def fixStartChars(self, paragraph, num):
lines = paragraph.getchildren()
# the first 'non large' word on the page
# has the x value which is the same start
# value as the lines affected by the large start letter
first = lines[0]
tokens = first.getchildren()
if (num >= len(tokens)):
return False
first_normal = tokens[num]
first_info = common.tokenExtract(first_normal)
first_x = first_info['x']
# figure out indent by looking for the first line
# with a different indent
indent = None
for i in xrange(1, len(lines)):
line_info = common.lineExtract(lines[i])
if (line_info['left'] != first_x):
indent = line_info['left']
break
# if we found an indent fix line indents
if (indent != None):
lines[0].set('left', unicode(indent))
for i in xrange(1, len(lines)):
line_info = common.lineExtract(lines[i])
if (line_info['left'] == first_x):
lines[i].set('left', unicode(indent))
return True
else:
return False
def apply(self, page, c):
output = common.copyElementAttributes(page)
line_info = None
for text in page.iter('TEXT'):
for token in text.iter('TOKEN'):
token_info = common.tokenExtract(token)
if (line_info == None):
line_info = token_info
line = etree.Element('LINE')
if (common.looseCompare(token_info['y'], line_info['y'],
c['y_height_diff'])):
# same line
line.append(token)
else:
# new line
output.append(line)
# reset
line = etree.Element('LINE')
line.append(token)
line_info = token_info
# handle last line
if (line_info != None):
output.append(line)
for line in output.iter('LINE'):
self.lineSummary(line)
return output
def apply(self,page,c):
output = common.copyElementAttributes(page)
line_info = None
for text in page.iter('TEXT'):
for token in text.iter('TOKEN'):
token_info = common.tokenExtract(token)
if (line_info == None):
line_info = token_info
line = etree.Element('LINE')
if (common.looseCompare(token_info['y'],line_info['y'],c['y_height_diff'])):
# same line
line.append(token)
else:
# new line
output.append(line)
# reset
line = etree.Element('LINE')
line.append(token)
line_info = token_info
# handle last line
if (line_info != None):
output.append(line)
for line in output.iter('LINE'):
self.lineSummary(line)
return output
def fixStartChars(self, paragraph, num):
lines = paragraph.getchildren()
# the first 'non large' word on the page
# has the x value which is the same start
# value as the lines affected by the large start letter
first = lines[0]
tokens = first.getchildren()
if (num >= len(tokens)):
return False
first_normal = tokens[num]
first_info = common.tokenExtract(first_normal)
first_x = first_info['x']
# figure out indent by looking for the first line
# with a different indent
indent = None
for i in xrange(1, len(lines)):
line_info = common.lineExtract(lines[i])
if (line_info['left'] != first_x):
indent = line_info['left']
break
# if we found an indent fix line indents
if (indent != None):
lines[0].set('left', unicode(indent))
for i in xrange(1, len(lines)):
line_info = common.lineExtract(lines[i])
if (line_info['left'] == first_x):
lines[i].set('left', unicode(indent))
return True
else:
return False
def numberOfStartChars(self, paragraph):
lines = paragraph.getchildren()
count = 0
if (len(lines) > 1):
for token in lines[0].iter('TOKEN'):
info = common.tokenExtract(token)
if (info['chars'] == 1):
# single character
count = count + 1
else:
break
return count
def apply(self, token, c):
info = common.tokenExtract(token)
sz = info['font-size']
s = "unknown"
if (sz < self.smallest):
s = 'xx-small'
elif (sz > self.largest):
s = 'xx-large'
else:
v = bisect.bisect_right(self.sizes, sz)
s = self.table[self.sizes[v]]
token.set('size', unicode(s))
def numberOfStartChars(self, paragraph):
lines = paragraph.getchildren()
count = 0
if (len(lines) > 1):
for token in lines[0].iter('TOKEN'):
info = common.tokenExtract(token)
if (info['chars'] == 1):
# single character
count = count + 1
else:
break
return count
def apply(self, token, c):
info = common.tokenExtract(token)
sz = info['font-size']
s = "unknown"
if (sz < self.smallest):
s = 'xx-small'
elif (sz > self.largest):
s = 'xx-large'
else:
v = bisect.bisect_right(self.sizes, sz)
s = self.table[self.sizes[v]]
token.set('size', unicode(s))
def lineSummary(self, line):
base = []
top = []
height = []
left = None
right = None
chars = 0
for token in line.iter('TOKEN'):
info = common.tokenExtract(token)
base.append(info['base'])
top.append(info['top'])
height.append(info['height'])
if (left == None):
left = info['left']
else:
if (info['left'] < left):
left = info['left']
if (right == None):
right = info['right']
else:
if (info['right'] > right):
right = info['right']
chars = chars + info['chars']
# apply summary
if (len(base) <= 2):
line.set('base', unicode(common.largest(base)))
else:
line.set('base', unicode(common.mostCommon(base)))
if (len(top) <= 2):
line.set('top', unicode(common.smallest(top)))
else:
line.set('top', unicode(common.mostCommon(top)))
line.set('left', unicode(left))
line.set('right', unicode(right))
if (len(height) <= 2):
line.set('height', unicode(common.largest(height)))
else:
line.set('height', unicode(common.mostCommon(height)))
line.set('chars', unicode(chars))
def lineSummary(self, line):
base = []
top = []
height = []
left = None
right = None
chars = 0
for token in line.iter('TOKEN'):
info = common.tokenExtract(token)
base.append(info['base'])
top.append(info['top'])
height.append(info['height'])
if (left == None):
left = info['left']
else:
if (info['left'] < left):
left = info['left']
if (right == None):
right = info['right']
else:
if (info['right'] > right):
right = info['right']
chars = chars + info['chars']
# apply summary
if (len(base) <= 2):
line.set('base', unicode(common.largest(base)))
else:
line.set('base', unicode(common.mostCommon(base)))
if (len(top) <= 2):
line.set('top', unicode(common.smallest(top)))
else:
line.set('top', unicode(common.mostCommon(top)))
line.set('left', unicode(left))
line.set('right', unicode(right))
if (len(height) <= 2):
line.set('height', unicode(common.largest(height)))
else:
line.set('height', unicode(common.mostCommon(height)))
line.set('chars', unicode(chars))