X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/b6ec0976a383cc1823f4a199bc3e6dc40880b049..cc0a99f82b1780ed842188fc22fbe18af6a9bc3b:/librarian/pdf.py diff --git a/librarian/pdf.py b/librarian/pdf.py index 02438a6..c899afa 100644 --- a/librarian/pdf.py +++ b/librarian/pdf.py @@ -3,6 +3,12 @@ # 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 @@ -21,6 +27,8 @@ from librarian.dcparser import Person from librarian.parser import WLDocument from librarian import ParseError, DCNS, get_resource, OutputFile from librarian import functions +from librarian.cover import DefaultEbookCover +from .sponsor import sponsor_logo functions.reg_substitute_entities() @@ -33,19 +41,21 @@ 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 - >>> t = etree.fromstring('A-B-CX-Y-Z'); - >>> insert_tags(t, re.compile('-'), 'd'); + >>> t = etree.fromstring('A-B-CX-Y-Z') + >>> insert_tags(t, re.compile('-'), 'd') >>> print etree.tostring(t) ABCXYZ """ @@ -87,12 +97,26 @@ def fix_hanging(doc): ) +def fix_tables(doc): + for kol in doc.iter(tag='kol'): + if kol.tail is not None: + if not kol.tail.strip(): + kol.tail = None + for table in doc.iter(tag='tabela'): + if table.get('ramka') == '1' or table.get('ramki') == '1': + table.set('_format', '|' + 'X|' * len(table[0])) + else: + table.set('_format', 'X' * len(table[0])) + + def move_motifs_inside(doc): """ moves motifs to be into block elements """ - for master in doc.xpath('//powiesc|//opowiadanie|//liryka_l|//liryka_lp|//dramat_wierszowany_l|//dramat_wierszowany_lp|//dramat_wspolczesny'): + for master in doc.xpath('//powiesc|//opowiadanie|//liryka_l|//liryka_lp|' + '//dramat_wierszowany_l|//dramat_wierszowany_lp|//dramat_wspolczesny'): for motif in master.xpath('motyw'): for sib in motif.itersiblings(): - if sib.tag not in ('sekcja_swiatlo', 'sekcja_asterysk', 'separator_linia', 'begin', 'end', 'motyw', 'extra', 'uwaga'): + if sib.tag not in ('sekcja_swiatlo', 'sekcja_asterysk', 'separator_linia', + 'begin', 'end', 'motyw', 'extra', 'uwaga'): # motif shouldn't have a tail - it would be untagged text motif.tail = None motif.getparent().remove(motif) @@ -133,10 +157,13 @@ def hack_motifs(doc): def parse_creator(doc): - """ find all dc:creator and dc.contributor tags and add *_parsed versions with forenames first """ - for person in doc.xpath("|".join('//dc:'+(tag) for tag in ( - 'creator', 'contributor.translator', 'contributor.editor', 'contributor.technical_editor')), - namespaces = {'dc': str(DCNS)})[::-1]: + """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')), + namespaces={'dc': str(DCNS)})[::-1]: if not person.text: continue p = Person.from_text(person.text) @@ -179,31 +206,51 @@ def transform(wldoc, verbose=False, save_tex=None, morefloats=None, 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: + 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, width=1200) + 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 + editors = document.editors() + if editors: + root.set('editors', u', '.join(sorted( + editor.readable() for editor in 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) @@ -211,22 +258,35 @@ def transform(wldoc, verbose=False, save_tex=None, morefloats=None, parse_creator(document.edoc) substitute_hyphens(document.edoc) fix_hanging(document.edoc) + fix_tables(document.edoc) # wl -> TeXML style_filename = get_stylesheet("wl2tex") style = etree.parse(style_filename) - - texml = document.transform(style) + functions.reg_mathml_latex() # TeXML -> LaTeX temp = mkdtemp('-wl2pdf') + for sponsor in book_info.sponsors: + ins = etree.Element("data-sponsor", name=sponsor) + logo = sponsor_logo(sponsor) + if logo: + fname = 'sponsor-%s' % os.path.basename(logo) + shutil.copy(logo, os.path.join(temp, fname)) + ins.set('src', fname) + root.insert(0, ins) + + if book_info.sponsor_note: + root.set("sponsor-note", book_info.sponsor_note) + + texml = document.transform(style) + 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, quality=80) - del document # no longer needed large object :) + del document # no longer needed large object :) tex_path = os.path.join(temp, 'doc.tex') fout = open(tex_path, 'w') @@ -241,7 +301,10 @@ def transform(wldoc, verbose=False, save_tex=None, morefloats=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: @@ -251,7 +314,8 @@ def transform(wldoc, verbose=False, save_tex=None, morefloats=None, 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') @@ -281,7 +345,7 @@ def load_including_children(wldoc=None, provider=None, uri=None): text = re.sub(ur"([\u0400-\u04ff]+)", ur"\1", text) - document = WLDocument.from_string(text, parse_dublincore=True) + document = WLDocument.from_string(text, parse_dublincore=True, provider=provider) document.swap_endlines() for child_uri in document.book_info.parts: