Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Conversion pipeline: If <body> tag is not under <html> move it to the correct place. LIT Input: Strip embedded <metadata> and <guide> elements. Fixes #4712 (Unable to convert .rtf and .lit files to .EPUB)

Browse Source
This commit is contained in:
Kovid Goyal 2010-01-28 09:43:14 -07:00
parent 68beb72bbd
commit 5e93ea1da2
4 changed files with 166 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/calibre/ebooks/lit/input.py
View File

@ -26,6 +26,11 @@ class LITInput(InputFormatPlugin):
for item in oeb.spine: for item in oeb.spine:
root = item.data root = item.data
if not hasattr(root, 'xpath'): continue if not hasattr(root, 'xpath'): continue
for bad in ('metadata', 'guide'):
metadata = XPath('//h:'+bad)(root)
if metadata:
for x in metadata:
x.getparent().remove(x)
body = XPath('//h:body')(root) body = XPath('//h:body')(root)
if body: if body:
body = body[0] body = body[0]

6
src/calibre/ebooks/oeb/base.py
View File

@ -909,6 +909,12 @@ class Manifest(object):
'content': '%s; charset=utf-8' % XHTML_NS}) 'content': '%s; charset=utf-8' % XHTML_NS})
# Ensure has a <body/> # Ensure has a <body/>
if not xpath(data, '/h:html/h:body'): if not xpath(data, '/h:html/h:body'):
body = xpath(data, '//h:body')
if body:
body = body[0]
body.getparent().remove(body)
data.append(body)
else:
self.oeb.logger.warn( self.oeb.logger.warn(
'File %r missing <body/> element' % self.href) 'File %r missing <body/> element' % self.href)
etree.SubElement(data, XHTML('body')) etree.SubElement(data, XHTML('body'))

158
src/calibre/ebooks/pdf/reflow.py
View File

@ -43,6 +43,10 @@ class Image(Element):
self.bottom = self.top + self.height self.bottom = self.top + self.height
self.right = self.left + self.width self.right = self.left + self.width
def to_html(self):
return '<img src="%s" width="%dpx" height="%dpx"/>' % \
(self.src, int(self.width), int(self.height))
class Text(Element): class Text(Element):
@ -66,8 +70,6 @@ class Text(Element):
self.raw = text.text if text.text else u'' self.raw = text.text if text.text else u''
for x in text.iterchildren(): for x in text.iterchildren():
self.raw += etree.tostring(x, method='xml', encoding=unicode) self.raw += etree.tostring(x, method='xml', encoding=unicode)
if x.tail:
self.raw += x.tail
self.average_character_width = self.width/len(self.text_as_string) self.average_character_width = self.width/len(self.text_as_string)
def coalesce(self, other, page_number): def coalesce(self, other, page_number):
@ -86,6 +88,9 @@ class Text(Element):
self.average_character_width = (self.average_character_width + self.average_character_width = (self.average_character_width +
other.average_character_width)/2.0 other.average_character_width)/2.0
def to_html(self):
return self.raw
class FontSizeStats(dict): class FontSizeStats(dict):
def __init__(self, stats): def __init__(self, stats):
@ -108,6 +113,11 @@ class Interval(object):
right = min(self.right, other.right) right = min(self.right, other.right)
return Interval(left, right) return Interval(left, right)
def centered_in(self, parent):
left = abs(self.left - parent.left)
right = abs(self.right - parent.right)
return abs(left-right) < 3
def __nonzero__(self): def __nonzero__(self):
return self.width > 0 return self.width > 0
@ -146,6 +156,9 @@ class Column(object):
for x in self.elements: for x in self.elements:
yield x yield x
def __len__(self):
return len(self.elements)
def contains(self, elem): def contains(self, elem):
return elem.left > self.left - self.HFUZZ*self.width and \ return elem.left > self.left - self.HFUZZ*self.width and \
elem.right < self.right + self.HFUZZ*self.width elem.right < self.right + self.HFUZZ*self.width
@ -174,17 +187,42 @@ class Column(object):
class Box(list): class Box(list):
def __init__(self, type='p'): def __init__(self, type='p'):
self.type = type self.tag = type
def to_html(self):
ans = ['<%s>'%self.tag]
for elem in self:
if isinstance(elem, int):
ans.append('<a name="page_%d"/>'%elem)
else:
ans.append(elem.to_html()+' ')
ans.append('</%s>'%self.tag)
return ans
class ImageBox(Box): class ImageBox(Box):
def __init__(self, img): def __init__(self, img):
Box.__init__(self, type='img') Box.__init__(self)
self.img = img self.img = img
def to_html(self):
ans = ['<div style="text-align:center">']
ans.append(self.img.to_html())
if len(self) > 0:
ans.append('<br/>')
for elem in self:
if isinstance(elem, int):
ans.append('<a name="page_%d"/>'%elem)
else:
ans.append(elem.to_html()+' ')
ans.append('</div>')
return ans
class Region(object): class Region(object):
def __init__(self): def __init__(self, opts, log):
self.opts, self.log = opts, log
self.columns = [] self.columns = []
self.top = self.bottom = self.left = self.right = self.width = self.height = 0 self.top = self.bottom = self.left = self.right = self.width = self.height = 0
@ -217,6 +255,40 @@ class Region(object):
def is_empty(self): def is_empty(self):
return len(self.columns) == 0 return len(self.columns) == 0
@property
def is_small(self):
max_lines = 0
for c in self.columns:
max_lines = max(max_lines, len(c))
return max_lines > 2
def absorb(self, singleton):
def most_suitable_column(elem):
mc, mw = None, 0
for c in self.columns:
i = Interval(c.left, c.right)
e = Interval(elem.left, elem.right)
w = i.intersection(e).width
if w > mw:
mc, mw = c, w
if mc is None:
self.log.warn('No suitable column for singleton',
elem.to_html())
mc = self.columns[0]
return mc
print
for c in singleton.columns:
for elem in c:
col = most_suitable_column(elem)
if self.opts.verbose > 3:
idx = self.columns.index(col)
self.log.debug(u'Absorbing singleton %s into column'%elem.to_html(),
idx)
col.add(elem)
def collect_stats(self): def collect_stats(self):
for column in self.columns: for column in self.columns:
column.collect_stats() column.collect_stats()
@ -231,7 +303,6 @@ class Region(object):
self.elements = [] self.elements = []
for x in self.columns: for x in self.columns:
self.elements.extend(x) self.elements.extend(x)
self.boxes = [Box()] self.boxes = [Box()]
for i, elem in enumerate(self.elements): for i, elem in enumerate(self.elements):
if isinstance(elem, Image): if isinstance(elem, Image):
@ -341,7 +412,7 @@ class Page(object):
return return
for i, x in enumerate(self.elements): for i, x in enumerate(self.elements):
x.idx = i x.idx = i
current_region = Region() current_region = Region(self.opts, self.log)
processed = set([]) processed = set([])
for x in self.elements: for x in self.elements:
if x in processed: continue if x in processed: continue
@ -350,12 +421,42 @@ class Page(object):
processed.update(elems) processed.update(elems)
if not current_region.contains(columns): if not current_region.contains(columns):
self.regions.append(current_region) self.regions.append(current_region)
current_region = Region() current_region = Region(self.opts, self.log)
current_region.add(columns) current_region.add(columns)
if not current_region.is_empty: if not current_region.is_empty:
self.regions.append(current_region) self.regions.append(current_region)
self.coalesce_regions()
def coalesce_regions(self):
# find contiguous sets of small regions
# absorb into a neighboring region (prefer the one with number of cols
# closer to the avg number of cols in the set, if equal use large
# region)
# merge contiguous regions that can contain each other
absorbed = set([])
found = True
while found:
found = False
for i, region in enumerate(self.regions):
if region.is_small:
found = True
regions = []
for j in range(i+1, len(self.regions)):
if self.regions[j].is_small:
regions.append(self.regions[j])
else:
break
prev = None if i == 0 else i-1
next = j if self.regions[j] not in regions else None
def sort_into_columns(self, elem, neighbors): def sort_into_columns(self, elem, neighbors):
neighbors.add(elem)
neighbors = sorted(neighbors, cmp=lambda x,y:cmp(x.left, y.left))
if self.opts.verbose > 3:
self.log.debug('Neighbors:', [x.to_html() for x in neighbors])
columns = [Column()] columns = [Column()]
columns[0].add(elem) columns[0].add(elem)
for x in neighbors: for x in neighbors:
@ -421,6 +522,9 @@ class PDFDocument(object):
page.first_pass() page.first_pass()
page.second_pass() page.second_pass()
self.linearize()
self.render()
def collect_font_statistics(self): def collect_font_statistics(self):
self.font_size_stats = {} self.font_size_stats = {}
for p in self.pages: for p in self.pages:
@ -432,5 +536,43 @@ class PDFDocument(object):
self.font_size_stats = FontSizeStats(self.font_size_stats) self.font_size_stats = FontSizeStats(self.font_size_stats)
def linearize(self):
self.elements = []
last_region = last_block = None
for page in self.pages:
page_number_inserted = False
for region in page.regions:
merge_first_block = last_region is not None and \
len(last_region.columns) == len(region.columns) and \
not hasattr(last_block, 'img')
for i, block in enumerate(region.boxes):
if merge_first_block:
merge_first_block = False
if not page_number_inserted:
last_block.append(page.number)
page_number_inserted = True
for elem in block:
last_block.append(elem)
else:
if not page_number_inserted:
block.insert(0, page.number)
page_number_inserted = True
self.elements.append(block)
last_block = block
last_region = region
def render(self):
html = ['<?xml version="1.0" encoding="UTF-8"?>',
'<html xmlns="http://www.w3.org/1999/xhtml">', '<head>',
'<title>PDF Reflow conversion</title>', '</head>', '<body>',
'<div>']
for elem in self.elements:
html.extend(elem.to_html())
html += ['</body>', '</html>']
with open('index.html', 'wb') as f:
f.write((u'\n'.join(html)).encode('utf-8'))

4
src/calibre/ebooks/rtf/input.py
View File

@ -195,9 +195,9 @@ class RTFInput(InputFormatPlugin):
fname = self.preprocess(stream.name) fname = self.preprocess(stream.name)
try: try:
xml = self.generate_xml(fname) xml = self.generate_xml(fname)
except RtfInvalidCodeException: except RtfInvalidCodeException, e:
raise ValueError(_('This RTF file has a feature calibre does not ' raise ValueError(_('This RTF file has a feature calibre does not '
'support. Convert it to HTML first and then try it.')) 'support. Convert it to HTML first and then try it.\n%s')%e)
d = glob.glob(os.path.join('*_rtf_pict_dir', 'picts.rtf')) d = glob.glob(os.path.join('*_rtf_pict_dir', 'picts.rtf'))
if d: if d:
imap = {} imap = {}