rec10/trunk/src/xml2db_dom.py

   1 #!/usr/bin/python
   2 # coding: UTF-8
   3 # Rec10 TS Recording Tools
   4 # Copyright (C) 2009-2010 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 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
  42     dtb=datetime.datetime.now()
  43     dom=xml.dom.minidom.parse(file(xmlpath))
  44     chlist=[]
  45     tvlist=[]
  46     bayes={}
  47     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
  48     for ch in dom.getElementsByTagName('channel'):
  49         chtxtt=ch.getAttribute("id")
  50         chname=ch.getElementsByTagName("display-name").item(0).childNodes[0].data
  51         chlist.append([chtxtt,chname])
  52     for tv in dom.getElementsByTagName('programme'):
  53         channel=zenhan.toHankaku_ABC123(tv.getAttribute("channel"))
  54         start=tv.getAttribute("start").replace(" +0900", "")
  55         stop=tv.getAttribute("stop").replace(" +0900", "")
  56         title=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("title"))).replace(",", "_")
  57         desc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("desc")))
  58         longdesc=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("longdesc")))
  59         category=zenhan.toHankaku_ABC123(getText_item(tv.getElementsByTagName("category")))
  60         title=addTitle_Subtitle(rHisch,1,title,desc)
  61         tch=chdb.chtxtsearch(channel)
  62         if tch!=None:
  63             if tch.get('bctype',"")==bctype:
  64                 bt=bayes.get(tch['chtxt'],["","",0])
  65                 bt2=[bt[0]+title+" ",bt[1]+desc+" "+longdesc+" ",bt[2]+1]
  66                 bayes[tch['chtxt']]=bt2
  67                 tvlist.append([channel,start,stop,title,desc,longdesc,category])
  68                 p=auto_rec.calc_key(tch['chtxt'], title,desc+" "+longdesc)
  69                 if p>2000:
  70                     bttt=datetime.datetime.strptime(start,"%Y%m%d%H%M%S")
  71                     bttt=bttt-datetime.timedelta(seconds=1200)
  72                     bttime=bttt.strftime("%Y-%m-%d %H:%M:%S")
  73                     ettt=datetime.datetime.strptime(stop,"%Y%m%d%H%M%S")
  74                     ettt=ettt+datetime.timedelta(seconds=1200)
  75                     ettime=ettt.strftime("%Y-%m-%d %H:%M:%S")
  76                     chs=rec10d.rec10db.select_bytime_bychtxt_all_timeline(bttime,ettime,tch['chtxt'])
  77                     p2=0
  78                     if len(chs)>0:
  79                         for ch in chs:
  80                             p2t=n_gram.trigram(ch['title'],title)
  81                             if p2t>p2:
  82                                 p2=p2t
  83                     if p2<350:
  84                         if status.getSettings_auto_bayes()==1:
  85                             if p2<200:
  86                                 maxnum=0
  87                                 if len(ch['ch'])>2:
  88                                     maxnum=epgdb.count_schedule_timeline(bttime, ettime)[1]
  89                                     maxnum=int(configreader.getenv("bscs_max"))-maxnum
  90                                 else:
  91                                     maxnum=epgdb.count_schedule_timeline(bttime, ettime)[0]
  92                                     maxnum=int(configreader.getenv("te_max"))-maxnum
  93                                 if maxnum>0:
  94                                     topt=status.getSettings_auto_opt()
  95                                     recdb.recreserv(title,tch['chtxt'], bttime, ettime,topt)
  96                         else:
  97                             recdb.add_auto_bayes(tch['chtxt'],title,start,stop,p)
  98
  99     dom.unlink()
 100     if len(chlist) > 0:
 101         writemultichDB(chlist)
 102         chlist = []
 103         if len(tvlist) > 0:
 104             writemultitvDB(bctype,tvlist)
 105         tvlist = []
 106         for ct,list in bayes.items():
 107             auto_rec.update_recall(ct,list[0],list[1],list[2])
 108             auto_rec.update_recall("ALL",list[0],list[1],list[2])
 109     dtb=datetime.datetime.now()-dtb
 110     recdblist.printutf8(bctype + u" epg取り出し終了")
 111     recdblist.printutf8(str(dtb.days * 24 * 60 * 60 + dtb.seconds)+u"seconds taken for updating bayes-auto DB .")
 112 def addTitle_Subtitle(recompiled,num,title,exp):
 113     try:
 114         tST=recompiled.match(exp)
 115         newtitle=u""
 116         if tST:
 117             newtitle=title+u"『"+tST.group(num)+u"』"
 118         else:
 119             newtitle=title
 120     except Exception, inst:
 121         recdblist.Commonlogex("Error", "addTitle_Subtitle(xml2db_dob.py)", str(type(inst)),str(inst)+traceback.format_exc(),log_level=200)
 122         newtitle=title
 123     return newtitle