rec10/trunk/src/xml2db_dom.py

   1 from timerec import epgdb
   2 import status
   3 #!/usr/bin/python
   4 # coding: UTF-8
   5 # Rec10 TS Recording Tools
   6 # Copyright (C) 2009 Yukikaze
   7
   8 import xml.dom.minidom
   9 import datetime
  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 def getText(elm):
  19     nodelist = elm.childNodes
  20     rc = ""
  21     # 全てのノードに対して
  22     for node in nodelist:
  23         # テキストノードなら値を取得
  24         if node.nodeType == node.TEXT_NODE:
  25             rc = rc + node.data
  26     return rc
  27 def getText_item(obj):
  28     rc=""
  29     for o in obj:
  30         rc=rc+getText(o)
  31     return rc
  32 def writemultitvDB(bctype,tvlists):
  33     rec10d.rec10db.new_epg_timeline(bctype)
  34     for channel,start,stop,title,desc,longdesc,category in tvlists:
  35         rec10d.rec10db.add_epg_timeline(bctype, channel, start, stop, title, desc, longdesc, category)
  36 def writemultichDB(chlists):
  37     for ontv,dn in chlists:
  38         rec10d.rec10db.update_chname_by_ontv_epg_ch(ontv, dn)
  39 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
  40     dtb=datetime.datetime.now()
  41     dom=xml.dom.minidom.parse(file(xmlpath))
  42     chlist=[]
  43     tvlist=[]
  44     bayes={}
  45     for ch in dom.getElementsByTagName('channel'):
  46         ontv=ch.getAttribute("id")
  47         chname=ch.getElementsByTagName("display-name").item(0).childNodes[0].data
  48         chlist.append([ontv,chname])
  49     for tv in dom.getElementsByTagName('programme'):
  50         channel=zenhan.toHankaku_ABC123(tv.getAttribute("channel"))
  51         start=tv.getAttribute("start").replace(" +0900", "")
  52         stop=tv.getAttribute("stop").replace(" +0900", "")
  53         title=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("title"))).replace(",", "_")
  54         desc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("desc")))
  55         longdesc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("longdesc")))
  56         category=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("category")))
  57         tch=chdb.ontvsearch(channel)
  58         if tch!=None:
  59             if tch.get('bctype',"")==bctype:
  60                 bt=bayes.get(tch['chtxt'],["","",0])
  61                 bt2=[bt[0]+title+" ",bt[1]+desc+" "+longdesc+" ",bt[2]+1]
  62                 bayes[tch['chtxt']]=bt2
  63                 tvlist.append([channel,start,stop,title,desc,longdesc,category])
  64                 p=auto_rec.calc_key(tch['chtxt'], title,desc+" "+longdesc)
  65                 if p>2000:
  66                     bttt=datetime.datetime.strptime(start,"%Y%m%d%H%M%S")
  67                     bttt=bttt-datetime.timedelta(seconds=1200)
  68                     bttime=bttt.strftime("%Y-%m-%d %H:%M:%S")
  69                     ettt=datetime.datetime.strptime(stop,"%Y%m%d%H%M%S")
  70                     ettt=ettt+datetime.timedelta(seconds=1200)
  71                     ettime=ettt.strftime("%Y-%m-%d %H:%M:%S")
  72                     chs=rec10d.rec10db.select_bytime_bychtxt_all_timeline(bttime,ettime,tch['chtxt'])
  73                     p2=0
  74                     if len(chs)>0:
  75                         for ch in chs:
  76                             p2t=n_gram.trigram(ch['title'],title)
  77                             if p2t>p2:
  78                                 p2=p2t
  79                     if p2<350:
  80                         if status.getSettings_auto_bayes()==1:
  81                             if p2<200:
  82                                 maxnum=0
  83                                 if len(ch['ch'])>2:
  84                                     maxnum=epgdb.count_schedule_timeline(bttime, ettime)[1]
  85                                     maxnum=int(configreader.getenv("bscs_max"))-maxnum
  86                                 else:
  87                                     maxnum=epgdb.count_schedule_timeline(bttime, ettime)[0]
  88                                     maxnum=int(configreader.getenv("te_max"))-maxnum
  89                                 if maxnum>0:
  90                                     topt=status.getSettings_auto_opt()
  91                                     recdb.recreserv(title,tch['chtxt'], bttime, ettime,topt)
  92                         else:
  93                             recdb.add_auto_bayes(tch['chtxt'],title,start,stop,p)
  94
  95     dom.unlink()
  96     if len(chlist) > 0:
  97         writemultichDB(chlist)
  98         chlist = []
  99         if len(tvlist) > 0:
 100             writemultitvDB(bctype,tvlist)
 101         tvlist = []
 102         for ct,list in bayes.items():
 103             auto_rec.update_recall(ct,list[0],list[1],list[2])
 104             auto_rec.update_recall("ALL",list[0],list[1],list[2])
 105     dtb=datetime.datetime.now()-dtb
 106     recdblist.printutf8(bctype + u" epg取り出し終了")
 107     recdblist.printutf8(str(dtb.days * 24 * 60 * 60 + dtb.seconds)+u"seconds taken for updating bayes-auto DB .")
 108 def getChdata(xmlpath):
 109     dom=xml.dom.minidom.parse(file(xmlpath))
 110     chlist=[]
 111     for ch in dom.getElementsByTagName('channel'):
 112         ontv=ch.getAttribute("id")
 113         chname=ch.getElementsByTagName("display-name").item(0).childNodes[0].data
 114         chlist.append([ontv,chname])
 115     dom.unlink()
 116     if len(chlist)>0:
 117         return chlist[0]
 118     else:
 119         return ["",""]