X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/cba76114f031d47c93e1af947a350230cbef0a1f..f4cbb15b0bcd2b876872983238a1656fbd990700:/librarian/epub.py?ds=sidebyside
diff --git a/librarian/epub.py b/librarian/epub.py
index bbeb3d7..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 WLCover
-
+from librarian.cover import WLCover, FutureOfCopyrightCover
+from librarian.latex import LatexFragment
from librarian import functions, get_resource
functions.reg_person_name()
@@ -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):
@@ -285,9 +330,33 @@ def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_s
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)
@@ -311,7 +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"Strona tytuÅowa", "title.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:
@@ -335,6 +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
@@ -392,28 +463,48 @@ def transform(wldoc, verbose=False,
'media-type="application/oebps-package+xml" />' \
'')
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(
+ '