+++ /dev/null
-#!/usr/bin/python
-# coding: UTF-8
-# Rec10 TS Recording Tools
-# Copyright (C) 2009-2011 Yukikaze
-
-import os
-import os.path
-import re
-import time
-import os
-import datetime
-import n_gram
-import recdblist
-import zenhan
-import auto_move
-
-def detSeriesNum(execpath):
- files=os.listdir(execpath)
- ss={}
- for file in files:
- ftitle=os.path.splitext(file)
- fname=ftitle[0]
- fname=zenhan.toHankaku(unicode(fname,"utf-8"))
- nt=detNameType(fname, execpath)
- st=nt['title']
- if not ss.has_key(st):
- ss[st]={}
- ss[st][nt['num']]=ftitle
- print ss
- return ss
-def detDeltaNum(title,movepath):
- """
- #番組の話数を推測する。
- """
- files=os.listdir(movepath)
- ff=[]
- maxnum=0
- for file in files:
- recdblist.printutf8(unicode(file,'utf-8'),verbose_level=800)
- file=os.path.join(movepath,file)
- if os.path.isfile(file):
- name=os.path.splitext(os.path.split(file)[1])
- name=name[0]
- name=zenhan.toHankaku(unicode(name,'utf-8'))
- p1=detNameType(name,movepath)
- #recdblist.printutf8(p1['title']+" "+str(p1['num']))
- time1=time.localtime(os.path.getmtime(file))
- time1=datetime.datetime.fromtimestamp(os.path.getmtime(file))
- if p1['num']!=-1:
- ff.append([p1['num'],p1['title'],time1])
- if maxnum<p1['num']+1:
- maxnum=p1['num']+1
- return detMultiDeltaDays(ff)
-def detMultiDeltaDays(num_with_title_with_dates):
- maxnum=0
- for ft in num_with_title_with_dates:
- if maxnum<ft[0]+1:
- maxnum=ft[0]+1
- ff=num_with_title_with_dates
- f3=[None]*maxnum
- fret={}
- for i in range(0, maxnum, 1):
- f3[i]=[]
- for f2 in ff:
- f3[f2[0]].append(f2)
- for i in range(maxnum):
- for j in range(i+1,maxnum,1):
- for ft3 in f3[i]:
- for ft4 in f3[j]:
- for gdd in detSingleDeltaDay(i, ft3[2], j, ft4[2]):
- #print gdd
- if fret.get(gdd[0])!=None:
- #print fret
- fret[gdd[0]]=fret[gdd[0]]+gdd[1]
- else:
- fret[gdd[0]]=gdd[1]
- #print fret
- maxk=0
- maxp=0
- for i in range(maxnum):
- if fret.get(i)!=None:
- if maxp<fret[i]:
- maxk=i
- maxp=fret[i]
- #print maxk
- #print maxp
- return maxk
-def detSingleDeltaDay(num1,date1,num2,date2):
- dd=date1-date2
- if date1<date2:
- dd=date2-date1
- dn=dd.days
- d=num2-num1
- if d<0:
- d=-1*d
- dp=dd+datetime.timedelta(hours=6)
- dp=dp.days
- dm=dd-datetime.timedelta(hours=6)
- dm=dm.days
- ret=[]
- if dn%d*2>d:
- dn=dn+d
- if dm%d*2>d:
- dm=dm+d
- if dp%d*2>d:
- dp=dp+d
- #recdblist.printutf8(str(d)+":"+str(dn))
- if dp!=dn:
- ret.append([dn/d,60])
- ret.append([dp/d,40])
- elif dm!=dn:
- ret.append([dn/d,60])
- ret.append([dm/d,40])
- else:
- ret.append([dn/d,100])
- return ret
-def detNameType(title,path):
- """
- type A ---title#<number>
- type B ---title#<number>subtitle
- type C ---title subtitle
- type D ---title(without number)
- type Aj ---title第<number>話
- path --search reflexively
- """
- new=0
- if re.search(u"[新]",title) or re.search(u" 新",title):
- title=title.replace(u"[新]","")
- title=title.replace(u" 新","")
- new=1
- recdblist.printutf8(title,verbose_level=800)
- title=auto_move.getTitle(title)##titleから日時を除く
- title=title.replace(u"無料≫","")
- #rA=re.compile(".+(?P<title>)#\d(?P<num>)\s[0,10]\z")
- rA=re.compile("(.+)#(\d*)\s*\Z")
- tA=rA.match(title)
- rB=re.compile("(.+)#(\d*)\s*(\D*)")
- tB=rB.match(title)
- rAj=re.compile("(.+)第(\d*)話\s*\Z")
- tAj=rAj.match(title)
- ret={'title':"",'type':"",'num':0,'subtitle':"",'folder':""}
- if tA:
- #recdblist.printutf8("typeA")
- #recdblist.printutf8("title="+tA.group(1))
- #recdblist.printutf8("num="+tA.group(2))
- ret['type']="A"
- ret['title']=tA.group(1).replace(" ","")
- ret['num']=int(tA.group(2))
- ret['folder']=searchFolder(tA.group(1),unicode(path,'utf-8'))
- if tAj:
- #recdblist.printutf8("typeA")
- #recdblist.printutf8("title="+tAj.group(1))
- #recdblist.printutf8("num="+tAj.group(2))
- ret['type']="Aj"
- ret['title']=tAj.group(1).replace(" ","")
- ret['num']=int(tAj.group(2))
- ret['folder']=searchFolder(tAj.group(1),unicode(path,'utf-8'))
- elif tB:
- #recdblist.printutf8("typeB")
- #recdblist.printutf8("title="+tB.group(1))
- #recdblist.printutf8("num="+tB.group(2))
- #recdblist.printutf8("subtitle="+tB.group(3))
- ret['type']="B"
- ret['title']=tB.group(1).replace(" ","")
- ret['num']=int(tB.group(2))
- ret['folder']=searchFolder(tB.group(1),unicode(path,'utf-8'))
- ret['subtitle']=tB.group(3)
- else:#type C or type D
- #fold=searchFolder(title, path)
- ts=title.split(" ")
- tt=""
- rt=["",0,""]
- for t in ts:
- tt=tt+" "+t
- ft1=searchFolder(tt,unicode(path,'utf-8'))
- #recdblist.printutf8(tt)
- #print ft1
- if ft1!="":
- #recdblist.printutf8(rt)
- #recdblist.printutf8(ft1[0]+" : "+str(ft1[1]))
- if ft1[1]>rt[1]:
- rt[0]=tt
- rt[1]=ft1[1]
- rt[2]=ft1[0]
- #recdblist.printutf8(rt)
- #recdblist.printutf8("title="+rt[0][1:]+"/")
- #recdblist.printutf8("subtitle = "+title.replace(rt[0][1:],"")[1:])
- ret['title']=rt[0][1:].replace(" ","")
- ret['num']=-1
- ret['folder']=rt[2]
- ret['subtitle']=title.replace(rt[0][1:],"")[1:]
- if ret['subtitle'].replace(" ","")=="":
- ret['type']="D"
- else:
- ret['type']="C"
- if new==1:
- ret['num']=1
- return ret
-def searchFolder(title,path,threshold=500):
- """
- titleからフォルダーを探す
- """
- folderpath=os.listdir(path)
- lfpath=[]
- ngram=[]
- for ft in folderpath:
- fullpath=os.path.join(path, ft)
- if os.path.isdir(fullpath):
- lfpath.append(fullpath)
- ftt=os.listdir(fullpath)
- if len(ftt)>0:
- for ft2 in ftt:
- folderpath.append(os.path.join(fullpath, ft2))
- else:
- lfpath.append(fullpath)
- for dirp in lfpath:
- cmpp=""
- appp=""
- if os.path.isdir(dirp):
- cmpp=os.path.dirname(dirp)
- appp=dirp
- else:
- cmpp=os.path.basename(dirp)
- appp=os.path.dirname(dirp)
- ntitle=auto_move.getTitle(title)
- ncmpp=auto_move.getTitle(cmpp)
- p=n_gram.trigram(ntitle,ncmpp)
- if p>0:
- ngram.append((p,appp))
- ngram=list(set(ngram))
- ngram.sort()
- ngram.reverse()
- if len(ngram)>0:
- #recdblist.printutf8(title + ngram[0][1] + " : "+str(ngram[0][0]))
- if ngram[0][0]>threshold:
- return ngram[0][1]
- else:
- return ""
- else:
- return ""
\ No newline at end of file
OSDN Git Service
