def __init__(self,raw=None,docid=None,oid=None,d=None):
if oid:
# get by mongo oid
d=Docs.find_one({"_id": oid})
elif docid:
# get by docid
d=Docs.find_one({"docid": docid})
if d:
# load the values
self.__dict__.update(d)
elif raw:
# create a new document
self.__dict__.update({
'docid' : docid,
'pippies' : [],
'pippiDocs' : [],
'pippiDocsLen' : 0,
'rawid' : None,
})
if not 'type' in self.__dict__:
self.__dict__['type']='raw'
if not 'metadata' in self.__dict__:
self.__dict__['metadata']={}
if raw:
self.raw=raw
self.lang=guessLanguage(" ".join(self.text))
self.save()
else:
raise KeyError('empty docid')
def __init__(self,raw=None,docid=None,oid=None,d=None,owner=None):
if oid:
# get by mongo oid
d=Docs.find_one({"_id": oid})
elif docid:
# get by docid
d=Docs.find_one({"docid": docid})
if d:
# load the values
self.__dict__.update(d)
elif raw:
# create a new document
self.__dict__.update({
'docid' : docid,
'owner': unicode(owner),
'pippies' : [],
'pippiDocs' : [],
'pippiDocsLen' : 0,
'rawid' : None,
'title': docid,
})
if not 'type' in self.__dict__:
self.__dict__['type']='raw'
if not 'metadata' in self.__dict__:
self.__dict__['metadata']={}
self.raw=raw
self.__dict__['lang']=guessLanguage(" ".join(self.text))
self.stems # for caching
self.save()
else:
raise KeyError('empty docid')
def __init__(self, docid=None, *args,**kwargs):
self.__dict__['type'] = 'etherpad'
if docid:
hostValidator = PADRE.search(docid)
if hostValidator:
if hostValidator.group(2) and hostValidator.group(3):
docid=("%s/%s" % (hostValidator.group(2), hostValidator.group(3))).encode('utf8')
kwargs['docid']=docid
url="%s%s/ep/pad/export/%s/latest?format=html" % (hostValidator.group(1) or 'http://', hostValidator.group(2), hostValidator.group(3))
if not Docs.find_one({"docid": docid}):
context = urllib2.urlopen(url).read()
soup = BeautifulSoup(context)
self.__dict__['title']=unescape(unicode(''.join(soup.title.findAll(text=True)))).strip().encode('utf8')
doc='<html><head><title>%s</title><meta http-equiv="content-type" content="text/html; charset=utf-8" /></head>%s</html>' % (self.title, unescape(unicode(soup.body)).encode('utf8'))
raw=str(tidy.parseString(doc, **{'output_xhtml' : 1,
'add_xml_decl' : 0,
'indent' : 0,
'tidy_mark' : 0,
'doctype' : "strict",
'wrap' : 0}))
kwargs['raw'] = raw
kwargs['docid']=docid
super(Etherpad,self).__init__(*args, **kwargs)
if not 'stems' in self.__dict__ or not self.stems:
# let's calculate and cache the results
models.tfidf.add_input_document(self.termcnt.keys())
self.save()
return
kwargs['docid']=docid
super(Etherpad,self).__init__(*args, **kwargs)
def __init__(self, docid=None, *args,**kwargs):
self.__dict__['type'] = 'co-ment'
if docid:
hostValidator = CMTRE.search(docid)
if hostValidator:
if hostValidator.group(1) or hostValidator.group(3) or hostValidator.group(5):
docid=("%s%s" % (hostValidator.group(2), hostValidator.group(4))).encode('utf8')
kwargs['docid']=docid
url="https://%s/text/%s/view/" % (hostValidator.group(2), hostValidator.group(4))
if not Docs.find_one({"docid": docid}):
context = urllib2.urlopen(url).read()
soup = BeautifulSoup(context)
self.__dict__['title']=unescape(unicode(''.join(soup.title.findAll(text=True)))).strip().encode('utf8')
dataurl = "https://%s/text%s/comments/" % (hostValidator.group(2), hostValidator.group(4))
data = urllib2.urlopen(dataurl).read()
soup = BeautifulSoup(data)
kwargs['raw'] = '<html><head><title>%s</title><meta http-equiv="content-type" content="text/html; charset=utf-8" /></head><body>%s</body></html>' % (self.title, unescape(unicode(soup.find(attrs={'id' : 'textcontainer'}))).encode('utf8'))
kwargs['docid']=docid
super(Coment,self).__init__(*args, **kwargs)
if not 'stems' in self.__dict__ or not self.stems:
# let's calculate and cache the results
models.tfidf.add_input_document(self.termcnt.keys())
self.save()
return
kwargs['docid']=docid
super(Coment,self).__init__(*args, **kwargs)
def __init__(self, docid=None, *args,**kwargs):
setTitle=False
if docid:
alias=re.match(SHORTCUTRE,docid)
if alias:
self.__dict__['sector'] = '3'
self.__dict__['year'] = alias.group(2)
self.__dict__['doctype'] = SHORTCUTMAP[alias.group(1)]
self.__dict__['refno'] = "%04d" % int(alias.group(3))
self.__dict__['lang'] = 'EN' # assuming default
else:
(code,lang)=docid.split(":")[1:3]
st=7 if code[6].isalpha() else 6
self.__dict__['sector'] = code[0]
self.__dict__['year'] = code[1:5]
self.__dict__['doctype'] = code[5:st]
self.__dict__['refno'] = code[st:]
self.__dict__['lang'] = lang
self.__dict__['type'] = 'eurlex'
kwargs['docid']=self.docid
if not Docs.find_one({"docid": self.docid}):
retries=4
while True:
raw=CACHE.fetchUrl(EURLEXURL+self.docid+":HTML")
soup=BeautifulSoup(raw)
# TODO handle empty or invalid celex ids - also handle other languages besides english!!!
# <TITLE>Request Error</TITLE>
# <h1>The parameters of the link are incorrect.</h1>
if soup.title and soup.title.string == "Request Error":
if retries>0:
retries=retries-1
continue
else:
raise ValueError, "Request Error"
if soup.h1 and soup.h1.string == 'The parameters of the link are incorrect.':
if retries>0:
retries=retries-1
continue
else:
raise ValueError, "Parameter Error"
# no errors found, continue, nothing to see here
break
# > /* There is no English version of this document available since it was not included in the English Special Edition.
content=soup.find(id='TexteOnly')
if (content and
content.findAll('p') and
len(content.findAll('p'))>1 and
'string' in dir(content.findAll('p')[1]) and
content.findAll('p')[1].string.strip().startswith('/* There is no English version of this document available since it was not included in the English Special Edition.')):
raise ValueError, "Language Error"
kwargs['raw']=anchorArticles(raw)
self.__dict__['metadata'] = self.extractMetadata()
setTitle=True
super(Eurlex,self).__init__(*args, **kwargs)
if setTitle:
self.__dict__['title']=self._gettitle()
self.save()
def frags(request):
filtr={}
template_vars={}
docfilter=None
cutoff=None
pippifilter=None
try:
docfilter = ObjectId(cgi.escape(request.GET.get('doc','')))
except:
pass
if docfilter:
filtr['doc']=docfilter
try:
pippifilter = ObjectId(cgi.escape(request.GET.get('pippi','')))
except:
pass
if pippifilter:
filtr['pippi']=pippifilter
else:
try:
cutoff = int(cgi.escape(request.GET.get('cutoff','7')))
except:
pass
if cutoff: filtr['l']={ '$gte': cutoff }
orderBy = 'l'
orderDesc = True
template_vars=pager(request,Frags.find(filtr),orderBy,orderDesc)
prevDoc=None
template_vars['frags']=[]
for frag in template_vars['data']:
p=Pippi('',oid=frag['pippi'])
d=Doc(oid=frag['doc'])
if pippifilter:
frag['txt']=diffFrag(prevDoc,frag['txt'])
prevDoc=frag['txt']
template_vars['frags'].append({'_id': frag['_id'],
'pos': frag['pos'],
'txt': " ".join(frag['txt']),
'len': frag['l'],
'score': sum([d.tfidf.get(t,0) for t in p.pippi]),
'pippi': p,
'doc': d,
})
template_vars['pippi']=pippifilter
template_vars['doc']=docfilter
if docfilter: template_vars['docTitle']=Docs.find_one({'_id': docfilter},['docid'])['docid']
if pippifilter: template_vars['pippiFilter']=1 #" ".join(Pippies.find_one({'_id': pippifilter},['pippi'])['pippi'])
return render_to_response('frags.html', template_vars, context_instance=RequestContext(request))
def Doc(*args, **kwargs):
if 'docid' in kwargs:
for (t,c,r) in DOCTYPES:
if re.match(r,kwargs['docid']):
return c(*args,**kwargs)
if 'd' in kwargs:
for (t,c,r) in DOCTYPES:
if kwargs['d'].get('type','') == t or re.match(r,kwargs['d'].get('docid','')):
return c(*args,**kwargs)
if 'oid' in kwargs:
dt=Docs.find_one({"_id": kwargs['oid']},['type'])['type']
for (t,c,r) in DOCTYPES:
if dt == t:
return c(*args,**kwargs)
return DOC(*args,**kwargs)
def Doc(*args, **kwargs):
if 'docid' in kwargs:
for (t,c,r) in DOCTYPES:
if re.match(r,kwargs['docid']):
return c(*args,**kwargs)
if 'd' in kwargs:
for (t,c,r) in DOCTYPES:
if kwargs['d'].get('type','') == t or re.match(r,kwargs['d'].get('docid','')):
return c(*args,**kwargs)
if 'oid' in kwargs:
dt=Docs.find_one({"_id": kwargs['oid']},['type'])['type']
for (t,c,r) in DOCTYPES:
if dt == t:
return c(*args,**kwargs)
return DOC(*args,**kwargs)
def pippies(request):
filtr = {}
template_vars = {}
docfilter = None
relfilter = None
cutoff = None
try:
cutoff = int(cgi.escape(request.GET.get('cutoff', '7')))
except:
pass
if cutoff: filtr['len'] = {'$gte': cutoff}
try:
docfilter = ObjectId(cgi.escape(request.GET.get('doc', '')))
except:
pass
if docfilter:
filtr['docs'] = docfilter
try:
relfilter = int(cgi.escape(request.GET.get('relevance', '')))
except:
pass
if relfilter: filtr['relevance'] = relfilter
# todo add sortable column headers ala http://djangosnippets.org/snippets/308/
orderBy = cgi.escape(request.GET.get('orderby', 'relevance'))
orderDesc = True if '1' == cgi.escape(request.GET.get('desc',
'1')) else False
template_vars = pager(request, Pippies.find(filtr), orderBy, orderDesc)
template_vars['pippies'] = [{
'id':
pippi['_id'],
'pippi':
' '.join([p if p else '*' for p in pippi['pippi'].split(' ')]),
'docslen':
len(pippi['docs']),
'relevance':
pippi.get('relevance', 0),
} for pippi in template_vars['data']]
template_vars['doc'] = docfilter
if docfilter:
doc = Docs.find_one({'_id': docfilter}, ['docid', 'title'])
template_vars[
'docTitle'] = doc['title'] if 'title' in doc else doc['docid']
return render_to_response('pippies.html',
template_vars,
context_instance=RequestContext(request))
def __init__(self, docid=None, *args, **kwargs):
self.__dict__['type'] = 'co-ment'
if docid:
hostValidator = CMTRE.search(docid)
if hostValidator:
if hostValidator.group(1) or hostValidator.group(
3) or hostValidator.group(5):
docid = ("%s%s" % (hostValidator.group(2),
hostValidator.group(4))).encode('utf8')
kwargs['docid'] = docid
url = "https://%s/text/%s/view/" % (hostValidator.group(2),
hostValidator.group(4))
if not Docs.find_one({"docid": docid}):
context = urllib2.urlopen(url).read()
soup = BeautifulSoup(context)
self.__dict__['title'] = unescape(
unicode(''.join(soup.title.findAll(
text=True)))).strip().encode('utf8')
dataurl = "https://%s/text%s/comments/" % (
hostValidator.group(2), hostValidator.group(4))
data = urllib2.urlopen(dataurl).read()
soup = BeautifulSoup(data)
kwargs[
'raw'] = '<html><head><title>%s</title><meta http-equiv="content-type" content="text/html; charset=utf-8" /></head><body>%s</body></html>' % (
self.title,
unescape(
unicode(
soup.find(attrs={'id': 'textcontainer'
}))).encode('utf8'))
kwargs['docid'] = docid
super(Coment, self).__init__(*args, **kwargs)
if not 'stems' in self.__dict__ or not self.stems:
# let's calculate and cache the results
models.tfidf.add_input_document(self.termcnt.keys())
self.save()
return
kwargs['docid'] = docid
super(Coment, self).__init__(*args, **kwargs)
def pippies(request):
filtr={}
template_vars={}
docfilter=None
relfilter=None
cutoff=None
try:
cutoff = int(cgi.escape(request.GET.get('cutoff','7')))
except:
pass
if cutoff: filtr['len']={ '$gte': cutoff }
try:
docfilter = ObjectId(cgi.escape(request.GET.get('doc','')))
except:
pass
if docfilter:
filtr['docs']=docfilter
try:
relfilter = int(cgi.escape(request.GET.get('relevance','')))
except:
pass
if relfilter: filtr['relevance']=relfilter
# todo add sortable column headers ala http://djangosnippets.org/snippets/308/
orderBy = cgi.escape(request.GET.get('orderby','relevance'))
orderDesc = True if '1'==cgi.escape(request.GET.get('desc','1')) else False
template_vars=pager(request,Pippies.find(filtr),orderBy,orderDesc)
template_vars['pippies']=[{'id': pippi['_id'],
'pippi': ' '.join([p if p else '*' for p in pippi['pippi'].split(' ')]),
'docslen':len(pippi['docs']),
'relevance':pippi.get('relevance',0),}
for pippi in template_vars['data']]
template_vars['doc']=docfilter
if docfilter:
doc=Docs.find_one({'_id': docfilter},['docid', 'title'])
template_vars['docTitle']=doc['title'] if 'title' in doc else doc['docid']
return render_to_response('pippies.html', template_vars, context_instance=RequestContext(request))
def frags(request):
filtr = {}
template_vars = {}
docfilter = None
cutoff = None
pippifilter = None
try:
docfilter = ObjectId(cgi.escape(request.GET.get('doc', '')))
except:
pass
if docfilter:
filtr['doc'] = docfilter
try:
pippifilter = ObjectId(cgi.escape(request.GET.get('pippi', '')))
except:
pass
if pippifilter:
filtr['pippi'] = pippifilter
else:
try:
cutoff = int(cgi.escape(request.GET.get('cutoff', '7')))
except:
pass
if cutoff: filtr['l'] = {'$gte': cutoff}
orderBy = 'l'
orderDesc = True
template_vars = pager(request, Frags.find(filtr), orderBy, orderDesc)
prevDoc = None
template_vars['frags'] = []
for frag in template_vars['data']:
p = Pippi('', oid=frag['pippi'])
d = Doc(oid=frag['doc'])
if pippifilter:
frag['txt'] = diffFrag(prevDoc, frag['txt'])
prevDoc = frag['txt']
template_vars['frags'].append({
'_id':
frag['_id'],
'pos':
frag['pos'],
'txt':
" ".join(frag['txt']),
'len':
frag['l'],
'score':
sum([d.tfidf.get(t, 0) for t in p.pippi]),
'pippi':
p,
'doc':
d,
})
template_vars['pippi'] = pippifilter
template_vars['doc'] = docfilter
if docfilter:
template_vars['docTitle'] = Docs.find_one({'_id': docfilter},
['docid'])['docid']
if pippifilter:
template_vars[
'pippiFilter'] = 1 #" ".join(Pippies.find_one({'_id': pippifilter},['pippi'])['pippi'])
return render_to_response('frags.html',
template_vars,
context_instance=RequestContext(request))
