# This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
#
+"""PDF creation library.
+
+Creates one big XML from the book and its children, converts it to LaTeX
+with TeXML, then runs it by XeLaTeX.
+
+"""
from __future__ import with_statement
import os
import os.path
from librarian.dcparser import Person
from librarian.parser import WLDocument
-from librarian import ParseError, DCNS, get_resource, OutputFile
+from librarian import ParseError, DCNS, get_resource, IOFile, Format
from librarian import functions
-from librarian.cover import WLCover
functions.reg_substitute_entities()
'wl2tex': 'pdf/wl2tex.xslt',
}
-#CUSTOMIZATIONS = [
-# 'nofootnotes',
-# 'nothemes',
-# 'defaultleading',
-# 'onehalfleading',
-# 'doubleleading',
-# 'nowlfont',
-# ]
-
def insert_tags(doc, split_re, tagname, exclude=None):
""" inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree
insert_tags(doc,
re.compile("(?<=[^-\s])-(?=[^-\s])"),
"dywiz",
- exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+ exclude=[DCNS("identifier.url"), DCNS("rights.license"), 'www']
)
def parse_creator(doc):
- """ find all dc:creator and dc.contributor tags and add *_parsed versions with forenames first """
+ """Generates readable versions of creator and translator tags.
+
+ Finds all dc:creator and dc.contributor.translator tags
+ and adds *_parsed versions with forenames first.
+ """
for person in doc.xpath("|".join('//dc:'+(tag) for tag in (
- 'creator', 'contributor.translator', 'contributor.editor', 'contributor.technical_editor')),
+ 'creator', 'contributor.translator')),
namespaces = {'dc': str(DCNS)})[::-1]:
if not person.text:
continue
return p == 0
-def transform(wldoc, verbose=False, save_tex=None, morefloats=None,
- cover=None, flags=None, customizations=None):
- """ produces a PDF file with XeLaTeX
+def load_including_children(wldoc=None, provider=None, uri=None):
+ """ Makes one big xml file with children inserted at end.
+
+ Either wldoc or provider and URI must be provided.
+ """
+
+ if uri and provider:
+ f = provider.by_uri(uri)
+ text = f.read().decode('utf-8')
+ f.close()
+ elif wldoc is not None:
+ text = etree.tostring(wldoc.edoc, encoding=unicode)
+ provider = wldoc.provider
+ else:
+ raise ValueError('Neither a WLDocument, nor provider and URI were provided.')
+
+ text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
+
+ document = WLDocument.from_string(text,
+ parse_dublincore=True, provider=provider)
+ document.swap_endlines()
+
+ for child_uri in document.book_info.parts:
+ child = load_including_children(provider=provider, uri=child_uri)
+ document.edoc.getroot().append(child.edoc.getroot())
+ return document
+
+
+class PDFFormat(Format):
+ """ Base PDF format.
+
+ Available customization:
+ nofootnotes: Doesn't do footnotes.
+ nothemes: Doesn't do themes.
+ defaultleading: Default leading.
+ onehalfleading: Bigger leading.
+ doubleleading: Big leading.
+ nowlfont: Uses standard TeX font instead of JUnicodeWL.
- wldoc: a WLDocument
- verbose: prints all output from LaTeX
- save_tex: path to save the intermediary LaTeX file to
- morefloats (old/new/none): force specific morefloats
- cover: a cover.Cover object
- flags: less-advertising,
- customizations: user requested customizations regarding various formatting parameters (passed to wl LaTeX class)
"""
- # Parse XSLT
- try:
- document = load_including_children(wldoc)
-
- if cover:
- if cover is True:
- cover = WLCover
- document.edoc.getroot().set('data-cover-width', str(cover.width))
- document.edoc.getroot().set('data-cover-height', str(cover.height))
- if cover.uses_dc_cover:
- if document.book_info.cover_by:
- document.edoc.getroot().set('data-cover-by', document.book_info.cover_by)
- if document.book_info.cover_source:
- document.edoc.getroot().set('data-cover-source', document.book_info.cover_source)
- if flags:
- for flag in flags:
- document.edoc.getroot().set('flag-' + flag, 'yes')
-
- # check for LaTeX packages
- if morefloats:
- document.edoc.getroot().set('morefloats', morefloats.lower())
- elif package_available('morefloats', 'maxfloats=19'):
- document.edoc.getroot().set('morefloats', 'new')
-
- # add customizations
- if customizations is not None:
- document.edoc.getroot().set('customizations', u','.join(customizations))
+ cover_class = None
+ tex_passes = 1
+ style = get_resource('pdf/default.sty')
+ cover = None
+
+ @property
+ def has_cover(self):
+ """ For use in XSLT. """
+ return self.cover is not None
+
+ @property
+ def customization_str(self):
+ """ For use in XSLT. """
+ return u','.join(k for k, v in self.customization.items() if v)
+
+ def get_document(self):
+ document = load_including_children(self.wldoc)
+ root = document.edoc.getroot()
+ root.set('editors', u', '.join(sorted(
+ editor.readable() for editor in document.editors())))
# hack the tree
move_motifs_inside(document.edoc)
parse_creator(document.edoc)
substitute_hyphens(document.edoc)
fix_hanging(document.edoc)
+ return document
- # wl -> TeXML
+ def get_texml(self):
style_filename = get_stylesheet("wl2tex")
- style = etree.parse(style_filename)
-
- texml = document.transform(style)
-
- # TeXML -> LaTeX
+ functions.reg_get(self)
+ try:
+ style = etree.parse(style_filename)
+ texml = self.get_document().transform(style)
+ return texml
+ except (XMLSyntaxError, XSLTApplyError), e:
+ raise ParseError(e)
+
+ def get_tex_dir(self):
+ texml = self.get_texml()
temp = mkdtemp('-wl2pdf')
-
- if cover:
- c = cover(document.book_info)
- with open(os.path.join(temp, 'cover.png'), 'w') as f:
- c.save(f)
-
- del document # no longer needed large object :)
-
+ # Save TeX file
tex_path = os.path.join(temp, 'doc.tex')
- fout = open(tex_path, 'w')
- process(StringIO(texml), fout, 'utf-8')
- fout.close()
- del texml
-
- if save_tex:
- shutil.copy(tex_path, save_tex)
-
- # LaTeX -> PDF
+ with open(tex_path, 'w') as fout:
+ process(StringIO(texml), fout, 'utf-8')
+ if self.save_tex:
+ shutil.copy(tex_path, self.save_tex)
+ # Copy style
shutil.copy(get_resource('pdf/wl.cls'), temp)
- shutil.copy(get_resource('res/wl-logo.png'), temp)
+ shutil.copy(self.style, os.path.join(temp, 'style.sty'))
+ for sfile in ['wasysym.sty', 'uwasyvar.fd', 'uwasy.fd']:
+ shutil.copy(get_resource(os.path.join('res/wasysym', sfile)), temp)
- cwd = os.getcwd()
+ # Save attachments
+ if self.cover:
+ self.cover.for_pdf().dump_to(os.path.join(temp, 'makecover.sty'))
+ return temp
+
+ def get_pdf(self):
+ temp = self.get_tex_dir()
+ tex_path = os.path.join(temp, 'doc.tex')
+ try:
+ cwd = os.getcwd()
+ except OSError:
+ cwd = None
os.chdir(temp)
- if verbose:
- p = call(['xelatex', tex_path])
+ if self.verbose:
+ for i in range(self.tex_passes):
+ p = call(['xelatex', tex_path])
else:
- p = call(['xelatex', '-interaction=batchmode', tex_path], stdout=PIPE, stderr=PIPE)
+ for i in range(self.tex_passes):
+ p = call(['xelatex', '-interaction=batchmode', tex_path],
+ stdout=PIPE, stderr=PIPE)
if p:
raise ParseError("Error parsing .tex file")
- os.chdir(cwd)
+ if cwd is not None:
+ os.chdir(cwd)
output_file = NamedTemporaryFile(prefix='librarian', suffix='.pdf', delete=False)
pdf_path = os.path.join(temp, 'doc.pdf')
shutil.move(pdf_path, output_file.name)
shutil.rmtree(temp)
- return OutputFile.from_filename(output_file.name)
+ return IOFile.from_filename(output_file.name)
- except (XMLSyntaxError, XSLTApplyError), e:
- raise ParseError(e)
+ def build(self, verbose=False, save_tex=None, morefloats=None):
+ """ morefloats: new/old/none
+ """
+ self.verbose = verbose
+ self.save_tex = save_tex
+ if morefloats is None and package_available('morefloats', 'maxfloats=19'):
+ morefloats = 'new'
+ self.morefloats = morefloats
-def load_including_children(wldoc=None, provider=None, uri=None):
- """ Makes one big xml file with children inserted at end.
-
- Either wldoc or provider and URI must be provided.
- """
+ book_info = self.wldoc.book_info
+ if self.cover_class:
+ self.cover = self.cover_class(book_info)
- if uri and provider:
- f = provider.by_uri(uri)
- text = f.read().decode('utf-8')
- f.close()
- elif wldoc is not None:
- text = etree.tostring(wldoc.edoc, encoding=unicode)
- provider = wldoc.provider
- else:
- raise ValueError('Neither a WLDocument, nor provider and URI were provided.')
-
- text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
-
- document = WLDocument.from_string(text, parse_dublincore=True)
- document.swap_endlines()
-
- for child_uri in document.book_info.parts:
- child = load_including_children(provider=provider, uri=child_uri)
- document.edoc.getroot().append(child.edoc.getroot())
- return document
+ return self.get_pdf()
fnp / librarian.git / blobdiff