X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/cec81dc04a23808077505265a3f8bb0e9b6053cb..6791e72f5bb378973798e1d410f7bd42a346cc9e:/librarian/fb2.py?ds=inline diff --git a/librarian/fb2.py b/librarian/fb2.py index b0ad410..d979566 100644 --- a/librarian/fb2.py +++ b/librarian/fb2.py @@ -8,9 +8,32 @@ from copy import deepcopy from lxml import etree from librarian import functions, OutputFile +from .epub import replace_by_verse functions.reg_substitute_entities() +functions.reg_person_name() + + +def sectionify(tree): + """Finds section headers and adds a tree of _section tags.""" + sections = ['naglowek_czesc', + 'naglowek_akt', 'naglowek_rozdzial', 'naglowek_scena', + 'naglowek_podrozdzial'] + section_level = dict((v,k) for (k,v) in enumerate(sections)) + + # We can assume there are just subelements an no text at section level. + for level, section_name in reversed(list(enumerate(sections))): + for header in tree.findall('//' + section_name): + section = header.makeelement("_section") + header.addprevious(section) + section.append(header) + sibling = section.getnext() + while (sibling is not None and + section_level.get(sibling.tag, 1000) > level): + section.append(sibling) + sibling = section.getnext() + def transform(wldoc, verbose=False, cover=None, flags=None): @@ -30,6 +53,11 @@ def transform(wldoc, verbose=False, style_filename = os.path.join(os.path.dirname(__file__), 'fb2/fb2.xslt') style = etree.parse(style_filename) + replace_by_verse(document.edoc) + sectionify(document.edoc) + result = document.transform(style) return OutputFile.from_string(unicode(result).encode('utf-8')) + +# vim:et