def save(self, fn):
start = datetime.datetime.now()
r = BinReader(fn)
# self.skipto(r, 'zhilian_144361202250012_1446732064')
while True:
fnpath = r.fn #todo
offset = r.fd.tell()
print "fn: %s, offset: %d" % (fnpath, offset)
(n, v) = r.readone()
if n is None:
break
print fn, n
(getime, jdid) = self.parse_name(n)
if getime is None or jdid is None:
continue
getime = int(getime)
pp = self.get_pagestore()
if pp.save(getime, jdid, self.get_url(jdid), v, fnpath, offset):
global saved_count
saved_count += 1
timeused = str(datetime.datetime.now() - start)
if saved_count % 1000 == 1:
print "====== saved:%d [%.2f%%][%s] =====" % (
saved_count, r.progress() * 100, timeused)
def save_bj_word(url, ch):
m = re.search('^(.*):\/\/(.*)', url)
channel = m.group(1)
docid = m.group(2)
print channel
dburl = 'mongodb://*****:*****@localhost/'
client = pymongo.MongoClient(dburl)
c = client.admin['page_store_' + channel]
page = c.find({'indexUrl': url})
count = 0
success = False
for p in page:
count += 1
print p['pageContentPath']
(ft, ofn, pos) = p['pageContentPath'].split('::')
reader = BinReader(ofn)
content = reader.readone_at(int(pos))
print content[0]
print content[1]
print datetime.datetime.fromtimestamp(
long(p['crawlerUpdateTime']) / 1000)
res = ch.parse(p['indexUrl'], content[1])
if res:
print res['name'], docid
success = True
# print res['value']
else:
print 'failed', docid
print 'count:', count
return success
def getRegNo(self, begin, end):
rs_path = "/home/peiyuan/data/qichacha/sum/qichacha/39/qichacha.9139.bin"
binReader = BinReader(rs_path)
i = begin
k = 0
while k< i:
binReader.readone()
k += 1
while i < end:
line = binReader.readone()
if line[0] == None:
break
i += 1
m = re.search(r"<li><label>注册号： </label>(.*?)</li>", line[1])
if m:
regNo = m.group(1)
print regNo
if regNo.strip()!="":
self.dsfile.write(regNo.strip()+"\n")
self.dsfile.flush()
else:
self.fail_file.write(line[0]+" has no regno!\n")
else:
self.fail_file.write(line[0]+" has no regno!\n")
def handle_item(self, item):
(ft, ofn, pos) = item['pageContentPath'].split('::')
if self.bin_reader is None or self.bin_reader.fd.name != ofn:
self.bin_reader = BinReader(ofn)
item['content'] = self.bin_reader.readone_at(int(pos))
items = self.parse_item(item)
for ii in items:
self.process_child_item(ii)
def read_bin_file():
t = BinReader('nacao_captcha_image.bin')
count = 0
while True:
(a, b) = t.readone()
if a is None or b is None:
break
count += 1
save_file("/home/windy/codeimg/nacao/"+ a +".jpeg", b)
print count, a
def read_all(self):
res = []
c = self.mongo.admin[self.source]
bin_reader = None
for i in c.find():
(ft, ofn, pos) = i['pageContentPath'].split('::')
if bin_reader is None or bin_reader.fd.name != ofn:
bin_reader = BinReader(ofn)
i['content'] = bin_reader.readone_at(int(pos))
res.append(i)
return res
def read_all(self):
res = []
c = self.client[self.database][self.collection]
bin_reader = None
for i in c.find():
(ft, ofn, pos) = i['pageContentPath'].split('::')
if bin_reader is None or bin_reader.fd.name != ofn:
bin_reader = BinReader(ofn)
i['content'] = bin_reader.readone_at(int(pos))
res.append(i)
return res
def read_all(self):
res = []
if len(self._docs) > 0:
bin_reader = None
for i in self._docs:
doc = copy.deepcopy(i)
(ft, ofn, pos) = doc['pageContentPath'].split('::')
if bin_reader is None or bin_reader.fd.name != ofn:
bin_reader = BinReader(ofn)
doc['content'] = bin_reader.readone_at(int(pos))
res.append(doc)
return res
def dispatch(self):
for binname in self.bin_list:
bs = BinReader("./jobdata/"+binname)
#bs = BinReader("./data/"+binname)
while True:
(a,b) = bs.readone()
if a is None:
break
job = {"id":a,"content":b}
self.add_job(job, True)
self.wait_q_breakable()
self.add_job(None, True)
def dojob(self):
self.init()
reader = None
for item in self._docs:
(ft, ofn, pos) = item['pageContentPath'].split('::')
if reader is None or reader.fd.name != ofn:
reader = BinReader(ofn)
content = reader.readone_at(int(pos))
if 'HTTP 错误 404.0' in content[1]:
self.client[self.database][self.collection].delete_one({
'indexUrl':
item['indexUrl'],
'contentSign':
item['contentSign']
})
class CWPParser(AbstractParser):
"""Consuming while Producing Parser"""
def __init__(self,
channel,
name,
db='admin',
url='mongodb://*****:*****@localhost/'):
AbstractParser.__init__(self, channel, name, db, url)
self.bin_reader = None
@abc.abstractmethod
def parse_item(self, page):
raise NotImplementedError('virtual function called')
def init(self):
return self.client[self.database][self.collection]
def iter_results(self, res):
return res.find()
def handle_item(self, item):
(ft, ofn, pos) = item['pageContentPath'].split('::')
if self.bin_reader is None or self.bin_reader.fd.name != ofn:
self.bin_reader = BinReader(ofn)
item['content'] = self.bin_reader.readone_at(int(pos))
items = self.parse_item(item)
for ii in items:
self.process_child_item(ii)
@abc.abstractmethod
def process_child_item(self, item):
pass
def read_bin_file():
t = BinReader('gsinfo_Guangdong_pic.bin')
count = 0
while True:
(a, b) = t.readone()
if a is None or b is None:
break
count += 1
imgtype = imghdr.what(None, b)
if imgtype in ['gif', 'jpeg', 'jpg', 'png', 'bmp']:
spider.util.FS.dbg_save_file(
"./captcha/" + spider.util.utf8str(a) + "." + imgtype, b)
#print a, "save suceess..."
else:
print a, "验证码格式无效，可能内容已经损坏..."
continue
def read_next(self, count=1000):
res = []
if len(self._docs) > 0:
bin_reader = None
end = self.__current + count
if end > len(self._docs):
end = len(self._docs)
for i in self._docs[self.__current:end]:
doc = copy.deepcopy(i)
(ft, ofn, pos) = doc['pageContentPath'].split('::')
if bin_reader is None or bin_reader.fd.name != ofn:
bin_reader = BinReader(ofn)
doc['content'] = bin_reader.readone_at(int(pos))
res.append(doc)
self.__current = end
return res
def random_check(self, start=0, limit=10):
binreader = BinReader(self.check_file)
opts, args = getopt.getopt(sys.argv[1:], "n:")
if len(opts) is 0 or opts[0][1] is "":
randomindex = random.randint(start, start + limit)
else:
randomindex = int(opts[0][1])
line = binreader.readone()
i = 1
while i < randomindex and line[0] is not None:
line = binreader.readone()
i += 1
if line[0] is None:
print "None!!!"
return
f = open(line[0] + ".html", "w+b")
f.write(line[1])
f.close()
def dispatch(self):
for binname in self.bin_list:
bin = BinReader("./jobdata/" + binname)
while True:
(a, b) = self.bs.readone()
if a is None:
break
job = {"index": a, "html": b}
self.add_job(job, True)
self.wait_q_breakable()
self.add_job(None, True)
def main():
try:
opts, args = getopt.gnu_getopt(sys.argv[1:], 'o:m:pi:')
except getopt.GetoptError as e:
showusage()
return 1
outfile = None
matchstr = ''
printout = False
index = -1
for (n, v) in opts:
if n == '-o':
outfile = v
if n == '-m':
matchstr = v
if n == '-p':
printout = True
if n == '-i':
index = int(v)
if len(args) == 0:
showusage()
return 1
if outfile:
fo = BinSaver(outfile)
for fn in args:
r = BinReader(fn)
while True:
(n, v) = r.readone()
if n is None:
break
if matchstr in v:
fo.append(n, v)
else:
for fn in args:
if printout or index != -1:
r = BinReader(fn)
else:
r = BinReader1(fn)
findex = 0
while True:
(n, v) = r.readone()
if n is None:
break
if index != -1:
if findex == index:
if printout:
print v
else:
print n
elif findex > index:
break
elif printout:
print n, v
else:
print n
findex += 1
def sum(self):
filelist = self._get_file_path()
totalcnt = 0
rsinfo = []
for i in range(len(filelist)):
if os.path.isdir(dirpath + filelist[i]): continue
binreader = BinReader(dirpath + filelist[i])
line = binreader.readone()
endline = 10000000
cnt = 0
emptycnt = 0
skipcnt = 0
while line[0]:
print "reading", filelist[i], ",line", cnt + 1
channel, cpid, gettime = line[0].split(".")
if self.pagestore.check_should_fetch(cpid):
if self.pagestore.save(
int(gettime), cpid,
"http://qichacha.com/firm_CN_" + cpid, line[1]):
print filelist[i], "line", cnt + 1, "saved."
else:
print "skip", filelist[i], "line", cnt + 1
skipcnt += 1
if line[1] is None:
emptycnt += 1
line = binreader.readone()
cnt += 1
totalcnt += 1
if cnt >= endline: break
infostr = "" + str(filelist[i]) + " total: " + str(
cnt) + " empty: " + str(emptycnt) + " skip: " + str(skipcnt)
print infostr
rsinfo.append(infostr)
for info in rsinfo:
print info
print len(filelist), "files, total:", totalcnt, "results"
return totalcnt
def go(self):
fns = ['fo.bin']
for fn in fns:
fr = BinReader(fn)
self.go_(fr)
if __name__ == '__main__':
# delete_doc()
url = 'abstract://2005100739298'
channel = re.search('^(.*):\/\/(.*)', url).group(1)
print channel
store = CourtStore(channel)
res = store.find_new(url)
print res
dburl = 'mongodb://*****:*****@localhost/'
client = pymongo.MongoClient(dburl)
c = client.zhuanli['page_store_' + channel]
page = c.find({'indexUrl': url})
count = 0
for p in page:
count += 1
print p['pageContentPath']
(ft, ofn, pos) = p['pageContentPath'].split('::')
reader = BinReader(ofn)
content = reader.readone_at(int(pos))
print content[0]
print content[1]
print p['indexUrl']
print p['realUrl']
print datetime.datetime.fromtimestamp(
long(p['crawlerUpdateTime']) / 1000)
# text=eval(content[1].replace('null','None'))
# doc=text['FileContent']
# save_to_word_2(doc,url[11:])
print 'count:', count
