X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/55e8586118a522d262a7b337ccf8a83cfb458054..134f8c2e9895aae155922f2b6bef9144ded7dfdc:/librarian/epub.py?ds=sidebyside
diff --git a/librarian/epub.py b/librarian/epub.py
index 034d82d..de136ea 100644
--- a/librarian/epub.py
+++ b/librarian/epub.py
@@ -7,6 +7,7 @@ from __future__ import with_statement
import os
import os.path
+import re
import subprocess
from StringIO import StringIO
from copy import deepcopy
@@ -14,10 +15,11 @@ from lxml import etree
import zipfile
from tempfile import mkdtemp, NamedTemporaryFile
from shutil import rmtree
+from mimetypes import guess_type
from librarian import RDFNS, WLNS, NCXNS, OPFNS, XHTMLNS, OutputFile
-from librarian.cover import ImageCover as WLCover
-
+from librarian.cover import WLCover, FutureOfCopyrightCover
+from librarian.latex import LatexFragment
from librarian import functions, get_resource
functions.reg_person_name()
@@ -79,7 +81,7 @@ def replace_characters(node):
return text.replace(u"\ufeff", u"")\
.replace("---", u"\u2014")\
.replace("--", u"\u2013")\
- .replace(",,", u"â")\
+ .replace(",,", u"\u201E")\
.replace('"', u"\u201D")\
.replace("'", u"\u2019")
if node.tag in ('uwaga', 'extra'):
@@ -109,31 +111,74 @@ def find_annotations(annotations, source, part_no):
find_annotations(annotations, child, part_no)
+class Stanza(object):
+ """
+ Converts / verse endings into verse elements in a stanza.
+
+ Slashes may only occur directly in the stanza. Any slashes in subelements
+ will be ignored, and the subelements will be put inside verse elements.
+
+ >>> s = etree.fromstring("a c c/\\nbx/\\nyc/ \\nd")
+ >>> Stanza(s).versify()
+ >>> print etree.tostring(s)
+ a c cbx/
+ ycd
+
+ """
+ def __init__(self, stanza_elem):
+ self.stanza = stanza_elem
+ self.verses = []
+ self.open_verse = None
+
+ def versify(self):
+ self.push_text(self.stanza.text)
+ for elem in self.stanza:
+ self.push_elem(elem)
+ self.push_text(elem.tail)
+ tail = self.stanza.tail
+ self.stanza.clear()
+ self.stanza.tail = tail
+ self.stanza.extend(self.verses)
+
+ def open_normal_verse(self):
+ self.open_verse = self.stanza.makeelement("wers_normalny")
+ self.verses.append(self.open_verse)
+
+ def get_open_verse(self):
+ if self.open_verse is None:
+ self.open_normal_verse()
+ return self.open_verse
+
+ def push_text(self, text):
+ if not text:
+ return
+ for i, verse_text in enumerate(re.split(r"/\s*\n", text)):
+ if i:
+ self.open_normal_verse()
+ verse = self.get_open_verse()
+ if len(verse):
+ verse[-1].tail = (verse[-1].tail or "") + verse_text
+ else:
+ verse.text = (verse.text or "") + verse_text
+
+ def push_elem(self, elem):
+ if elem.tag.startswith("wers"):
+ verse = deepcopy(elem)
+ verse.tail = None
+ self.verses.append(verse)
+ self.open_verse = verse
+ else:
+ appended = deepcopy(elem)
+ appended.tail = None
+ self.get_open_verse().append(appended)
+
+
def replace_by_verse(tree):
""" Find stanzas and create new verses in place of a '/' character """
stanzas = tree.findall('.//' + WLNS('strofa'))
- for node in stanzas:
- for child_node in node:
- if child_node.tag in ('slowo_obce', 'wyroznienie'):
- foreign_verses = inner_xml(child_node).split('/\n')
- if len(foreign_verses) > 1:
- new_foreign = ''
- for foreign_verse in foreign_verses:
- if foreign_verse.startswith('', foreign_verse, ''))
- set_inner_xml(child_node, new_foreign)
- verses = inner_xml(node).split('/\n')
- if len(verses) > 1:
- modified_inner_xml = ''
- for verse in verses:
- if verse.startswith('', verse, ''))
- set_inner_xml(node, modified_inner_xml)
+ for stanza in stanzas:
+ Stanza(stanza).versify()
def add_to_manifest(manifest, partno):
@@ -248,15 +293,14 @@ def chop(main_text):
last_node_part = False
for one_part in main_text:
name = one_part.tag
- #if name == 'naglowek_czesc':
- # yield part_xml
- # last_node_part = True
- # main_xml_part[:] = [deepcopy(one_part)]
- #elif not last_node_part and name in ("naglowek_rozdzial", "naglowek_akt", "srodtytul"):
- # yield part_xml
- # main_xml_part[:] = [deepcopy(one_part)]
- #else:
- if True:
+ if name == 'naglowek_czesc':
+ yield part_xml
+ last_node_part = True
+ main_xml_part[:] = [deepcopy(one_part)]
+ elif not last_node_part and name in ("naglowek_rozdzial", "naglowek_akt", "srodtytul"):
+ yield part_xml
+ main_xml_part[:] = [deepcopy(one_part)]
+ else:
main_xml_part.append(deepcopy(one_part))
last_node_part = False
yield part_xml
@@ -266,36 +310,57 @@ def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_s
""" transforms one chunk, returns a HTML string, a TOC object and a set of used characters """
toc = TOC()
- #for element in chunk_xml[0]:
- # if element.tag in ("naglowek_czesc", "naglowek_rozdzial", "naglowek_akt", "srodtytul"):
- # toc.add(node_name(element), "part%d.html" % chunk_no)
- # elif element.tag in ('naglowek_podrozdzial', 'naglowek_scena'):
- # subnumber = toc.add(node_name(element), "part%d.html" % chunk_no, level=1, is_part=False)
- # element.set('sub', str(subnumber))
+ for element in chunk_xml[0]:
+ if element.tag in ("naglowek_czesc", "naglowek_rozdzial", "naglowek_akt", "srodtytul"):
+ toc.add(node_name(element), "part%d.html" % chunk_no)
+ elif element.tag in ('naglowek_podrozdzial', 'naglowek_scena'):
+ subnumber = toc.add(node_name(element), "part%d.html" % chunk_no, level=1, is_part=False)
+ element.set('sub', str(subnumber))
if empty:
if not _empty_html_static:
_empty_html_static.append(open(get_resource('epub/emptyChunk.html')).read())
chars = set()
output_html = _empty_html_static[0]
else:
- if chunk_no == 1:
- html_tree = xslt(chunk_xml, get_resource('epub/xsltScheme-FoC.xsl'))
- else:
- find_annotations(annotations, chunk_xml, chunk_no)
- replace_by_verse(chunk_xml)
- html_tree = xslt(chunk_xml, get_resource('epub/xsltScheme.xsl'))
+ find_annotations(annotations, chunk_xml, chunk_no)
+ replace_by_verse(chunk_xml)
+ html_tree = xslt(chunk_xml, get_resource('epub/xsltScheme.xsl'))
chars = used_chars(html_tree.getroot())
output_html = etree.tostring(html_tree, method="html", pretty_print=True)
return output_html, toc, chars
+def flatten_image_paths(wldoc):
+ root = wldoc.edoc.getroot()
+ for node in root.findall(".//ilustr"):
+ node.attrib['src'] = os.path.basename(node.attrib['src'])
+ return wldoc
+
+def render_latex(wldoc, prefix="latex"):
+ """
+ Renders CODE as images and returns
+ (changed_wldoc, [ (epub_filepath1, latexfragment_object1), ... ]
+"""
+ root = wldoc.edoc.getroot()
+ latex_nodes = root.findall(".//latex")
+ images = []
+ for ln in latex_nodes:
+ fragment = LatexFragment(ln.text, resize=40)
+ images.append((os.path.join(prefix, fragment.filename), fragment))
+ ln.tag = "img"
+ ln.text = os.path.join(prefix, fragment.filename)
+
+ return wldoc, images
+
+
def transform(wldoc, verbose=False,
style=None, html_toc=False,
- sample=None, cover=None, flags=None):
+ sample=None, cover=None, flags=None, resources=None,
+ intro_file=None, cover_file=None):
""" produces a EPUB file
sample=n: generate sample e-book (with at least n paragraphs)
- cover: a cover.Cover object or True for default
+ cover: a cover.Cover factory or True for default
flags: less-advertising, without-fonts, working-copy
"""
@@ -306,16 +371,7 @@ def transform(wldoc, verbose=False,
# every input file will have a TOC entry,
# pointing to starting chunk
-
- # hack for FoC:
- if wldoc.book_info.author is not None:
- toc_title = "%s, %s" % (wldoc.book_info.author.readable(), wldoc.book_info.title)
- note = wldoc.edoc.find('//dzielo_nadrzedne')
- if note is not None:
- toc_title += " (%s)" % note.text
- else:
- toc_title = wldoc.book_info.title
- toc = TOC(toc_title, "part%d.html" % chunk_counter)
+ toc = TOC(wldoc.book_info.title, "part%d.html" % chunk_counter)
chars = set()
if first:
# write book title page
@@ -324,8 +380,7 @@ def transform(wldoc, verbose=False,
zip.writestr('OPS/title.html',
etree.tostring(html_tree, method="html", pretty_print=True))
# add a title page TOC entry
- toc.add(u"Title page", "title.html")
- toc.add(u"Dear readers!", "part1.html")
+ toc.add(u"TytuÅ", "title.html")
elif wldoc.book_info.parts:
# write title page for every parent
if sample is not None and sample <= 0:
@@ -349,6 +404,8 @@ def transform(wldoc, verbose=False,
if main_text.tag == RDFNS('RDF'):
main_text = None
+ flatten_image_paths(wldoc)
+
if main_text is not None:
for chunk_xml in chop(main_text):
empty = False
@@ -382,6 +439,10 @@ def transform(wldoc, verbose=False,
for flag in flags:
document.edoc.getroot().set(flag, 'yes')
+ # add editors info
+ document.edoc.getroot().set('editors', u', '.join(sorted(
+ editor.readable() for editor in document.editors())))
+
opf = xslt(document.book_info.to_etree(), get_resource('epub/xsltContent.xsl'))
manifest = opf.find('.//' + OPFNS('manifest'))
guide = opf.find('.//' + OPFNS('guide'))
@@ -401,27 +462,38 @@ def transform(wldoc, verbose=False,
'' \
'')
- #zip.write(get_resource('res/wl-logo-small.png'), os.path.join('OPS', 'logo_wolnelektury.png'))
- #zip.write(get_resource('res/jedenprocent.png'), os.path.join('OPS', 'jedenprocent.png'))
- zip.write('logo.png', os.path.join('OPS', 'logo.png'))
+ zip.write(get_resource('res/wl-logo-small.png'), os.path.join('OPS', 'logo_wolnelektury.png'))
+ zip.write(get_resource('res/logo.png'), os.path.join('OPS', 'logo.png'))
+ zip.write(get_resource('res/jedenprocent.png'), os.path.join('OPS', 'jedenprocent.png'))
if not style:
style = get_resource('epub/style.css')
zip.write(style, os.path.join('OPS', 'style.css'))
+ document, latex_images = render_latex(document)
+ for image in latex_images:
+ zip.write(image[1].path, os.path.join('OPS', image[0]))
+ image[1].remove()
+
+ if resources:
+ if os.path.isdir(resources):
+ for dp, dirs, files in os.walk(resources):
+ for fname in files:
+ fpath = os.path.join(dp, fname)
+ if os.path.isfile(fpath):
+ zip.write(fpath, os.path.join('OPS', fname))
+ manifest.append(etree.fromstring(
+ '