+++ /dev/null
-#!/usr/bin/python
-# coding: UTF-8
-# Rec10 TS Recording Tools
-# Copyright (C) 2009-2011 Yukikaze
-
-import xml.dom.minidom
-import datetime
-import re
-import traceback
-
-import zenhan
-import chdb
-import auto_rec
-import recdb
-import rec10d
-import n_gram
-import recdblist
-import epgdb
-import status
-def getText(elm):
- nodelist = elm.childNodes
- rc = ""
- # 全てのノードに対して
- for node in nodelist:
- # テキストノードなら値を取得
- if node.nodeType == node.TEXT_NODE:
- rc = rc + node.data
- return rc
-def getText_item(obj):
- rc=""
- for o in obj:
- rc=rc+getText(o)
- return rc
-def writeMultiTVDB(bctype,tvlists):
- rec10d.rec10db.new_epg_timeline(bctype)
- for channel,start,stop,title,desc,longdesc,category in tvlists:
- rec10d.rec10db.add_epg_timeline(bctype, channel, start, stop, title, desc, longdesc, category)
-def writeMultiCHDB(chlists):
- for chtxt,dn in chlists:
- rec10d.rec10db.update_chname_by_chtxt_epg_ch(chtxt,dn)
- if dn==None:
- chdb.changeCHShow(chtxt,"0")
- else:
- if len(dn)<1:
- chdb.changeCHShow(chtxt,"0")
-def writeMultiLogoDB(logolists):
- #[type,sv,pngdata]
- for type,sv,pngdata in logolists:
- chtxt=chdb.searchCHFromCsch(sv)['chtxt']
- if len(chtxt)>0:
- rec10d.rec10db.change_logodata_epg_ch(chtxt,type,pngdata)
-def xml2db_dom(xmlpath, bctype):#bctypeは放送種別で'TE'(地デジ)BS,CSがある。地デジの場合は te数字 が入る
- dtb=datetime.datetime.now()
- dom=xml.dom.minidom.parse(file(xmlpath))
- chlist=[]
- tvlist=[]
- bayes={}
- rHisch=re.compile(u"『([^『]+)』(.+)\Z")##History CHのタイトル#01 exp:「#01説明」をタイトルに含める
- for ch in dom.getElementsByTagName('channel'):
- chtxtt=ch.getAttribute("id")
- chname=ch.getElementsByTagName("display-name").item(0).childNodes[0].data
- chlist.append([chtxtt,chname])
- for tv in dom.getElementsByTagName('programme'):
- channel=zenhan.toHankaku_ABC123(tv.getAttribute("channel"))
- start=tv.getAttribute("start").replace(" +0900", "")
- stop=tv.getAttribute("stop").replace(" +0900", "")
- title=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("title"))).replace(",", "_")
- desc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("desc")))
- longdesc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("longdesc")))
- category=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("category")))
- title=addTitle_Subtitle(rHisch,1,title,desc)
- tch=chdb.searchCHFromChtxt(channel)
- if tch!=None:
- if tch.get('bctype',"")==bctype and not tch['visible']==0:
- bt=bayes.get(tch['chtxt'],["","",0])
- bt2=[bt[0]+title+" ",bt[1]+desc+" "+longdesc+" ",bt[2]+1]
- bayes[tch['chtxt']]=bt2
- tvlist.append([channel,start,stop,title,desc,longdesc,category])
- p=auto_rec.calcKey(tch['chtxt'], title,desc+" "+longdesc)
- if p>2000:
- bttt=datetime.datetime.strptime(start,"%Y%m%d%H%M%S")
- bttt=bttt-datetime.timedelta(seconds=1200)
- bttime=bttt.strftime("%Y-%m-%d %H:%M:%S")
- ettt=datetime.datetime.strptime(stop,"%Y%m%d%H%M%S")
- ettt=ettt+datetime.timedelta(seconds=1200)
- ettime=ettt.strftime("%Y-%m-%d %H:%M:%S")
- chs=rec10d.rec10db.select_bytime_bychtxt_all_timeline(bttime,ettime,tch['chtxt'])
- p2=0
- if len(chs)>0:
- for ch in chs:
- p2t=n_gram.trigram(ch['title'],title)
- if p2t>p2:
- p2=p2t
- if p2<350:
- if status.getSettings_auto_bayes()==1:
- if p2<200:
- maxnum=0
- if len(ch['ch'])>2:
- maxnum=epgdb.count_schedule_timeline(bttime, ettime)[1]
- maxnum=int(configreader.getConfEnv("bscs_max"))-maxnum
- else:
- maxnum=epgdb.count_schedule_timeline(bttime, ettime)[0]
- maxnum=int(configreader.getConfEnv("te_max"))-maxnum
- if maxnum>0:
- topt=status.getSettings_auto_opt()
- recdb.reserveReckey(recdblist.REC_RESERVE,title,tch['chtxt'], bttime, ettime,topt)
- else:
- recdb.addAutoBayesKeyword(tch['chtxt'],title,start,stop,p)
- dom.unlink()
- if len(chlist) > 0:
- writeMultiCHDB(chlist)
- chlist = []
- if len(tvlist) > 0:
- writeMultiTVDB(bctype,tvlist)
- tvlist = []
- for ct,list in bayes.items():
- auto_rec.updateRecall(ct,list[0],list[1],list[2])
- auto_rec.updateRecall("ALL",list[0],list[1],list[2])
- dtb=datetime.datetime.now()-dtb
- recdblist.printutf8(bctype + u" epg取り出し終了")
- recdblist.printutf8(str(dtb.days * 24 * 60 * 60 + dtb.seconds)+u"seconds taken for updating bayes-auto DB .")
-def xml2db_dom_logo(xmlpath, bctype):#bctypeは放送種別で'TE'(地デジ)BS,CSがある。地デジの場合は te数字 が入る
- dtb=datetime.datetime.now()
- dom=xml.dom.minidom.parse(file(xmlpath))
- logolist=[]
- for logo in dom.getElementsByTagName('logo'):
- type=int(logo.getAttribute("type"))
- sv=logo.getAttribute("sv")
- pngdata=logo.childNodes[0].data
- logolist.append([type,sv,pngdata])
- dom.unlink()
- if len(logolist) > 0:
- writeMultiLogoDB(logolist)
- dtb=datetime.datetime.now()-dtb
- recdblist.printutf8(bctype + u" epg取り出し終了")
- recdblist.printutf8(str(dtb.days * 24 * 60 * 60 + dtb.seconds)+u"seconds taken for updating logo DB .")
-def addTitle_Subtitle(recompiled,num,title,exp):
- try:
- tST=recompiled.match(exp)
- newtitle=u""
- if tST:
- newtitle=title+u"『"+tST.group(num)+u"』"
- else:
- newtitle=title
- except Exception, inst:
- recdblist.addCommonlogEX("Error", "addTitle_Subtitle(xml2db_dob.py)", str(type(inst)),str(inst)+traceback.format_exc(),log_level=200)
- newtitle=title
- return newtitle
OSDN Git Service
