Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-07 18:24:30 -04:00
Code Issues Packages Projects Releases Wiki Activity

Slightly improve handling of OEB metadata.

Browse Source
This commit is contained in:
Marshall T. Vandegrift 2009-01-22 22:30:00 -05:00
parent d633507c8c
commit 5060e619ba
1 changed files with 16 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
src/calibre/ebooks/oeb/base.py
View File

@ -21,6 +21,7 @@ from lxml import etree
from lxml import html from lxml import html
from calibre import LoggingInterface from calibre import LoggingInterface
from calibre.translations.dynamic import translate from calibre.translations.dynamic import translate
from calibre.startup import get_lang
XML_PARSER = etree.XMLParser(recover=True) XML_PARSER = etree.XMLParser(recover=True)
XML_NS = 'http://www.w3.org/XML/1998/namespace' XML_NS = 'http://www.w3.org/XML/1998/namespace'
@ -30,6 +31,7 @@ OPF2_NS = 'http://www.idpf.org/2007/opf'
DC09_NS = 'http://purl.org/metadata/dublin_core' DC09_NS = 'http://purl.org/metadata/dublin_core'
DC10_NS = 'http://purl.org/dc/elements/1.0/' DC10_NS = 'http://purl.org/dc/elements/1.0/'
DC11_NS = 'http://purl.org/dc/elements/1.1/' DC11_NS = 'http://purl.org/dc/elements/1.1/'
DC_NSES = set([DC09_NS, DC10_NS, DC11_NS])
XSI_NS = 'http://www.w3.org/2001/XMLSchema-instance' XSI_NS = 'http://www.w3.org/2001/XMLSchema-instance'
DCTERMS_NS = 'http://purl.org/dc/terms/' DCTERMS_NS = 'http://purl.org/dc/terms/'
NCX_NS = 'http://www.daisy.org/z3986/2005/ncx/' NCX_NS = 'http://www.daisy.org/z3986/2005/ncx/'
@ -194,8 +196,12 @@ class Metadata(object):
if term == OPF('meta') and not value: if term == OPF('meta') and not value:
term = self.fq_attrib.pop('name') term = self.fq_attrib.pop('name')
value = self.fq_attrib.pop('content') value = self.fq_attrib.pop('content')
elif term in Metadata.TERMS and not namespace(term): elif barename(term).lower() in Metadata.TERMS and \
term = DC(term) (not namespace(term) or namespace(term) in DC_NSES):
# Anything looking like Dublin Core is coerced
term = DC(barename(term).lower())
elif namespace(term) == OPF2_NS:
term = barename(term)
self.term = term self.term = term
self.value = value self.value = value
self.attrib = attrib = {} self.attrib = attrib = {}
@ -814,7 +820,7 @@ class OEBBook(object):
break break
if not metadata.language: if not metadata.language:
self.logger.warn(u'Language not specified.') self.logger.warn(u'Language not specified.')
metadata.add('language', 'en') metadata.add('language', get_lang())
if not metadata.creator: if not metadata.creator:
self.logger.warn(u'Creator not specified.') self.logger.warn(u'Creator not specified.')
metadata.add('creator', _('Unknown')) metadata.add('creator', _('Unknown'))
@ -857,6 +863,8 @@ class OEBBook(object):
extras.sort() extras.sort()
for item in extras: for item in extras:
spine.add(item, False) spine.add(item, False)
if len(spine) == 0:
raise OEBError("Spine is empty")
def _guide_from_opf(self, opf): def _guide_from_opf(self, opf):
self.guide = guide = Guide(self) self.guide = guide = Guide(self)
@ -886,8 +894,11 @@ class OEBBook(object):
if len(result) != 1: if len(result) != 1:
return False return False
id = result[0] id = result[0]
ncx = self.manifest[id].data if id not in self.manifest.ids:
self.manifest.remove(id) return False
item = self.manifest.ids[id]
ncx = item.data
self.manifest.remove(item)
title = xpath(ncx, 'ncx:docTitle/ncx:text/text()')[0] title = xpath(ncx, 'ncx:docTitle/ncx:text/text()')[0]
self.toc = toc = TOC(title) self.toc = toc = TOC(title)
navmaps = xpath(ncx, 'ncx:navMap') navmaps = xpath(ncx, 'ncx:navMap')