def lineDiff(self,lines):
values = []
for i in xrange(0, len(lines) - 1):
a = lines[i]
b = lines[i + 1]
a_info = common.lineExtract(a)
b_info = common.lineExtract(b)
values.append(b_info['base'] - a_info['base'])
return common.mostCommon(values)
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 split(self, paragraph, c):
output = []
lines = paragraph.getchildren()
common_diff = self.lineDiff(lines)
tmp = etree.Element('PARAGRAPH')
if (len(lines) > 0):
tmp.append(lines[0])
for i in xrange(0, len(lines) - 1):
a = lines[i]
b = lines[i + 1]
a_values = common.lineExtract(a)
b_values = common.lineExtract(b)
diff = b_values['base'] - a_values['base']
if (common.looseCompare(common_diff, diff, c['vertical_diff'])):
# same
tmp.append(b)
else:
# split
output.append(tmp)
tmp = etree.Element('PARAGRAPH')
tmp.append(b)
output.append(tmp)
return output
def apply(self, page, c):
output = common.copyElementAttributes(page)
paragraph = etree.Element('PARAGRAPH')
lines = page.getchildren()
if (len(lines) > 1):
paragraph.append(lines[0])
for i in xrange(0, len(lines) - 1):
current = lines[i]
nxt = lines[i + 1]
c_info = common.lineExtract(current)
n_info = common.lineExtract(nxt)
if (common.looseCompare(c_info['height'], n_info['height'],
c['line_height_diff'])):
# same para
paragraph.append(nxt)
else:
# new para
output.append(paragraph)
paragraph = etree.Element('PARAGRAPH')
paragraph.append(nxt)
# add final paragraph
output.append(paragraph)
return output
def apply(self, page, c):
output = common.copyElementAttributes(page)
paragraph = etree.Element('PARAGRAPH')
lines = page.getchildren()
if (len(lines) > 1):
paragraph.append(lines[0])
for i in xrange(0, len(lines) - 1):
current = lines[i]
nxt = lines[i + 1]
c_info = common.lineExtract(current)
n_info = common.lineExtract(nxt)
if (common.looseCompare(c_info['height'],n_info['height'],c['line_height_diff'])):
# same para
paragraph.append(nxt)
else:
# new para
output.append(paragraph)
paragraph = etree.Element('PARAGRAPH')
paragraph.append(nxt)
# add final paragraph
output.append(paragraph)
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 split(self, paragraph, c):
output = []
lines = paragraph.getchildren()
common_diff = self.lineDiff(lines)
tmp = etree.Element('PARAGRAPH')
if (len(lines) > 0):
tmp.append(lines[0])
for i in xrange(0,len(lines) - 1):
a = lines[i]
b = lines[i + 1]
a_values = common.lineExtract(a)
b_values = common.lineExtract(b)
diff = b_values['base'] - a_values['base']
if (common.looseCompare(common_diff,diff,c['vertical_diff'])):
# same
tmp.append(b)
else:
# split
output.append(tmp)
tmp = etree.Element('PARAGRAPH')
tmp.append(b)
output.append(tmp)
return output
def lineDiff(self, lines):
values = []
for i in xrange(0, len(lines) - 1):
a = lines[i]
b = lines[i + 1]
a_info = common.lineExtract(a)
b_info = common.lineExtract(b)
values.append(b_info['base'] - a_info['base'])
return common.mostCommon(values)
def split(self, paragraph, c):
lines = paragraph.getchildren()
indent = self.estimateIndent(paragraph, c)
justify = self.estimateJustify(paragraph, c)
output = []
tmp = etree.Element('PARAGRAPH')
info = None
for i in xrange(0, len(lines)):
info = common.lineExtract(lines[i])
if (common.looseCompare(info['left'], indent,
c['para_indent_diff'])):
# same para
tmp.append(lines[i])
elif (i == 0):
# indented first line
tmp.append(lines[i])
else:
# new para
tmp.set('complete', 'yes')
output.append(tmp)
tmp = etree.Element('PARAGRAPH')
tmp.append(lines[i])
if (info != None):
if (common.looseCompare(info['right'], justify,
c['para_indent_diff'])):
if (len(tmp) == 1):
tmp.set('complete', 'yes')
else:
tmp.set('complete', 'no')
else:
tmp.set('complete', 'yes')
output.append(tmp)
return output
def split(self, paragraph, c):
lines = paragraph.getchildren()
indent = self.estimateIndent(paragraph, c)
justify = self.estimateJustify(paragraph, c)
output = []
tmp = etree.Element('PARAGRAPH')
info = None
for i in xrange(0,len(lines)):
info = common.lineExtract(lines[i])
if (common.looseCompare(info['left'],indent,c['para_indent_diff'])):
# same para
tmp.append(lines[i])
elif (i == 0):
# indented first line
tmp.append(lines[i])
else:
# new para
tmp.set('complete', 'yes')
output.append(tmp)
tmp = etree.Element('PARAGRAPH')
tmp.append(lines[i])
if (info != None):
if (common.looseCompare(info['right'],justify,c['para_indent_diff'])):
if (len(tmp) == 1):
tmp.set('complete', 'yes')
else:
tmp.set('complete', 'no')
else:
tmp.set('complete', 'yes')
output.append(tmp)
return output
def estimateJustify(self, paragraph, c):
lines = paragraph.getchildren()
right = []
for i in xrange(0, len(lines)):
info = common.lineExtract(lines[i])
right.append(info['right'])
if (len(right) <= 2):
indent = common.largest(right)
else:
indent = common.mostCommon(right)
return indent
def estimateJustify(self, paragraph, c):
lines = paragraph.getchildren()
right = []
for i in xrange(0,len(lines)):
info = common.lineExtract(lines[i])
right.append(info['right'])
if (len(right) <= 2):
indent = common.largest(right)
else:
indent = common.mostCommon(right)
return indent
def paragraphSummary(self, para):
base = None
top = None
left = None
right = None
chars = 0
lines = 0
height = []
for line in para.iter('LINE'):
info = common.lineExtract(line)
if (base == None):
base = info['base']
else:
if (info['base'] > base):
base = info['base']
if (top == None):
top = info['top']
else:
if (info['top'] < top):
top = info['top']
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']
height.append(info['height'])
lines += 1
# apply summary
para.set('base', unicode(base))
para.set('top', unicode(top))
para.set('left', unicode(left))
para.set('right', unicode(right))
para.set('chars', unicode(chars))
para.set('height', unicode(common.mostCommon(height)))
para.set('lines', unicode(lines))
def paragraphSummary(self, para):
base = None
top = None
left = None
right = None
chars = 0
lines = 0
height = []
for line in para.iter('LINE'):
info = common.lineExtract(line)
if (base == None):
base = info['base']
else:
if (info['base'] > base):
base = info['base']
if (top == None):
top = info['top']
else:
if (info['top'] < top):
top = info['top']
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']
height.append(info['height'])
lines += 1
# apply summary
para.set('base', unicode(base))
para.set('top', unicode(top))
para.set('left', unicode(left))
para.set('right', unicode(right))
para.set('chars', unicode(chars))
para.set('height', unicode(common.mostCommon(height)))
para.set('lines', unicode(lines))