1
0
mirror of https://github.com/krateng/maloja.git synced 2023-08-10 21:12:55 +03:00

Associated artists now only show up if relevant

This commit is contained in:
Krateng 2019-04-08 17:32:31 +02:00
parent 7842b247ad
commit 400c920958
2 changed files with 50 additions and 22 deletions

View File

@ -43,10 +43,12 @@ class CleanerAgent:
confirmed = self.rules_belongtogether + [self.rules_replaceartist[r] for r in self.rules_replaceartist] confirmed = self.rules_belongtogether + [self.rules_replaceartist[r] for r in self.rules_replaceartist]
return (a in confirmed) return (a in confirmed)
delimiters_feat = ["ft.","ft","feat.","feat","featuring","Ft.","Ft","Feat.","Feat","Featuring"] #Delimiters used for extra artists, even when in the title field #Delimiters used for extra artists, even when in the title field
delimiters = ["vs.","vs","&"] #Delimiters in informal artist strings, spaces expected around them delimiters_feat = ["ft.","ft","feat.","feat","featuring","Ft.","Ft","Feat.","Feat","Featuring"]
delimiters_formal = ["; ",";","/"] #Delimiters used specifically to tag multiple artists when only one tag field is available, no spaces used #Delimiters in informal artist strings, spaces expected around them
delimiters = ["vs.","vs","&"]
#Delimiters used specifically to tag multiple artists when only one tag field is available, no spaces used
delimiters_formal = ["; ",";","/"]
def parseArtists(self,a): def parseArtists(self,a):
@ -68,7 +70,8 @@ class CleanerAgent:
for d in self.delimiters_feat: for d in self.delimiters_feat:
if re.match(r"(.*) \(" + d + " (.*)\)",a) is not None: if re.match(r"(.*) \(" + d + " (.*)\)",a) is not None:
return self.parseArtists(re.sub(r"(.*) \(" + d + " (.*)\)",r"\1",a)) + self.parseArtists(re.sub(r"(.*) \(" + d + " (.*)\)",r"\2",a)) return self.parseArtists(re.sub(r"(.*) \(" + d + " (.*)\)",r"\1",a)) + \
self.parseArtists(re.sub(r"(.*) \(" + d + " (.*)\)",r"\2",a))
for d in self.delimiters_formal: for d in self.delimiters_formal:
if (d in a): if (d in a):
@ -127,41 +130,57 @@ class CollectorAgent:
def __init__(self): def __init__(self):
self.updateRules() self.updateRules()
# rules_countas dict: real artist -> credited artist
# rules_countas_id dict: real artist ID -> credited artist ID
# rules_include dict: credited artist -> all real artists
def updateRules(self): def updateRules(self):
raw = tsv.parse_all("rules","string","string","string") raw = tsv.parse_all("rules","string","string","string")
self.rules_countas = {b:c for [a,b,c] in raw if a=="countas"} self.rules_countas = {b:c for [a,b,c] in raw if a=="countas"}
self.rules_include = {} #Twice the memory, double the performance! (Yes, we're saving redundant information here, but it's not unelegant if it's within a closed object!) self.rules_countas_id = {}
self.rules_include = {} #Twice the memory, double the performance!
# (Yes, we're saving redundant information here, but it's not unelegant if it's within a closed object!)
for a in self.rules_countas: for a in self.rules_countas:
self.rules_include[self.rules_countas[a]] = self.rules_include.setdefault(self.rules_countas[a],[]) + [a] self.rules_include[self.rules_countas[a]] = self.rules_include.setdefault(self.rules_countas[a],[]) + [a]
# this agent needs to be aware of the current id assignment in the main program. unelegant, but the best way i can think of # this agent needs to be aware of the current id assignment in the main program
# unelegant, but the best way i can think of
def updateIDs(self,artistlist): def updateIDs(self,artistlist):
self.rules_countas_id = {artistlist.index(a):artistlist.index(self.rules_countas[a]) for a in self.rules_countas} self.rules_countas_id = {artistlist.index(a):artistlist.index(self.rules_countas[a]) for a in self.rules_countas if a in artistlist}
#self.rules_include_id = {artistlist.index(a):artistlist.index(self.rules_include[a]) for a in self.rules_include} #self.rules_include_id = {artistlist.index(a):artistlist.index(self.rules_include[a]) for a in self.rules_include}
#this needs to take lists into account #this needs to take lists into account
# get who is credited for this artist
def getCredited(self,artist): def getCredited(self,artist):
if artist in self.rules_countas_id:
return self.rules_countas_id[artist]
if artist in self.rules_countas: if artist in self.rules_countas:
return self.rules_countas[artist] return self.rules_countas[artist]
if artist in self.rules_countas_id:
return self.rules_countas_id[artist]
else: else:
return artist return artist
# get all credited artists for the artists given
def getCreditedList(self,artists): def getCreditedList(self,artists):
updatedArtists = [] updatedArtists = []
for artist in artists: for artist in artists:
updatedArtists.append(self.getCredited(artist)) updatedArtists.append(self.getCredited(artist))
return list(set(updatedArtists)) return list(set(updatedArtists))
# get artists who the given artist is given credit for
def getAllAssociated(self,artist): def getAllAssociated(self,artist):
return self.rules_include.get(artist,[]) return self.rules_include.get(artist,[])
# this function is there to check for artists that we should include in the database even though they never have any scrobble. important to avoid bugs when # this function is there to check for artists that we should include in the
# countas rules are declared preemptively # database even though they never have any scrobble.
def getAllArtists(self): def getAllArtists(self):
return list(set([a for a in self.rules_countas] + [self.rules_countas[a] for a in self.rules_countas])) return list(set([self.rules_countas[a] for a in self.rules_countas]))
# artists that count can be nonexisting (counting HyunA as 4Minute even
# though 4Minute has never been listened to)
# but artists that are counted as someone else are only relevant if they
# exist (so we can preemptively declare lots of rules just in case)
#return list(set([a for a in self.rules_countas] + [self.rules_countas[a] for a in self.rules_countas]))

View File

@ -135,6 +135,13 @@ def getArtistID(name):
ARTISTS.append(obj) ARTISTS.append(obj)
ARTIST_SET.add(objlower) ARTIST_SET.add(objlower)
ARTISTS_LOWER.append(objlower) ARTISTS_LOWER.append(objlower)
# with a new artist added, we might also get new artists that they are credited as
cr = coa.getCredited(name)
getArtistID(cr)
coa.updateIDs(ARTISTS)
return i return i
def getTrackID(artists,title): def getTrackID(artists,title):
@ -473,7 +480,7 @@ def artistInfo(artist):
scrobbles = len(db_query(artists=[artist])) #we cant take the scrobble number from the charts because that includes all countas scrobbles scrobbles = len(db_query(artists=[artist])) #we cant take the scrobble number from the charts because that includes all countas scrobbles
try: try:
c = [e for e in charts if e["artist"] == artist][0] c = [e for e in charts if e["artist"] == artist][0]
others = coa.getAllAssociated(artist) others = [a for a in coa.getAllAssociated(artist) if a in ARTISTS]
position = c["rank"] position = c["rank"]
return {"scrobbles":scrobbles,"position":position,"associated":others,"medals":MEDALS.get(artist)} return {"scrobbles":scrobbles,"position":position,"associated":others,"medals":MEDALS.get(artist)}
except: except:
@ -785,12 +792,14 @@ def build_db():
# inform malojatime module about earliest scrobble # inform malojatime module about earliest scrobble
register_scrobbletime(STAMPS[0]) register_scrobbletime(STAMPS[0])
# get extra artists with zero scrobbles from countas rules # NOT NEEDED BECAUSE WE DO THAT ON ADDING EVERY ARTIST ANYWAY
for artist in coa.getAllArtists(): # get extra artists with no real scrobbles from countas rules
if artist not in ARTISTS: #for artist in coa.getAllArtists():
ARTISTS.append(artist) #for artist in coa.getCreditedList(ARTISTS):
# if artist not in ARTISTS:
coa.updateIDs(ARTISTS) # log(artist + " is added to database because of countas rules",module="debug")
# ARTISTS.append(artist)
# coa.updateIDs(ARTISTS)
#start regular tasks #start regular tasks
update_medals() update_medals()
@ -968,7 +977,7 @@ def db_aggregate_full(by=None,since=None,to=None,within=None,artist=None):
# this either creates the new entry or increments the existing one # this either creates the new entry or increments the existing one
charts[a] = charts.setdefault(a,0) + 1 charts[a] = charts.setdefault(a,0) + 1
ls = [{"artist":get_artist_dict(ARTISTS[a]),"scrobbles":charts[a],"counting":coa.getAllAssociated(ARTISTS[a])} for a in charts] ls = [{"artist":get_artist_dict(ARTISTS[a]),"scrobbles":charts[a],"counting":[arti for arti in coa.getAllAssociated(ARTISTS[a]) if arti in ARTISTS]} for a in charts]
ls.sort(key=lambda k:k["scrobbles"],reverse=True) ls.sort(key=lambda k:k["scrobbles"],reverse=True)
# add ranks # add ranks
for rnk in range(len(ls)): for rnk in range(len(ls)):