' or not self.out[-1]:
if self.params['skip-empty-lines']:
self.out.pop()
else:
self.out[-1] = "
') self.mark_end_tag('p') def mark_start_tag(self, tag, attrs=None): ''' Remember open tag and put it to output ''' try: flag = self._TAGS[tag] except: flag = 0 if tag in self.nstack[0]: self.mark_end_tag(tag) self.nstack[0].append(tag) self.nstack[1].append(attrs) if flag & _TAG_INP: self.data += self.tag_repr(tag, attrs) else: self.out.append(self.tag_repr(tag, attrs)) def mark_end_tag(self, tag): ''' Close corresponding tags. If tag is not last tag was outed, close all previously opened tags. I.e. text -> text ''' if tag not in self.nstack[0]: return while self.nstack[0]: v = self.nstack[0].pop() a = self.nstack[1].pop() try: flag = self._TAGS[v] except: flag = 0 if flag & _TAG_INP: et=self.tag_repr(v,a) if self.data.rstrip().endswith(et): self.data=self.data.rstrip()[:-len(et)] if v=='a': try: self.ids[a[0][1]][1]-=1 except: pass else: self.data += "" % v else: self.process_data() if self.out[-1]=="<%s>" % v: ## duplicate tag detected, often from replacing embedded
inside ... ## FIXME not really sure if this is 100% appropriate, is this ONLY for missing target links? self.msg("!!!!\n") self.msg('DEBUG** ' + repr(v) + ' --- ' + repr(self.out[-10:]) + ' ' + repr(v), 1) self.out.pop() else: self.out.append("" % v) if tag == v: break def process_data(self): ''' Handle accomulated data when close paragraph. ''' if not self.data.strip(): return if not self.nofill: self.data=_CH_FLOW+self.data.strip() self.out.append(self.data) else: self.data = self.process_pre(self.data) self.data = self.detect_headers(self.data) try: if self.data[0]=='
' and self.out[-1]=='': self.out.pop() self.data = self.data[1:] msg("WoW! I must be impossible!!!") except IndexError: pass self.out.extend(self.data) self.data = '' # --- Parsed data processing methods def pre_process(self, data): data=data.replace(u'\x0e',u'
and
']
while i < len(data)-1:
empty0 = not data[i]
try:
empty1 = not data[i+1]
empty2 = not data[i+2]
empty3 = not data[i+3]
except IndexError:
empty1 = empty2 = empty3 = False
if empty0 and empty1 and not empty2 and empty3:
res.append(data[pstart:i])
header[5]=_CH_FLOW + data[i+1].strip()
res.extend(header)
i+=2
pstart = i+2
else:
istitle = (
empty0 and
not empty1 and
empty2 and
(
empty3 or
data[i+1].startswith(' '*8) or
data[i+1].isupper() or
(
data[i+1].lstrip()[0] not in _DIAL_START and
data[i+1][-1] not in _SENTENCE_FIN
) or
data[i+1].lstrip()[0] in _HEAD_CHARS or
self.is_roman_number(data[i+1])
)
)
istitle = istitle or \
not empty1 and \
self.header_re and \
self.header_re.match(data[i+1])
if istitle:
res.append(data[pstart:i])
header[5]=_CH_FLOW + data[i+1].strip()
res.extend(header)
i+=1
while i < len(data)-1 and not data[i+1]:
i+=1
pstart = i+1
i+=1
if pstart < len(data):
res.append(data[pstart:])
return res
def detect_epigraphs(self):
'''
Detect epigraphs (in plain text)
'''
if not self.params['detect-epigraphs']:
return
sect_found = 0
i = 0
while i < len(self.out):
if type(self.out[i]) != ListType:
if self.out[i] == ' ',raw[j:eep-author]])
if author and self.clean_str(raw[eep-1]).lstrip()[0].isupper():
res.extend(['
': istart-=1 if j < len(raw)-1: res.extend(['
',raw[j:]]) elif i < len(self.out)-1 and self.out[i+1] == '
': iend+=1 self.out[istart:iend] = res i = istart + len(res)-1 sect_found = 0 i += 1 def detect_verses(self): ''' Detect verses in plain text ''' if not self.params['detect-verses']: return i=0 while i < len(self.out): if type(self.out[i]) == ListType: res=[] raw=[self.clean_str(x).rstrip() for x in self.out[i]] lraw=len(raw) pfound = jstart = j = 0 while j < lraw-3: if raw[j] and len(raw[j]) < 60 and \ raw[j+1] and len(raw[j+1]) < 80 and \ raw[j+2] and len(raw[j+2]) < 80 and \ raw[j+3] and len(raw[j+3]) < 80: fl = len(raw[j]) k = j while k < lraw: if raw[k].strip() and ( abs(len(raw[k])-fl) > 15 or \ raw[k].lstrip()[0] in _DIAL_START ): break k += 1 if k - j > 3: pfound += 1 if jstart: res.append('') if jstart != j: res.extend([self.out[i][jstart:j],'
']) res.extend(['',self.out[i][jstart:]])
istart = i
iend = i+1
try:
if res[0] == ' ':
istart -= 1
except:
pass
try:
if res[-1] == '
',_CH_FLOW + '\n'.join(raw[jstart:j+1]),'
']) #res.extend(['',_CH_FLOW + ''.join(raw[jstart:j+1]),'
']) jstart = j+1 j+=1 if pfound > 0: self.out[i:i+1]=res[1:-1] i+=len(res)-2 else: self.out[i]='\n'.join(raw).lstrip() i+=1 def detect_italic(self, text, arg): signs='_.,!?:' istart=-1 res='' while True: istart = text.find('_') if istart >= 0: iend = sys.maxint for i in signs: try: iend=min(iend, text.index(i, istart+1)) except: pass if iend == sys.maxint: iend=0 if text[istart+1:iend or None]: emp = 'andtag or plain text file) Determine text format. ''' data = [x.rstrip() for x in data.splitlines()] if type(self.asline) == TupleType: count,G80,L80 = self.asline for i in data: if len(i) > 80: G80+=1 else: L80+=1 count+=1 if count > 2000: self.asline = G80 > L80 break if type(self.asline) == TupleType: self.asline=(count, G80, L80) return data def process_paragraph(self, paragraph, id): ''' Process paragraph. Find id, normalize quotes. ''' paragraph=paragraph.strip() startp = paragraph.find('