X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/ab4401272d1b8507effd51da00f40bafe81f2b44..90a4097e55f34b00e81724ccbdf6893a8991f810:/librarian/pdf.py?ds=inline diff --git a/librarian/pdf.py b/librarian/pdf.py index 1bfd949..af3d1df 100644 --- a/librarian/pdf.py +++ b/librarian/pdf.py @@ -8,21 +8,20 @@ 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 ImageCover as WLCover functions.reg_substitute_entities() @@ -35,13 +34,14 @@ 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 @@ -173,17 +173,12 @@ 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, - cover=None, flags=None, customizations=None): +def transform(wldoc, verbose=False, save_tex=None, morefloats=None, + cover=None, flags=None, customizations=None, + imgdir=""): """ 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 @@ -194,18 +189,19 @@ def transform(provider, slug=None, file_path=None, # 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) + document = load_including_children(wldoc) 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 = WLCover + the_cover = cover(document.book_info) + document.edoc.getroot().set('data-cover-width', str(the_cover.width)) + document.edoc.getroot().set('data-cover-height', str(the_cover.height)) + if the_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') @@ -221,30 +217,32 @@ def transform(provider, slug=None, file_path=None, document.edoc.getroot().set('customizations', u','.join(customizations)) # hack the tree - move_motifs_inside(document.edoc) - hack_motifs(document.edoc) + #move_motifs_inside(document.edoc) + #hack_motifs(document.edoc) parse_creator(document.edoc) - substitute_hyphens(document.edoc) + if document.book_info.language == 'pol': + 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) texml = document.transform(style) - + # etree.dump(texml.getroot()) # TeXML -> LaTeX 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) + with open(os.path.join(temp, 'cover.jpg'), 'w') as f: + the_cover.save(f) + + shutil.copy("cce_trust.eps", temp) + shutil.copy("logo.eps", temp) + for img in document.edoc.findall('//ilustr'): + # print "--->> %s %s %s" % (imgdir, img, img.get('src')) + shutil.copy(os.path.join(imgdir, img.get('src')), temp) + del document # no longer needed large object :) @@ -260,6 +258,7 @@ def transform(provider, slug=None, file_path=None, # LaTeX -> PDF shutil.copy(get_resource('pdf/wl.cls'), temp) shutil.copy(get_resource('res/wl-logo.png'), temp) + shutil.copy('logo.eps', temp) cwd = os.getcwd() os.chdir(temp) @@ -273,56 +272,39 @@ def transform(provider, slug=None, file_path=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: + print 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) + 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