extract_info now expects ie.extract to return a list in the format proposed in issue 608.

Each element should have a '_type' key specifying if it's a video, an url or a playlist.
`extract_info` will process each element to get the full info
This commit is contained in:
Jaime Marquínez Ferrándiz 2013-03-05 20:14:32 +01:00
parent f6e6da9525
commit 6ac7f082c4

View File

@ -410,12 +410,9 @@ def extract_info(self, url):
# Extract information from URL and process it
try:
videos = ie.extract(url)
for video in videos or []:
if not 'extractor' in video:
#The extractor has already been set somewher else
video['extractor'] = ie.IE_NAME
return videos
ie_results = ie.extract(url)
results = self.process_ie_results(ie_results, ie)
return results
except ExtractorError as de: # An error we somewhat expected
self.trouble(u'ERROR: ' + compat_str(de), de.format_traceback())
break
@ -435,6 +432,29 @@ def extract_info_iterable(self, urls):
for url in urls:
results.extend(self.extract_info(url))
return results
def process_ie_results(self, ie_results, ie):
"""
Take the results of the ie and return a list of videos.
For url elements it will seartch the suitable ie and get the videos
For playlist elements it will process each of the elements of the 'entries' key
"""
results = []
for result in ie_results or []:
result_type = result.get('_type', 'video') #If not given we suppose it's a video, support the dafault old system
if result_type == 'video':
if not 'extractor' in result:
#The extractor has already been set somewhere else
result['extractor'] = ie.IE_NAME
results.append(result)
elif result_type == 'url':
#We get the videos pointed by the url
results.extend(self.extract_info(result['url']))
elif result_type == 'playlist':
#We process each entry in the playlist
entries_result = self.process_ie_results(result['entries'], ie)
results.extend(entries_result)
return results
def process_info(self, info_dict):
"""Process a single dictionary returned by an InfoExtractor."""