midgard/youtube-dl
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

[utils] Don't attempt to coerce JS strings to numbers in js_to_json (#26851)

Browse source 
The current logic in `js_to_json` tries to rewrite octal/hex numbers to
decimal. However, when the logic actually happens the `"` or `'` have
already been trimmed off. This causes what were originally strings, that
happen to look like octal/hex numbers, to get rewritten to decimal and
returned as a number rather than a string.

In practive something like:

```js
{
  "0x40": "foo",
  "040": "bar",
}
```

would get rewritten as:

```json
{
  64: "foo",
  32: "bar
}
```

This is problematic since this isn't valid JSON as you cannot have
non-string keys.
This commit is contained in:
Kevin O'Connor 2020-10-17 13:10:41 -04:00 committed by GitHub
parent 605535776a
commit 4eda10499e
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 12 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
test/test_utils.py
View file

@ -994,6 +994,12 @@ class TestUtil(unittest.TestCase):
on = js_to_json('{42:4.2e1}') on = js_to_json('{42:4.2e1}')
self.assertEqual(json.loads(on), {'42': 42.0}) self.assertEqual(json.loads(on), {'42': 42.0})
on = js_to_json('{ "0x40": "0x40" }')
self.assertEqual(json.loads(on), {'0x40': '0x40'})
on = js_to_json('{ "040": "040" }')
self.assertEqual(json.loads(on), {'040': '040'})
def test_js_to_json_malformed(self): def test_js_to_json_malformed(self):
self.assertEqual(js_to_json('42a1'), '42"a1"') self.assertEqual(js_to_json('42a1'), '42"a1"')
self.assertEqual(js_to_json('42a-1'), '42"a"-1') self.assertEqual(js_to_json('42a-1'), '42"a"-1')

12
youtube_dl/utils.py
View file

@ -4088,12 +4088,12 @@ def js_to_json(code):
'\\\n': '', '\\\n': '',
'\\x': '\\u00', '\\x': '\\u00',
}.get(m.group(0), m.group(0)), v[1:-1]) }.get(m.group(0), m.group(0)), v[1:-1])
else:
for regex, base in INTEGER_TABLE: for regex, base in INTEGER_TABLE:
im = re.match(regex, v) im = re.match(regex, v)
if im: if im:
i = int(im.group(1), base) i = int(im.group(1), base)
return '"%d":' % i if v.endswith(':') else '%d' % i return '"%d":' % i if v.endswith(':') else '%d' % i
return '"%s"' % v return '"%s"' % v