from lxml import etree
from lxml.etree import XMLSyntaxError, XSLTApplyError
+from librarian.dcparser import Person
from librarian.parser import WLDocument
-from librarian import ParseError
+from librarian import ParseError, DCNS, get_resource
from librarian import functions
-
functions.reg_substitute_entities()
-functions.reg_person_name()
functions.reg_strip()
functions.reg_starts_white()
functions.reg_ends_white()
+functions.reg_texcommand()
STYLESHEETS = {
'wl2tex': 'xslt/wl2tex.xslt',
}
-def insert_tags(doc, split_re, tagname):
+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>');
<a><b>A<d/>B<d/>C</b>X<d/>Y<d/>Z</a>
"""
- for elem in doc.iter():
- try:
- if elem.text:
- chunks = split_re.split(elem.text)
- while len(chunks) > 1:
- ins = etree.Element(tagname)
- ins.tail = chunks.pop()
- elem.insert(0, ins)
- elem.text = chunks.pop(0)
- if elem.tail:
- chunks = split_re.split(elem.tail)
- parent = elem.getparent()
- ins_index = parent.index(elem) + 1
- while len(chunks) > 1:
- ins = etree.Element(tagname)
- ins.tail = chunks.pop()
- parent.insert(ins_index, ins)
- elem.tail = chunks.pop(0)
- except TypeError, e:
- # element with no children, like comment
- pass
+ 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:
+ ins = etree.Element(tagname)
+ ins.tail = chunks.pop()
+ elem.insert(0, ins)
+ elem.text = chunks.pop(0)
+ if elem.tail:
+ chunks = split_re.split(elem.tail)
+ parent = elem.getparent()
+ ins_index = parent.index(elem) + 1
+ while len(chunks) > 1:
+ ins = etree.Element(tagname)
+ ins.tail = chunks.pop()
+ parent.insert(ins_index, ins)
+ elem.tail = chunks.pop(0)
def substitute_hyphens(doc):
insert_tags(doc,
re.compile("(?<=[^-\s])-(?=[^-\s])"),
- "dywiz")
+ "dywiz",
+ exclude=[DCNS("identifier.url"), DCNS("rights.license")]
+ )
def fix_hanging(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
break
-def get_resource(path):
- return os.path.join(os.path.dirname(__file__), path)
+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)
+
def get_stylesheet(name):
return get_resource(STYLESHEETS[name])
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(provider, slug=None, file_path=None,
+ output_file=None, output_dir=None, make_dir=False, verbose=False, save_tex=None, morefloats=None):
""" produces a PDF file with XeLaTeX
provider: a DocProvider
make_dir: writes output to <output_dir>/<author>/<slug>.pdf istead of <output_dir>/<slug>.pdf
verbose: prints all output from LaTeX
save_tex: path to save the intermediary LaTeX file to
+ morefloats (old/new/none): force specific morefloats
"""
# Parse XSLT
document = load_including_children(provider, slug=slug)
# 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')
# hack the tree
move_motifs_inside(document.edoc)
hack_motifs(document.edoc)
+ parse_creator(document.edoc)
substitute_hyphens(document.edoc)
fix_hanging(document.edoc)
else:
raise ValueError('Neither slug, URI nor file path provided for a book.')
- document = WLDocument.from_file(f, True,
- parse_dublincore=True,
- preserve_lines=False)
+ text = f.read().decode('utf-8')
+ text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
+
+ document = WLDocument.from_string(text, True,
+ parse_dublincore=True)
f.close()
fnp / librarian.git / blobdiff