From d305dd73a3d6927f0a2c63d08662a183fa173833 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Wed, 18 Feb 2015 23:59:50 +0100
Subject: [PATCH] [utils] Fix js_to_json
Previously, the runtime could be atrocious for longer inputs.
---
test/test_utils.py | 4 ++++
youtube_dl/utils.py | 4 ++--
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/test/test_utils.py b/test/test_utils.py
index 1c29d0889..c7373af1e 100644
--- a/test/test_utils.py
+++ b/test/test_utils.py
@@ -370,6 +370,10 @@ class TestUtil(unittest.TestCase):
"playlist":[{"controls":{"all":null}}]
}''')
+ inp = '"SAND Number: SAND 2013-7800P\\nPresenter: Tom Russo\\nHabanero Software Training - Xyce Software\\nXyce, Sandia\\u0027s"'
+ json_code = js_to_json(inp)
+ self.assertEqual(json.loads(json_code), json.loads(inp))
+
def test_js_to_json_edgecases(self):
on = js_to_json("{abc_def:'1\\'\\\\2\\\\\\'3\"4'}")
self.assertEqual(json.loads(on), {"abc_def": "1'\\2\\'3\"4"})
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 3eb6bc6d4..4358137a0 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1560,8 +1560,8 @@ def js_to_json(code):
return '"%s"' % v
res = re.sub(r'''(?x)
- "(?:[^"\\]*(?:\\\\|\\")?)*"|
- '(?:[^'\\]*(?:\\\\|\\')?)*'|
+ "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
+ '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
[a-zA-Z_][.a-zA-Z_0-9]*
''', fix_kv, code)
res = re.sub(r',(\s*\])', lambda m: m.group(1), res)