X-Git-Url: https://git.mdrn.pl/librarian.git/blobdiff_plain/17a9ed3b7ef12e0786ddf46bf8a52b1087224762..1c58de768c37f60fb6207374cbf523537bfca9ae:/librarian/epub.py

diff --git a/librarian/epub.py b/librarian/epub.py
index bbeb3d7..712863c 100644
--- a/librarian/epub.py
+++ b/librarian/epub.py
@@ -7,9 +7,12 @@ from __future__ import with_statement
 
 import os
 import os.path
+import re
 import subprocess
 from StringIO import StringIO
 from copy import deepcopy
+from mimetypes import guess_type
+
 from lxml import etree
 import zipfile
 from tempfile import mkdtemp, NamedTemporaryFile
@@ -109,31 +112,74 @@ 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("<strofa>a/\\nb<x>x/\\ny</x>c/ \\nd</strofa>")
+    >>> Stanza(s).versify()
+    >>> print etree.tostring(s)
+    <strofa><wers_normalny>a</wers_normalny><wers_normalny>b<x>x/
+    y</x>c</wers_normalny><wers_normalny>d</wers_normalny></strofa>
+    
+    """
+    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 or not text.strip():
+            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.strip()
+            else:
+                verse.text = (verse.text or "") + verse_text.strip()
+
+    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('<wers'):
-                            new_foreign += foreign_verse
-                        else:
-                            new_foreign += ''.join(('<wers_normalny>', foreign_verse, '</wers_normalny>'))
-                    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('<wers') or verse.startswith('<extra'):
-                    modified_inner_xml += verse
-                else:
-                    modified_inner_xml += ''.join(('<wers_normalny>', verse, '</wers_normalny>'))
-            set_inner_xml(node, modified_inner_xml)
+    for stanza in stanzas:
+        Stanza(stanza).versify()
 
 
 def add_to_manifest(manifest, partno):
@@ -285,9 +331,8 @@ def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_s
     return output_html, toc, chars
 
 
-def transform(wldoc, verbose=False,
-              style=None, html_toc=False,
-              sample=None, cover=None, flags=None):
+def transform(wldoc, verbose=False, style=None, html_toc=False,
+              sample=None, cover=None, flags=None, ilustr_path=''):
     """ produces a EPUB file
 
     sample=n: generate sample e-book (with at least n paragraphs)
@@ -380,6 +425,14 @@ def transform(wldoc, verbose=False,
     output_file = NamedTemporaryFile(prefix='librarian', suffix='.epub', delete=False)
     zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
 
+    if os.path.isdir(ilustr_path):
+        for i, filename in enumerate(os.listdir(ilustr_path)):
+            file_path = os.path.join(ilustr_path, filename)
+            zip.write(file_path, os.path.join('OPS', filename))
+            image_id = 'image%s' % i
+            manifest.append(etree.fromstring(
+                '<item id="%s" href="%s" media-type="%s" />' % (image_id, filename, guess_type(file_path)[0])))
+
     # write static elements
     mime = zipfile.ZipInfo()
     mime.filename = 'mimetype'
@@ -391,12 +444,15 @@ def transform(wldoc, verbose=False,
                        '<rootfiles><rootfile full-path="OPS/content.opf" ' \
                        'media-type="application/oebps-package+xml" />' \
                        '</rootfiles></container>')
-    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(get_resource('res/wl-logo-small.png'), os.path.join('OPS', 'logo_wolnelektury.png'))
+    # zip.write(get_resource('res/koedlogo.png'), os.path.join('OPS', 'logo_koed.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'))
 
+    if cover is None:
+        cover = WLCover
     if cover:
         if cover is True:
             cover = WLCover
@@ -462,15 +518,15 @@ def transform(wldoc, verbose=False,
         zip.writestr('OPS/annotations.html', etree.tostring(
                             html_tree, method="html", pretty_print=True))
 
-    toc.add("Strona redakcyjna", "last.html")
-    manifest.append(etree.fromstring(
-        '<item id="last" href="last.html" media-type="application/xhtml+xml" />'))
-    spine.append(etree.fromstring(
-        '<itemref idref="last" />'))
-    html_tree = xslt(document.edoc, get_resource('epub/xsltLast.xsl'))
-    chars.update(used_chars(html_tree.getroot()))
-    zip.writestr('OPS/last.html', etree.tostring(
-                        html_tree, method="html", pretty_print=True))
+    # toc.add("Strona redakcyjna", "last.html")
+    # manifest.append(etree.fromstring(
+    #     '<item id="last" href="last.html" media-type="application/xhtml+xml" />'))
+    # spine.append(etree.fromstring(
+    #     '<itemref idref="last" />'))
+    # html_tree = xslt(document.edoc, get_resource('epub/xsltLast.xsl'))
+    # chars.update(used_chars(html_tree.getroot()))
+    # zip.writestr('OPS/last.html', etree.tostring(
+    #                     html_tree, method="html", pretty_print=True))
 
     if not flags or not 'without-fonts' in flags:
         # strip fonts