dirty workaround for python unicode problems

[librarian.git] / librarian / epub.py

diff --git a/librarian/epub.py b/librarian/epub.py
index 2584029..0f23923 100644 (file)
--- a/librarian/epub.py
+++ b/librarian/epub.py
@@ -1,13 +1,14 @@
 # -*- coding: utf-8 -*-
 #
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # -*- coding: utf-8 -*-
 #
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.

-# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.  
+# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.

 #
 from __future__ import with_statement
 
 import os
 import os.path
 import subprocess
 #
 from __future__ import with_statement
 
 import os
 import os.path
 import subprocess

+from StringIO import StringIO

 from copy import deepcopy
 from lxml import etree
 import zipfile
 from copy import deepcopy
 from lxml import etree
 import zipfile


@@ -15,11 +16,14 @@ from tempfile import mkdtemp
 from shutil import rmtree
 
 import sys
 from shutil import rmtree
 
 import sys

-sys.path.append('..') # for running from working copy

 
 

-from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, NoDublinCore
+from librarian import XMLNamespace, RDFNS, DCNS, WLNS, NCXNS, OPFNS, XHTMLNS, NoDublinCore

 from librarian.dcparser import BookInfo
 
 from librarian.dcparser import BookInfo
 

+from librarian import functions, get_resource
+
+functions.reg_person_name()
+

 
 def inner_xml(node):
     """ returns node's text and children as a string
 
 def inner_xml(node):
     """ returns node's text and children as a string


@@ -29,7 +33,7 @@ def inner_xml(node):
     """
 
     nt = node.text if node.text is not None else ''
     """
 
     nt = node.text if node.text is not None else ''

-    return ''.join([nt] + [etree.tostring(child) for child in node]) 
+    return ''.join([nt] + [etree.tostring(child) for child in node])

 
 def set_inner_xml(node, text):
     """ sets node's text and children from a string
 
 def set_inner_xml(node, text):
     """ sets node's text and children from a string


@@ -70,27 +74,24 @@ def xslt(xml, sheet):
         return xml.xslt(etree.parse(xsltf))
 
 
         return xml.xslt(etree.parse(xsltf))
 
 

-_resdir = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'epub')
-def res(fname):
-    return os.path.join(_resdir, fname)
-
-

 def replace_characters(node):
     def replace_chars(text):
         if text is None:
             return None
 def replace_characters(node):
     def replace_chars(text):
         if text is None:
             return None

-        return text.replace("---", u"\u2014")\
+        return text.replace(u"\ufeff", u"")\
+                   .replace("---", u"\u2014")\

                    .replace("--", u"\u2013")\
                    .replace(",,", u"\u201E")\
                    .replace('"', u"\u201D")\
                    .replace("'", u"\u2019")
                    .replace("--", u"\u2013")\
                    .replace(",,", u"\u201E")\
                    .replace('"', u"\u201D")\
                    .replace("'", u"\u2019")

-    if node.tag == 'extra':
+    if node.tag in ('uwaga', 'extra'):
+        t = node.tail

         node.clear()
         node.clear()

-    else:
-        node.text = replace_chars(node.text)
-        node.tail = replace_chars(node.tail)
-        for child in node:
-            replace_characters(child)
+        node.tail = t
+    node.text = replace_chars(node.text)
+    node.tail = replace_chars(node.tail)
+    for child in node:
+        replace_characters(child)

 
 
 def find_annotations(annotations, source, part_no):
 
 
 def find_annotations(annotations, source, part_no):


@@ -106,7 +107,7 @@ def find_annotations(annotations, source, part_no):
             child.clear()
             child.tail = tail
             child.text = number
             child.clear()
             child.tail = tail
             child.text = number

-        if child.tag not in ('extra', 'podtytul'):
+        if child.tag not in ('extra', 'uwaga'):

             find_annotations(annotations, child, part_no)
 
 
             find_annotations(annotations, child, part_no)
 
 


@@ -211,7 +212,6 @@ class TOC(object):
 
 def used_chars(element):
     """ Lists characters used in an ETree Element """
 
 def used_chars(element):
     """ Lists characters used in an ETree Element """

-    print (element.text or '') + (element.tail or '')

     chars = set((element.text or '') + (element.tail or ''))
     for child in element:
         chars = chars.union(used_chars(child))
     chars = set((element.text or '') + (element.tail or ''))
     for child in element:
         chars = chars.union(used_chars(child))


@@ -242,7 +242,7 @@ def chop(main_text):
     yield part_xml
 
 
     yield part_xml
 
 

-def transform_chunk(chunk_xml, chunk_no, annotations):
+def transform_chunk(chunk_xml, chunk_no, annotations, empty=False, _empty_html_static=[]):

     """ transforms one chunk, returns a HTML string, a TOC object and a set of used characters """
 
     toc = TOC()
     """ transforms one chunk, returns a HTML string, a TOC object and a set of used characters """
 
     toc = TOC()


@@ -252,25 +252,39 @@ def transform_chunk(chunk_xml, chunk_no, annotations):
         elif element.tag in ('naglowek_podrozdzial', 'naglowek_scena'):
             subnumber = toc.add(node_name(element), chunk_no, level=1, is_part=False)
             element.set('sub', str(subnumber))
         elif element.tag in ('naglowek_podrozdzial', 'naglowek_scena'):
             subnumber = toc.add(node_name(element), chunk_no, level=1, is_part=False)
             element.set('sub', str(subnumber))

-    find_annotations(annotations, chunk_xml, chunk_no)
-    replace_by_verse(chunk_xml)
-    html_tree = xslt(chunk_xml, res('xsltScheme.xsl'))
-    chars = used_chars(html_tree.getroot())
-    output_html = etree.tostring(html_tree, pretty_print=True)
+    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:
+        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
 
 
     return output_html, toc, chars
 
 

-def transform(provider, slug, output_file=None, output_dir=None):
-    """ produces an epub
+def transform(provider, slug=None, file_path=None, output_file=None, output_dir=None, make_dir=False, verbose=False,
+              sample=None, cover=None, flags=None):
+    """ produces a EPUB file

 
 

-    provider is a DocProvider
-    either output_file (a file-like object) or output_dir (path to file/dir) should be specified
-    if output_dir is specified, file will be written to <output_dir>/<author>/<slug>.epub
+    provider: a DocProvider
+    slug: slug of file to process, available by provider
+    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 <output_dir>/<author>/<slug>.epub instead of <output_dir>/<slug>.epub
+    sample=n: generate sample e-book (with at least n paragraphs)
+    cover: a cover.Cover object
+    flags: less-advertising,

     """
 
     """
 

-    def transform_file(input_xml, chunk_counter=1, first=True):
+    def transform_file(input_xml, chunk_counter=1, first=True, sample=None):

         """ processes one input file and proceeds to its children """
 
         """ processes one input file and proceeds to its children """
 

+        replace_characters(input_xml.getroot())
+

         children = [child.text for child in input_xml.findall('.//'+DCNS('relation.hasPart'))]
 
         # every input file will have a TOC entry,
         children = [child.text for child in input_xml.findall('.//'+DCNS('relation.hasPart'))]
 
         # every input file will have a TOC entry,


@@ -279,16 +293,20 @@ def transform(provider, slug, output_file=None, output_dir=None):
         chars = set()
         if first:
             # write book title page
         chars = set()
         if first:
             # write book title page

-            html_tree = xslt(input_xml, res('xsltTitle.xsl'))
+            html_tree = xslt(input_xml, get_resource('epub/xsltTitle.xsl'))

             chars = used_chars(html_tree.getroot())
             zip.writestr('OPS/title.html',
             chars = used_chars(html_tree.getroot())
             zip.writestr('OPS/title.html',

-                 etree.tostring(html_tree, pretty_print=True))
+                 etree.tostring(html_tree, method="html", pretty_print=True))

         elif children:
             # write title page for every parent
         elif children:
             # write title page for every parent

-            html_tree = xslt(input_xml, res('xsltChunkTitle.xsl'))
-            chars = used_chars(html_tree.getroot())
-            zip.writestr('OPS/part%d.html' % chunk_counter, 
-                etree.tostring(html_tree, pretty_print=True))
+            if sample is not None and sample <= 0:
+                chars = set()
+                html_string = open(get_resource('epub/emptyChunk.html')).read()
+            else:
+                html_tree = xslt(input_xml, get_resource('epub/xsltChunkTitle.xsl'))
+                chars = used_chars(html_tree.getroot())
+                html_string = etree.tostring(html_tree, method="html", pretty_print=True)
+            zip.writestr('OPS/part%d.html' % chunk_counter, html_string)

             add_to_manifest(manifest, chunk_counter)
             add_to_spine(spine, chunk_counter)
             chunk_counter += 1
             add_to_manifest(manifest, chunk_counter)
             add_to_spine(spine, chunk_counter)
             chunk_counter += 1


@@ -303,10 +321,15 @@ def transform(provider, slug, output_file=None, output_dir=None):
                 main_text = None
 
         if main_text is not None:
                 main_text = None
 
         if main_text is not None:

-            replace_characters(main_text)
-

             for chunk_xml in chop(main_text):
             for chunk_xml in chop(main_text):

-                chunk_html, chunk_toc, chunk_chars = transform_chunk(chunk_xml, chunk_counter, annotations)
+                empty = False
+                if sample is not None:
+                    if sample <= 0:
+                        empty = True
+                    else:
+                        sample -= len(chunk_xml.xpath('//strofa|//akap|//akap_cd|//akap_dialog'))
+                chunk_html, chunk_toc, chunk_chars = transform_chunk(chunk_xml, chunk_counter, annotations, empty)
+

                 toc.extend(chunk_toc)
                 chars = chars.union(chunk_chars)
                 zip.writestr('OPS/part%d.html' % chunk_counter, chunk_html)
                 toc.extend(chunk_toc)
                 chars = chars.union(chunk_chars)
                 zip.writestr('OPS/part%d.html' % chunk_counter, chunk_html)


@@ -317,14 +340,28 @@ def transform(provider, slug, output_file=None, output_dir=None):
         if children:
             for child in children:
                 child_xml = etree.parse(provider.by_uri(child))
         if children:
             for child in children:
                 child_xml = etree.parse(provider.by_uri(child))

-                child_toc, chunk_counter, chunk_chars = transform_file(child_xml, chunk_counter, first=False)
+                child_toc, chunk_counter, chunk_chars, sample = transform_file(child_xml, chunk_counter, first=False, sample=sample)

                 toc.append(child_toc)
                 chars = chars.union(chunk_chars)
 
                 toc.append(child_toc)
                 chars = chars.union(chunk_chars)
 

-        return toc, chunk_counter, chars
+        return toc, chunk_counter, chars, sample

 
     # read metadata from the first file
 
     # read metadata from the first file

-    input_xml = etree.parse(provider[slug])
+    if file_path:
+        if slug:
+            raise ValueError('slug or file_path should be specified, not both')
+        f = open(file_path, 'r')
+        input_xml = etree.parse(f)
+        f.close()
+    else:
+        if not slug:
+            raise ValueError('either slug or file_path should be specified')
+        input_xml = etree.parse(provider[slug])
+
+    if flags:
+        for flag in flags:
+            input_xml.getroot().set(flag, 'yes')
+

     metadata = input_xml.find('.//'+RDFNS('Description'))
     if metadata is None:
         raise NoDublinCore('Document has no DublinCore - which is required.')
     metadata = input_xml.find('.//'+RDFNS('Description'))
     if metadata is None:
         raise NoDublinCore('Document has no DublinCore - which is required.')


@@ -333,14 +370,17 @@ def transform(provider, slug, output_file=None, output_dir=None):
 
     # if output to dir, create the file
     if output_dir is not None:
 
     # if output to dir, create the file
     if output_dir is not None:

-        author = unicode(book_info.author)
-        author_dir = os.path.join(output_dir, author)
-        try:
-            os.makedirs(author_dir)
-        except OSError:
-            pass
-        output_file = open(os.path.join(author_dir, '%s.epub' % slug), 'w')
-
+        if make_dir:
+            author = unicode(book_info.author)
+            output_dir = os.path.join(output_dir, author)
+            try:
+                os.makedirs(output_dir)
+            except OSError:
+                pass
+        if slug:
+            output_file = open(os.path.join(output_dir, '%s.epub' % slug), 'w')
+        else:
+            output_file = open(os.path.join(output_dir, os.path.splitext(os.path.basename(file_path))[0] + '.epub'), 'w')

 
     zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
 
 
     zip = zipfile.ZipFile(output_file, 'w', zipfile.ZIP_DEFLATED)
 


@@ -355,13 +395,35 @@ def transform(provider, slug, output_file=None, output_dir=None):
                        '<rootfiles><rootfile full-path="OPS/content.opf" ' \
                        'media-type="application/oebps-package+xml" />' \
                        '</rootfiles></container>')
                        '<rootfiles><rootfile full-path="OPS/content.opf" ' \
                        'media-type="application/oebps-package+xml" />' \
                        '</rootfiles></container>')

-    for fname in 'style.css', 'logo_wolnelektury.png':
-        zip.write(res(fname), os.path.join('OPS', fname))
+    zip.write(get_resource('epub/style.css'), os.path.join('OPS', 'style.css'))
+    zip.write(get_resource('res/wl-logo-small.png'), os.path.join('OPS', 'logo_wolnelektury.png'))

 
 

-    opf = xslt(metadata, res('xsltContent.xsl'))
+    opf = xslt(metadata, get_resource('epub/xsltContent.xsl'))

     manifest = opf.find('.//' + OPFNS('manifest'))
     spine = opf.find('.//' + OPFNS('spine'))
 
     manifest = opf.find('.//' + OPFNS('manifest'))
     spine = opf.find('.//' + OPFNS('spine'))
 

+    if cover:
+        cover_file = StringIO()
+        c = cover(book_info.author.readable(), book_info.title)
+        c.save(cover_file)
+        c_name = 'cover.%s' % c.ext()
+        zip.writestr(os.path.join('OPS', c_name), cover_file.getvalue())
+        del cover_file
+
+        cover_tree = etree.parse(get_resource('epub/cover.html'))
+        cover_tree.find('//' + XHTMLNS('img')).set('src', c_name)
+        zip.writestr('OPS/cover.html', etree.tostring(
+                        cover_tree, method="html", pretty_print=True))
+
+        manifest.append(etree.fromstring(
+            '<item id="cover" href="cover.html" media-type="application/xhtml+xml" />'))
+        manifest.append(etree.fromstring(
+            '<item id="cover-image" href="%s" media-type="%s" />' % (c_name, c.mime_type())))
+        spine.insert(0, etree.fromstring('<itemref idref="cover" />'))
+        opf.getroot()[0].append(etree.fromstring('<meta name="cover" content="cover-image"/>'))
+        opf.getroot().append(etree.fromstring('<guide><reference href="cover.html" type="cover" title="Okładka"/></guide>'))
+
+

     annotations = etree.Element('annotations')
 
     toc_file = etree.fromstring('<?xml version="1.0" encoding="utf-8"?><!DOCTYPE ncx PUBLIC ' \
     annotations = etree.Element('annotations')
 
     toc_file = etree.fromstring('<?xml version="1.0" encoding="utf-8"?><!DOCTYPE ncx PUBLIC ' \


@@ -373,7 +435,7 @@ def transform(provider, slug, output_file=None, output_dir=None):
                                '</navPoint></navMap></ncx>')
     nav_map = toc_file[-1]
 
                                '</navPoint></navMap></ncx>')
     nav_map = toc_file[-1]
 

-    toc, chunk_counter, chars = transform_file(input_xml)
+    toc, chunk_counter, chars, sample = transform_file(input_xml, sample=sample)

 
     if not toc.children:
         toc.add(u"Początek utworu", 1)
 
     if not toc.children:
         toc.add(u"Początek utworu", 1)


@@ -384,23 +446,42 @@ def transform(provider, slug, output_file=None, output_dir=None):
         nav_map.append(etree.fromstring(
             '<navPoint id="NavPoint-%(i)d" playOrder="%(i)d" ><navLabel><text>Przypisy</text>'\
             '</navLabel><content src="annotations.html" /></navPoint>' % {'i': toc_counter}))
         nav_map.append(etree.fromstring(
             '<navPoint id="NavPoint-%(i)d" playOrder="%(i)d" ><navLabel><text>Przypisy</text>'\
             '</navLabel><content src="annotations.html" /></navPoint>' % {'i': toc_counter}))

+        toc_counter += 1

         manifest.append(etree.fromstring(
             '<item id="annotations" href="annotations.html" media-type="application/xhtml+xml" />'))
         spine.append(etree.fromstring(
             '<itemref idref="annotations" />'))
         replace_by_verse(annotations)
         manifest.append(etree.fromstring(
             '<item id="annotations" href="annotations.html" media-type="application/xhtml+xml" />'))
         spine.append(etree.fromstring(
             '<itemref idref="annotations" />'))
         replace_by_verse(annotations)

-        html_tree = xslt(annotations, res("xsltAnnotations.xsl"))
+        html_tree = xslt(annotations, get_resource('epub/xsltAnnotations.xsl'))

         chars = chars.union(used_chars(html_tree.getroot()))
         zip.writestr('OPS/annotations.html', etree.tostring(
         chars = chars.union(used_chars(html_tree.getroot()))
         zip.writestr('OPS/annotations.html', etree.tostring(

-                            html_tree, pretty_print=True))
+                            html_tree, method="html", pretty_print=True))
+
+    nav_map.append(etree.fromstring(
+        '<navPoint id="NavPoint-%(i)d" playOrder="%(i)d" ><navLabel><text>Strona redakcyjna</text>'\
+        '</navLabel><content src="last.html" /></navPoint>' % {'i': toc_counter}))
+    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(input_xml, 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))

 
     # strip fonts
     tmpdir = mkdtemp('-librarian-epub')
     cwd = os.getcwd()
 
 
     # strip fonts
     tmpdir = mkdtemp('-librarian-epub')
     cwd = os.getcwd()
 

-    os.chdir(os.path.join(os.path.dirname(os.path.realpath(__file__)), '../font-optimizer'))
+    os.chdir(os.path.join(os.path.dirname(os.path.realpath(__file__)), 'font-optimizer'))

     for fname in 'DejaVuSerif.ttf', 'DejaVuSerif-Bold.ttf', 'DejaVuSerif-Italic.ttf', 'DejaVuSerif-BoldItalic.ttf':
     for fname in 'DejaVuSerif.ttf', 'DejaVuSerif-Bold.ttf', 'DejaVuSerif-Italic.ttf', 'DejaVuSerif-BoldItalic.ttf':

-        subprocess.check_call(['./subset.pl', '--chars', ''.join(chars), res('../fonts/' + fname), os.path.join(tmpdir, fname)])
+        optimizer_call = ['perl', 'subset.pl', '--chars', ''.join(chars).encode('utf-8'),
+                          get_resource('fonts/' + fname), os.path.join(tmpdir, fname)]
+        if verbose:
+            print "Running font-optimizer"
+            subprocess.check_call(optimizer_call)
+        else:
+            subprocess.check_call(optimizer_call, stdout=subprocess.PIPE, stderr=subprocess.PIPE)

         zip.write(os.path.join(tmpdir, fname), os.path.join('OPS', fname))
     rmtree(tmpdir)
     os.chdir(cwd)
         zip.write(os.path.join(tmpdir, fname), os.path.join('OPS', fname))
     rmtree(tmpdir)
     os.chdir(cwd)


@@ -419,18 +500,3 @@ def transform(provider, slug, output_file=None, output_dir=None):
     set_inner_xml(toc_file[1], ''.join(('<text>', title, '</text>')))
     zip.writestr('OPS/toc.ncx', etree.tostring(toc_file, pretty_print=True))
     zip.close()
     set_inner_xml(toc_file[1], ''.join(('<text>', title, '</text>')))
     zip.writestr('OPS/toc.ncx', etree.tostring(toc_file, pretty_print=True))
     zip.close()

-
-
-if __name__ == '__main__':
-    from librarian import DirDocProvider
-
-    if len(sys.argv) < 2:
-        print >> sys.stderr, 'Usage: python epub.py <input file>'
-        sys.exit(1)
-
-    main_input = sys.argv[1]
-    basepath, ext = os.path.splitext(main_input)
-    path, slug = os.path.realpath(basepath).rsplit('/', 1)
-    provider = DirDocProvider(path)
-    transform(provider, slug, output_dir=path)
-