Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-08 10:44:09 -04:00
Code Issues Packages Projects Releases Wiki Activity

Speed up import of builtin input plugins

Browse Source
This commit is contained in:
Kovid Goyal 2012-02-05 22:37:07 +05:30
parent a9cedb0a0b
commit 4b9d1e40fc
28 changed files with 1042 additions and 944 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

42
src/calibre/customize/builtins.py
View File

@ -501,27 +501,27 @@ class TXTZMetadataWriter(MetadataWriterPlugin):
# }}}
from calibre.ebooks.comic.input import ComicInput
from calibre.ebooks.djvu.input import DJVUInput
from calibre.ebooks.epub.input import EPUBInput
from calibre.ebooks.fb2.input import FB2Input
from calibre.ebooks.html.input import HTMLInput
from calibre.ebooks.htmlz.input import HTMLZInput
from calibre.ebooks.lit.input import LITInput
from calibre.ebooks.mobi.input import MOBIInput
from calibre.ebooks.odt.input import ODTInput
from calibre.ebooks.pdb.input import PDBInput
from calibre.ebooks.azw4.input import AZW4Input
from calibre.ebooks.pdf.input import PDFInput
from calibre.ebooks.pml.input import PMLInput
from calibre.ebooks.rb.input import RBInput
from calibre.web.feeds.input import RecipeInput
from calibre.ebooks.rtf.input import RTFInput
from calibre.ebooks.tcr.input import TCRInput
from calibre.ebooks.txt.input import TXTInput
from calibre.ebooks.lrf.input import LRFInput
from calibre.ebooks.chm.input import CHMInput
from calibre.ebooks.snb.input import SNBInput
from calibre.ebooks.conversion.plugins.comic_input import ComicInput
from calibre.ebooks.conversion.plugins.djvu_input import DJVUInput
from calibre.ebooks.conversion.plugins.epub_input import EPUBInput
from calibre.ebooks.conversion.plugins.fb2_input import FB2Input
from calibre.ebooks.conversion.plugins.html_input import HTMLInput
from calibre.ebooks.conversion.plugins.htmlz_input import HTMLZInput
from calibre.ebooks.conversion.plugins.lit_input import LITInput
from calibre.ebooks.conversion.plugins.mobi_input import MOBIInput
from calibre.ebooks.conversion.plugins.odt_input import ODTInput
from calibre.ebooks.conversion.plugins.pdb_input import PDBInput
from calibre.ebooks.conversion.plugins.azw4_input import AZW4Input
from calibre.ebooks.conversion.plugins.pdf_input import PDFInput
from calibre.ebooks.conversion.plugins.pml_input import PMLInput
from calibre.ebooks.conversion.plugins.rb_input import RBInput
from calibre.ebooks.conversion.plugins.recipe_input import RecipeInput
from calibre.ebooks.conversion.plugins.rtf_input import RTFInput
from calibre.ebooks.conversion.plugins.tcr_input import TCRInput
from calibre.ebooks.conversion.plugins.txt_input import TXTInput
from calibre.ebooks.conversion.plugins.lrf_input import LRFInput
from calibre.ebooks.conversion.plugins.chm_input import CHMInput
from calibre.ebooks.conversion.plugins.snb_input import SNBInput
from calibre.ebooks.epub.output import EPUBOutput
from calibre.ebooks.fb2.output import FB2Output

246
src/calibre/ebooks/comic/input.py
View File

@ -7,11 +7,10 @@ __docformat__ = 'restructuredtext en'
Based on ideas from comiclrf created by FangornUK.
'''
import os, shutil, traceback, textwrap, time, codecs
import os, traceback, time
from Queue import Empty
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
from calibre import extract, CurrentDir, prints, walk
from calibre import extract, prints, walk
from calibre.constants import filesystem_encoding
from calibre.ptempfile import PersistentTemporaryDirectory
from calibre.utils.ipc.server import Server
@ -273,245 +272,4 @@ def process_pages(pages, opts, update, tdir):
return ans, failures
class ComicInput(InputFormatPlugin):
name = 'Comic Input'
author = 'Kovid Goyal'
description = 'Optimize comic files (.cbz, .cbr, .cbc) for viewing on portable devices'
file_types = set(['cbz', 'cbr', 'cbc'])
is_image_collection = True
core_usage = -1
options = set([
OptionRecommendation(name='colors', recommended_value=256,
help=_('Number of colors for grayscale image conversion. Default: '
'%default. Values of less than 256 may result in blurred text '
'on your device if you are creating your comics in EPUB format.')),
OptionRecommendation(name='dont_normalize', recommended_value=False,
help=_('Disable normalize (improve contrast) color range '
'for pictures. Default: False')),
OptionRecommendation(name='keep_aspect_ratio', recommended_value=False,
help=_('Maintain picture aspect ratio. Default is to fill the screen.')),
OptionRecommendation(name='dont_sharpen', recommended_value=False,
help=_('Disable sharpening.')),
OptionRecommendation(name='disable_trim', recommended_value=False,
help=_('Disable trimming of comic pages. For some comics, '
'trimming might remove content as well as borders.')),
OptionRecommendation(name='landscape', recommended_value=False,
help=_("Don't split landscape images into two portrait images")),
OptionRecommendation(name='wide', recommended_value=False,
help=_("Keep aspect ratio and scale image using screen height as "
"image width for viewing in landscape mode.")),
OptionRecommendation(name='right2left', recommended_value=False,
help=_('Used for right-to-left publications like manga. '
'Causes landscape pages to be split into portrait pages '
'from right to left.')),
OptionRecommendation(name='despeckle', recommended_value=False,
help=_('Enable Despeckle. Reduces speckle noise. '
'May greatly increase processing time.')),
OptionRecommendation(name='no_sort', recommended_value=False,
help=_("Don't sort the files found in the comic "
"alphabetically by name. Instead use the order they were "
"added to the comic.")),
OptionRecommendation(name='output_format', choices=['png', 'jpg'],
recommended_value='png', help=_('The format that images in the created ebook '
'are converted to. You can experiment to see which format gives '
'you optimal size and look on your device.')),
OptionRecommendation(name='no_process', recommended_value=False,
help=_("Apply no processing to the image")),
OptionRecommendation(name='dont_grayscale', recommended_value=False,
help=_('Do not convert the image to grayscale (black and white)')),
OptionRecommendation(name='comic_image_size', recommended_value=None,
help=_('Specify the image size as widthxheight pixels. Normally,'
' an image size is automatically calculated from the output '
'profile, this option overrides it.')),
OptionRecommendation(name='dont_add_comic_pages_to_toc', recommended_value=False,
help=_('When converting a CBC do not add links to each page to'
' the TOC. Note this only applies if the TOC has more than one'
' section')),
])
recommendations = set([
('margin_left', 0, OptionRecommendation.HIGH),
('margin_top', 0, OptionRecommendation.HIGH),
('margin_right', 0, OptionRecommendation.HIGH),
('margin_bottom', 0, OptionRecommendation.HIGH),
('insert_blank_line', False, OptionRecommendation.HIGH),
('remove_paragraph_spacing', False, OptionRecommendation.HIGH),
('change_justification', 'left', OptionRecommendation.HIGH),
('dont_split_on_pagebreaks', True, OptionRecommendation.HIGH),
('chapter', None, OptionRecommendation.HIGH),
('page_breaks_brefore', None, OptionRecommendation.HIGH),
('use_auto_toc', False, OptionRecommendation.HIGH),
('page_breaks_before', None, OptionRecommendation.HIGH),
('disable_font_rescaling', True, OptionRecommendation.HIGH),
('linearize_tables', False, OptionRecommendation.HIGH),
])
def get_comics_from_collection(self, stream):
from calibre.libunzip import extract as zipextract
tdir = PersistentTemporaryDirectory('_comic_collection')
zipextract(stream, tdir)
comics = []
with CurrentDir(tdir):
if not os.path.exists('comics.txt'):
raise ValueError((
'%s is not a valid comic collection'
' no comics.txt was found in the file')
%stream.name)
raw = open('comics.txt', 'rb').read()
if raw.startswith(codecs.BOM_UTF16_BE):
raw = raw.decode('utf-16-be')[1:]
elif raw.startswith(codecs.BOM_UTF16_LE):
raw = raw.decode('utf-16-le')[1:]
elif raw.startswith(codecs.BOM_UTF8):
raw = raw.decode('utf-8')[1:]
else:
raw = raw.decode('utf-8')
for line in raw.splitlines():
line = line.strip()
if not line:
continue
fname, title = line.partition(':')[0], line.partition(':')[-1]
fname = fname.replace('#', '_')
fname = os.path.join(tdir, *fname.split('/'))
if not title:
title = os.path.basename(fname).rpartition('.')[0]
if os.access(fname, os.R_OK):
comics.append([title, fname])
if not comics:
raise ValueError('%s has no comics'%stream.name)
return comics
def get_pages(self, comic, tdir2):
tdir = extract_comic(comic)
new_pages = find_pages(tdir, sort_on_mtime=self.opts.no_sort,
verbose=self.opts.verbose)
thumbnail = None
if not new_pages:
raise ValueError('Could not find any pages in the comic: %s'
%comic)
if self.opts.no_process:
n2 = []
for page in new_pages:
n2.append(os.path.join(tdir2, os.path.basename(page)))
shutil.copyfile(page, n2[-1])
new_pages = n2
else:
new_pages, failures = process_pages(new_pages, self.opts,
self.report_progress, tdir2)
if failures:
self.log.warning('Could not process the following pages '
'(run with --verbose to see why):')
for f in failures:
self.log.warning('\t', f)
if not new_pages:
raise ValueError('Could not find any valid pages in comic: %s'
% comic)
thumbnail = os.path.join(tdir2,
'thumbnail.'+self.opts.output_format.lower())
if not os.access(thumbnail, os.R_OK):
thumbnail = None
return new_pages
def get_images(self):
return self._images
def convert(self, stream, opts, file_ext, log, accelerators):
from calibre.ebooks.metadata import MetaInformation
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.ebooks.metadata.toc import TOC
self.opts, self.log= opts, log
if file_ext == 'cbc':
comics_ = self.get_comics_from_collection(stream)
else:
comics_ = [['Comic', os.path.abspath(stream.name)]]
stream.close()
comics = []
for i, x in enumerate(comics_):
title, fname = x
cdir = 'comic_%d'%(i+1) if len(comics_) > 1 else '.'
cdir = os.path.abspath(cdir)
if not os.path.exists(cdir):
os.makedirs(cdir)
pages = self.get_pages(fname, cdir)
if not pages: continue
wrappers = self.create_wrappers(pages)
comics.append((title, pages, wrappers))
if not comics:
raise ValueError('No comic pages found in %s'%stream.name)
mi = MetaInformation(os.path.basename(stream.name).rpartition('.')[0],
[_('Unknown')])
opf = OPFCreator(os.path.abspath('.'), mi)
entries = []
def href(x):
if len(comics) == 1: return os.path.basename(x)
return '/'.join(x.split(os.sep)[-2:])
for comic in comics:
pages, wrappers = comic[1:]
entries += [(w, None) for w in map(href, wrappers)] + \
[(x, None) for x in map(href, pages)]
opf.create_manifest(entries)
spine = []
for comic in comics:
spine.extend(map(href, comic[2]))
self._images = []
for comic in comics:
self._images.extend(comic[1])
opf.create_spine(spine)
toc = TOC()
if len(comics) == 1:
wrappers = comics[0][2]
for i, x in enumerate(wrappers):
toc.add_item(href(x), None, _('Page')+' %d'%(i+1),
play_order=i)
else:
po = 0
for comic in comics:
po += 1
wrappers = comic[2]
stoc = toc.add_item(href(wrappers[0]),
None, comic[0], play_order=po)
if not opts.dont_add_comic_pages_to_toc:
for i, x in enumerate(wrappers):
stoc.add_item(href(x), None,
_('Page')+' %d'%(i+1), play_order=po)
po += 1
opf.set_toc(toc)
m, n = open('metadata.opf', 'wb'), open('toc.ncx', 'wb')
opf.render(m, n, 'toc.ncx')
return os.path.abspath('metadata.opf')
def create_wrappers(self, pages):
from calibre.ebooks.oeb.base import XHTML_NS
wrappers = []
WRAPPER = textwrap.dedent('''\
<html xmlns="%s">
<head>
<title>Page #%d</title>
<style type="text/css">
@page { margin:0pt; padding: 0pt}
body { margin: 0pt; padding: 0pt}
div { text-align: center }
</style>
</head>
<body>
<div>
<img src="%s" alt="comic page #%d" />
</div>
</body>
</html>
''')
dir = os.path.dirname(pages[0])
for i, page in enumerate(pages):
wrapper = WRAPPER%(XHTML_NS, i+1, os.path.basename(page), i+1)
page = os.path.join(dir, 'page_%d.xhtml'%(i+1))
open(page, 'wb').write(wrapper)
wrappers.append(page)
return wrappers

11
src/calibre/ebooks/conversion/plugins/__init__.py Normal file
View File

@ -0,0 +1,11 @@
#!/usr/bin/env python
# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
from __future__ import (unicode_literals, division, absolute_import,
print_function)
__license__ = 'GPL v3'
__copyright__ = '2012, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'

5
src/calibre/ebooks/azw4/input.py → src/calibre/ebooks/conversion/plugins/azw4_input.py
View File

@ -7,8 +7,6 @@ __docformat__ = 'restructuredtext en'
import os
from calibre.customize.conversion import InputFormatPlugin
from calibre.ebooks.pdb.header import PdbHeaderReader
from calibre.ebooks.azw4.reader import Reader
class AZW4Input(InputFormatPlugin):
@ -19,6 +17,9 @@ class AZW4Input(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.pdb.header import PdbHeaderReader
from calibre.ebooks.azw4.reader import Reader
header = PdbHeaderReader(stream)
reader = Reader(header, stream, log, options)
opf = reader.extract_content(os.getcwd())

9
src/calibre/ebooks/chm/input.py → src/calibre/ebooks/conversion/plugins/chm_input.py
View File

@ -3,9 +3,7 @@ __license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>,' \
' and Alex Bramley <a.bramley at gmail.com>.'
import os, uuid
from lxml import html
import os
from calibre.customize.conversion import InputFormatPlugin
from calibre.ptempfile import TemporaryDirectory
@ -77,7 +75,7 @@ class CHMInput(InputFormatPlugin):
def _create_oebbook_html(self, htmlpath, basedir, opts, log, mi):
# use HTMLInput plugin to generate book
from calibre.ebooks.html.input import HTMLInput
from calibre.customize.builtins import HTMLInput
opts.breadth_first = True
htmlinput = HTMLInput(None)
oeb = htmlinput.create_oebbook(htmlpath, basedir, opts, log, mi)
@ -85,6 +83,8 @@ class CHMInput(InputFormatPlugin):
def _create_oebbook(self, hhcpath, basedir, opts, log, mi):
import uuid
from lxml import html
from calibre.ebooks.conversion.plumber import create_oebbook
from calibre.ebooks.oeb.base import DirContainer
oeb = create_oebbook(log, None, opts,
@ -142,6 +142,7 @@ class CHMInput(InputFormatPlugin):
return oeb
def _create_html_root(self, hhcpath, log):
from lxml import html
hhcdata = self._read_file(hhcpath)
hhcroot = html.fromstring(hhcdata)
chapters = self._process_nodes(hhcroot)

259
src/calibre/ebooks/conversion/plugins/comic_input.py Normal file
View File

@ -0,0 +1,259 @@
from __future__ import with_statement
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal kovid@kovidgoyal.net'
__docformat__ = 'restructuredtext en'
'''
Based on ideas from comiclrf created by FangornUK.
'''
import shutil, textwrap, codecs, os
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
from calibre import CurrentDir
from calibre.ptempfile import PersistentTemporaryDirectory
class ComicInput(InputFormatPlugin):
name = 'Comic Input'
author = 'Kovid Goyal'
description = 'Optimize comic files (.cbz, .cbr, .cbc) for viewing on portable devices'
file_types = set(['cbz', 'cbr', 'cbc'])
is_image_collection = True
core_usage = -1
options = set([
OptionRecommendation(name='colors', recommended_value=256,
help=_('Number of colors for grayscale image conversion. Default: '
'%default. Values of less than 256 may result in blurred text '
'on your device if you are creating your comics in EPUB format.')),
OptionRecommendation(name='dont_normalize', recommended_value=False,
help=_('Disable normalize (improve contrast) color range '
'for pictures. Default: False')),
OptionRecommendation(name='keep_aspect_ratio', recommended_value=False,
help=_('Maintain picture aspect ratio. Default is to fill the screen.')),
OptionRecommendation(name='dont_sharpen', recommended_value=False,
help=_('Disable sharpening.')),
OptionRecommendation(name='disable_trim', recommended_value=False,
help=_('Disable trimming of comic pages. For some comics, '
'trimming might remove content as well as borders.')),
OptionRecommendation(name='landscape', recommended_value=False,
help=_("Don't split landscape images into two portrait images")),
OptionRecommendation(name='wide', recommended_value=False,
help=_("Keep aspect ratio and scale image using screen height as "
"image width for viewing in landscape mode.")),
OptionRecommendation(name='right2left', recommended_value=False,
help=_('Used for right-to-left publications like manga. '
'Causes landscape pages to be split into portrait pages '
'from right to left.')),
OptionRecommendation(name='despeckle', recommended_value=False,
help=_('Enable Despeckle. Reduces speckle noise. '
'May greatly increase processing time.')),
OptionRecommendation(name='no_sort', recommended_value=False,
help=_("Don't sort the files found in the comic "
"alphabetically by name. Instead use the order they were "
"added to the comic.")),
OptionRecommendation(name='output_format', choices=['png', 'jpg'],
recommended_value='png', help=_('The format that images in the created ebook '
'are converted to. You can experiment to see which format gives '
'you optimal size and look on your device.')),
OptionRecommendation(name='no_process', recommended_value=False,
help=_("Apply no processing to the image")),
OptionRecommendation(name='dont_grayscale', recommended_value=False,
help=_('Do not convert the image to grayscale (black and white)')),
OptionRecommendation(name='comic_image_size', recommended_value=None,
help=_('Specify the image size as widthxheight pixels. Normally,'
' an image size is automatically calculated from the output '
'profile, this option overrides it.')),
OptionRecommendation(name='dont_add_comic_pages_to_toc', recommended_value=False,
help=_('When converting a CBC do not add links to each page to'
' the TOC. Note this only applies if the TOC has more than one'
' section')),
])
recommendations = set([
('margin_left', 0, OptionRecommendation.HIGH),
('margin_top', 0, OptionRecommendation.HIGH),
('margin_right', 0, OptionRecommendation.HIGH),
('margin_bottom', 0, OptionRecommendation.HIGH),
('insert_blank_line', False, OptionRecommendation.HIGH),
('remove_paragraph_spacing', False, OptionRecommendation.HIGH),
('change_justification', 'left', OptionRecommendation.HIGH),
('dont_split_on_pagebreaks', True, OptionRecommendation.HIGH),
('chapter', None, OptionRecommendation.HIGH),
('page_breaks_brefore', None, OptionRecommendation.HIGH),
('use_auto_toc', False, OptionRecommendation.HIGH),
('page_breaks_before', None, OptionRecommendation.HIGH),
('disable_font_rescaling', True, OptionRecommendation.HIGH),
('linearize_tables', False, OptionRecommendation.HIGH),
])
def get_comics_from_collection(self, stream):
from calibre.libunzip import extract as zipextract
tdir = PersistentTemporaryDirectory('_comic_collection')
zipextract(stream, tdir)
comics = []
with CurrentDir(tdir):
if not os.path.exists('comics.txt'):
raise ValueError((
'%s is not a valid comic collection'
' no comics.txt was found in the file')
%stream.name)
raw = open('comics.txt', 'rb').read()
if raw.startswith(codecs.BOM_UTF16_BE):
raw = raw.decode('utf-16-be')[1:]
elif raw.startswith(codecs.BOM_UTF16_LE):
raw = raw.decode('utf-16-le')[1:]
elif raw.startswith(codecs.BOM_UTF8):
raw = raw.decode('utf-8')[1:]
else:
raw = raw.decode('utf-8')
for line in raw.splitlines():
line = line.strip()
if not line:
continue
fname, title = line.partition(':')[0], line.partition(':')[-1]
fname = fname.replace('#', '_')
fname = os.path.join(tdir, *fname.split('/'))
if not title:
title = os.path.basename(fname).rpartition('.')[0]
if os.access(fname, os.R_OK):
comics.append([title, fname])
if not comics:
raise ValueError('%s has no comics'%stream.name)
return comics
def get_pages(self, comic, tdir2):
from calibre.ebooks.comic.input import (extract_comic, process_pages,
find_pages)
tdir = extract_comic(comic)
new_pages = find_pages(tdir, sort_on_mtime=self.opts.no_sort,
verbose=self.opts.verbose)
thumbnail = None
if not new_pages:
raise ValueError('Could not find any pages in the comic: %s'
%comic)
if self.opts.no_process:
n2 = []
for page in new_pages:
n2.append(os.path.join(tdir2, os.path.basename(page)))
shutil.copyfile(page, n2[-1])
new_pages = n2
else:
new_pages, failures = process_pages(new_pages, self.opts,
self.report_progress, tdir2)
if failures:
self.log.warning('Could not process the following pages '
'(run with --verbose to see why):')
for f in failures:
self.log.warning('\t', f)
if not new_pages:
raise ValueError('Could not find any valid pages in comic: %s'
% comic)
thumbnail = os.path.join(tdir2,
'thumbnail.'+self.opts.output_format.lower())
if not os.access(thumbnail, os.R_OK):
thumbnail = None
return new_pages
def get_images(self):
return self._images
def convert(self, stream, opts, file_ext, log, accelerators):
from calibre.ebooks.metadata import MetaInformation
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.ebooks.metadata.toc import TOC
self.opts, self.log= opts, log
if file_ext == 'cbc':
comics_ = self.get_comics_from_collection(stream)
else:
comics_ = [['Comic', os.path.abspath(stream.name)]]
stream.close()
comics = []
for i, x in enumerate(comics_):
title, fname = x
cdir = 'comic_%d'%(i+1) if len(comics_) > 1 else '.'
cdir = os.path.abspath(cdir)
if not os.path.exists(cdir):
os.makedirs(cdir)
pages = self.get_pages(fname, cdir)
if not pages: continue
wrappers = self.create_wrappers(pages)
comics.append((title, pages, wrappers))
if not comics:
raise ValueError('No comic pages found in %s'%stream.name)
mi = MetaInformation(os.path.basename(stream.name).rpartition('.')[0],
[_('Unknown')])
opf = OPFCreator(os.path.abspath('.'), mi)
entries = []
def href(x):
if len(comics) == 1: return os.path.basename(x)
return '/'.join(x.split(os.sep)[-2:])
for comic in comics:
pages, wrappers = comic[1:]
entries += [(w, None) for w in map(href, wrappers)] + \
[(x, None) for x in map(href, pages)]
opf.create_manifest(entries)
spine = []
for comic in comics:
spine.extend(map(href, comic[2]))
self._images = []
for comic in comics:
self._images.extend(comic[1])
opf.create_spine(spine)
toc = TOC()
if len(comics) == 1:
wrappers = comics[0][2]
for i, x in enumerate(wrappers):
toc.add_item(href(x), None, _('Page')+' %d'%(i+1),
play_order=i)
else:
po = 0
for comic in comics:
po += 1
wrappers = comic[2]
stoc = toc.add_item(href(wrappers[0]),
None, comic[0], play_order=po)
if not opts.dont_add_comic_pages_to_toc:
for i, x in enumerate(wrappers):
stoc.add_item(href(x), None,
_('Page')+' %d'%(i+1), play_order=po)
po += 1
opf.set_toc(toc)
m, n = open('metadata.opf', 'wb'), open('toc.ncx', 'wb')
opf.render(m, n, 'toc.ncx')
return os.path.abspath('metadata.opf')
def create_wrappers(self, pages):
from calibre.ebooks.oeb.base import XHTML_NS
wrappers = []
WRAPPER = textwrap.dedent('''\
<html xmlns="%s">
<head>
<title>Page #%d</title>
<style type="text/css">
@page { margin:0pt; padding: 0pt}
body { margin: 0pt; padding: 0pt}
div { text-align: center }
</style>
</head>
<body>
<div>
<img src="%s" alt="comic page #%d" />
</div>
</body>
</html>
''')
dir = os.path.dirname(pages[0])
for i, page in enumerate(pages):
wrapper = WRAPPER%(XHTML_NS, i+1, os.path.basename(page), i+1)
page = os.path.join(dir, 'page_%d.xhtml'%(i+1))
open(page, 'wb').write(wrapper)
wrappers.append(page)
return wrappers

3
src/calibre/ebooks/djvu/input.py → src/calibre/ebooks/conversion/plugins/djvu_input.py
View File

@ -12,7 +12,6 @@ from subprocess import Popen, PIPE
from cStringIO import StringIO
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
from calibre.ebooks.txt.processor import convert_basic
class DJVUInput(InputFormatPlugin):
@ -28,6 +27,8 @@ class DJVUInput(InputFormatPlugin):
])
def convert(self, stream, options, file_ext, log, accelerators):
from calibre.ebooks.txt.processor import convert_basic
stdout = StringIO()
ppdjvu = True
# using djvutxt is MUCH faster, should make it an option

8
src/calibre/ebooks/epub/input.py → src/calibre/ebooks/conversion/plugins/epub_input.py
View File

@ -3,11 +3,9 @@ __license__ = 'GPL 3'
__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'
import os, uuid
import os
from itertools import cycle
from lxml import etree
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
class EPUBInput(InputFormatPlugin):
@ -30,6 +28,8 @@ class EPUBInput(InputFormatPlugin):
f.write(raw[1024:])
def process_encryption(self, encfile, opf, log):
from lxml import etree
import uuid
key = None
for item in opf.identifier_iter():
scheme = None
@ -65,6 +65,7 @@ class EPUBInput(InputFormatPlugin):
return False
def rationalize_cover(self, opf, log):
from lxml import etree
guide_cover, guide_elem = None, None
for guide_elem in opf.iterguide():
if guide_elem.get('type', '').lower() == 'cover':
@ -110,6 +111,7 @@ class EPUBInput(InputFormatPlugin):
renderer)
def find_opf(self):
from lxml import etree
def attr(n, attr):
for k, v in n.attrib.items():
if k.endswith(attr):

2
src/calibre/ebooks/fb2/input.py → src/calibre/ebooks/conversion/plugins/fb2_input.py
View File

@ -6,7 +6,6 @@ Convert .fb2 files to .lrf
"""
import os, re
from base64 import b64decode
from lxml import etree
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
from calibre import guess_type
@ -38,6 +37,7 @@ class FB2Input(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from lxml import etree
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.ebooks.metadata.meta import get_metadata
from calibre.ebooks.oeb.base import XLINK_NS, XHTML_NS, RECOVER_PARSER

283
src/calibre/ebooks/conversion/plugins/html_input.py Normal file
View File

@ -0,0 +1,283 @@
#!/usr/bin/env python
# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
from __future__ import (unicode_literals, division, absolute_import,
print_function)
__license__ = 'GPL v3'
__copyright__ = '2012, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'
import re, tempfile, os
from functools import partial
from itertools import izip
from urllib import quote
from calibre.constants import islinux, isbsd
from calibre.customize.conversion import (InputFormatPlugin,
OptionRecommendation)
from calibre.utils.localization import get_lang
from calibre.utils.filenames import ascii_filename
class HTMLInput(InputFormatPlugin):
name = 'HTML Input'
author = 'Kovid Goyal'
description = 'Convert HTML and OPF files to an OEB'
file_types = set(['opf', 'html', 'htm', 'xhtml', 'xhtm', 'shtm', 'shtml'])
options = set([
OptionRecommendation(name='breadth_first',
recommended_value=False, level=OptionRecommendation.LOW,
help=_('Traverse links in HTML files breadth first. Normally, '
'they are traversed depth first.'
)
),
OptionRecommendation(name='max_levels',
recommended_value=5, level=OptionRecommendation.LOW,
help=_('Maximum levels of recursion when following links in '
'HTML files. Must be non-negative. 0 implies that no '
'links in the root HTML file are followed. Default is '
'%default.'
)
),
OptionRecommendation(name='dont_package',
recommended_value=False, level=OptionRecommendation.LOW,
help=_('Normally this input plugin re-arranges all the input '
'files into a standard folder hierarchy. Only use this option '
'if you know what you are doing as it can result in various '
'nasty side effects in the rest of the conversion pipeline.'
)
),
])
def convert(self, stream, opts, file_ext, log,
accelerators):
self._is_case_sensitive = None
basedir = os.getcwd()
self.opts = opts
fname = None
if hasattr(stream, 'name'):
basedir = os.path.dirname(stream.name)
fname = os.path.basename(stream.name)
if file_ext != 'opf':
if opts.dont_package:
raise ValueError('The --dont-package option is not supported for an HTML input file')
from calibre.ebooks.metadata.html import get_metadata
mi = get_metadata(stream)
if fname:
from calibre.ebooks.metadata.meta import metadata_from_filename
fmi = metadata_from_filename(fname)
fmi.smart_update(mi)
mi = fmi
oeb = self.create_oebbook(stream.name, basedir, opts, log, mi)
return oeb
from calibre.ebooks.conversion.plumber import create_oebbook
return create_oebbook(log, stream.name, opts,
encoding=opts.input_encoding)
def is_case_sensitive(self, path):
if getattr(self, '_is_case_sensitive', None) is not None:
return self._is_case_sensitive
if not path or not os.path.exists(path):
return islinux or isbsd
self._is_case_sensitive = not (os.path.exists(path.lower()) \
and os.path.exists(path.upper()))
return self._is_case_sensitive
def create_oebbook(self, htmlpath, basedir, opts, log, mi):
import uuid
from calibre.ebooks.conversion.plumber import create_oebbook
from calibre.ebooks.oeb.base import (DirContainer,
rewrite_links, urlnormalize, urldefrag, BINARY_MIME, OEB_STYLES,
xpath)
from calibre import guess_type
from calibre.ebooks.oeb.transforms.metadata import \
meta_info_to_oeb_metadata
from calibre.ebooks.html.input import get_filelist
import cssutils, logging
cssutils.log.setLevel(logging.WARN)
self.OEB_STYLES = OEB_STYLES
oeb = create_oebbook(log, None, opts, self,
encoding=opts.input_encoding, populate=False)
self.oeb = oeb
metadata = oeb.metadata
meta_info_to_oeb_metadata(mi, metadata, log)
if not metadata.language:
oeb.logger.warn(u'Language not specified')
metadata.add('language', get_lang().replace('_', '-'))
if not metadata.creator:
oeb.logger.warn('Creator not specified')
metadata.add('creator', self.oeb.translate(__('Unknown')))
if not metadata.title:
oeb.logger.warn('Title not specified')
metadata.add('title', self.oeb.translate(__('Unknown')))
bookid = str(uuid.uuid4())
metadata.add('identifier', bookid, id='uuid_id', scheme='uuid')
for ident in metadata.identifier:
if 'id' in ident.attrib:
self.oeb.uid = metadata.identifier[0]
break
filelist = get_filelist(htmlpath, basedir, opts, log)
filelist = [f for f in filelist if not f.is_binary]
htmlfile_map = {}
for f in filelist:
path = f.path
oeb.container = DirContainer(os.path.dirname(path), log,
ignore_opf=True)
bname = os.path.basename(path)
id, href = oeb.manifest.generate(id='html',
href=ascii_filename(bname))
htmlfile_map[path] = href
item = oeb.manifest.add(id, href, 'text/html')
item.html_input_href = bname
oeb.spine.add(item, True)
self.added_resources = {}
self.log = log
self.log('Normalizing filename cases')
for path, href in htmlfile_map.items():
if not self.is_case_sensitive(path):
path = path.lower()
self.added_resources[path] = href
self.urlnormalize, self.DirContainer = urlnormalize, DirContainer
self.urldefrag = urldefrag
self.guess_type, self.BINARY_MIME = guess_type, BINARY_MIME
self.log('Rewriting HTML links')
for f in filelist:
path = f.path
dpath = os.path.dirname(path)
oeb.container = DirContainer(dpath, log, ignore_opf=True)
item = oeb.manifest.hrefs[htmlfile_map[path]]
rewrite_links(item.data, partial(self.resource_adder, base=dpath))
for item in oeb.manifest.values():
if item.media_type in self.OEB_STYLES:
dpath = None
for path, href in self.added_resources.items():
if href == item.href:
dpath = os.path.dirname(path)
break
cssutils.replaceUrls(item.data,
partial(self.resource_adder, base=dpath))
toc = self.oeb.toc
self.oeb.auto_generated_toc = True
titles = []
headers = []
for item in self.oeb.spine:
if not item.linear: continue
html = item.data
title = ''.join(xpath(html, '/h:html/h:head/h:title/text()'))
title = re.sub(r'\s+', ' ', title.strip())
if title:
titles.append(title)
headers.append('(unlabled)')
for tag in ('h1', 'h2', 'h3', 'h4', 'h5', 'strong'):
expr = '/h:html/h:body//h:%s[position()=1]/text()'
header = ''.join(xpath(html, expr % tag))
header = re.sub(r'\s+', ' ', header.strip())
if header:
headers[-1] = header
break
use = titles
if len(titles) > len(set(titles)):
use = headers
for title, item in izip(use, self.oeb.spine):
if not item.linear: continue
toc.add(title, item.href)
oeb.container = DirContainer(os.getcwdu(), oeb.log, ignore_opf=True)
return oeb
def link_to_local_path(self, link_, base=None):
from calibre.ebooks.html.input import Link
if not isinstance(link_, unicode):
try:
link_ = link_.decode('utf-8', 'error')
except:
self.log.warn('Failed to decode link %r. Ignoring'%link_)
return None, None
try:
l = Link(link_, base if base else os.getcwdu())
except:
self.log.exception('Failed to process link: %r'%link_)
return None, None
if l.path is None:
# Not a local resource
return None, None
link = l.path.replace('/', os.sep).strip()
frag = l.fragment
if not link:
return None, None
return link, frag
def resource_adder(self, link_, base=None):
link, frag = self.link_to_local_path(link_, base=base)
if link is None:
return link_
try:
if base and not os.path.isabs(link):
link = os.path.join(base, link)
link = os.path.abspath(link)
except:
return link_
if not os.access(link, os.R_OK):
return link_
if os.path.isdir(link):
self.log.warn(link_, 'is a link to a directory. Ignoring.')
return link_
if not self.is_case_sensitive(tempfile.gettempdir()):
link = link.lower()
if link not in self.added_resources:
bhref = os.path.basename(link)
id, href = self.oeb.manifest.generate(id='added',
href=bhref)
guessed = self.guess_type(href)[0]
media_type = guessed or self.BINARY_MIME
if media_type == 'text/plain':
self.log.warn('Ignoring link to text file %r'%link_)
return None
self.oeb.log.debug('Added', link)
self.oeb.container = self.DirContainer(os.path.dirname(link),
self.oeb.log, ignore_opf=True)
# Load into memory
item = self.oeb.manifest.add(id, href, media_type)
# bhref refers to an already existing file. The read() method of
# DirContainer will call unquote on it before trying to read the
# file, therefore we quote it here.
if isinstance(bhref, unicode):
bhref = bhref.encode('utf-8')
item.html_input_href = quote(bhref).decode('utf-8')
if guessed in self.OEB_STYLES:
item.override_css_fetch = partial(
self.css_import_handler, os.path.dirname(link))
item.data
self.added_resources[link] = href
nlink = self.added_resources[link]
if frag:
nlink = '#'.join((nlink, frag))
return nlink
def css_import_handler(self, base, href):
link, frag = self.link_to_local_path(href, base=base)
if link is None or not os.access(link, os.R_OK) or os.path.isdir(link):
return (None, None)
try:
raw = open(link, 'rb').read().decode('utf-8', 'replace')
raw = self.oeb.css_preprocessor(raw, add_namespace=True)
except:
self.log.exception('Failed to read CSS file: %r'%link)
return (None, None)
return (None, raw)

7
src/calibre/ebooks/htmlz/input.py → src/calibre/ebooks/conversion/plugins/htmlz_input.py
View File

@ -10,9 +10,6 @@ import os
from calibre import guess_type
from calibre.customize.conversion import InputFormatPlugin
from calibre.ebooks.chardet import xml_to_unicode
from calibre.ebooks.metadata.opf2 import OPF
from calibre.utils.zipfile import ZipFile
class HTMLZInput(InputFormatPlugin):
@ -23,6 +20,10 @@ class HTMLZInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.chardet import xml_to_unicode
from calibre.ebooks.metadata.opf2 import OPF
from calibre.utils.zipfile import ZipFile
self.log = log
html = u''
top_levels = []

0
src/calibre/ebooks/lit/input.py → src/calibre/ebooks/conversion/plugins/lit_input.py
View File

87
src/calibre/ebooks/conversion/plugins/lrf_input.py Normal file
View File

@ -0,0 +1,87 @@
#!/usr/bin/env python
# vim:fileencoding=UTF-8:ts=4:sw=4:sta:et:sts=4:ai
from __future__ import with_statement
__license__ = 'GPL v3'
__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'
import os, sys
from calibre.customize.conversion import InputFormatPlugin
class LRFInput(InputFormatPlugin):
name = 'LRF Input'
author = 'Kovid Goyal'
description = 'Convert LRF files to HTML'
file_types = set(['lrf'])
def convert(self, stream, options, file_ext, log,
accelerators):
from lxml import etree
from calibre.ebooks.lrf.input import (MediaType, Styles, TextBlock,
Canvas, ImageBlock, RuledLine)
self.log = log
self.log('Generating XML')
from calibre.ebooks.lrf.lrfparser import LRFDocument
d = LRFDocument(stream)
d.parse()
xml = d.to_xml(write_files=True)
if options.verbose > 2:
open('lrs.xml', 'wb').write(xml.encode('utf-8'))
parser = etree.XMLParser(no_network=True, huge_tree=True)
try:
doc = etree.fromstring(xml, parser=parser)
except:
self.log.warn('Failed to parse XML. Trying to recover')
parser = etree.XMLParser(no_network=True, huge_tree=True,
recover=True)
doc = etree.fromstring(xml, parser=parser)
char_button_map = {}
for x in doc.xpath('//CharButton[@refobj]'):
ro = x.get('refobj')
jump_button = doc.xpath('//*[@objid="%s"]'%ro)
if jump_button:
jump_to = jump_button[0].xpath('descendant::JumpTo[@refpage and @refobj]')
if jump_to:
char_button_map[ro] = '%s.xhtml#%s'%(jump_to[0].get('refpage'),
jump_to[0].get('refobj'))
plot_map = {}
for x in doc.xpath('//Plot[@refobj]'):
ro = x.get('refobj')
image = doc.xpath('//Image[@objid="%s" and @refstream]'%ro)
if image:
imgstr = doc.xpath('//ImageStream[@objid="%s" and @file]'%
image[0].get('refstream'))
if imgstr:
plot_map[ro] = imgstr[0].get('file')
self.log('Converting XML to HTML...')
styledoc = etree.fromstring(P('templates/lrf.xsl', data=True))
media_type = MediaType()
styles = Styles()
text_block = TextBlock(styles, char_button_map, plot_map, log)
canvas = Canvas(doc, styles, text_block, log)
image_block = ImageBlock(canvas)
ruled_line = RuledLine()
extensions = {
('calibre', 'media-type') : media_type,
('calibre', 'text-block') : text_block,
('calibre', 'ruled-line') : ruled_line,
('calibre', 'styles') : styles,
('calibre', 'canvas') : canvas,
('calibre', 'image-block'): image_block,
}
transform = etree.XSLT(styledoc, extensions=extensions)
try:
result = transform(doc)
except RuntimeError:
sys.setrecursionlimit(5000)
result = transform(doc)
with open('content.opf', 'wb') as f:
f.write(result)
styles.write()
return os.path.abspath('content.opf')

0
src/calibre/ebooks/mobi/input.py → src/calibre/ebooks/conversion/plugins/mobi_input.py
View File

25
src/calibre/ebooks/conversion/plugins/odt_input.py Normal file
View File

@ -0,0 +1,25 @@
from __future__ import with_statement
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal kovid@kovidgoyal.net'
__docformat__ = 'restructuredtext en'
'''
Convert an ODT file into a Open Ebook
'''
from calibre.customize.conversion import InputFormatPlugin
class ODTInput(InputFormatPlugin):
name = 'ODT Input'
author = 'Kovid Goyal'
description = 'Convert ODT (OpenOffice) files to HTML'
file_types = set(['odt'])
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.odt.input import Extract
return Extract()(stream, '.', log)

5
src/calibre/ebooks/pdb/input.py → src/calibre/ebooks/conversion/plugins/pdb_input.py
View File

@ -7,8 +7,6 @@ __docformat__ = 'restructuredtext en'
import os
from calibre.customize.conversion import InputFormatPlugin
from calibre.ebooks.pdb.header import PdbHeaderReader
from calibre.ebooks.pdb import PDBError, IDENTITY_TO_NAME, get_reader
class PDBInput(InputFormatPlugin):
@ -19,6 +17,9 @@ class PDBInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.pdb.header import PdbHeaderReader
from calibre.ebooks.pdb import PDBError, IDENTITY_TO_NAME, get_reader
header = PdbHeaderReader(stream)
Reader = get_reader(header.ident)

5
src/calibre/ebooks/pdf/input.py → src/calibre/ebooks/conversion/plugins/pdf_input.py
View File

@ -7,8 +7,6 @@ __docformat__ = 'restructuredtext en'
import os
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
from calibre.ebooks.pdf.pdftohtml import pdftohtml
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.constants import plugins
pdfreflow, pdfreflow_err = plugins['pdfreflow']
@ -43,6 +41,9 @@ class PDFInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.ebooks.pdf.pdftohtml import pdftohtml
log.debug('Converting file to html...')
# The main html file will be named index.html
self.opts, self.log = options, log

8
src/calibre/ebooks/pml/input.py → src/calibre/ebooks/conversion/plugins/pml_input.py
View File

@ -11,9 +11,6 @@ import shutil
from calibre.customize.conversion import InputFormatPlugin
from calibre.ptempfile import TemporaryDirectory
from calibre.utils.zipfile import ZipFile
from calibre.ebooks.pml.pmlconverter import PML_HTMLizer
from calibre.ebooks.metadata.toc import TOC
from calibre.ebooks.metadata.opf2 import OPFCreator
class PMLInput(InputFormatPlugin):
@ -24,6 +21,8 @@ class PMLInput(InputFormatPlugin):
file_types = set(['pml', 'pmlz'])
def process_pml(self, pml_path, html_path, close_all=False):
from calibre.ebooks.pml.pmlconverter import PML_HTMLizer
pclose = False
hclose = False
@ -85,6 +84,9 @@ class PMLInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.metadata.toc import TOC
from calibre.ebooks.metadata.opf2 import OPFCreator
self.options = options
self.log = log
pages, images = [], []

3
src/calibre/ebooks/rb/input.py → src/calibre/ebooks/conversion/plugins/rb_input.py
View File

@ -6,7 +6,6 @@ __docformat__ = 'restructuredtext en'
import os
from calibre.ebooks.rb.reader import Reader
from calibre.customize.conversion import InputFormatPlugin
class RBInput(InputFormatPlugin):
@ -18,6 +17,8 @@ class RBInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.rb.reader import Reader
reader = Reader(stream, log, options.input_encoding)
opf = reader.extract_content(os.getcwd())

0
src/calibre/web/feeds/input.py → src/calibre/ebooks/conversion/plugins/recipe_input.py
View File

298
src/calibre/ebooks/conversion/plugins/rtf_input.py Normal file
View File

@ -0,0 +1,298 @@
from __future__ import with_statement
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
import os, glob, re, textwrap
from calibre.customize.conversion import InputFormatPlugin
border_style_map = {
'single' : 'solid',
'double-thickness-border' : 'double',
'shadowed-border': 'outset',
'double-border': 'double',
'dotted-border': 'dotted',
'dashed': 'dashed',
'hairline': 'solid',
'inset': 'inset',
'dash-small': 'dashed',
'dot-dash': 'dotted',
'dot-dot-dash': 'dotted',
'outset': 'outset',
'tripple': 'double',
'triple': 'double',
'thick-thin-small': 'solid',
'thin-thick-small': 'solid',
'thin-thick-thin-small': 'solid',
'thick-thin-medium': 'solid',
'thin-thick-medium': 'solid',
'thin-thick-thin-medium': 'solid',
'thick-thin-large': 'solid',
'thin-thick-thin-large': 'solid',
'wavy': 'ridge',
'double-wavy': 'ridge',
'striped': 'ridge',
'emboss': 'inset',
'engrave': 'inset',
'frame': 'ridge',
}
class RTFInput(InputFormatPlugin):
name = 'RTF Input'
author = 'Kovid Goyal'
description = 'Convert RTF files to HTML'
file_types = set(['rtf'])
def generate_xml(self, stream):
from calibre.ebooks.rtf2xml.ParseRtf import ParseRtf
ofile = 'dataxml.xml'
run_lev, debug_dir, indent_out = 1, None, 0
if getattr(self.opts, 'debug_pipeline', None) is not None:
try:
os.mkdir('rtfdebug')
debug_dir = 'rtfdebug'
run_lev = 4
indent_out = 1
self.log('Running RTFParser in debug mode')
except:
self.log.warn('Impossible to run RTFParser in debug mode')
parser = ParseRtf(
in_file = stream,
out_file = ofile,
# Convert symbol fonts to unicode equivalents. Default
# is 1
convert_symbol = 1,
# Convert Zapf fonts to unicode equivalents. Default
# is 1.
convert_zapf = 1,
# Convert Wingding fonts to unicode equivalents.
# Default is 1.
convert_wingdings = 1,
# Convert RTF caps to real caps.
# Default is 1.
convert_caps = 1,
# Indent resulting XML.
# Default is 0 (no indent).
indent = indent_out,
# Form lists from RTF. Default is 1.
form_lists = 1,
# Convert headings to sections. Default is 0.
headings_to_sections = 1,
# Group paragraphs with the same style name. Default is 1.
group_styles = 1,
# Group borders. Default is 1.
group_borders = 1,
# Write or do not write paragraphs. Default is 0.
empty_paragraphs = 1,
#debug
deb_dir = debug_dir,
run_level = run_lev,
)
parser.parse_rtf()
with open(ofile, 'rb') as f:
return f.read()
def extract_images(self, picts):
import imghdr
self.log('Extracting images...')
with open(picts, 'rb') as f:
raw = f.read()
picts = filter(len, re.findall(r'\{\\pict([^}]+)\}', raw))
hex = re.compile(r'[^a-fA-F0-9]')
encs = [hex.sub('', pict) for pict in picts]
count = 0
imap = {}
for enc in encs:
if len(enc) % 2 == 1:
enc = enc[:-1]
data = enc.decode('hex')
fmt = imghdr.what(None, data)
if fmt is None:
fmt = 'wmf'
count += 1
name = '%04d.%s' % (count, fmt)
with open(name, 'wb') as f:
f.write(data)
imap[count] = name
# with open(name+'.hex', 'wb') as f:
# f.write(enc)
return self.convert_images(imap)
def convert_images(self, imap):
self.default_img = None
for count, val in imap.iteritems():
try:
imap[count] = self.convert_image(val)
except:
self.log.exception('Failed to convert', val)
return imap
def convert_image(self, name):
if not name.endswith('.wmf'):
return name
try:
return self.rasterize_wmf(name)
except:
self.log.exception('Failed to convert WMF image %r'%name)
return self.replace_wmf(name)
def replace_wmf(self, name):
from calibre.ebooks import calibre_cover
if self.default_img is None:
self.default_img = calibre_cover('Conversion of WMF images is not supported',
'Use Microsoft Word or OpenOffice to save this RTF file'
' as HTML and convert that in calibre.', title_size=36,
author_size=20)
name = name.replace('.wmf', '.jpg')
with open(name, 'wb') as f:
f.write(self.default_img)
return name
def rasterize_wmf(self, name):
from calibre.utils.wmf.parse import wmf_unwrap
with open(name, 'rb') as f:
data = f.read()
data = wmf_unwrap(data)
name = name.replace('.wmf', '.png')
with open(name, 'wb') as f:
f.write(data)
return name
def write_inline_css(self, ic, border_styles):
font_size_classes = ['span.fs%d { font-size: %spt }'%(i, x) for i, x in
enumerate(ic.font_sizes)]
color_classes = ['span.col%d { color: %s }'%(i, x) for i, x in
enumerate(ic.colors)]
css = textwrap.dedent('''
span.none {
text-decoration: none; font-weight: normal;
font-style: normal; font-variant: normal
}
span.italics { font-style: italic }
span.bold { font-weight: bold }
span.small-caps { font-variant: small-caps }
span.underlined { text-decoration: underline }
span.strike-through { text-decoration: line-through }
''')
css += '\n'+'\n'.join(font_size_classes)
css += '\n' +'\n'.join(color_classes)
for cls, val in border_styles.iteritems():
css += '\n\n.%s {\n%s\n}'%(cls, val)
with open('styles.css', 'ab') as f:
f.write(css)
def convert_borders(self, doc):
border_styles = []
style_map = {}
for elem in doc.xpath(r'//*[local-name()="cell"]'):
style = ['border-style: hidden', 'border-width: 1px',
'border-color: black']
for x in ('bottom', 'top', 'left', 'right'):
bs = elem.get('border-cell-%s-style'%x, None)
if bs:
cbs = border_style_map.get(bs, 'solid')
style.append('border-%s-style: %s'%(x, cbs))
bw = elem.get('border-cell-%s-line-width'%x, None)
if bw:
style.append('border-%s-width: %spt'%(x, bw))
bc = elem.get('border-cell-%s-color'%x, None)
if bc:
style.append('border-%s-color: %s'%(x, bc))
style = ';\n'.join(style)
if style not in border_styles:
border_styles.append(style)
idx = border_styles.index(style)
cls = 'border_style%d'%idx
style_map[cls] = style
elem.set('class', cls)
return style_map
def convert(self, stream, options, file_ext, log,
accelerators):
from lxml import etree
from calibre.ebooks.metadata.meta import get_metadata
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.ebooks.rtf2xml.ParseRtf import RtfInvalidCodeException
from calibre.ebooks.rtf.input import InlineClass
self.opts = options
self.log = log
self.log('Converting RTF to XML...')
try:
xml = self.generate_xml(stream.name)
except RtfInvalidCodeException as e:
raise ValueError(_('This RTF file has a feature calibre does not '
'support. Convert it to HTML first and then try it.\n%s')%e)
d = glob.glob(os.path.join('*_rtf_pict_dir', 'picts.rtf'))
if d:
imap = {}
try:
imap = self.extract_images(d[0])
except:
self.log.exception('Failed to extract images...')
self.log('Parsing XML...')
parser = etree.XMLParser(recover=True, no_network=True)
doc = etree.fromstring(xml, parser=parser)
border_styles = self.convert_borders(doc)
for pict in doc.xpath('//rtf:pict[@num]',
namespaces={'rtf':'http://rtf2xml.sourceforge.net/'}):
num = int(pict.get('num'))
name = imap.get(num, None)
if name is not None:
pict.set('num', name)
self.log('Converting XML to HTML...')
inline_class = InlineClass(self.log)
styledoc = etree.fromstring(P('templates/rtf.xsl', data=True))
extensions = { ('calibre', 'inline-class') : inline_class }
transform = etree.XSLT(styledoc, extensions=extensions)
result = transform(doc)
html = 'index.xhtml'
with open(html, 'wb') as f:
res = transform.tostring(result)
# res = res[:100].replace('xmlns:html', 'xmlns') + res[100:]
#clean multiple \n
res = re.sub('\n+', '\n', res)
# Replace newlines inserted by the 'empty_paragraphs' option in rtf2xml with html blank lines
# res = re.sub('\s*<body>', '<body>', res)
# res = re.sub('(?<=\n)\n{2}',
# u'<p>\u00a0</p>\n'.encode('utf-8'), res)
f.write(res)
self.write_inline_css(inline_class, border_styles)
stream.seek(0)
mi = get_metadata(stream, 'rtf')
if not mi.title:
mi.title = _('Unknown')
if not mi.authors:
mi.authors = [_('Unknown')]
opf = OPFCreator(os.getcwd(), mi)
opf.create_manifest([('index.xhtml', None)])
opf.create_spine(['index.xhtml'])
opf.render(open('metadata.opf', 'wb'))
return os.path.abspath('metadata.opf')

9
src/calibre/ebooks/snb/input.py → src/calibre/ebooks/conversion/plugins/snb_input.py
View File

@ -4,13 +4,11 @@ __license__ = 'GPL 3'
__copyright__ = '2010, Li Fanxi <lifanxi@freemindworld.com>'
__docformat__ = 'restructuredtext en'
import os, uuid
import os
from calibre.customize.conversion import InputFormatPlugin
from calibre.ebooks.snb.snbfile import SNBFile
from calibre.ptempfile import TemporaryDirectory
from calibre.utils.filenames import ascii_filename
from lxml import etree
HTML_TEMPLATE = u'<html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"/><title>%s</title></head><body>\n%s\n</body></html>'
@ -29,7 +27,12 @@ class SNBInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
import uuid
from lxml import etree
from calibre.ebooks.oeb.base import DirContainer
from calibre.ebooks.snb.snbfile import SNBFile
log.debug("Parsing SNB file...")
snbFile = SNBFile()
try:

3
src/calibre/ebooks/tcr/input.py → src/calibre/ebooks/conversion/plugins/tcr_input.py
View File

@ -7,7 +7,6 @@ __docformat__ = 'restructuredtext en'
from cStringIO import StringIO
from calibre.customize.conversion import InputFormatPlugin
from calibre.ebooks.compression.tcr import decompress
class TCRInput(InputFormatPlugin):
@ -17,6 +16,8 @@ class TCRInput(InputFormatPlugin):
file_types = set(['tcr'])
def convert(self, stream, options, file_ext, log, accelerators):
from calibre.ebooks.compression.tcr import decompress
log.info('Decompressing text...')
raw_txt = decompress(stream)

19
src/calibre/ebooks/txt/input.py → src/calibre/ebooks/conversion/plugins/txt_input.py
View File

@ -8,14 +8,6 @@ import os
from calibre import _ent_pat, walk, xml_entity_to_unicode
from calibre.customize.conversion import InputFormatPlugin, OptionRecommendation
from calibre.ebooks.conversion.preprocess import DocAnalysis, Dehyphenator
from calibre.ebooks.chardet import detect
from calibre.ebooks.txt.processor import convert_basic, convert_markdown, \
separate_paragraphs_single_line, separate_paragraphs_print_formatted, \
preserve_spaces, detect_paragraph_type, detect_formatting_type, \
normalize_line_endings, convert_textile, remove_indents, block_to_single_line, \
separate_hard_scene_breaks
from calibre.utils.zipfile import ZipFile
class TXTInput(InputFormatPlugin):
@ -61,6 +53,17 @@ class TXTInput(InputFormatPlugin):
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.conversion.preprocess import DocAnalysis, Dehyphenator
from calibre.ebooks.chardet import detect
from calibre.utils.zipfile import ZipFile
from calibre.ebooks.txt.processor import (convert_basic,
convert_markdown, separate_paragraphs_single_line,
separate_paragraphs_print_formatted, preserve_spaces,
detect_paragraph_type, detect_formatting_type,
normalize_line_endings, convert_textile, remove_indents,
block_to_single_line, separate_hard_scene_breaks)
self.log = log
txt = ''
log.debug('Reading text from file...')

270
src/calibre/ebooks/html/input.py
View File

@ -11,19 +11,13 @@ __docformat__ = 'restructuredtext en'
Input plugin for HTML or OPF ebooks.
'''
import os, re, sys, uuid, tempfile, errno as gerrno
import os, re, sys, errno as gerrno
from urlparse import urlparse, urlunparse
from urllib import unquote, quote
from functools import partial
from itertools import izip
from urllib import unquote
from calibre.customize.conversion import InputFormatPlugin
from calibre.ebooks.chardet import detect_xml_encoding
from calibre.customize.conversion import OptionRecommendation
from calibre.constants import islinux, isbsd, iswindows
from calibre.constants import iswindows
from calibre import unicode_path, as_unicode
from calibre.utils.localization import get_lang
from calibre.utils.filenames import ascii_filename
class Link(object):
'''
@ -241,262 +235,4 @@ def get_filelist(htmlfile, dir, opts, log):
return filelist
class HTMLInput(InputFormatPlugin):
name = 'HTML Input'
author = 'Kovid Goyal'
description = 'Convert HTML and OPF files to an OEB'
file_types = set(['opf', 'html', 'htm', 'xhtml', 'xhtm', 'shtm', 'shtml'])
options = set([
OptionRecommendation(name='breadth_first',
recommended_value=False, level=OptionRecommendation.LOW,
help=_('Traverse links in HTML files breadth first. Normally, '
'they are traversed depth first.'
)
),
OptionRecommendation(name='max_levels',
recommended_value=5, level=OptionRecommendation.LOW,
help=_('Maximum levels of recursion when following links in '
'HTML files. Must be non-negative. 0 implies that no '
'links in the root HTML file are followed. Default is '
'%default.'
)
),
OptionRecommendation(name='dont_package',
recommended_value=False, level=OptionRecommendation.LOW,
help=_('Normally this input plugin re-arranges all the input '
'files into a standard folder hierarchy. Only use this option '
'if you know what you are doing as it can result in various '
'nasty side effects in the rest of the conversion pipeline.'
)
),
])
def convert(self, stream, opts, file_ext, log,
accelerators):
self._is_case_sensitive = None
basedir = os.getcwd()
self.opts = opts
fname = None
if hasattr(stream, 'name'):
basedir = os.path.dirname(stream.name)
fname = os.path.basename(stream.name)
if file_ext != 'opf':
if opts.dont_package:
raise ValueError('The --dont-package option is not supported for an HTML input file')
from calibre.ebooks.metadata.html import get_metadata
mi = get_metadata(stream)
if fname:
from calibre.ebooks.metadata.meta import metadata_from_filename
fmi = metadata_from_filename(fname)
fmi.smart_update(mi)
mi = fmi
oeb = self.create_oebbook(stream.name, basedir, opts, log, mi)
return oeb
from calibre.ebooks.conversion.plumber import create_oebbook
return create_oebbook(log, stream.name, opts,
encoding=opts.input_encoding)
def is_case_sensitive(self, path):
if getattr(self, '_is_case_sensitive', None) is not None:
return self._is_case_sensitive
if not path or not os.path.exists(path):
return islinux or isbsd
self._is_case_sensitive = not (os.path.exists(path.lower()) \
and os.path.exists(path.upper()))
return self._is_case_sensitive
def create_oebbook(self, htmlpath, basedir, opts, log, mi):
from calibre.ebooks.conversion.plumber import create_oebbook
from calibre.ebooks.oeb.base import (DirContainer,
rewrite_links, urlnormalize, urldefrag, BINARY_MIME, OEB_STYLES,
xpath)
from calibre import guess_type
from calibre.ebooks.oeb.transforms.metadata import \
meta_info_to_oeb_metadata
import cssutils, logging
cssutils.log.setLevel(logging.WARN)
self.OEB_STYLES = OEB_STYLES
oeb = create_oebbook(log, None, opts, self,
encoding=opts.input_encoding, populate=False)
self.oeb = oeb
metadata = oeb.metadata
meta_info_to_oeb_metadata(mi, metadata, log)
if not metadata.language:
oeb.logger.warn(u'Language not specified')
metadata.add('language', get_lang().replace('_', '-'))
if not metadata.creator:
oeb.logger.warn('Creator not specified')
metadata.add('creator', self.oeb.translate(__('Unknown')))
if not metadata.title:
oeb.logger.warn('Title not specified')
metadata.add('title', self.oeb.translate(__('Unknown')))
bookid = str(uuid.uuid4())
metadata.add('identifier', bookid, id='uuid_id', scheme='uuid')
for ident in metadata.identifier:
if 'id' in ident.attrib:
self.oeb.uid = metadata.identifier[0]
break
filelist = get_filelist(htmlpath, basedir, opts, log)
filelist = [f for f in filelist if not f.is_binary]
htmlfile_map = {}
for f in filelist:
path = f.path
oeb.container = DirContainer(os.path.dirname(path), log,
ignore_opf=True)
bname = os.path.basename(path)
id, href = oeb.manifest.generate(id='html',
href=ascii_filename(bname))
htmlfile_map[path] = href
item = oeb.manifest.add(id, href, 'text/html')
item.html_input_href = bname
oeb.spine.add(item, True)
self.added_resources = {}
self.log = log
self.log('Normalizing filename cases')
for path, href in htmlfile_map.items():
if not self.is_case_sensitive(path):
path = path.lower()
self.added_resources[path] = href
self.urlnormalize, self.DirContainer = urlnormalize, DirContainer
self.urldefrag = urldefrag
self.guess_type, self.BINARY_MIME = guess_type, BINARY_MIME
self.log('Rewriting HTML links')
for f in filelist:
path = f.path
dpath = os.path.dirname(path)
oeb.container = DirContainer(dpath, log, ignore_opf=True)
item = oeb.manifest.hrefs[htmlfile_map[path]]
rewrite_links(item.data, partial(self.resource_adder, base=dpath))
for item in oeb.manifest.values():
if item.media_type in self.OEB_STYLES:
dpath = None
for path, href in self.added_resources.items():
if href == item.href:
dpath = os.path.dirname(path)
break
cssutils.replaceUrls(item.data,
partial(self.resource_adder, base=dpath))
toc = self.oeb.toc
self.oeb.auto_generated_toc = True
titles = []
headers = []
for item in self.oeb.spine:
if not item.linear: continue
html = item.data
title = ''.join(xpath(html, '/h:html/h:head/h:title/text()'))
title = re.sub(r'\s+', ' ', title.strip())
if title:
titles.append(title)
headers.append('(unlabled)')
for tag in ('h1', 'h2', 'h3', 'h4', 'h5', 'strong'):
expr = '/h:html/h:body//h:%s[position()=1]/text()'
header = ''.join(xpath(html, expr % tag))
header = re.sub(r'\s+', ' ', header.strip())
if header:
headers[-1] = header
break
use = titles
if len(titles) > len(set(titles)):
use = headers
for title, item in izip(use, self.oeb.spine):
if not item.linear: continue
toc.add(title, item.href)
oeb.container = DirContainer(os.getcwdu(), oeb.log, ignore_opf=True)
return oeb
def link_to_local_path(self, link_, base=None):
if not isinstance(link_, unicode):
try:
link_ = link_.decode('utf-8', 'error')
except:
self.log.warn('Failed to decode link %r. Ignoring'%link_)
return None, None
try:
l = Link(link_, base if base else os.getcwdu())
except:
self.log.exception('Failed to process link: %r'%link_)
return None, None
if l.path is None:
# Not a local resource
return None, None
link = l.path.replace('/', os.sep).strip()
frag = l.fragment
if not link:
return None, None
return link, frag
def resource_adder(self, link_, base=None):
link, frag = self.link_to_local_path(link_, base=base)
if link is None:
return link_
try:
if base and not os.path.isabs(link):
link = os.path.join(base, link)
link = os.path.abspath(link)
except:
return link_
if not os.access(link, os.R_OK):
return link_
if os.path.isdir(link):
self.log.warn(link_, 'is a link to a directory. Ignoring.')
return link_
if not self.is_case_sensitive(tempfile.gettempdir()):
link = link.lower()
if link not in self.added_resources:
bhref = os.path.basename(link)
id, href = self.oeb.manifest.generate(id='added',
href=bhref)
guessed = self.guess_type(href)[0]
media_type = guessed or self.BINARY_MIME
if media_type == 'text/plain':
self.log.warn('Ignoring link to text file %r'%link_)
return None
self.oeb.log.debug('Added', link)
self.oeb.container = self.DirContainer(os.path.dirname(link),
self.oeb.log, ignore_opf=True)
# Load into memory
item = self.oeb.manifest.add(id, href, media_type)
# bhref refers to an already existing file. The read() method of
# DirContainer will call unquote on it before trying to read the
# file, therefore we quote it here.
if isinstance(bhref, unicode):
bhref = bhref.encode('utf-8')
item.html_input_href = quote(bhref).decode('utf-8')
if guessed in self.OEB_STYLES:
item.override_css_fetch = partial(
self.css_import_handler, os.path.dirname(link))
item.data
self.added_resources[link] = href
nlink = self.added_resources[link]
if frag:
nlink = '#'.join((nlink, frag))
return nlink
def css_import_handler(self, base, href):
link, frag = self.link_to_local_path(href, base=base)
if link is None or not os.access(link, os.R_OK) or os.path.isdir(link):
return (None, None)
try:
raw = open(link, 'rb').read().decode('utf-8', 'replace')
raw = self.oeb.css_preprocessor(raw, add_namespace=True)
except:
self.log.exception('Failed to read CSS file: %r'%link)
return (None, None)
return (None, raw)

75
src/calibre/ebooks/lrf/input.py
View File

@ -6,12 +6,11 @@ __license__ = 'GPL v3'
__copyright__ = '2009, Kovid Goyal <kovid@kovidgoyal.net>'
__docformat__ = 'restructuredtext en'
import os, textwrap, sys, operator
import textwrap, operator
from copy import deepcopy, copy
from lxml import etree
from calibre.customize.conversion import InputFormatPlugin
from calibre import guess_type
class Canvas(etree.XSLTExtension):
@ -406,76 +405,4 @@ class Styles(etree.XSLTExtension):
class LRFInput(InputFormatPlugin):
name = 'LRF Input'
author = 'Kovid Goyal'
description = 'Convert LRF files to HTML'
file_types = set(['lrf'])
def convert(self, stream, options, file_ext, log,
accelerators):
self.log = log
self.log('Generating XML')
from calibre.ebooks.lrf.lrfparser import LRFDocument
d = LRFDocument(stream)
d.parse()
xml = d.to_xml(write_files=True)
if options.verbose > 2:
open('lrs.xml', 'wb').write(xml.encode('utf-8'))
parser = etree.XMLParser(no_network=True, huge_tree=True)
try:
doc = etree.fromstring(xml, parser=parser)
except:
self.log.warn('Failed to parse XML. Trying to recover')
parser = etree.XMLParser(no_network=True, huge_tree=True,
recover=True)
doc = etree.fromstring(xml, parser=parser)
char_button_map = {}
for x in doc.xpath('//CharButton[@refobj]'):
ro = x.get('refobj')
jump_button = doc.xpath('//*[@objid="%s"]'%ro)
if jump_button:
jump_to = jump_button[0].xpath('descendant::JumpTo[@refpage and @refobj]')
if jump_to:
char_button_map[ro] = '%s.xhtml#%s'%(jump_to[0].get('refpage'),
jump_to[0].get('refobj'))
plot_map = {}
for x in doc.xpath('//Plot[@refobj]'):
ro = x.get('refobj')
image = doc.xpath('//Image[@objid="%s" and @refstream]'%ro)
if image:
imgstr = doc.xpath('//ImageStream[@objid="%s" and @file]'%
image[0].get('refstream'))
if imgstr:
plot_map[ro] = imgstr[0].get('file')
self.log('Converting XML to HTML...')
styledoc = etree.fromstring(P('templates/lrf.xsl', data=True))
media_type = MediaType()
styles = Styles()
text_block = TextBlock(styles, char_button_map, plot_map, log)
canvas = Canvas(doc, styles, text_block, log)
image_block = ImageBlock(canvas)
ruled_line = RuledLine()
extensions = {
('calibre', 'media-type') : media_type,
('calibre', 'text-block') : text_block,
('calibre', 'ruled-line') : ruled_line,
('calibre', 'styles') : styles,
('calibre', 'canvas') : canvas,
('calibre', 'image-block'): image_block,
}
transform = etree.XSLT(styledoc, extensions=extensions)
try:
result = transform(doc)
except RuntimeError:
sys.setrecursionlimit(5000)
result = transform(doc)
with open('content.opf', 'wb') as f:
f.write(result)
styles.write()
return os.path.abspath('content.opf')

13
src/calibre/ebooks/odt/input.py
View File

@ -12,7 +12,6 @@ from lxml import etree
from odf.odf2xhtml import ODF2XHTML
from calibre import CurrentDir, walk
from calibre.customize.conversion import InputFormatPlugin
class Extract(ODF2XHTML):
@ -178,16 +177,4 @@ class Extract(ODF2XHTML):
return os.path.abspath('metadata.opf')
class ODTInput(InputFormatPlugin):
name = 'ODT Input'
author = 'Kovid Goyal'
description = 'Convert ODT (OpenOffice) files to HTML'
file_types = set(['odt'])
def convert(self, stream, options, file_ext, log,
accelerators):
return Extract()(stream, '.', log)

291
src/calibre/ebooks/rtf/input.py
View File

@ -2,42 +2,9 @@ from __future__ import with_statement
__license__ = 'GPL v3'
__copyright__ = '2008, Kovid Goyal <kovid at kovidgoyal.net>'
import os, glob, re, textwrap
from lxml import etree
from calibre.customize.conversion import InputFormatPlugin
border_style_map = {
'single' : 'solid',
'double-thickness-border' : 'double',
'shadowed-border': 'outset',
'double-border': 'double',
'dotted-border': 'dotted',
'dashed': 'dashed',
'hairline': 'solid',
'inset': 'inset',
'dash-small': 'dashed',
'dot-dash': 'dotted',
'dot-dot-dash': 'dotted',
'outset': 'outset',
'tripple': 'double',
'triple': 'double',
'thick-thin-small': 'solid',
'thin-thick-small': 'solid',
'thin-thick-thin-small': 'solid',
'thick-thin-medium': 'solid',
'thin-thick-medium': 'solid',
'thin-thick-thin-medium': 'solid',
'thick-thin-large': 'solid',
'thin-thick-thin-large': 'solid',
'wavy': 'ridge',
'double-wavy': 'ridge',
'striped': 'ridge',
'emboss': 'inset',
'engrave': 'inset',
'frame': 'ridge',
}
class InlineClass(etree.XSLTExtension):
@ -71,261 +38,3 @@ class InlineClass(etree.XSLTExtension):
output_parent.text = ' '.join(classes)
class RTFInput(InputFormatPlugin):
name = 'RTF Input'
author = 'Kovid Goyal'
description = 'Convert RTF files to HTML'
file_types = set(['rtf'])
def generate_xml(self, stream):
from calibre.ebooks.rtf2xml.ParseRtf import ParseRtf
ofile = 'dataxml.xml'
run_lev, debug_dir, indent_out = 1, None, 0
if getattr(self.opts, 'debug_pipeline', None) is not None:
try:
os.mkdir('rtfdebug')
debug_dir = 'rtfdebug'
run_lev = 4
indent_out = 1
self.log('Running RTFParser in debug mode')
except:
self.log.warn('Impossible to run RTFParser in debug mode')
parser = ParseRtf(
in_file = stream,
out_file = ofile,
# Convert symbol fonts to unicode equivalents. Default
# is 1
convert_symbol = 1,
# Convert Zapf fonts to unicode equivalents. Default
# is 1.
convert_zapf = 1,
# Convert Wingding fonts to unicode equivalents.
# Default is 1.
convert_wingdings = 1,
# Convert RTF caps to real caps.
# Default is 1.
convert_caps = 1,
# Indent resulting XML.
# Default is 0 (no indent).
indent = indent_out,
# Form lists from RTF. Default is 1.
form_lists = 1,
# Convert headings to sections. Default is 0.
headings_to_sections = 1,
# Group paragraphs with the same style name. Default is 1.
group_styles = 1,
# Group borders. Default is 1.
group_borders = 1,
# Write or do not write paragraphs. Default is 0.
empty_paragraphs = 1,
#debug
deb_dir = debug_dir,
run_level = run_lev,
)
parser.parse_rtf()
with open(ofile, 'rb') as f:
return f.read()
def extract_images(self, picts):
import imghdr
self.log('Extracting images...')
with open(picts, 'rb') as f:
raw = f.read()
picts = filter(len, re.findall(r'\{\\pict([^}]+)\}', raw))
hex = re.compile(r'[^a-fA-F0-9]')
encs = [hex.sub('', pict) for pict in picts]
count = 0
imap = {}
for enc in encs:
if len(enc) % 2 == 1:
enc = enc[:-1]
data = enc.decode('hex')
fmt = imghdr.what(None, data)
if fmt is None:
fmt = 'wmf'
count += 1
name = '%04d.%s' % (count, fmt)
with open(name, 'wb') as f:
f.write(data)
imap[count] = name
# with open(name+'.hex', 'wb') as f:
# f.write(enc)
return self.convert_images(imap)
def convert_images(self, imap):
self.default_img = None
for count, val in imap.iteritems():
try:
imap[count] = self.convert_image(val)
except:
self.log.exception('Failed to convert', val)
return imap
def convert_image(self, name):
if not name.endswith('.wmf'):
return name
try:
return self.rasterize_wmf(name)
except:
self.log.exception('Failed to convert WMF image %r'%name)
return self.replace_wmf(name)
def replace_wmf(self, name):
from calibre.ebooks import calibre_cover
if self.default_img is None:
self.default_img = calibre_cover('Conversion of WMF images is not supported',
'Use Microsoft Word or OpenOffice to save this RTF file'
' as HTML and convert that in calibre.', title_size=36,
author_size=20)
name = name.replace('.wmf', '.jpg')
with open(name, 'wb') as f:
f.write(self.default_img)
return name
def rasterize_wmf(self, name):
from calibre.utils.wmf.parse import wmf_unwrap
with open(name, 'rb') as f:
data = f.read()
data = wmf_unwrap(data)
name = name.replace('.wmf', '.png')
with open(name, 'wb') as f:
f.write(data)
return name
def write_inline_css(self, ic, border_styles):
font_size_classes = ['span.fs%d { font-size: %spt }'%(i, x) for i, x in
enumerate(ic.font_sizes)]
color_classes = ['span.col%d { color: %s }'%(i, x) for i, x in
enumerate(ic.colors)]
css = textwrap.dedent('''
span.none {
text-decoration: none; font-weight: normal;
font-style: normal; font-variant: normal
}
span.italics { font-style: italic }
span.bold { font-weight: bold }
span.small-caps { font-variant: small-caps }
span.underlined { text-decoration: underline }
span.strike-through { text-decoration: line-through }
''')
css += '\n'+'\n'.join(font_size_classes)
css += '\n' +'\n'.join(color_classes)
for cls, val in border_styles.iteritems():
css += '\n\n.%s {\n%s\n}'%(cls, val)
with open('styles.css', 'ab') as f:
f.write(css)
def convert_borders(self, doc):
border_styles = []
style_map = {}
for elem in doc.xpath(r'//*[local-name()="cell"]'):
style = ['border-style: hidden', 'border-width: 1px',
'border-color: black']
for x in ('bottom', 'top', 'left', 'right'):
bs = elem.get('border-cell-%s-style'%x, None)
if bs:
cbs = border_style_map.get(bs, 'solid')
style.append('border-%s-style: %s'%(x, cbs))
bw = elem.get('border-cell-%s-line-width'%x, None)
if bw:
style.append('border-%s-width: %spt'%(x, bw))
bc = elem.get('border-cell-%s-color'%x, None)
if bc:
style.append('border-%s-color: %s'%(x, bc))
style = ';\n'.join(style)
if style not in border_styles:
border_styles.append(style)
idx = border_styles.index(style)
cls = 'border_style%d'%idx
style_map[cls] = style
elem.set('class', cls)
return style_map
def convert(self, stream, options, file_ext, log,
accelerators):
from calibre.ebooks.metadata.meta import get_metadata
from calibre.ebooks.metadata.opf2 import OPFCreator
from calibre.ebooks.rtf2xml.ParseRtf import RtfInvalidCodeException
self.opts = options
self.log = log
self.log('Converting RTF to XML...')
try:
xml = self.generate_xml(stream.name)
except RtfInvalidCodeException as e:
raise ValueError(_('This RTF file has a feature calibre does not '
'support. Convert it to HTML first and then try it.\n%s')%e)
d = glob.glob(os.path.join('*_rtf_pict_dir', 'picts.rtf'))
if d:
imap = {}
try:
imap = self.extract_images(d[0])
except:
self.log.exception('Failed to extract images...')
self.log('Parsing XML...')
parser = etree.XMLParser(recover=True, no_network=True)
doc = etree.fromstring(xml, parser=parser)
border_styles = self.convert_borders(doc)
for pict in doc.xpath('//rtf:pict[@num]',
namespaces={'rtf':'http://rtf2xml.sourceforge.net/'}):
num = int(pict.get('num'))
name = imap.get(num, None)
if name is not None:
pict.set('num', name)
self.log('Converting XML to HTML...')
inline_class = InlineClass(self.log)
styledoc = etree.fromstring(P('templates/rtf.xsl', data=True))
extensions = { ('calibre', 'inline-class') : inline_class }
transform = etree.XSLT(styledoc, extensions=extensions)
result = transform(doc)
html = 'index.xhtml'
with open(html, 'wb') as f:
res = transform.tostring(result)
# res = res[:100].replace('xmlns:html', 'xmlns') + res[100:]
#clean multiple \n
res = re.sub('\n+', '\n', res)
# Replace newlines inserted by the 'empty_paragraphs' option in rtf2xml with html blank lines
# res = re.sub('\s*<body>', '<body>', res)
# res = re.sub('(?<=\n)\n{2}',
# u'<p>\u00a0</p>\n'.encode('utf-8'), res)
f.write(res)
self.write_inline_css(inline_class, border_styles)
stream.seek(0)
mi = get_metadata(stream, 'rtf')
if not mi.title:
mi.title = _('Unknown')
if not mi.authors:
mi.authors = [_('Unknown')]
opf = OPFCreator(os.getcwd(), mi)
opf.create_manifest([('index.xhtml', None)])
opf.create_spine(['index.xhtml'])
opf.render(open('metadata.opf', 'wb'))
return os.path.abspath('metadata.opf')
#ebook-convert "bad.rtf" test.epub -v -d "E:\Mes eBooks\Developpement\debug"
# os.makedirs("E:\\Mes eBooks\\Developpement\\rtfdebug")
# debug_dir = "E:\\Mes eBooks\\Developpement\\rtfdebug"