def got_results (self, result, callback, *data):
if result is None:
callback (None, *data)
return
results = re.sub('\n', '', re.sub('\r', '', result))
if re.search('(<tr><td bgcolor="#BBBBBB".*)(More Songs >)', results) is not None:
body = re.split('(<tr><td bgcolor="#BBBBBB".*)(More Songs >)', results)[1]
entries = re.split('<tr><td bgcolor="#BBBBBB"', body)
entries.pop(0)
print "found %d entries; looking for [%s,%s]" % (len(entries), self.title, self.artist)
for entry in entries:
url = re.split('(\/display[^"]*)', entry)[1]
artist = re.split('(Artist:.*html">)([^<]*)', entry)[2].strip()
title = re.split('(\/display[^>]*)([^<]*)', entry)[2][1:].strip()
if self.artist != "":
artist_str = string_match(self.artist, artist)
else:
artist_str = artist_match + 0.1
title_str = string_match(self.title, title)
print "checking [%s,%s]: match strengths [%f,%f]" % (title.strip(), artist.strip(), title_str, artist_str)
if title_str > title_match and artist_str > artist_match:
loader = rb.Loader()
loader.get_url ('http://display.lyrics.astraweb.com' + url, self.parse_lyrics, callback, *data)
return
callback (None, *data)
return
def search_artist(self, artist_page, callback, *data):
"""Search for the link to the page of artist in artists_page
"""
if artist_page is None:
callback (None, *data)
return
link_section = re.split ('<SCRIPT LANGUAGE="javascript" src="tban2.js"></SCRIPT>',
artist_page, 1)[1]
pattern_link = '<a href="'
pattern_artist = '([^"]*)">*([^<]*)</a><br><br>'
links = re.split (pattern_link, link_section.lower())
links.pop(0)
best_match = ()
for line in links:
artist = re.findall(pattern_artist, line)
if len(artist) == 0:
continue
artist_link, artist_name = artist[0]
artist_url = 'http://www.darklyrics.com/%s' % (artist_link)
if artist_link[:5] == 'http:':
continue
artist_name = artist_name.strip()
smvalue = string_match (artist_name, self.artist_ascii)
if smvalue > min_artist_match:
best_match = (smvalue, artist_url, artist_name)
if not best_match:
# Lyrics are located in external site
callback (None, *data)
return
loader = rb.Loader ()
self.artist = best_match[2]
loader.get_url (best_match[1], self.search_song, callback, *data)
def got_lyrics(self, lyrics, callback, *data):
if lyrics is None:
callback(None, *data)
return
element = cElementTree.fromstring(lyrics)
if element.find("response").attrib['code'] is not '0':
print "got failed response:" + lyrics
callback(None, *data)
return
match = None
matches = element.find("searchResults").findall("result")
print "got %d result(s)" % (len(matches))
for m in matches:
matchtitle = m.findtext("title")
matchartist = m.findtext("artist/name")
# if we don't know the artist, then anyone will do
if self.artist != "":
artist_str = string_match(self.artist, matchartist)
else:
artist_str = artist_match + 0.1
title_str = string_match(self.title, matchtitle)
if artist_str > artist_match and title_str > title_match:
print "found acceptable match, artist: %s (%f), title: %s (%f)" % (
matchartist, artist_str, matchtitle, title_str)
match = m
break
else:
print "skipping match, artist: %s (%f), title: %s (%f)" % (
matchartist, artist_str, matchtitle, title_str)
if match is not None:
hid = m.attrib['hid'].encode('utf-8')
lurl = "http://api.leoslyrics.com/api_lyrics.php?auth=Rhythmbox&hid=%s" % (
urllib.quote(hid))
loader = rb.Loader()
loader.get_url(lurl, self.parse_lyrics, callback, *data)
else:
print "no acceptable match found"
callback(None, *data)
def got_lyrics (self, lyrics, callback, *data):
if lyrics is None:
callback (None, *data)
return
element = cElementTree.fromstring(lyrics)
if element.find("response").attrib['code'] is not '0':
print "got failed response:" + lyrics
callback (None, *data)
return
match = None
matches = element.find("searchResults").findall("result")
print "got %d result(s)" % (len(matches))
for m in matches:
matchtitle = m.findtext("title")
matchartist = m.findtext("artist/name")
# if we don't know the artist, then anyone will do
if self.artist != "":
artist_str = string_match(self.artist, matchartist)
else:
artist_str = artist_match + 0.1
title_str = string_match(self.title, matchtitle)
if artist_str > artist_match and title_str > title_match:
print "found acceptable match, artist: %s (%f), title: %s (%f)" % (matchartist, artist_str, matchtitle, title_str)
match = m
break
else:
print "skipping match, artist: %s (%f), title: %s (%f)" % (matchartist, artist_str, matchtitle, title_str)
if match is not None:
hid = m.attrib['hid'].encode('utf-8')
lurl = "http://api.leoslyrics.com/api_lyrics.php?auth=Rhythmbox&hid=%s" % (urllib.quote(hid))
loader = rb.Loader()
loader.get_url (lurl, self.parse_lyrics, callback, *data)
else:
print "no acceptable match found"
callback (None, *data)
def search_song (self, songlist, callback, *data):
"""If artist's page is found, search_song looks for the song.
The artist page contains a list of all the albums and
links to the songs lyrics from this.
"""
if songlist is None:
callback (None, *data)
return
# Search for all the <a>
# filter for those that has the artist name string_match
# and for those which its content is artist string_match
# Sort by values given from string_match
# and get the best
link_section = re.split('LYRICS<BR></FONT>', songlist)[1]
link_section = link_section.lower()
pattern_song = '<a href="../lyrics/(.*)/(.*).html#([^"]+)" target="_blank"><FONT COLOR="#CCCCCC">(.*)</FONT></a><br>'
matches = re.findall (pattern_song.lower(), link_section)
best_match = ""
for line in matches:
artist, album, number, title = line
smvalue = string_match (title.lower().replace(' ', '' ),
self.title.lower().replace(' ', ''))
if smvalue > min_song_match:
best_match = self.SongFound(smvalue,
title,
number,
album,
artist)
if not best_match:
callback (None, *data)
return
loader = rb.Loader ()
url = 'http://www.darklyrics.com/lyrics/%s/%s.html' % (best_match.artist, best_match.album)
self.title = best_match.title
self.titlenumber = best_match.number
loader.get_url (url, self.parse_lyrics, callback, *data)
def get_best_match_urls (self, search_results):
best = None
best_match = 0.0
print "attempting to find best search result from %d sets" % len(search_results)
for result in search_results:
if not hasattr (result, "Item"):
# Search was unsuccessful, try next batch job
continue
valid = filter(self.__valid_match, result.Item)
print "%d valid results in this set" % len(valid)
for item in valid:
album_match = DEFAULT_MATCH
item_album = str(getattr(item.ItemAttributes, "Title", ""))
if item_album != "":
album_match = string_match (self.search_album, item_album)
# match against all returned artist names, taking the best result
item_artists = getattr(item.ItemAttributes, "Artist", [])
if len(item_artists) > 0:
artist_match = 0.0
for artist in item_artists:
# special case 'various artists' for great justice
if artist == "Various Artists":
m = DEFAULT_MATCH
else:
m = string_match (self.search_artist, artist)
if m > artist_match:
artist_match = m
else:
artist_match = DEFAULT_MATCH
# figure out the match strength
if self.search_album == "":
this_match = artist_match
elif self.search_artist == "":
this_match = album_match
else:
# this probably isn't the best way to combine match strengths.
# extremely low values for one match should disqualify a result completely,
# so we don't end up with the wrong album for the right artist.
this_match = (artist_match + album_match) / 2
# reject results with a match strength below a given floor
if album_match < REJECT_MATCH or artist_match < REJECT_MATCH:
result = "rejected"
elif this_match > best_match:
best = item
best_match = this_match
result = "best"
else:
result = "discarded"
print "search result: \"%s\" (%f) by \"%s\" (%f): %f, %s" % (item_album, album_match, str(item_artists), artist_match, this_match, result)
print "best result has match strength %f" % best_match
if best_match > MINIMUM_MATCH:
urls = [getattr (best, size).URL for size in ("LargeImage", "MediumImage") if hasattr (best, size)]
if urls:
print "got urls: %s" % urls
return urls
# No search was successful
print "no search results to return"
return []
def get_best_match_urls(self, search_results):
best = None
best_match = 0.0
print "attempting to find best search result from %d sets" % len(
search_results)
for result in search_results:
if not hasattr(result, "Item"):
# Search was unsuccessful, try next batch job
continue
valid = filter(self.__valid_match, result.Item)
print "%d valid results in this set" % len(valid)
for item in valid:
album_match = DEFAULT_MATCH
item_album = str(getattr(item.ItemAttributes, "Title", ""))
if item_album != "":
album_match = string_match(self.search_album, item_album)
# match against all returned artist names, taking the best result
item_artists = getattr(item.ItemAttributes, "Artist", [])
if len(item_artists) > 0:
artist_match = 0.0
for artist in item_artists:
# special case 'various artists' for great justice
if artist == "Various Artists":
m = DEFAULT_MATCH
else:
m = string_match(self.search_artist, artist)
if m > artist_match:
artist_match = m
else:
artist_match = DEFAULT_MATCH
# figure out the match strength
if self.search_album == "":
this_match = artist_match
elif self.search_artist == "":
this_match = album_match
else:
# this probably isn't the best way to combine match strengths.
# extremely low values for one match should disqualify a result completely,
# so we don't end up with the wrong album for the right artist.
this_match = (artist_match + album_match) / 2
# reject results with a match strength below a given floor
if album_match < REJECT_MATCH or artist_match < REJECT_MATCH:
result = "rejected"
elif this_match > best_match:
best = item
best_match = this_match
result = "best"
else:
result = "discarded"
print "search result: \"%s\" (%f) by \"%s\" (%f): %f, %s" % (
item_album, album_match, str(item_artists), artist_match,
this_match, result)
print "best result has match strength %f" % best_match
if best_match > MINIMUM_MATCH:
urls = [
getattr(best, size).URL
for size in ("LargeImage", "MediumImage")
if hasattr(best, size)
]
if urls:
print "got urls: %s" % urls
return urls
# No search was successful
print "no search results to return"
return []
