X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/09dded3d8606e8e4406fffcf477ceb4a1c97fee2..0859d20e39ee940904f2bfdc36572000109b2214:/librarian/pdf.py
diff --git a/librarian/pdf.py b/librarian/pdf.py
index 94441db..bb1ba64 100644
--- a/librarian/pdf.py
+++ b/librarian/pdf.py
@@ -19,14 +19,13 @@ 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
+from librarian import ParseError, DCNS
from librarian import functions
-
functions.reg_substitute_entities()
-functions.reg_person_name()
functions.reg_strip()
functions.reg_starts_white()
functions.reg_ends_white()
@@ -45,27 +44,23 @@ def insert_tags(doc, split_re, tagname):
ABCXYZ
"""
- 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 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):
@@ -84,7 +79,6 @@ 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
@@ -126,6 +120,16 @@ def hack_motifs(doc):
break
+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_resource(path):
return os.path.join(os.path.dirname(__file__), path)
@@ -153,11 +157,12 @@ def package_available(package, args='', verbose=False):
return p == 0
-def transform(provider, slug, 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):
""" 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
@@ -167,20 +172,24 @@ def transform(provider, slug, output_file=None, output_dir=None, make_dir=False,
# Parse XSLT
try:
- # check for latex packages
- if not package_available('morefloats', 'maxfloats=19', verbose=verbose):
- document.edoc.getroot().set('old-morefloats', 'yes')
- print >> sys.stderr, """
-==============================================================================
-LaTeX `morefloats' package is older than v.1.0c or not available at all.
-Some documents with many motifs in long stanzas or paragraphs may not compile.
-=============================================================================="""
+ 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(provider, 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')
# hack the tree
move_motifs_inside(document.edoc)
hack_motifs(document.edoc)
+ parse_creator(document.edoc)
substitute_hyphens(document.edoc)
fix_hanging(document.edoc)
@@ -229,7 +238,10 @@ Some documents with many motifs in long stanzas or paragraphs may not compile.
os.makedirs(output_dir)
except OSError:
pass
- output_path = os.path.join(output_dir, '%s.pdf' % slug)
+ 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'):
@@ -246,7 +258,7 @@ Some documents with many motifs in long stanzas or paragraphs may not compile.
raise ParseError(e)
-def load_including_children(provider, slug=None, uri=None):
+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
"""
@@ -255,13 +267,17 @@ def load_including_children(provider, slug=None, uri=None):
f = provider.by_uri(uri)
elif slug:
f = provider[slug]
+ elif file_path:
+ f = open(file_path, 'r')
else:
- raise ValueError('Neither slug nor URI provided for a book.')
+ raise ValueError('Neither slug, URI nor file path provided for a book.')
document = WLDocument.from_file(f, True,
parse_dublincore=True,
preserve_lines=False)
+ f.close()
+
for child_uri in document.book_info.parts:
child = load_including_children(provider, uri=child_uri)
document.edoc.getroot().append(child.edoc.getroot())