Improve Minify_Lines algorithm and add tests

Fixes #531
Fixes #534

tests/MinifyLinesTest.php

@@ -4,8 +4,6 @@ class MinifyLinesTest extends TestCase
 {
     public function test_lines()
     {
-        $exp = file_get_contents(self::$test_files . "/minify/lines_output.js");
-
         $env = new Minify_Env(array(
             'server' => array(
                 'DOCUMENT_ROOT' => dirname(__DIR__),
@@ -15,15 +13,27 @@ class MinifyLinesTest extends TestCase
         $controller = new Minify_Controller_Files($env, $sourceFactory);
         $minify = new Minify(new Minify_Cache_Null());
 
-        $ret = $minify->serve($controller, array(
-            'debug' => true
-            ,'quiet' => true
-            ,'encodeOutput' => false
-            ,'files' => array(
-                self::$test_files . "/js/before.js"
-            )
-        ));
+        $files = glob(self::$test_files . "/lines/*.in.js");
 
-        $this->assertEquals($exp, $ret['content']);
+        // uncomment to debug one
+        //$files = array(self::$test_files . "/lines/basic.in.js");
+
+        foreach ($files as $file) {
+            $ret = $minify->serve($controller, array(
+                'debug' => true,
+                'quiet' => true,
+                'encodeOutput' => false,
+                'files' => array($file),
+            ));
+
+            $outFile = str_replace('.in.js', '.out.js', $file);
+
+            $exp = file_get_contents($outFile);
+
+            // uncomment to set up expected output
+            //file_put_contents($outFile, $ret['content']);
+
+            $this->assertEquals($exp, $ret['content'], "Did not match: " . basename($outFile));
+        }
     }
 }

tests/_test_files/lines/basic.in.js

@@ -0,0 +1,66 @@
+/*! is.js
+
+ (c) 2001 Douglas Crockford
+ 2001 June 3
+*/
+
+// is
+
+// The -is- object is used to identify the browser.  Every browser edition
+// identifies itself, but there is no standard way of doing it, and some of
+// the identification is deceptive. This is because the authors of web
+// browsers are liars. For example, Microsoft's IE browsers claim to be
+// Mozilla 4. Netscape 6 claims to be version 5.
+
+var is = {
+    ie:      navigator.appName == 'Microsoft Internet Explorer',
+    java:    navigator.javaEnabled(),
+    ns:      navigator.appName == 'Netscape',
+    ua:      navigator.userAgent.toLowerCase(),
+    version: parseFloat(navigator.appVersion.substr(21)) ||
+             parseFloat(navigator.appVersion),
+    win:     navigator.platform == 'Win32'
+}
+/*!*
+ * preserve this comment, too
+ */
+is.mac = is.ua.indexOf('mac') >= 0;
+if (is.ua.indexOf('opera') >= 0) {
+    is.ie = is.ns = false;
+    is.opera = true;
+}
+if (is.ua.indexOf('gecko') >= 0) {
+    is.ie = is.ns = false;
+    is.gecko = true;
+}
+
+/*@cc_on
+   /*@if (@_win32)
+    if (is.ie && is.win)
+        document.write("PASS: IE/win honored conditional comment.<br>");
+   @else @*/
+    if (is.ie && is.win)
+        document.write("FAIL: IE/win did not honor multi-line conditional comment.<br>");
+    else
+        document.write("PASS: Non-IE/win browser ignores multi-line conditional comment.<br>");
+   /*@end
+@*/
+
+var recognizesCondComm = true;
+//@cc_on/*
+recognizesCondComm = false;
+//@cc_on*/
+
+if ((is.ie && is.win) == recognizesCondComm)
+    document.write("PASS: IE/win honored single-line conditional comment.<br>");
+else
+    document.write("FAIL: Non-IE/win browser did not ignore single-line conditional comment.<br>");
+
+// hello
+//@cc_on/*
+// world
+//@cc_on*/
+//@cc_on/*
+'hello';
+/*!* preserved */
+/*!* preserved */

tests/_test_files/lines/basic.out.js

@@ -1,20 +1,20 @@
 
-/* before.js */
+/* basic.in.js */
 
 /* 1  */ /*! is.js
-/* 2  *| 
+/* 2  *|
 /* 3  *|  (c) 2001 Douglas Crockford
 /* 4  *|  2001 June 3
 /* 5  *| */
-/* 6  */ 
+/* 6  */
 /* 7  */ // is
-/* 8  */ 
+/* 8  */
 /* 9  */ // The -is- object is used to identify the browser.  Every browser edition
 /* 10 */ // identifies itself, but there is no standard way of doing it, and some of
 /* 11 */ // the identification is deceptive. This is because the authors of web
 /* 12 */ // browsers are liars. For example, Microsoft's IE browsers claim to be
 /* 13 */ // Mozilla 4. Netscape 6 claims to be version 5.
-/* 14 */ 
+/* 14 */
 /* 15 */ var is = {
 /* 16 */     ie:      navigator.appName == 'Microsoft Internet Explorer',
 /* 17 */     java:    navigator.javaEnabled(),
@@ -36,7 +36,7 @@
 /* 33 */     is.ie = is.ns = false;
 /* 34 */     is.gecko = true;
 /* 35 */ }
-/* 36 */ 
+/* 36 */
 /* 37 */ /*@cc_on
 /* 38 *|    /*@if (@_win32)
 /* 39 *|     if (is.ie && is.win)
@@ -44,24 +44,24 @@
 /* 41 *|    @else @*/
 /* 42 */     if (is.ie && is.win)
 /* 43 */         document.write("FAIL: IE/win did not honor multi-line conditional comment.<br>");
-/* 44 */     else 
+/* 44 */     else
 /* 45 */         document.write("PASS: Non-IE/win browser ignores multi-line conditional comment.<br>");
 /* 46 */    /*@end
 /* 47 *| @*/
-/* 48 */ 
+/* 48 */
 /* 49 */ var recognizesCondComm = true;
 /* 50 */ //@cc_on/*
 
-/* before.js */
+/* basic.in.js */
 
 /* 51 */ recognizesCondComm = false;
 /* 52 */ //@cc_on*/
-/* 53 */ 
+/* 53 */
 /* 54 */ if ((is.ie && is.win) == recognizesCondComm)
 /* 55 */     document.write("PASS: IE/win honored single-line conditional comment.<br>");
-/* 56 */ else 
+/* 56 */ else
 /* 57 */     document.write("FAIL: Non-IE/win browser did not ignore single-line conditional comment.<br>");
-/* 58 */ 
+/* 58 */
 /* 59 */ // hello
 /* 60 */ //@cc_on/*
 /* 61 */ // world

tests/_test_files/lines/misc.in.js

@@ -1,10 +1,14 @@
-// sections from Prototype 1.6.1
-var xpath = ".//*[local-name()='ul' or local-name()='UL']" +
-          "//*[local-name()='li' or local-name()='LI']";
-this.matcher = ['.//*'];
-xpath = {
-    descendant:   "//*",
-    child:        "/*",
-    f: 0
-};
-document._getElementsByXPath('.//*' + cond, element);
+// sections from Prototype 1.6.1
+var xpath = ".//*[local-name()='ul' or local-name()='UL']" +
+          "//*[local-name()='li' or local-name()='LI']";
+this.matcher = ['.//*'];
+xpath = {
+    descendant:   "//*",
+    child:        "/*",
+    f: 0
+};
+document._getElementsByXPath('.//*' + cond, element);
+
+// from angular 1.4.8
+var URL_REGEXP = /^[a-z][a-z\d.+-]*:\/*(?:[^:@]+(?::[^@]+)?@)?(?:[^\s:/?#]+|\[[a-f\d:]+\])(?::\d+)?(?:\/[^?#]*)?(?:\?[^#]*)?(?:#.*)?$/i;
+

tests/_test_files/lines/misc.out.js

@@ -0,0 +1,18 @@
+
+/* misc.in.js */
+
+/* 1  */ // sections from Prototype 1.6.1
+/* 2  */ var xpath = ".//*[local-name()='ul' or local-name()='UL']" +
+/* 3  */           "//*[local-name()='li' or local-name()='LI']";
+/* 4  */ this.matcher = ['.//*'];
+/* 5  */ xpath = {
+/* 6  */     descendant:   "//*",
+/* 7  */     child:        "/*",
+/* 8  */     f: 0
+/* 9  */ };
+/* 10 */ document._getElementsByXPath('.//*' + cond, element);
+/* 11 */
+/* 12 */ // from angular 1.4.8
+/* 13 */ var URL_REGEXP = /^[a-z][a-z\d.+-]*:\/*(?:[^:@]+(?::[^@]+)?@)?(?:[^\s:/?#]+|\[[a-f\d:]+\])(?::\d+)?(?:\/[^?#]*)?(?:\?[^#]*)?(?:#.*)?$/i;
+/* 14 */
+/* 15 */

tests/_test_files/lines/url.in.js

@@ -0,0 +1,13 @@
+foo; /* http://example.com */
+bar;
+
+foo; /*
+    http://example.com */
+bar;
+
+foo = "http://example.com"; /* hello */
+bar;
+
+foo = "http://example.com"; /*
+hello */
+bar;

tests/_test_files/lines/url.out.js

@@ -0,0 +1,17 @@
+
+/* url.in.js */
+
+/* 1  */ foo; /* http://example.com */
+/* 2  */ bar;
+/* 3  */
+/* 4  */ foo; /*
+/* 5  *|     http://example.com */
+/* 6  */ bar;
+/* 7  */
+/* 8  */ foo = "http://example.com"; /* hello */
+/* 9  */ bar;
+/* 10 */
+/* 11 */ foo = "http://example.com"; /*
+/* 12 *| hello */
+/* 13 */ bar;
+/* 14 */