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
+from librarian import ParseError, DCNS, get_resource, OutputFile
from librarian import functions
+from librarian.cover import ImageCover as WLCover
functions.reg_substitute_entities()
functions.reg_strip()
functions.reg_starts_white()
functions.reg_ends_white()
+functions.reg_texcommand()
STYLESHEETS = {
- 'wl2tex': 'xslt/wl2tex.xslt',
+ 'wl2tex': 'pdf/wl2tex.xslt',
}
+#CUSTOMIZATIONS = [
+# 'nofootnotes',
+# 'nothemes',
+# 'defaultleading',
+# 'onehalfleading',
+# 'doubleleading',
+# 'nowlfont',
+# ]
-def insert_tags(doc, split_re, tagname):
- """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree
+def insert_tags(doc, split_re, tagname, exclude=None):
+ """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree
>>> t = etree.fromstring('<a><b>A-B-C</b>X-Y-Z</a>');
>>> insert_tags(t, re.compile('-'), 'd');
"""
for elem in doc.iter(tag=etree.Element):
+ if exclude and elem.tag in exclude:
+ continue
if elem.text:
chunks = split_re.split(elem.text)
while len(chunks) > 1:
def substitute_hyphens(doc):
- insert_tags(doc,
+ insert_tags(doc,
re.compile("(?<=[^-\s])-(?=[^-\s])"),
- "dywiz")
+ "dywiz",
+ exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+ )
def fix_hanging(doc):
- insert_tags(doc,
+ insert_tags(doc,
re.compile("(?<=\s\w)\s+"),
- "nbsp")
+ "nbsp",
+ exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+ )
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 motif in master.xpath('motyw'):
- print motif.text
for sib in motif.itersiblings():
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
def parse_creator(doc):
- """ find all dc:creator tags and add dc:creator_parsed with forenames first """
- for creator in doc.findall('//'+DCNS('creator')):
- p = Person.from_text(creator.text)
- creator_parsed = deepcopy(creator)
- creator_parsed.tag = DCNS('creator_parsed')
- creator_parsed.text = ' '.join(p.first_names + (p.last_name,))
- creator.getparent().insert(0, creator_parsed)
-
+ """ 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]:
+ if not person.text:
+ continue
+ p = Person.from_text(person.text)
+ person_parsed = deepcopy(person)
+ person_parsed.tag = person.tag + '_parsed'
+ person_parsed.set('sortkey', person.text)
+ person_parsed.text = p.readable()
+ person.getparent().insert(0, person_parsed)
-def get_resource(path):
- return os.path.join(os.path.dirname(__file__), path)
def get_stylesheet(name):
return get_resource(STYLESHEETS[name])
def package_available(package, args='', verbose=False):
- """ check if a verion of a latex package accepting given args is available """
+ """ check if a verion of a latex package accepting given args is available """
tempdir = mkdtemp('-wl2pdf-test')
fpath = os.path.join(tempdir, 'test.tex')
f = open(fpath, 'w')
f.write(r"""
- \documentclass{book}
+ \documentclass{wl}
\usepackage[%s]{%s}
\begin{document}
\end{document}
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):
+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 <output_dir>/<author>/<slug>.pdf istead of <output_dir>/<slug>.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
+ 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)
+ document = load_including_children(wldoc)
+
+ if cover:
+ 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')
# check for LaTeX packages
- if not package_available('morefloats', 'maxfloats=19'):
- # using old morefloats or none at all
- document.edoc.getroot().set('old-morefloats', 'yes')
+ 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))
# 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)
- del document # no longer needed large object :)
+ texml = document.transform(style)
+ etree.dump(texml.getroot())
# TeXML -> LaTeX
temp = mkdtemp('-wl2pdf')
+
+ if cover:
+ with open(os.path.join(temp, 'cover.jpg'), 'w') as f:
+ the_cover.save(f)
+
+ shutil.copy("logo.eps", temp)
+ for img in document.edoc.findall('//ilustr'):
+ shutil.copy(os.path.join(imgdir, img.get('src')), temp)
+
+
+ del document # no longer needed large object :)
+
tex_path = os.path.join(temp, 'doc.tex')
fout = open(tex_path, 'w')
process(StringIO(texml), fout, 'utf-8')
shutil.copy(tex_path, save_tex)
# LaTeX -> PDF
- shutil.copy(get_resource('pdf/wl.sty'), temp)
- shutil.copy(get_resource('pdf/wl-logo.png'), temp)
+ 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)
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.')
- document = WLDocument.from_file(f, True,
- parse_dublincore=True,
- preserve_lines=False)
+ text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
- f.close()
+ document = WLDocument.from_string(text, parse_dublincore=True)
+ document.swap_endlines()
for child_uri in document.book_info.parts:
- 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
fnp / librarian.git / blobdiff