def __init__(self,context,querykey):
threading.Thread.__init__(self)
self.querykey = querykey
self.context = context.decode("utf-8")
self.splitmode=1 # 1 or 2
self.gramword = []
self.companword = []
self.dicword = []
self.ascword = []
self.wordtypedic = {}
self.textx = None
self.text = None
self.debugmsg = None
self.db = db.DB()
dicdir=config.defaultpath+'/dic/'
self.utf8doc=dicdir+"utf-8doc.db"
self.db.open(self.utf8doc, None, db.DB_HASH, db.DB_DIRTY_READ)
self.bign = None
self.litn = None
self.word = None
markworddic=dicdir+"markword.pick"
self.wordclass=cPickle.load(open(markworddic,'r'))
self.logger = logger.infologger()
self.xsplit=[]
def OriginalHTMLprocess(listsplit):
OriginalHTMLdb = OriginalPage()
ilog = infologger()
purei = Purecontent("c")
pat = re.compile("<([^>]|\n)*>")
space = re.compile("\ \;|\©\;|\r|\t")
stmk = stopmarks()
md5urllist = {}
for i in listsplit:
md5url = md5hex(i)
md5urllist[md5url] = [i]
word = ""
st = time.time()
purei.url_md5 = md5url
if purei.checkexist():
OriginalHTMLdb.url = i
parser = html2txt()
try:
parser.feed(OriginalHTMLdb.queryoriginalct())
charset = parser.charset # charset detector
parser.close()
except:
charset = ""
Originaltext = langconvert(OriginalHTMLdb.queryoriginalct(), charset)
Originaltext = Originaltext.decode("utf-8")
ilog.sentence_split_info(time.time() - st)
try: # If this page is normal html format
parser = ""
parser = html2txt()
parser.feed(Originaltext)
word = word + parser.text
if len(word) == 0:
word = word + space.sub(chr(32), pat.sub(chr(32), Originaltext))
contenttitle = clearspace(parser.title)
parser.close()
# print contenttitle,i,charset
purei.title = contenttitle.encode("utf-8")
except:
try:
parser = html2txt()
parser.feed(Originaltext)
contenttitle = clearspace(parser.title)
parser.close()
except:
contenttitle = ""
purei.title = contenttitle.encode("utf-8")
word = word + space.sub(chr(32), pat.sub(chr(32), Originaltext))
context = ""
ilog.sentence_split_info(time.time() - st)
n = 0
for xw in word:
if ord(xw) >= 32 or ord(xw) in [9, 10, 13]:
context = context + xw
n += 1
if n > 40000000: # may over 65535 line of a document.
break
ilog.sentence_split_info(purei.title + str(len(context)) + i + charset)
context = context + chr(32)
contline = []
contline.append("")
i = 0 # line of contline list
# for x in xrange(len(context)):
x = 0 # word number
msl = 260
while x < len(context):
ordx = ord(context[x])
contline[i] = contline[i] + context[x]
sentencecount = len(clearspace((contline[i])))
# sentencecount=len(contline[i])
if (
sentencecount > msl
and stmk.atypestopmarks(ordx)
or sentencecount > msl
and context[x : x + 2] == ". "
or sentencecount > msl + 20
and stmk.btypestopmarks(ordx)
or sentencecount > msl + 20
and ordx == 10
and ord(context[x + 1 : x + 2]) < 65
):
nextword = context[x + 1 : x + 2]
if nextword:
if punctuationmarks(ord(nextword)):
# at some case, chinese word will use two marks.
x += 1
contline[i] = contline[i] + context[x]
contline.append("")
i += 1
if msl <= 16640 and i % 2:
msl = msl + msl # Dobule it, Until this value bigger then 4000.
x += 1
if sentencecount < msl:
contline[i] = contline[i] + context[x : x + msl]
x = x + msl
contcleanline = []
i = 0
ilog.sentence_split_info(time.time() - st)
for x in contline:
cont = clearspace(x)
if len(cont) > 1:
if cont[0] == chr(32) and cont[-1] == chr(32):
cont = cont[1:-1]
elif cont[-1] == chr(32):
cont = cont[:-1]
elif cont[0] == chr(32):
cont = cont[1:]
if len(cont) < 65025 and cont != chr(32):
contcleanline.append(cont.encode("utf-8"))
i = i + 1
ilog.sentence_split_info(time.time() - st)
purei.purecotentinline = contcleanline
purei.content = clearspace(context).encode("utf-8")
purei.insertPurecontent()
stderr.write(".")
OriginalHTMLdb.close()
purei.close()
return md5urllist