X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/ab4401272d1b8507effd51da00f40bafe81f2b44..0081ccb8e51026b5af824b3c641648d23155f1e1:/librarian/pdf.py diff --git a/librarian/pdf.py b/librarian/pdf.py index 1bfd949..2154985 100644 --- a/librarian/pdf.py +++ b/librarian/pdf.py @@ -3,26 +3,31 @@ # 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 import shutil from StringIO import StringIO -from tempfile import mkdtemp +from tempfile import mkdtemp, NamedTemporaryFile import re from copy import deepcopy from subprocess import call, PIPE -import sys - from Texml.processor import process from lxml import etree from lxml.etree import XMLSyntaxError, XSLTApplyError from librarian.dcparser import Person from librarian.parser import WLDocument -from librarian import ParseError, DCNS, get_resource +from librarian import ParseError, DCNS, get_resource, OutputFile from librarian import functions +from librarian.cover import DefaultEbookCover functions.reg_substitute_entities() @@ -35,13 +40,14 @@ STYLESHEETS = { 'wl2tex': 'pdf/wl2tex.xslt', } -CUSTOMIZATIONS = [ - 'nofootnotes', - 'nothemes', - 'onehalfleading', - 'doubleleading', - 'nowlfont', - ] +#CUSTOMIZATIONS = [ +# 'nofootnotes', +# 'nothemes', +# 'defaultleading', +# 'onehalfleading', +# 'doubleleading', +# 'nowlfont', +# ] def insert_tags(doc, split_re, tagname, exclude=None): """ inserts for every occurence of `split_re' in text nodes in the `doc' tree @@ -135,9 +141,13 @@ def hack_motifs(doc): 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 @@ -173,52 +183,58 @@ def package_available(package, args='', verbose=False): return p == 0 -def transform(provider, slug=None, file_path=None, - output_file=None, output_dir=None, make_dir=False, verbose=False, save_tex=None, morefloats=None, +def transform(wldoc, verbose=False, save_tex=None, morefloats=None, cover=None, flags=None, customizations=None): """ produces a PDF file with XeLaTeX - provider: a DocProvider - slug: slug of file to process, available by provider - file_path can be provided instead of a slug - output_file: file-like object or path to output file - output_dir: path to directory to save output file to; either this or output_file must be present - make_dir: writes output to //.pdf istead of /.pdf + 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 + cover: a cover.Cover factory or True for default flags: less-advertising, customizations: user requested customizations regarding various formatting parameters (passed to wl LaTeX class) """ # Parse XSLT try: - if file_path: - if slug: - raise ValueError('slug or file_path should be specified, not both') - document = load_including_children(provider, file_path=file_path) - else: - if not slug: - raise ValueError('either slug or file_path should be specified') - document = load_including_children(provider, slug=slug) + book_info = wldoc.book_info + document = load_including_children(wldoc) + root = document.edoc.getroot() if cover: - document.edoc.getroot().set('data-cover-width', str(cover.width)) - document.edoc.getroot().set('data-cover-height', str(cover.height)) + if cover is True: + cover = DefaultEbookCover + bound_cover = cover(book_info) + root.set('data-cover-width', str(bound_cover.width)) + root.set('data-cover-height', str(bound_cover.height)) + if bound_cover.uses_dc_cover: + if book_info.cover_by: + root.set('data-cover-by', book_info.cover_by) + if book_info.cover_source: + root.set('data-cover-source', + book_info.cover_source) if flags: for flag in flags: - document.edoc.getroot().set('flag-' + flag, 'yes') + root.set('flag-' + flag, 'yes') # check for LaTeX packages if morefloats: - document.edoc.getroot().set('morefloats', morefloats.lower()) + root.set('morefloats', morefloats.lower()) elif package_available('morefloats', 'maxfloats=19'): - document.edoc.getroot().set('morefloats', 'new') + root.set('morefloats', 'new') # add customizations if customizations is not None: - document.edoc.getroot().set('customizations', u','.join(customizations)) + root.set('customizations', u','.join(customizations)) + + # add editors info + root.set('editors', u', '.join(sorted( + editor.readable() for editor in document.editors()))) + if document.book_info.funders: + root.set('funders', u', '.join(document.book_info.funders)) + if document.book_info.thanks: + root.set('thanks', document.book_info.thanks) # hack the tree move_motifs_inside(document.edoc) @@ -227,11 +243,6 @@ def transform(provider, slug=None, file_path=None, substitute_hyphens(document.edoc) fix_hanging(document.edoc) - # find output dir - if make_dir and output_dir is not None: - author = unicode(document.book_info.author) - output_dir = os.path.join(output_dir, author) - # wl -> TeXML style_filename = get_stylesheet("wl2tex") style = etree.parse(style_filename) @@ -242,9 +253,8 @@ def transform(provider, slug=None, file_path=None, temp = mkdtemp('-wl2pdf') if cover: - c = cover(document.book_info.author.readable(), document.book_info.title) with open(os.path.join(temp, 'cover.png'), 'w') as f: - c.save(f) + bound_cover.save(f) del document # no longer needed large object :) @@ -261,7 +271,10 @@ def transform(provider, slug=None, file_path=None, shutil.copy(get_resource('pdf/wl.cls'), temp) shutil.copy(get_resource('res/wl-logo.png'), temp) - cwd = os.getcwd() + try: + cwd = os.getcwd() + except OSError: + cwd = None os.chdir(temp) if verbose: @@ -271,58 +284,42 @@ def transform(provider, slug=None, file_path=None, if p: raise ParseError("Error parsing .tex file") - os.chdir(cwd) + if cwd is not None: + os.chdir(cwd) - # save the PDF + output_file = NamedTemporaryFile(prefix='librarian', suffix='.pdf', delete=False) pdf_path = os.path.join(temp, 'doc.pdf') - if output_dir is not None: - try: - os.makedirs(output_dir) - except OSError: - pass - if slug: - output_path = os.path.join(output_dir, '%s.pdf' % slug) - else: - output_path = os.path.join(output_dir, os.path.splitext(os.path.basename(file_path))[0] + '.pdf') - shutil.move(pdf_path, output_path) - else: - if hasattr(output_file, 'write'): - # file-like object - with open(pdf_path) as f: - output_file.write(f.read()) - output_file.close() - else: - # path to output file - shutil.copy(pdf_path, output_file) + shutil.move(pdf_path, output_file.name) shutil.rmtree(temp) + return OutputFile.from_filename(output_file.name) except (XMLSyntaxError, XSLTApplyError), e: raise ParseError(e) -def load_including_children(provider, slug=None, uri=None, file_path=None): - """ makes one big xml file with children inserted at end - either slug or uri must be provided +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: + if uri and provider: f = provider.by_uri(uri) - elif slug: - f = provider[slug] - elif file_path: - f = open(file_path, 'r') + 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 slug, URI nor file path provided for a book.') + raise ValueError('Neither a WLDocument, nor provider and URI were provided.') - text = f.read().decode('utf-8') text = re.sub(ur"([\u0400-\u04ff]+)", ur"\1", text) - document = WLDocument.from_string(text, True, - parse_dublincore=True) + document = WLDocument.from_string(text, + parse_dublincore=True, provider=provider) + document.swap_endlines() - f.close() for child_uri in document.book_info.parts: - print child_uri - child = load_including_children(provider, uri=child_uri) + child = load_including_children(provider=provider, uri=child_uri) document.edoc.getroot().append(child.edoc.getroot()) return document