def temporary_file(self, suffix):
"""
Return a file-like object that is a temporary file on the file system.
This file will remain available even after being closed and will only
be removed on interpreter shutdown. Use the ``name`` member of the
returned object to access the full path to the created temporary file.
:param suffix: The suffix that the temporary file will have.
"""
return PersistentTemporaryFile(suffix)
def process_image(self, data):
if not self.process_images:
return data
func = mobify_image if self.opts.mobi_keep_original_images else rescale_image
try:
return func(data)
except Exception:
ext = what(None, data)
if ext not in ('png', 'gif'):
raise
if ext == 'gif':
with PersistentTemporaryFile(suffix='.gif') as pt:
pt.write(data)
return mobify_image(data)
with PersistentTemporaryFile(suffix='.png') as pt:
pt.write(data)
try:
from ebook_converter.utils.img import optimize_png
optimize_png(pt.name)
data = open(pt.name, 'rb').read()
finally:
os.remove(pt.name)
return func(data)
def extract_content(self, output_dir):
self.log.info('Extracting PDF...')
pdf = PersistentTemporaryFile('.pdf')
pdf.close()
pdf = open(pdf, 'wb')
for x in range(self.header.section_count()):
pdf.write(self.header.section_data(x))
pdf.close()
from ebook_converter.customize.ui import plugin_for_input_format
pdf_plugin = plugin_for_input_format('pdf')
for opt in pdf_plugin.options:
if not hasattr(self.options, opt.option.name):
setattr(self.options, opt.option.name, opt.recommended_value)
return pdf_plugin.convert(open(pdf, 'rb'), self.options, 'pdf', self.log, {})
def dataize_svg(self, item, svg=None):
if svg is None:
svg = item.data
hrefs = self.oeb.manifest.hrefs
for elem in xpath(svg, '//svg:*[@xl:href]'):
href = urlnormalize(elem.attrib[base.tag('xlink', 'href')])
path = urllib.parse.urldefrag(href)[0]
if not path:
continue
abshref = item.abshref(path)
if abshref not in hrefs:
continue
linkee = hrefs[abshref]
data = linkee.bytes_representation
ext = what(None, data) or 'jpg'
with PersistentTemporaryFile(suffix='.' + ext) as pt:
pt.write(data)
self.temp_files.append(pt.name)
elem.attrib[base.tag('xlink', 'href')] = pt.name
return svg
def pdftohtml(output_dir, pdf_path, no_images, as_xml=False):
'''
Convert the pdf into html using the pdftohtml app.
This will write the html as index.html into output_dir.
It will also write all extracted images to the output_dir
'''
pdfsrc = os.path.join(output_dir, 'src.pdf')
index = os.path.join(output_dir, 'index.' + ('xml' if as_xml else 'html'))
with open(pdf_path, 'rb') as src, open(pdfsrc, 'wb') as dest:
shutil.copyfileobj(src, dest)
with CurrentDir(output_dir):
def a(x):
return os.path.basename(x)
cmd = [
'pdftohtml', '-enc', 'UTF-8', '-noframes', '-p', '-nomerge',
'-nodrm',
a(pdfsrc),
a(index)
]
if no_images:
cmd.append('-i')
if as_xml:
cmd.append('-xml')
logf = PersistentTemporaryFile('pdftohtml_log')
try:
p = popen(cmd,
stderr=logf._fd,
stdout=logf._fd,
stdin=subprocess.PIPE)
except OSError as err:
if err.errno == errno.ENOENT:
raise ConversionError('Could not find pdftohtml, check it is '
'in your PATH')
else:
raise
ret = eintr_retry_call(p.wait)
logf.flush()
logf.close()
out = open(logf.name, 'rb').read().decode('utf-8', 'replace').strip()
if ret != 0:
raise ConversionError('pdftohtml failed with return code: '
'%d\n%s' % (ret, out))
if out:
print("pdftohtml log:")
print(out)
if not os.path.exists(index) or os.stat(index).st_size < 100:
raise DRMError()
if not as_xml:
with open(index, 'r+b') as i:
raw = i.read().decode('utf-8', 'replace')
raw = flip_images(raw)
raw = raw.replace(
'<head', '<!-- created by ebook-converter\'s'
' pdftohtml -->\n <head', 1)
i.seek(0)
i.truncate()
# versions of pdftohtml >= 0.20 output self closing <br> tags,
# this breaks the pdf heuristics regexps, so replace them
raw = raw.replace('<br/>', '<br>')
raw = re.sub(r'<a\s+name=(\d+)',
r'<a id="\1"',
raw,
flags=re.I)
raw = re.sub(r'<a id="(\d+)"', r'<a id="p\1"', raw, flags=re.I)
raw = re.sub(r'<a href="index.html#(\d+)"',
r'<a href="#p\1"',
raw,
flags=re.I)
raw = xml_replace_entities(raw)
raw = raw.replace('\u00a0', ' ')
i.write(raw.encode('utf-8'))
cmd = [
'pdftohtml', '-f', '1', '-l', '1', '-xml', '-i', '-enc',
'UTF-8', '-noframes', '-p', '-nomerge', '-nodrm', '-q',
'-stdout',
a(pdfsrc)
]
p = popen(cmd, stdout=subprocess.PIPE)
raw = p.stdout.read().strip()
if p.wait() == 0 and raw:
parse_outline(raw, output_dir)
try:
os.remove(pdfsrc)
except Exception:
pass