Parser\Tokens::' . $name)) {
                     // Other tokens can be mapped directly
                     $tokenMap[$i] = constant($name);
                 }
@@ -282,7 +367,11 @@ class Lexer
 
         // HHVM uses a special token for numbers that overflow to double
         if (defined('T_ONUMBER')) {
-            $tokenMap[T_ONUMBER] = Parser::T_DNUMBER;
+            $tokenMap[T_ONUMBER] = Tokens::T_DNUMBER;
+        }
+        // HHVM also has a separate token for the __COMPILER_HALT_OFFSET__ constant
+        if (defined('T_COMPILER_HALT_OFFSET')) {
+            $tokenMap[T_COMPILER_HALT_OFFSET] = Tokens::T_STRING;
         }
 
         return $tokenMap;

lib/PhpParser/Lexer/Emulative.php

@@ -2,41 +2,29 @@
 
 namespace PhpParser\Lexer;
 
-use PhpParser\Parser;
+use PhpParser\ErrorHandler;
+use PhpParser\Parser\Tokens;
 
-/**
- * ATTENTION: This code is WRITE-ONLY. Do not try to read it.
- */
 class Emulative extends \PhpParser\Lexer
 {
     protected $newKeywords;
     protected $inObjectAccess;
 
-    const T_ELLIPSIS  = 1001;
-    const T_POW       = 1002;
-    const T_POW_EQUAL = 1003;
-    const T_COALESCE  = 1004;
-    const T_SPACESHIP = 1005;
+    const T_ELLIPSIS   = 1001;
+    const T_POW        = 1002;
+    const T_POW_EQUAL  = 1003;
+    const T_COALESCE   = 1004;
+    const T_SPACESHIP  = 1005;
+    const T_YIELD_FROM = 1006;
 
     const PHP_7_0 = '7.0.0dev';
     const PHP_5_6 = '5.6.0rc1';
-    const PHP_5_5 = '5.5.0beta1';
-    const PHP_5_4 = '5.4.0beta1';
 
     public function __construct(array $options = array()) {
         parent::__construct($options);
 
         $newKeywordsPerVersion = array(
-            self::PHP_5_5 => array(
-                'finally'       => Parser::T_FINALLY,
-                'yield'         => Parser::T_YIELD,
-            ),
-            self::PHP_5_4 => array(
-                'callable'      => Parser::T_CALLABLE,
-                'insteadof'     => Parser::T_INSTEADOF,
-                'trait'         => Parser::T_TRAIT,
-                '__trait__'     => Parser::T_TRAIT_C,
-            ),
+            // No new keywords since PHP 5.5
         );
 
         $this->newKeywords = array();
@@ -51,145 +39,118 @@ class Emulative extends \PhpParser\Lexer
         if (version_compare(PHP_VERSION, self::PHP_7_0, '>=')) {
             return;
         }
-        $this->tokenMap[self::T_COALESCE] = Parser::T_COALESCE;
-        $this->tokenMap[self::T_SPACESHIP] = Parser::T_SPACESHIP;
+        $this->tokenMap[self::T_COALESCE]   = Tokens::T_COALESCE;
+        $this->tokenMap[self::T_SPACESHIP]  = Tokens::T_SPACESHIP;
+        $this->tokenMap[self::T_YIELD_FROM] = Tokens::T_YIELD_FROM;
 
         if (version_compare(PHP_VERSION, self::PHP_5_6, '>=')) {
             return;
         }
-        $this->tokenMap[self::T_ELLIPSIS]  = Parser::T_ELLIPSIS;
-        $this->tokenMap[self::T_POW]       = Parser::T_POW;
-        $this->tokenMap[self::T_POW_EQUAL] = Parser::T_POW_EQUAL;
+        $this->tokenMap[self::T_ELLIPSIS]  = Tokens::T_ELLIPSIS;
+        $this->tokenMap[self::T_POW]       = Tokens::T_POW;
+        $this->tokenMap[self::T_POW_EQUAL] = Tokens::T_POW_EQUAL;
     }
 
-    public function startLexing($code) {
+    public function startLexing($code, ErrorHandler $errorHandler = null) {
         $this->inObjectAccess = false;
 
-        $preprocessedCode = $this->preprocessCode($code);
-        parent::startLexing($preprocessedCode);
-        if ($preprocessedCode !== $code) {
-            $this->postprocessTokens();
+        parent::startLexing($code, $errorHandler);
+        if ($this->requiresEmulation($code)) {
+            $this->emulateTokens();
         }
-
-        // Set code property back to the original code, so __halt_compiler()
-        // handling and (start|end)FilePos attributes use the correct offsets
-        $this->code = $code;
     }
 
     /*
-     * Replaces new features in the code by ~__EMU__{NAME}__{DATA}__~ sequences.
-     * ~LABEL~ is never valid PHP code, that's why we can (to some degree) safely
-     * use it here.
-     * Later when preprocessing the tokens these sequences will either be replaced
-     * by real tokens or replaced with their original content (e.g. if they occurred
-     * inside a string, i.e. a place where they don't have a special meaning).
+     * Checks if the code is potentially using features that require emulation.
      */
-    protected function preprocessCode($code) {
+    protected function requiresEmulation($code) {
         if (version_compare(PHP_VERSION, self::PHP_7_0, '>=')) {
-            return $code;
+            return false;
         }
 
-        $code = str_replace('??', '~__EMU__COALESCE__~', $code);
-        $code = str_replace('<=>', '~__EMU__SPACESHIP__~', $code);
+        if (preg_match('(\?\?|<=>|yield[ \n\r\t]+from)', $code)) {
+            return true;
+        }
 
         if (version_compare(PHP_VERSION, self::PHP_5_6, '>=')) {
-            return $code;
+            return false;
         }
 
-        $code = str_replace('...', '~__EMU__ELLIPSIS__~', $code);
-        $code = preg_replace('((?<!/)\*\*=)', '~__EMU__POWEQUAL__~', $code);
-        $code = preg_replace('((?<!/)\*\*(?!/))', '~__EMU__POW__~', $code);
-
-        if (version_compare(PHP_VERSION, self::PHP_5_4, '>=')) {
-            return $code;
-        }
-
-        // binary notation (0b010101101001...)
-        return preg_replace('(\b0b[01]+\b)', '~__EMU__BINARY__$0__~', $code);
+        return preg_match('(\.\.\.|(?<!/)\*\*(?!/))', $code);
     }
 
     /*
-     * Replaces the ~__EMU__...~ sequences with real tokens or their original
-     * value.
+     * Emulates tokens for newer PHP versions.
      */
-    protected function postprocessTokens() {
-        // we need to manually iterate and manage a count because we'll change
+    protected function emulateTokens() {
+        // We need to manually iterate and manage a count because we'll change
         // the tokens array on the way
+        $line = 1;
         for ($i = 0, $c = count($this->tokens); $i < $c; ++$i) {
-            // first check that the following tokens are of form ~LABEL~,
-            // then match the __EMU__... sequence.
-            if ('~' === $this->tokens[$i]
-                && isset($this->tokens[$i + 2])
-                && '~' === $this->tokens[$i + 2]
-                && T_STRING === $this->tokens[$i + 1][0]
-                && preg_match('(^__EMU__([A-Z]++)__(?:([A-Za-z0-9]++)__)?$)', $this->tokens[$i + 1][1], $matches)
-            ) {
-                if ('BINARY' === $matches[1]) {
-                    // the binary number can either be an integer or a double, so return a LNUMBER
-                    // or DNUMBER respectively
-                    $replace = array(
-                        array(is_int(bindec($matches[2])) ? T_LNUMBER : T_DNUMBER, $matches[2], $this->tokens[$i + 1][2])
-                    );
-                } else if ('ELLIPSIS' === $matches[1]) {
-                    $replace = array(
-                        array(self::T_ELLIPSIS, '...', $this->tokens[$i + 1][2])
-                    );
-                } else if ('POW' === $matches[1]) {
-                    $replace = array(
-                        array(self::T_POW, '**', $this->tokens[$i + 1][2])
-                    );
-                } else if ('POWEQUAL' === $matches[1]) {
-                    $replace = array(
-                        array(self::T_POW_EQUAL, '**=', $this->tokens[$i + 1][2])
-                    );
-                } else if ('COALESCE' === $matches[1]) {
-                    $replace = array(
-                        array(self::T_COALESCE, '??', $this->tokens[$i + 1][2])
-                    );
-                } else if ('SPACESHIP' === $matches[1]) {
-                    $replace = array(
-                        array(self::T_SPACESHIP, '<=>', $this->tokens[$i + 1][2]),
-                    );
-                } else {
-                    // just ignore all other __EMU__ sequences
+            $replace = null;
+            if (isset($this->tokens[$i + 1])) {
+                if ($this->tokens[$i] === '?' && $this->tokens[$i + 1] === '?') {
+                    array_splice($this->tokens, $i, 2, array(
+                        array(self::T_COALESCE, '??', $line)
+                    ));
+                    $c--;
+                    continue;
+                }
+                if ($this->tokens[$i][0] === T_IS_SMALLER_OR_EQUAL
+                    && $this->tokens[$i + 1] === '>'
+                ) {
+                    array_splice($this->tokens, $i, 2, array(
+                        array(self::T_SPACESHIP, '<=>', $line)
+                    ));
+                    $c--;
+                    continue;
+                }
+                if ($this->tokens[$i] === '*' && $this->tokens[$i + 1] === '*') {
+                    array_splice($this->tokens, $i, 2, array(
+                        array(self::T_POW, '**', $line)
+                    ));
+                    $c--;
+                    continue;
+                }
+                if ($this->tokens[$i] === '*' && $this->tokens[$i + 1][0] === T_MUL_EQUAL) {
+                    array_splice($this->tokens, $i, 2, array(
+                        array(self::T_POW_EQUAL, '**=', $line)
+                    ));
+                    $c--;
                     continue;
                 }
-
-                array_splice($this->tokens, $i, 3, $replace);
-                $c -= 3 - count($replace);
-            // for multichar tokens (e.g. strings) replace any ~__EMU__...~ sequences
-            // in their content with the original character sequence
-            } elseif (is_array($this->tokens[$i])
-                      && 0 !== strpos($this->tokens[$i][1], '__EMU__')
-            ) {
-                $this->tokens[$i][1] = preg_replace_callback(
-                    '(~__EMU__([A-Z]++)__(?:([A-Za-z0-9]++)__)?~)',
-                    array($this, 'restoreContentCallback'),
-                    $this->tokens[$i][1]
-                );
             }
-        }
-    }
 
-    /*
-     * This method is a callback for restoring EMU sequences in
-     * multichar tokens (like strings) to their original value.
-     */
-    public function restoreContentCallback(array $matches) {
-        if ('BINARY' === $matches[1]) {
-            return $matches[2];
-        } else if ('ELLIPSIS' === $matches[1]) {
-            return '...';
-        } else if ('POW' === $matches[1]) {
-            return '**';
-        } else if ('POWEQUAL' === $matches[1]) {
-            return '**=';
-        } else if ('COALESCE' === $matches[1]) {
-            return '??';
-        } else if ('SPACESHIP' === $matches[1]) {
-            return '<=>';
-        } else {
-            return $matches[0];
+            if (isset($this->tokens[$i + 2])) {
+                if ($this->tokens[$i][0] === T_YIELD && $this->tokens[$i + 1][0] === T_WHITESPACE
+                    && $this->tokens[$i + 2][0] === T_STRING
+                    && !strcasecmp($this->tokens[$i + 2][1], 'from')
+                ) {
+                    array_splice($this->tokens, $i, 3, array(
+                        array(
+                            self::T_YIELD_FROM,
+                            $this->tokens[$i][1] . $this->tokens[$i + 1][1] . $this->tokens[$i + 2][1],
+                            $line
+                        )
+                    ));
+                    $c -= 2;
+                    $line += substr_count($this->tokens[$i][1], "\n");
+                    continue;
+                }
+                if ($this->tokens[$i] === '.' && $this->tokens[$i + 1] === '.'
+                    && $this->tokens[$i + 2] === '.'
+                ) {
+                    array_splice($this->tokens, $i, 3, array(
+                        array(self::T_ELLIPSIS, '...', $line)
+                    ));
+                    $c -= 2;
+                    continue;
+                }
+            }
+
+            if (\is_array($this->tokens[$i])) {
+                $line += substr_count($this->tokens[$i][1], "\n");
+            }
         }
     }
 
@@ -199,15 +160,13 @@ class Emulative extends \PhpParser\Lexer
         // replace new keywords by their respective tokens. This is not done
         // if we currently are in an object access (e.g. in $obj->namespace
         // "namespace" stays a T_STRING tokens and isn't converted to T_NAMESPACE)
-        if (Parser::T_STRING === $token && !$this->inObjectAccess) {
+        if (Tokens::T_STRING === $token && !$this->inObjectAccess) {
             if (isset($this->newKeywords[strtolower($value)])) {
                 return $this->newKeywords[strtolower($value)];
             }
-        // keep track of whether we currently are in an object access (after ->)
-        } elseif (Parser::T_OBJECT_OPERATOR === $token) {
-            $this->inObjectAccess = true;
         } else {
-            $this->inObjectAccess = false;
+            // keep track of whether we currently are in an object access (after ->)
+            $this->inObjectAccess = Tokens::T_OBJECT_OPERATOR === $token;
         }
 
         return $token;

lib/PhpParser/Node.php

@@ -41,6 +41,15 @@ interface Node
      */
     public function getDocComment();
 
+    /**
+     * Sets the doc comment of the node.
+     *
+     * This will either replace an existing doc comment or add it to the comments array.
+     *
+     * @param Comment\Doc $docComment Doc comment to set
+     */
+    public function setDocComment(Comment\Doc $docComment);
+
     /**
      * Sets an attribute on a node.
      *

lib/PhpParser/Node/Arg.php

@@ -22,7 +22,7 @@ class Arg extends NodeAbstract
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $value, $byRef = false, $unpack = false, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->value = $value;
         $this->byRef = $byRef;
         $this->unpack = $unpack;

lib/PhpParser/Node/Const_.php

@@ -19,7 +19,7 @@ class Const_ extends NodeAbstract
      * @param array   $attributes Additional attributes
      */
     public function __construct($name, Expr $value, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->name = $name;
         $this->value = $value;
     }

lib/PhpParser/Node/Expr/ArrayDimFetch.php

@@ -19,12 +19,12 @@ class ArrayDimFetch extends Expr
      * @param array     $attributes Additional attributes
      */
     public function __construct(Expr $var, Expr $dim = null, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->dim = $dim;
     }
 
-    public function getSubnodeNames() {
+    public function getSubNodeNames() {
         return array('var', 'dim');
     }
 }

lib/PhpParser/Node/Expr/ArrayItem.php

@@ -22,7 +22,7 @@ class ArrayItem extends Expr
      * @param array     $attributes Additional attributes
      */
     public function __construct(Expr $value, Expr $key = null, $byRef = false, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->key = $key;
         $this->value = $value;
         $this->byRef = $byRef;

lib/PhpParser/Node/Expr/Array_.php

@@ -6,6 +6,10 @@ use PhpParser\Node\Expr;
 
 class Array_ extends Expr
 {
+    // For use in "kind" attribute
+    const KIND_LONG = 1;  // array() syntax
+    const KIND_SHORT = 2; // [] syntax
+
     /** @var ArrayItem[] Items */
     public $items;
 
@@ -16,7 +20,7 @@ class Array_ extends Expr
      * @param array       $attributes Additional attributes
      */
     public function __construct(array $items = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->items = $items;
     }
 

lib/PhpParser/Node/Expr/Assign.php

@@ -19,7 +19,7 @@ class Assign extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->expr = $expr;
     }

lib/PhpParser/Node/Expr/AssignOp.php

@@ -4,10 +4,6 @@ namespace PhpParser\Node\Expr;
 
 use PhpParser\Node\Expr;
 
-/**
- * @property Expr $var  Variable
- * @property Expr $expr Expression
- */
 abstract class AssignOp extends Expr
 {
     /** @var Expr Variable */
@@ -23,7 +19,7 @@ abstract class AssignOp extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->expr = $expr;
     }

lib/PhpParser/Node/Expr/AssignRef.php

@@ -4,10 +4,6 @@ namespace PhpParser\Node\Expr;
 
 use PhpParser\Node\Expr;
 
-/**
- * @property Expr $var  Variable reference is assigned to
- * @property Expr $expr Variable which is referenced
- */
 class AssignRef extends Expr
 {
     /** @var Expr Variable reference is assigned to */
@@ -23,7 +19,7 @@ class AssignRef extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->expr = $expr;
     }

lib/PhpParser/Node/Expr/BinaryOp.php

@@ -19,7 +19,7 @@ abstract class BinaryOp extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $left, Expr $right, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->left = $left;
         $this->right = $right;
     }

lib/PhpParser/Node/Expr/BitwiseNot.php

@@ -16,7 +16,7 @@ class BitwiseNot extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/BooleanNot.php

@@ -16,7 +16,7 @@ class BooleanNot extends Expr
      * @param array               $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/Cast.php

@@ -16,7 +16,7 @@ abstract class Cast extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/ClassConstFetch.php

@@ -2,25 +2,25 @@
 
 namespace PhpParser\Node\Expr;
 
-use PhpParser\Node\Name;
 use PhpParser\Node\Expr;
+use PhpParser\Node\Name;
 
 class ClassConstFetch extends Expr
 {
     /** @var Name|Expr Class name */
     public $class;
-    /** @var string Constant name */
+    /** @var string|Error Constant name */
     public $name;
 
     /**
      * Constructs a class const fetch node.
      *
-     * @param Name|Expr $class      Class name
-     * @param string    $name       Constant name
-     * @param array     $attributes Additional attributes
+     * @param Name|Expr    $class      Class name
+     * @param string|Error $name       Constant name
+     * @param array        $attributes Additional attributes
      */
     public function __construct($class, $name, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->class = $class;
         $this->name = $name;
     }

lib/PhpParser/Node/Expr/Clone_.php

@@ -16,7 +16,7 @@ class Clone_ extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/Closure.php

@@ -4,8 +4,9 @@ namespace PhpParser\Node\Expr;
 
 use PhpParser\Node;
 use PhpParser\Node\Expr;
+use PhpParser\Node\FunctionLike;
 
-class Closure extends Expr
+class Closure extends Expr implements FunctionLike
 {
     /** @var bool Whether the closure is static */
     public $static;
@@ -15,7 +16,7 @@ class Closure extends Expr
     public $params;
     /** @var ClosureUse[] use()s */
     public $uses;
-    /** @var null|string|Node\Name[] Return type */
+    /** @var null|string|Node\Name|Node\NullableType Return type */
     public $returnType;
     /** @var Node[] Statements */
     public $stmts;
@@ -33,7 +34,7 @@ class Closure extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(array $subNodes = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->static = isset($subNodes['static']) ? $subNodes['static'] : false;
         $this->byRef = isset($subNodes['byRef']) ? $subNodes['byRef'] : false;
         $this->params = isset($subNodes['params']) ? $subNodes['params'] : array();
@@ -45,4 +46,20 @@ class Closure extends Expr
     public function getSubNodeNames() {
         return array('static', 'byRef', 'params', 'uses', 'returnType', 'stmts');
     }
+
+    public function returnsByRef() {
+        return $this->byRef;
+    }
+
+    public function getParams() {
+        return $this->params;
+    }
+
+    public function getReturnType() {
+        return $this->returnType;
+    }
+
+    public function getStmts() {
+        return $this->stmts;
+    }
 }

lib/PhpParser/Node/Expr/ClosureUse.php

@@ -19,7 +19,7 @@ class ClosureUse extends Expr
      * @param array       $attributes Additional attributes
      */
     public function __construct($var, $byRef = false, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->byRef = $byRef;
     }

lib/PhpParser/Node/Expr/ConstFetch.php

@@ -2,8 +2,8 @@
 
 namespace PhpParser\Node\Expr;
 
-use PhpParser\Node\Name;
 use PhpParser\Node\Expr;
+use PhpParser\Node\Name;
 
 class ConstFetch extends Expr
 {
@@ -17,7 +17,7 @@ class ConstFetch extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Name $name, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->name = $name;
     }
 

lib/PhpParser/Node/Expr/Empty_.php

@@ -16,7 +16,7 @@ class Empty_ extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/Error.php

@@ -0,0 +1,27 @@
+<?php
+
+namespace PhpParser\Node\Expr;
+
+use PhpParser\Node\Expr;
+
+/**
+ * Error node used during parsing with error recovery.
+ *
+ * An error node may be placed at a position where an expression is required, but an error occurred.
+ * Error nodes will not be present if the parser is run in throwOnError mode (the default).
+ */
+class Error extends Expr
+{
+    /**
+     * Constructs an error node.
+     *
+     * @param array $attributes Additional attributes
+     */
+    public function __construct(array $attributes = array()) {
+        parent::__construct($attributes);
+    }
+
+    public function getSubNodeNames() {
+        return array();
+    }
+}

lib/PhpParser/Node/Expr/ErrorSuppress.php

@@ -16,7 +16,7 @@ class ErrorSuppress extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/Eval_.php

@@ -16,7 +16,7 @@ class Eval_ extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/Exit_.php

@@ -6,6 +6,10 @@ use PhpParser\Node\Expr;
 
 class Exit_ extends Expr
 {
+    /* For use in "kind" attribute */
+    const KIND_EXIT = 1;
+    const KIND_DIE = 2;
+
     /** @var null|Expr Expression */
     public $expr;
 
@@ -16,7 +20,7 @@ class Exit_ extends Expr
      * @param array                    $attributes Additional attributes
      */
     public function __construct(Expr $expr = null, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/FuncCall.php

@@ -20,7 +20,7 @@ class FuncCall extends Expr
      * @param array                                   $attributes Additional attributes
      */
     public function __construct($name, array $args = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->name = $name;
         $this->args = $args;
     }

lib/PhpParser/Node/Expr/Include_.php

@@ -24,7 +24,7 @@ class Include_ extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, $type, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
         $this->type = $type;
     }

lib/PhpParser/Node/Expr/Instanceof_.php

@@ -2,8 +2,8 @@
 
 namespace PhpParser\Node\Expr;
 
-use PhpParser\Node\Name;
 use PhpParser\Node\Expr;
+use PhpParser\Node\Name;
 
 class Instanceof_ extends Expr
 {
@@ -20,7 +20,7 @@ class Instanceof_ extends Expr
      * @param array     $attributes Additional attributes
      */
     public function __construct(Expr $expr, $class, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
         $this->class = $class;
     }

lib/PhpParser/Node/Expr/Isset_.php

@@ -16,7 +16,7 @@ class Isset_ extends Expr
      * @param array  $attributes Additional attributes
      */
     public function __construct(array $vars, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->vars = $vars;
     }
 

lib/PhpParser/Node/Expr/List_.php

@@ -6,21 +6,21 @@ use PhpParser\Node\Expr;
 
 class List_ extends Expr
 {
-    /** @var Expr[] List of variables to assign to */
-    public $vars;
+    /** @var ArrayItem[] List of items to assign to */
+    public $items;
 
     /**
      * Constructs a list() destructuring node.
      *
-     * @param Expr[] $vars       List of variables to assign to
-     * @param array  $attributes Additional attributes
+     * @param ArrayItem[] $items      List of items to assign to
+     * @param array       $attributes Additional attributes
      */
-    public function __construct(array $vars, array $attributes = array()) {
-        parent::__construct(null, $attributes);
-        $this->vars = $vars;
+    public function __construct(array $items, array $attributes = array()) {
+        parent::__construct($attributes);
+        $this->items = $items;
     }
 
     public function getSubNodeNames() {
-        return array('vars');
+        return array('items');
     }
 }

lib/PhpParser/Node/Expr/MethodCall.php

@@ -23,7 +23,7 @@ class MethodCall extends Expr
      * @param array       $attributes Additional attributes
      */
     public function __construct(Expr $var, $name, array $args = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->name = $name;
         $this->args = $args;

lib/PhpParser/Node/Expr/New_.php

@@ -7,7 +7,7 @@ use PhpParser\Node\Expr;
 
 class New_ extends Expr
 {
-    /** @var Node\Name|Expr Class name */
+    /** @var Node\Name|Expr|Node\Stmt\Class_ Class name */
     public $class;
     /** @var Node\Arg[] Arguments */
     public $args;
@@ -15,12 +15,12 @@ class New_ extends Expr
     /**
      * Constructs a function call node.
      *
-     * @param Node\Name|Expr $class      Class name
-     * @param Node\Arg[]     $args       Arguments
-     * @param array          $attributes Additional attributes
+     * @param Node\Name|Expr|Node\Stmt\Class_ $class      Class name (or class node for anonymous classes)
+     * @param Node\Arg[]                      $args       Arguments
+     * @param array                           $attributes Additional attributes
      */
     public function __construct($class, array $args = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->class = $class;
         $this->args = $args;
     }

lib/PhpParser/Node/Expr/PostDec.php

@@ -16,7 +16,7 @@ class PostDec extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
     }
 

lib/PhpParser/Node/Expr/PostInc.php

@@ -16,7 +16,7 @@ class PostInc extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
     }
 

lib/PhpParser/Node/Expr/PreDec.php

@@ -16,7 +16,7 @@ class PreDec extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
     }
 

lib/PhpParser/Node/Expr/PreInc.php

@@ -16,7 +16,7 @@ class PreInc extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $var, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
     }
 

lib/PhpParser/Node/Expr/Print_.php

@@ -16,7 +16,7 @@ class Print_ extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/PropertyFetch.php

@@ -19,7 +19,7 @@ class PropertyFetch extends Expr
      * @param array       $attributes Additional attributes
      */
     public function __construct(Expr $var, $name, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->var = $var;
         $this->name = $name;
     }

lib/PhpParser/Node/Expr/ShellExec.php

@@ -15,8 +15,8 @@ class ShellExec extends Expr
      * @param array $parts      Encapsed string array
      * @param array $attributes Additional attributes
      */
-    public function __construct($parts, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+    public function __construct(array $parts, array $attributes = array()) {
+        parent::__construct($attributes);
         $this->parts = $parts;
     }
 

lib/PhpParser/Node/Expr/StaticCall.php

@@ -23,7 +23,7 @@ class StaticCall extends Expr
      * @param array          $attributes Additional attributes
      */
     public function __construct($class, $name, array $args = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->class = $class;
         $this->name = $name;
         $this->args = $args;

lib/PhpParser/Node/Expr/StaticPropertyFetch.php

@@ -2,8 +2,8 @@
 
 namespace PhpParser\Node\Expr;
 
-use PhpParser\Node\Name;
 use PhpParser\Node\Expr;
+use PhpParser\Node\Name;
 
 class StaticPropertyFetch extends Expr
 {
@@ -20,7 +20,7 @@ class StaticPropertyFetch extends Expr
      * @param array       $attributes Additional attributes
      */
     public function __construct($class, $name, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->class = $class;
         $this->name = $name;
     }

lib/PhpParser/Node/Expr/Ternary.php

@@ -22,7 +22,7 @@ class Ternary extends Expr
      * @param array                    $attributes Additional attributes
      */
     public function __construct(Expr $cond, $if, Expr $else, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->cond = $cond;
         $this->if = $if;
         $this->else = $else;

lib/PhpParser/Node/Expr/UnaryMinus.php

@@ -16,7 +16,7 @@ class UnaryMinus extends Expr
      * @param array $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/UnaryPlus.php

@@ -16,7 +16,7 @@ class UnaryPlus extends Expr
      * @param array               $attributes Additional attributes
      */
     public function __construct(Expr $expr, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->expr = $expr;
     }
 

lib/PhpParser/Node/Expr/Variable.php

@@ -16,7 +16,7 @@ class Variable extends Expr
      * @param array                      $attributes Additional attributes
      */
     public function __construct($name, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->name = $name;
     }
 

lib/PhpParser/Node/Expr/YieldFrom.php

@@ -0,0 +1,26 @@
+<?php
+
+namespace PhpParser\Node\Expr;
+
+use PhpParser\Node\Expr;
+
+class YieldFrom extends Expr
+{
+    /** @var Expr Expression to yield from */
+    public $expr;
+
+    /**
+     * Constructs an "yield from" node.
+     *
+     * @param Expr  $expr       Expression
+     * @param array $attributes Additional attributes
+     */
+    public function __construct(Expr $expr, array $attributes = array()) {
+        parent::__construct($attributes);
+        $this->expr = $expr;
+    }
+
+    public function getSubNodeNames() {
+        return array('expr');
+    }
+}

lib/PhpParser/Node/Expr/Yield_.php

@@ -19,7 +19,7 @@ class Yield_ extends Expr
      * @param array     $attributes Additional attributes
      */
     public function __construct(Expr $value = null, Expr $key = null, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
         $this->key = $key;
         $this->value = $value;
     }

lib/PhpParser/Node/FunctionLike.php

@@ -0,0 +1,36 @@
+<?php
+
+namespace PhpParser\Node;
+
+use PhpParser\Node;
+
+interface FunctionLike extends Node
+{
+    /**
+     * Whether to return by reference
+     *
+     * @return bool
+     */
+    public function returnsByRef();
+
+    /**
+     * List of parameters
+     *
+     * @return Node\Param[]
+     */
+    public function getParams();
+
+    /**
+     * Get the declared return type or null
+     * 
+     * @return null|string|Node\Name|Node\NullableType
+     */
+    public function getReturnType();
+
+    /**
+     * The function body
+     *
+     * @return Node\Stmt[]
+     */
+    public function getStmts();
+}

lib/PhpParser/Node/Name.php

@@ -6,22 +6,21 @@ use PhpParser\NodeAbstract;
 
 class Name extends NodeAbstract
 {
-    /** @var string[] Parts of the name */
+    /**
+     * @var string[] Parts of the name
+     * @deprecated Avoid directly accessing $parts, use methods instead.
+     */
     public $parts;
 
     /**
      * Constructs a name node.
      *
-     * @param string|array $parts      Parts of the name (or name as string)
-     * @param array        $attributes Additional attributes
+     * @param string|array|self $name       Name as string, part array or Name instance (copy ctor)
+     * @param array             $attributes Additional attributes
      */
-    public function __construct($parts, array $attributes = array()) {
-        if (!is_array($parts)) {
-            $parts = explode('\\', $parts);
-        }
-
-        parent::__construct(null, $attributes);
-        $this->parts = $parts;
+    public function __construct($name, array $attributes = array()) {
+        parent::__construct($attributes);
+        $this->parts = self::prepareName($name);
     }
 
     public function getSubNodeNames() {
@@ -83,14 +82,13 @@ class Name extends NodeAbstract
     }
 
     /**
-     * Returns a string representation of the name by imploding the namespace parts with a separator.
-     *
-     * @param string $separator The separator to use (defaults to the namespace separator \)
+     * Returns a string representation of the name by imploding the namespace parts with the
+     * namespace separator.
      *
      * @return string String representation
      */
-    public function toString($separator = '\\') {
-        return implode($separator, $this->parts);
+    public function toString() {
+        return implode('\\', $this->parts);
     }
 
     /**
@@ -104,48 +102,75 @@ class Name extends NodeAbstract
     }
 
     /**
-     * Sets the whole name.
+     * Gets a slice of a name (similar to array_slice).
      *
-     * @param string|array|self $name The name to set the whole name to
+     * This method returns a new instance of the same type as the original and with the same
+     * attributes.
+     *
+     * If the slice is empty, null is returned. The null value will be correctly handled in
+     * concatenations using concat().
+     *
+     * Offset and length have the same meaning as in array_slice().
+     *
+     * @param int      $offset Offset to start the slice at (may be negative)
+     * @param int|null $length Length of the slice (may be negative)
+     *
+     * @return static|null Sliced name
      */
-    public function set($name) {
-        $this->parts = $this->prepareName($name);
+    public function slice($offset, $length = null) {
+        $numParts = count($this->parts);
+
+        $realOffset = $offset < 0 ? $offset + $numParts : $offset;
+        if ($realOffset < 0 || $realOffset > $numParts) {
+            throw new \OutOfBoundsException(sprintf('Offset %d is out of bounds', $offset));
+        }
+
+        if (null === $length) {
+            $realLength = $numParts - $realOffset;
+        } else {
+            $realLength = $length < 0 ? $length + $numParts - $realOffset : $length;
+            if ($realLength < 0 || $realLength > $numParts) {
+                throw new \OutOfBoundsException(sprintf('Length %d is out of bounds', $length));
+            }
+        }
+
+        if ($realLength === 0) {
+            // Empty slice is represented as null
+            return null;
+        }
+
+        return new static(array_slice($this->parts, $realOffset, $realLength), $this->attributes);
     }
 
     /**
-     * Prepends a name to this name.
+     * Concatenate two names, yielding a new Name instance.
      *
-     * @param string|array|self $name Name to prepend
-     */
-    public function prepend($name) {
-        $this->parts = array_merge($this->prepareName($name), $this->parts);
-    }
-
-    /**
-     * Appends a name to this name.
+     * The type of the generated instance depends on which class this method is called on, for
+     * example Name\FullyQualified::concat() will yield a Name\FullyQualified instance.
      *
-     * @param string|array|self $name Name to append
-     */
-    public function append($name) {
-        $this->parts = array_merge($this->parts, $this->prepareName($name));
-    }
-
-    /**
-     * Sets the first part of the name.
+     * If one of the arguments is null, a new instance of the other name will be returned. If both
+     * arguments are null, null will be returned. As such, writing
+     *     Name::concat($namespace, $shortName)
+     * where $namespace is a Name node or null will work as expected.
      *
-     * @param string|array|self $name The name to set the first part to
-     */
-    public function setFirst($name) {
-        array_splice($this->parts, 0, 1, $this->prepareName($name));
-    }
-
-    /**
-     * Sets the last part of the name.
+     * @param string|array|self|null $name1      The first name
+     * @param string|array|self|null $name2      The second name
+     * @param array                  $attributes Attributes to assign to concatenated name
      *
-     * @param string|array|self $name The name to set the last part to
+     * @return static|null Concatenated name
      */
-    public function setLast($name) {
-        array_splice($this->parts, -1, 1, $this->prepareName($name));
+    public static function concat($name1, $name2, array $attributes = []) {
+        if (null === $name1 && null === $name2) {
+            return null;
+        } elseif (null === $name1) {
+            return new static(self::prepareName($name2), $attributes);
+        } else if (null === $name2) {
+            return new static(self::prepareName($name1), $attributes);
+        } else {
+            return new static(
+                array_merge(self::prepareName($name1), self::prepareName($name2)), $attributes
+            );
+        }
     }
 
     /**
@@ -156,17 +181,17 @@ class Name extends NodeAbstract
      *
      * @return array Prepared name
      */
-    protected function prepareName($name) {
-        if (is_string($name)) {
+    private static function prepareName($name) {
+        if (\is_string($name)) {
             return explode('\\', $name);
-        } elseif (is_array($name)) {
+        } elseif (\is_array($name)) {
             return $name;
         } elseif ($name instanceof self) {
             return $name->parts;
         }
 
         throw new \InvalidArgumentException(
-            'When changing a name you need to pass either a string, an array or a Name node'
+            'Expected string, array of parts or Name instance'
         );
     }
 }

lib/PhpParser/Node/NullableType.php

@@ -0,0 +1,26 @@
+<?php
+
+namespace PhpParser\Node;
+
+use PhpParser\NodeAbstract;
+
+class NullableType extends NodeAbstract
+{
+    /** @var string|Name Type */
+    public $type;
+
+    /**
+     * Constructs a nullable type (wrapping another type).
+     *
+     * @param string|Name $type       Type
+     * @param array       $attributes Additional attributes
+     */
+    public function __construct($type, array $attributes = array()) {
+        parent::__construct($attributes);
+        $this->type = $type;
+    }
+
+    public function getSubNodeNames() {
+        return array('type');
+    }
+}

lib/PhpParser/Node/Param.php

@@ -2,12 +2,11 @@
 
 namespace PhpParser\Node;
 
-use PhpParser\Error;
 use PhpParser\NodeAbstract;
 
 class Param extends NodeAbstract
 {
-    /** @var null|string|Name Typehint */
+    /** @var null|string|Name|NullableType Typehint */
     public $type;
     /** @var bool Whether parameter is passed by reference */
     public $byRef;
@@ -21,24 +20,20 @@ class Param extends NodeAbstract
     /**
      * Constructs a parameter node.
      *
-     * @param string           $name       Name
-     * @param null|Expr        $default    Default value
-     * @param null|string|Name $type       Typehint
-     * @param bool             $byRef      Whether is passed by reference
-     * @param bool             $variadic   Whether this is a variadic argument
-     * @param array            $attributes Additional attributes
+     * @param string                        $name       Name
+     * @param null|Expr                     $default    Default value
+     * @param null|string|Name|NullableType $type       Typehint
+     * @param bool                          $byRef      Whether is passed by reference
+     * @param bool                          $variadic   Whether this is a variadic argument
+     * @param array                         $attributes Additional attributes
      */
-    public function __construct($name, $default = null, $type = null, $byRef = false, $variadic = false, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+    public function __construct($name, Expr $default = null, $type = null, $byRef = false, $variadic = false, array $attributes = array()) {
+        parent::__construct($attributes);
         $this->type = $type;
         $this->byRef = $byRef;
         $this->variadic = $variadic;
         $this->name = $name;
         $this->default = $default;
-
-        if ($variadic && null !== $default) {
-            throw new Error('Variadic parameter cannot have a default value');
-        }
     }
 
     public function getSubNodeNames() {

lib/PhpParser/Node/Scalar/DNumber.php

@@ -15,8 +15,8 @@ class DNumber extends Scalar
      * @param float $value      Value of the number
      * @param array $attributes Additional attributes
      */
-    public function __construct($value = 0.0, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+    public function __construct($value, array $attributes = array()) {
+        parent::__construct($attributes);
         $this->value = $value;
     }
 

lib/PhpParser/Node/Scalar/Encapsed.php

@@ -2,11 +2,12 @@
 
 namespace PhpParser\Node\Scalar;
 
+use PhpParser\Node\Expr;
 use PhpParser\Node\Scalar;
 
 class Encapsed extends Scalar
 {
-    /** @var array Encaps list */
+    /** @var Expr[] list of string parts */
     public $parts;
 
     /**
@@ -15,8 +16,8 @@ class Encapsed extends Scalar
      * @param array $parts      Encaps list
      * @param array $attributes Additional attributes
      */
-    public function __construct(array $parts = array(), array $attributes = array()) {
-        parent::__construct(null, $attributes);
+    public function __construct(array $parts, array $attributes = array()) {
+        parent::__construct($attributes);
         $this->parts = $parts;
     }
 

lib/PhpParser/Node/Scalar/EncapsedStringPart.php

@@ -0,0 +1,26 @@
+<?php
+
+namespace PhpParser\Node\Scalar;
+
+use PhpParser\Node\Scalar;
+
+class EncapsedStringPart extends Scalar
+{
+    /** @var string String value */
+    public $value;
+
+    /**
+     * Constructs a node representing a string part of an encapsed string.
+     *
+     * @param string $value      String value
+     * @param array  $attributes Additional attributes
+     */
+    public function __construct($value, array $attributes = array()) {
+        parent::__construct($attributes);
+        $this->value = $value;
+    }
+
+    public function getSubNodeNames() {
+        return array('value');
+    }
+}

lib/PhpParser/Node/Scalar/LNumber.php

@@ -2,10 +2,17 @@
 
 namespace PhpParser\Node\Scalar;
 
+use PhpParser\Error;
 use PhpParser\Node\Scalar;
 
 class LNumber extends Scalar
 {
+    /* For use in "kind" attribute */
+    const KIND_BIN = 2;
+    const KIND_OCT = 8;
+    const KIND_DEC = 10;
+    const KIND_HEX = 16;
+
     /** @var int Number value */
     public $value;
 
@@ -15,8 +22,8 @@ class LNumber extends Scalar
      * @param int   $value      Value of the number
      * @param array $attributes Additional attributes
      */
-    public function __construct($value = 0, array $attributes = array()) {
-        parent::__construct(null, $attributes);
+    public function __construct($value, array $attributes = array()) {
+        parent::__construct($attributes);
         $this->value = $value;
     }
 
@@ -25,37 +32,36 @@ class LNumber extends Scalar
     }
 
     /**
-     * @internal
+     * Constructs an LNumber node from a string number literal.
      *
-     * Parses an LNUMBER token (dec, hex, oct and bin notations) like PHP would.
+     * @param string $str               String number literal (decimal, octal, hex or binary)
+     * @param array  $attributes        Additional attributes
+     * @param bool   $allowInvalidOctal Whether to allow invalid octal numbers (PHP 5)
      *
-     * @param string $str A string number
-     *
-     * @return int The parsed number
+     * @return LNumber The constructed LNumber, including kind attribute
      */
-    public static function parse($str) {
-        // handle plain 0 specially
-        if ('0' === $str) {
-            return 0;
+    public static function fromString($str, array $attributes = array(), $allowInvalidOctal = false) {
+        if ('0' !== $str[0] || '0' === $str) {
+            $attributes['kind'] = LNumber::KIND_DEC;
+            return new LNumber((int) $str, $attributes);
         }
 
-        // if first char is 0 (and number isn't 0) it's a special syntax
-        if ('0' === $str[0]) {
-            // hex
-            if ('x' === $str[1] || 'X' === $str[1]) {
-                return hexdec($str);
-            }
-
-            // bin
-            if ('b' === $str[1] || 'B' === $str[1]) {
-                return bindec($str);
-            }
-
-            // oct (intval instead of octdec to get proper cutting behavior with malformed numbers)
-            return intval($str, 8);
+        if ('x' === $str[1] || 'X' === $str[1]) {
+            $attributes['kind'] = LNumber::KIND_HEX;
+            return new LNumber(hexdec($str), $attributes);
         }
 
-        // dec
-        return (int) $str;
+        if ('b' === $str[1] || 'B' === $str[1]) {
+            $attributes['kind'] = LNumber::KIND_BIN;
+            return new LNumber(bindec($str), $attributes);
+        }
+
+        if (!$allowInvalidOctal && strpbrk($str, '89')) {
+            throw new Error('Invalid numeric literal', $attributes);
+        }
+
+        // use intval instead of octdec to get proper cutting behavior with malformed numbers
+        $attributes['kind'] = LNumber::KIND_OCT;
+        return new LNumber(intval($str, 8), $attributes);
     }
 }

lib/PhpParser/Node/Scalar/MagicConst.php

@@ -12,7 +12,7 @@ abstract class MagicConst extends Scalar
      * @param array $attributes Additional attributes
      */
     public function __construct(array $attributes = array()) {
-        parent::__construct(null, $attributes);
+        parent::__construct($attributes);
     }
 
     public function getSubNodeNames() {

lib/PhpParser/Node/Scalar/String_.php

@@ -2,10 +2,17 @@
 
 namespace PhpParser\Node\Scalar;
 
+use PhpParser\Error;
 use PhpParser\Node\Scalar;
 
 class String_ extends Scalar
 {
+    /* For use in "kind" attribute */
+    const KIND_SINGLE_QUOTED = 1;
+    const KIND_DOUBLE_QUOTED = 2;
+    const KIND_HEREDOC = 3;
+    const KIND_NOWDOC = 4;
+
     /** @var string String value */
     public $value;
 
@@ -26,8 +33,8 @@ class String_ extends Scalar
      * @param string $value      Value of the string
      * @param array  $attributes Additional attributes
      */
-    public function __construct($value = '', array $attributes = array()) {
-        parent::__construct(null, $attributes);
+    public function __construct($value, array $attributes = array()) {
+        parent::__construct($attributes);
         $this->value = $value;
     }
 
@@ -41,12 +48,13 @@ class String_ extends Scalar
      * Parses a string token.
      *
      * @param string $str String token content
+     * @param bool $parseUnicodeEscape Whether to parse PHP 7 \u escapes
      *
      * @return string The parsed string
      */
-    public static function parse($str) {
+    public static function parse($str, $parseUnicodeEscape = true) {
         $bLength = 0;
-        if ('b' === $str[0]) {
+        if ('b' === $str[0] || 'B' === $str[0]) {
             $bLength = 1;
         }
 
@@ -57,7 +65,9 @@ class String_ extends Scalar
                 substr($str, $bLength + 1, -1)
             );
         } else {
-            return self::parseEscapeSequences(substr($str, $bLength + 1, -1), '"');
+            return self::parseEscapeSequences(
+                substr($str, $bLength + 1, -1), '"', $parseUnicodeEscape
+            );
         }
     }
 
@@ -68,31 +78,54 @@ class String_ extends Scalar
      *
      * @param string      $str   String without quotes
      * @param null|string $quote Quote type
+     * @param bool $parseUnicodeEscape Whether to parse PHP 7 \u escapes
      *
      * @return string String with escape sequences parsed
      */
-    public static function parseEscapeSequences($str, $quote) {
+    public static function parseEscapeSequences($str, $quote, $parseUnicodeEscape = true) {
         if (null !== $quote) {
             $str = str_replace('\\' . $quote, $quote, $str);
         }
 
+        $extra = '';
+        if ($parseUnicodeEscape) {
+            $extra = '|u\{([0-9a-fA-F]+)\}';
+        }
+
         return preg_replace_callback(
-            '~\\\\([\\\\$nrtfve]|[xX][0-9a-fA-F]{1,2}|[0-7]{1,3})~',
-            array(__CLASS__, 'parseCallback'),
+            '~\\\\([\\\\$nrtfve]|[xX][0-9a-fA-F]{1,2}|[0-7]{1,3}' . $extra . ')~',
+            function($matches) {
+                $str = $matches[1];
+
+                if (isset(self::$replacements[$str])) {
+                    return self::$replacements[$str];
+                } elseif ('x' === $str[0] || 'X' === $str[0]) {
+                    return chr(hexdec($str));
+                } elseif ('u' === $str[0]) {
+                    return self::codePointToUtf8(hexdec($matches[2]));
+                } else {
+                    return chr(octdec($str));
+                }
+            },
             $str
         );
     }
 
-    private static function parseCallback($matches) {
-        $str = $matches[1];
-
-        if (isset(self::$replacements[$str])) {
-            return self::$replacements[$str];
-        } elseif ('x' === $str[0] || 'X' === $str[0]) {
-            return chr(hexdec($str));
-        } else {
-            return chr(octdec($str));
+    private static function codePointToUtf8($num) {
+        if ($num <= 0x7F) {
+            return chr($num);
         }
+        if ($num <= 0x7FF) {
+            return chr(($num>>6) + 0xC0) . chr(($num&0x3F) + 0x80);
+        }
+        if ($num <= 0xFFFF) {
+            return chr(($num>>12) + 0xE0) . chr((($num>>6)&0x3F) + 0x80) . chr(($num&0x3F) + 0x80);
+        }
+        if ($num <= 0x1FFFFF) {
+            return chr(($num>>18) + 0xF0) . chr((($num>>12)&0x3F) + 0x80)
+                 . chr((($num>>6)&0x3F) + 0x80) . chr(($num&0x3F) + 0x80);
+        }
+        throw new Error('Invalid UTF-8 codepoint escape sequence: Codepoint too large');
     }
 
     /**
@@ -102,18 +135,19 @@ class String_ extends Scalar
      *
      * @param string $startToken Doc string start token content (<<<SMTHG)
      * @param string $str        String token content
+     * @param bool $parseUnicodeEscape Whether to parse PHP 7 \u escapes
      *
      * @return string Parsed string
      */
-    public static function parseDocString($startToken, $str) {
+    public static function parseDocString($startToken, $str, $parseUnicodeEscape = true) {
         // strip last newline (thanks tokenizer for sticking it into the string!)
-        $str = preg_replace('~(\r\n|\n|\r)$~', '', $str);
+        $str = preg_replace('~(\r\n|\n|\r)\z~', '', $str);
 
         // nowdoc string
         if (false !== strpos($startToken, '\'')) {
             return $str;
         }
 
-        return self::parseEscapeSequences($str, null);
+        return self::parseEscapeSequences($str, null, $parseUnicodeEscape);
     }
 }