From db13c16ef8968613680e2bbc85f373c3e74faf98 Mon Sep 17 00:00:00 2001 From: =?utf8?q?Sergey=20M=E2=80=A4?= Date: Wed, 15 Feb 2017 23:12:10 +0700 Subject: [PATCH] [utils] Add support for quoted string literals in --match-filter (closes #8050, closes #12142, closes #12144) --- test/test_YoutubeDL.py | 24 ++++++++++++++++++++++++ youtube_dl/utils.py | 9 +++++++-- 2 files changed, 31 insertions(+), 2 deletions(-) diff --git a/test/test_YoutubeDL.py b/test/test_YoutubeDL.py index 8bf00bea9..d07c35be8 100644 --- a/test/test_YoutubeDL.py +++ b/test/test_YoutubeDL.py @@ -1,4 +1,5 @@ #!/usr/bin/env python +# coding: utf-8 from __future__ import unicode_literals @@ -606,6 +607,8 @@ class TestYoutubeDL(unittest.TestCase): 'duration': 30, 'filesize': 10 * 1024, 'playlist_id': '42', + 'uploader': "變態妍字幕版 太妍 тест", + 'creator': "тест ' 123 ' тест--", } second = { 'id': '2', @@ -616,6 +619,7 @@ class TestYoutubeDL(unittest.TestCase): 'description': 'foo', 'filesize': 5 * 1024, 'playlist_id': '43', + 'uploader': "тест 123", } videos = [first, second] @@ -656,6 +660,26 @@ class TestYoutubeDL(unittest.TestCase): res = get_videos(f) self.assertEqual(res, ['1']) + f = match_filter_func('uploader = "變態妍字幕版 太妍 тест"') + res = get_videos(f) + self.assertEqual(res, ['1']) + + f = match_filter_func('uploader != "變態妍字幕版 太妍 тест"') + res = get_videos(f) + self.assertEqual(res, ['2']) + + f = match_filter_func('creator = "тест \' 123 \' тест--"') + res = get_videos(f) + self.assertEqual(res, ['1']) + + f = match_filter_func("creator = 'тест \\' 123 \\' тест--'") + res = get_videos(f) + self.assertEqual(res, ['1']) + + f = match_filter_func(r"creator = 'тест \' 123 \' тест--' & duration > 30") + res = get_videos(f) + self.assertEqual(res, []) + def test_playlist_items_selection(self): entries = [{ 'id': compat_str(i), diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 1279a9042..07c07be6f 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -2383,6 +2383,7 @@ def _match_one(filter_part, dct): \s*(?P%s)(?P\s*\?)?\s* (?: (?P[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)| + (?P["\'])(?P(?:\\.|(?!(?P=quote)|\\).)+?)(?P=quote)| (?P(?![0-9.])[a-z0-9A-Z]*) ) \s*$ @@ -2391,7 +2392,8 @@ def _match_one(filter_part, dct): if m: op = COMPARISON_OPERATORS[m.group('op')] actual_value = dct.get(m.group('key')) - if (m.group('strval') is not None or + if (m.group('quotedstrval') is not None or + m.group('strval') is not None or # If the original field is a string and matching comparisonvalue is # a number we should respect the origin of the original field # and process comparison value as a string (see @@ -2401,7 +2403,10 @@ def _match_one(filter_part, dct): if m.group('op') not in ('=', '!='): raise ValueError( 'Operator %s does not support string values!' % m.group('op')) - comparison_value = m.group('strval') or m.group('intval') + comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval') + quote = m.group('quote') + if quote is not None: + comparison_value = comparison_value.replace(r'\%s' % quote, quote) else: try: comparison_value = int(m.group('intval')) -- 2.22.2