mirror/archivebox
1
0
Fork 0
mirror of https://github.com/pirate/ArchiveBox.git synced 2025-08-26 07:44:41 +02:00
Code Issues Releases Wiki Activity

move parsers to global instead of func

Browse Source
This commit is contained in:
Nick Sweeting
2019-02-27 04:49:02 -05:00
parent af7e8df0eb
commit b2c22a73e6
1 changed files with 14 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
archivebox/parse.py
View File

@@ -30,20 +30,6 @@ from util import (
) )
def get_parsers(file):
"""return all parsers that work on a given file, defaults to all of them"""
return OrderedDict([
('Pocket HTML', parse_pocket_html_export),
('Pinboard JSON', parse_pinboard_json_export),
('Netscape HTML', parse_netscape_html_export),
('RSS', parse_rss_export),
('Pinboard RSS', parse_pinboard_rss_export),
('Shaarli RSS', parse_shaarli_rss_export),
('Medium RSS', parse_medium_rss_export),
('Plain Text', parse_plain_text_export),
])
def parse_links(path): def parse_links(path):
"""parse a list of links dictionaries from a bookmark export file""" """parse a list of links dictionaries from a bookmark export file"""
@@ -55,15 +41,14 @@ def parse_links(path):
**ANSI, **ANSI,
)) ))
for parser_name, parser_func in get_parsers(file).items(): for parser_name, parser_func in PARSERS.items():
# otherwise try all parsers until one works
try: try:
links += list(parser_func(file)) links += list(parser_func(file))
if links: if links:
break break
except Exception as err: except Exception as err:
# we try each parser one by one, each parser will throw exeption an exception if unsupported # we try each parser one by one, wong parsers will throw exeptions
# so we accept the first one that # if unsupported and we accept the first one that passes
# uncomment the following line to see why the parser was unsupported for each attempted format # uncomment the following line to see why the parser was unsupported for each attempted format
# print('[!] Parser {} failed: {} {}'.format(parser_name, err.__class__.__name__, err)) # print('[!] Parser {} failed: {} {}'.format(parser_name, err.__class__.__name__, err))
pass pass
@@ -308,3 +293,14 @@ def parse_plain_text_export(text_file):
info['type'] = get_link_type(info) info['type'] = get_link_type(info)
yield info yield info
PARSERS = OrderedDict([
('Pocket HTML', parse_pocket_html_export),
('Pinboard JSON', parse_pinboard_json_export),
('Netscape HTML', parse_netscape_html_export),
('RSS', parse_rss_export),
('Pinboard RSS', parse_pinboard_rss_export),
('Shaarli RSS', parse_shaarli_rss_export),
('Medium RSS', parse_medium_rss_export),
('Plain Text', parse_plain_text_export),
])