rec10/xml2db_dom.py

   1 #!/usr/bin/python
   2 # coding: UTF-8
   3 # Rec10 TS Recording Tools
   4 # Copyright (C) 2009-2011 Yukikaze
   5
   6 import xml.dom.minidom
   7 import datetime
   8 import re
   9 import traceback
  10
  11 import zenhan
  12 import chdb
  13 import auto_rec
  14 import recdb
  15 import rec10d
  16 import n_gram
  17 import recdblist
  18 import epgdb
  19 import status
  20 def getText(elm):
  21     nodelist = elm.childNodes
  22     rc = ""
  23     # 全てのノードに対して
  24     for node in nodelist:
  25         # テキストノードなら値を取得
  26         if node.nodeType == node.TEXT_NODE:
  27             rc = rc + node.data
  28     return rc
  29 def getText_item(obj):
  30     rc=""
  31     for o in obj:
  32         rc=rc+getText(o)
  33     return rc
  34 def writeMultiTVDB(bctype,tvlists):
  35     rec10d.rec10db.new_epg_timeline(bctype)
  36     for channel,start,stop,title,desc,longdesc,category in tvlists:
  37         rec10d.rec10db.add_epg_timeline(bctype, channel, start, stop, title, desc, longdesc, category)
  38 def writeMultiCHDB(chlists):
  39     for chtxt,dn in chlists:
  40         rec10d.rec10db.update_chname_by_chtxt_epg_ch(chtxt,dn)
  41         if dn==None:
  42             chdb.changeCHShow(chtxt,"0")
  43         else:
  44             if len(dn)<1:
  45                 chdb.changeCHShow(chtxt,"0")
  46 def writeMultiLogoDB(logolists):
  47     #[type,sv,pngdata]
  48     for type,sv,pngdata in logolists:
  49         chtxt=chdb.searchCHFromCsch(sv)['chtxt']
  50         if len(chtxt)>0:
  51             rec10d.rec10db.change_logodata_epg_ch(chtxt,type,pngdata)
  52 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
  53     dtb=datetime.datetime.now()
  54     dom=xml.dom.minidom.parse(file(xmlpath))
  55     chlist=[]
  56     tvlist=[]
  57     bayes={}
  58     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
  59     for ch in dom.getElementsByTagName('channel'):
  60         chtxtt=ch.getAttribute("id")
  61         chname=ch.getElementsByTagName("display-name").item(0).childNodes[0].data
  62         chlist.append([chtxtt,chname])
  63     for tv in dom.getElementsByTagName('programme'):
  64         channel=zenhan.toHankaku_ABC123(tv.getAttribute("channel"))
  65         start=tv.getAttribute("start").replace(" +0900", "")
  66         stop=tv.getAttribute("stop").replace(" +0900", "")
  67         title=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("title"))).replace(",", "_")
  68         desc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("desc")))
  69         longdesc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("longdesc")))
  70         category=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("category")))
  71         title=addTitle_Subtitle(rHisch,1,title,desc)
  72         tch=chdb.searchCHFromChtxt(channel)
  73         if tch!=None:
  74             if tch.get('bctype',"")==bctype and not tch['visible']==0:
  75                 bt=bayes.get(tch['chtxt'],["","",0])
  76                 bt2=[bt[0]+title+" ",bt[1]+desc+" "+longdesc+" ",bt[2]+1]
  77                 bayes[tch['chtxt']]=bt2
  78                 tvlist.append([channel,start,stop,title,desc,longdesc,category])
  79                 p=auto_rec.calcKey(tch['chtxt'], title,desc+" "+longdesc)
  80                 if p>2000:
  81                     bttt=datetime.datetime.strptime(start,"%Y%m%d%H%M%S")
  82                     bttt=bttt-datetime.timedelta(seconds=1200)
  83                     bttime=bttt.strftime("%Y-%m-%d %H:%M:%S")
  84                     ettt=datetime.datetime.strptime(stop,"%Y%m%d%H%M%S")
  85                     ettt=ettt+datetime.timedelta(seconds=1200)
  86                     ettime=ettt.strftime("%Y-%m-%d %H:%M:%S")
  87                     chs=rec10d.rec10db.select_bytime_bychtxt_all_timeline(bttime,ettime,tch['chtxt'])
  88                     p2=0
  89                     if len(chs)>0:
  90                         for ch in chs:
  91                             p2t=n_gram.trigram(ch['title'],title)
  92                             if p2t>p2:
  93                                 p2=p2t
  94                     if p2<350:
  95                         if status.getSettings_auto_bayes()==1:
  96                             if p2<200:
  97                                 maxnum=0
  98                                 if len(ch['ch'])>2:
  99                                     maxnum=epgdb.count_schedule_timeline(bttime, ettime)[1]
 100                                     maxnum=int(configreader.getConfEnv("bscs_max"))-maxnum
 101                                 else:
 102                                     maxnum=epgdb.count_schedule_timeline(bttime, ettime)[0]
 103                                     maxnum=int(configreader.getConfEnv("te_max"))-maxnum
 104                                 if maxnum>0:
 105                                     topt=status.getSettings_auto_opt()
 106                                     recdb.reserveReckey(recdblist.REC_RESERVE,title,tch['chtxt'], bttime, ettime,topt)
 107                         else:
 108                             recdb.addAutoBayesKeyword(tch['chtxt'],title,start,stop,p)
 109     dom.unlink()
 110     if len(chlist) > 0:
 111         writeMultiCHDB(chlist)
 112         chlist = []
 113         if len(tvlist) > 0:
 114             writeMultiTVDB(bctype,tvlist)
 115         tvlist = []
 116         for ct,list in bayes.items():
 117             auto_rec.updateRecall(ct,list[0],list[1],list[2])
 118             auto_rec.updateRecall("ALL",list[0],list[1],list[2])
 119     dtb=datetime.datetime.now()-dtb
 120     recdblist.printutf8(bctype + u" epg取り出し終了")
 121     recdblist.printutf8(str(dtb.days * 24 * 60 * 60 + dtb.seconds)+u"seconds taken for updating bayes-auto DB .")
 122 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
 123     dtb=datetime.datetime.now()
 124     dom=xml.dom.minidom.parse(file(xmlpath))
 125     logolist=[]
 126     for logo in dom.getElementsByTagName('logo'):
 127         type=int(logo.getAttribute("type"))
 128         sv=logo.getAttribute("sv")
 129         pngdata=logo.childNodes[0].data
 130         logolist.append([type,sv,pngdata])
 131     dom.unlink()
 132     if len(logolist) > 0:
 133         writeMultiLogoDB(logolist)
 134     dtb=datetime.datetime.now()-dtb
 135     recdblist.printutf8(bctype + u" epg取り出し終了")
 136     recdblist.printutf8(str(dtb.days * 24 * 60 * 60 + dtb.seconds)+u"seconds taken for updating logo DB .")
 137 def addTitle_Subtitle(recompiled,num,title,exp):
 138     try:
 139         tST=recompiled.match(exp)
 140         newtitle=u""
 141         if tST:
 142             newtitle=title+u"『"+tST.group(num)+u"』"
 143         else:
 144             newtitle=title
 145     except Exception, inst:
 146         recdblist.addCommonlogEX("Error", "addTitle_Subtitle(xml2db_dob.py)", str(type(inst)),str(inst)+traceback.format_exc(),log_level=200)
 147         newtitle=title
 148     return newtitle