Some http addresses cannot be opened

[Carl Eugen Hoyos]

(vlc ticket 8155)
The following two addresses play fine with firefox:
​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres
​http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa
FFmpeg only decodes the second:

$ ffmpeg -loglevel 99 -i "http://translate.google.com/translate_tts?tl=fr&q=l%E8vres"
ffmpeg version N-49844-g1ac0fa5 Copyright (c) 2000-2013 the FFmpeg developers
  built on Feb 12 2013 23:51:40 with gcc 4.7 (SUSE Linux)
  configuration: --enable-gpl --disable-indev=jack
  libavutil      52. 17.101 / 52. 17.101
  libavcodec     54. 91.103 / 54. 91.103
  libavformat    54. 63.100 / 54. 63.100
  libavdevice    54.  3.103 / 54.  3.103
  libavfilter     3. 37.101 /  3. 37.101
  libswscale      2.  2.100 /  2.  2.100
  libswresample   0. 17.102 /  0. 17.102
  libpostproc    52.  2.100 / 52.  2.100
Splitting the commandline.
Reading option '-loglevel' ... matched as option 'loglevel' (set libav* logging level) with argument '99'.
Reading option '-i' ... matched as input file with argument 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vres'.
Finished splitting the commandline.
Parsing a group of options: global .
Applying option loglevel (set libav* logging level) with argument 99.
Successfully parsed a group of options.
Parsing a group of options: input file http://translate.google.com/translate_tts?tl=fr&q=l%E8vres.
Successfully parsed a group of options.
Opening an input file: http://translate.google.com/translate_tts?tl=fr&q=l%E8vres.
[AVIOContext @ 0x2113460] Statistics: 0 bytes read, 0 seeks
http://translate.google.com/translate_tts?tl=fr&q=l%E8vres: Invalid data found when processing input

$ ffmpeg -loglevel 99 -i "http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa"
ffmpeg version N-49844-g1ac0fa5 Copyright (c) 2000-2013 the FFmpeg developers
  built on Feb 12 2013 23:51:40 with gcc 4.7 (SUSE Linux)
  configuration: --enable-gpl --disable-indev=jack
  libavutil      52. 17.101 / 52. 17.101
  libavcodec     54. 91.103 / 54. 91.103
  libavformat    54. 63.100 / 54. 63.100
  libavdevice    54.  3.103 / 54.  3.103
  libavfilter     3. 37.101 /  3. 37.101
  libswscale      2.  2.100 /  2.  2.100
  libswresample   0. 17.102 /  0. 17.102
  libpostproc    52.  2.100 / 52.  2.100
Splitting the commandline.
Reading option '-loglevel' ... matched as option 'loglevel' (set libav* logging level) with argument '99'.
Reading option '-i' ... matched as input file with argument 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa'.
Finished splitting the commandline.
Parsing a group of options: global .
Applying option loglevel (set libav* logging level) with argument 99.
Successfully parsed a group of options.
Parsing a group of options: input file http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa.
Successfully parsed a group of options.
Opening an input file: http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa.
[mp3 @ 0x1ba1520] Format mp3 probed with size=2048 and score=51
[mp3 @ 0x1ba1520] File position before avformat_find_stream_info() is 0
[mp3 @ 0x1ba1520] Estimating duration from bitrate, this may be inaccurate
[mp3 @ 0x1ba1520] File position after avformat_find_stream_info() is 4752
Input #0, mp3, from 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa':
  Duration: 00:00:01.19, start: 0.000000, bitrate: 32 kb/s
    Stream #0:0, 33, 1/14112000: Audio: mp3, 16000 Hz, mono, s16p, 32 kb/s
Successfully opened the file.
At least one output file must be specified
Statistics: 4752 bytes read, 0 seeks

[Michael Niedermayer]

patch posted to ML

wget "​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres"
--2013-04-03 01:00:58-- ​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres
Resolving translate.google.com (translate.google.com)... 173.194.70.139, 173.194.70.102, 173.194.70.100, ...
Connecting to translate.google.com (translate.google.com)|173.194.70.139|:80... connected.
HTTP request sent, awaiting response... 403 Forbidden
2013-04-03 01:00:58 ERROR 403: Forbidden.

wget -U Random "​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres"
--2013-04-03 01:01:56-- ​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres
Resolving translate.google.com (translate.google.com)... 173.194.70.139, 173.194.70.102, 173.194.70.100, ...
Connecting to translate.google.com (translate.google.com)|173.194.70.139|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 0 [audio/mpeg]
Saving to: `translate_tts?tl=fr&q=l\350vres.2'

2013-04-03 01:01:56 (0.00 B/s) - `translate_tts?tl=fr&q=l\350vres.2' saved [0/0]

wget -U 'Mozilla/5.0 (X11; Ubuntu; Linux i686; rv:15.0) Gecko/20100101 Firefox/15.0.1' "​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres"
--2013-04-03 01:02:48-- ​http://translate.google.com/translate_tts?tl=fr&q=l%E8vres
Resolving translate.google.com (translate.google.com)... 173.194.70.139, 173.194.70.102, 173.194.70.100, ...
Connecting to translate.google.com (translate.google.com)|173.194.70.139|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 3168 (3.1K) [audio/mpeg]
Saving to: `translate_tts?tl=fr&q=l\350vres.3'

2013-04-03 01:02:48 (1.68 MB/s) - `translate_tts?tl=fr&q=l\350vres.3' saved [3168/3168]

[Clément Bœsch]

Fixed in 1fabd950.

Note that something is very funny:

This doesn't work:

ffplay -user-agent 'Foobar/1.0' 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vres'

This on the other hand does:

ffplay -user-agent 'Mozilla/5.0' 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vres'

Now the funny part:

This does actually "work":

ffplay -user-agent 'Foobar/1.0' 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa'

But sounds different from:

ffplay -user-agent 'Mozilla/5.0' 'http://translate.google.com/translate_tts?tl=fr&q=l%E8vresa'

Note that the difference between "levres" and "levresa" is that the later is not an existing word.

[Clément Bœsch]

Re-opened since de12b454f21675e18127cd8b549c53deec7cd048

[Clément Bœsch]

I believe you need to make FFmpeg behave as a browser for this, so -user-agent 'Mozilla/5.0' is the correct way to deal with the problem. Feel free to re-open if you disagree.