def get_fb2_data(stream):
from ebook_converter.utils.zipfile import ZipFile, BadZipfile
pos = stream.tell()
try:
zf = ZipFile(stream)
except BadZipfile:
stream.seek(pos)
ans = stream.read()
zip_file_name = None
else:
names = zf.namelist()
names = [x for x in names if x.lower().endswith('.fb2')] or names
zip_file_name = names[0]
ans = zf.open(zip_file_name).read()
return ans, zip_file_name
class DOCX(object):
def __init__(self, path_or_stream, log=None, extract=True):
self.docx_is_transitional = True
stream = path_or_stream
if not hasattr(path_or_stream, 'read'):
stream = open(path_or_stream, 'rb')
self.name = getattr(stream, 'name', None) or '<stream>'
self.log = log or default_log
if extract:
self.extract(stream)
else:
self.init_zipfile(stream)
self.read_content_types()
self.read_package_relationships()
self.namespace = DOCXNamespace(self.docx_is_transitional)
def init_zipfile(self, stream):
self.zipf = ZipFile(stream)
self.names = frozenset(self.zipf.namelist())
def extract(self, stream):
self.tdir = PersistentTemporaryDirectory('docx_container')
try:
zf = ZipFile(stream)
zf.extractall(self.tdir)
except Exception:
self.log.exception('DOCX appears to be invalid ZIP file, trying a'
' more forgiving ZIP parser')
from ebook_converter.utils.localunzip import extractall
stream.seek(0)
extractall(stream, self.tdir)
self.names = {}
for f in walk(self.tdir):
name = os.path.relpath(f, self.tdir).replace(os.sep, '/')
self.names[name] = f
def exists(self, name):
return name in self.names
def read(self, name):
if hasattr(self, 'zipf'):
return self.zipf.open(name).read()
path = self.names[name]
with open(path, 'rb') as f:
return f.read()
def read_content_types(self):
try:
raw = self.read('[Content_Types].xml')
except KeyError:
raise InvalidDOCX('The file %s docx file has no '
'[Content_Types].xml' % self.name)
root = etree.fromstring(raw)
self.content_types = {}
self.default_content_types = {}
for item in root.xpath('//*[local-name()="Types"]/*[local-name()='
'"Default" and @Extension and @ContentType]'):
self.default_content_types[item.get('Extension').lower()] = \
item.get('ContentType')
for item in root.xpath('//*[local-name()="Types"]/*[local-name()='
'"Override" and @PartName and @ContentType]'):
name = item.get('PartName').lstrip('/')
self.content_types[name] = item.get('ContentType')
def content_type(self, name):
if name in self.content_types:
return self.content_types[name]
ext = name.rpartition('.')[-1].lower()
if ext in self.default_content_types:
return self.default_content_types[ext]
return mimetypes.guess_type(name)[0]
def read_package_relationships(self):
try:
raw = self.read('_rels/.rels')
except KeyError:
raise InvalidDOCX('The file %s docx file has no _rels/.rels' %
self.name)
root = etree.fromstring(raw)
self.relationships = {}
self.relationships_rmap = {}
for item in root.xpath('//*[local-name()="Relationships"]/*[local-name'
'()="Relationship" and @Type and @Target]'):
target = item.get('Target').lstrip('/')
typ = item.get('Type')
if target == 'word/document.xml':
self.docx_is_transitional = (typ != 'http://purl.oclc.org/'
'ooxml/officeDocument/'
'relationships/officeDocument')
self.relationships[typ] = target
self.relationships_rmap[target] = typ
@property
def document_name(self):
name = self.relationships.get(self.namespace.names['DOCUMENT'], None)
if name is None:
names = tuple(
n for n in self.names
if n == 'document.xml' or n.endswith('/document.xml'))
if not names:
raise InvalidDOCX('The file %s docx file has no main '
'document' % self.name)
name = names[0]
return name
@property
def document(self):
return etree.fromstring(self.read(self.document_name))
@property
def document_relationships(self):
return self.get_relationships(self.document_name)
def get_relationships(self, name):
base = '/'.join(name.split('/')[:-1])
by_id, by_type = {}, {}
parts = name.split('/')
name = '/'.join(parts[:-1] + ['_rels', parts[-1] + '.rels'])
try:
raw = self.read(name)
except KeyError:
pass
else:
root = etree.fromstring(raw)
for item in root.xpath('//*[local-name()="Relationships"]/*'
'[local-name()="Relationship" and @Type '
'and @Target]'):
target = item.get('Target')
if (item.get('TargetMode', None) != 'External'
and not target.startswith('#')):
target = '/'.join((base, target.lstrip('/')))
typ = item.get('Type')
Id = item.get('Id')
by_id[Id] = by_type[typ] = target
return by_id, by_type
def get_document_properties_names(self):
name = self.relationships.get(self.namespace.names['DOCPROPS'], None)
if name is None:
names = tuple(n for n in self.names
if n.lower() == 'docprops/core.xml')
if names:
name = names[0]
yield name
name = self.relationships.get(self.namespace.names['APPPROPS'], None)
if name is None:
names = tuple(n for n in self.names
if n.lower() == 'docprops/app.xml')
if names:
name = names[0]
yield name
@property
def metadata(self):
mi = Metadata('Unknown')
dp_name, ap_name = self.get_document_properties_names()
if dp_name:
try:
raw = self.read(dp_name)
except KeyError:
pass
else:
read_doc_props(raw, mi, self.namespace.XPath)
if mi.is_null('language'):
try:
raw = self.read('word/styles.xml')
except KeyError:
pass
else:
read_default_style_language(raw, mi, self.namespace.XPath)
ap_name = self.relationships.get(self.namespace.names['APPPROPS'],
None)
if ap_name:
try:
raw = self.read(ap_name)
except KeyError:
pass
else:
read_app_props(raw, mi)
return mi
def close(self):
if hasattr(self, 'zipf'):
self.zipf.close()
else:
try:
shutil.rmtree(self.tdir)
except EnvironmentError:
pass