Fix for missing xml declaration of content.opf and toc.ncx

[librarian.git] / librarian / pdf.py

diff --git a/librarian/pdf.py b/librarian/pdf.py
index baf7be4..2154985 100644 (file)
--- a/librarian/pdf.py
+++ b/librarian/pdf.py
@@ -3,26 +3,31 @@
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
 # This file is part of Librarian, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #

+"""PDF creation library.
+
+Creates one big XML from the book and its children, converts it to LaTeX
+with TeXML, then runs it by XeLaTeX.
+
+"""

 from __future__ import with_statement
 import os
 import os.path
 import shutil
 from StringIO import StringIO
 from __future__ import with_statement
 import os
 import os.path
 import shutil
 from StringIO import StringIO

-from tempfile import mkdtemp
+from tempfile import mkdtemp, NamedTemporaryFile

 import re
 from copy import deepcopy
 from subprocess import call, PIPE
 
 import re
 from copy import deepcopy
 from subprocess import call, PIPE
 

-import sys
-

 from Texml.processor import process
 from lxml import etree
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 
 from librarian.dcparser import Person
 from librarian.parser import WLDocument
 from Texml.processor import process
 from lxml import etree
 from lxml.etree import XMLSyntaxError, XSLTApplyError
 
 from librarian.dcparser import Person
 from librarian.parser import WLDocument

-from librarian import ParseError, DCNS, get_resource
+from librarian import ParseError, DCNS, get_resource, OutputFile

 from librarian import functions
 from librarian import functions

+from librarian.cover import DefaultEbookCover

 
 
 functions.reg_substitute_entities()
 
 
 functions.reg_substitute_entities()


@@ -35,9 +40,17 @@ STYLESHEETS = {
     'wl2tex': 'pdf/wl2tex.xslt',
 }
 
     'wl2tex': 'pdf/wl2tex.xslt',
 }
 

+#CUSTOMIZATIONS = [
+#    'nofootnotes',
+#    'nothemes',
+#    'defaultleading',
+#    'onehalfleading',
+#    'doubleleading',
+#    'nowlfont',
+#    ]

 
 def insert_tags(doc, split_re, tagname, exclude=None):
 
 def insert_tags(doc, split_re, tagname, exclude=None):

-    """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree 
+    """ inserts <tagname> for every occurence of `split_re' in text nodes in the `doc' tree

 
     >>> t = etree.fromstring('<a><b>A-B-C</b>X-Y-Z</a>');
     >>> insert_tags(t, re.compile('-'), 'd');
 
     >>> t = etree.fromstring('<a><b>A-B-C</b>X-Y-Z</a>');
     >>> insert_tags(t, re.compile('-'), 'd');


@@ -67,7 +80,7 @@ def insert_tags(doc, split_re, tagname, exclude=None):
 
 
 def substitute_hyphens(doc):
 
 
 def substitute_hyphens(doc):

-    insert_tags(doc, 
+    insert_tags(doc,

                 re.compile("(?<=[^-\s])-(?=[^-\s])"),
                 "dywiz",
                 exclude=[DCNS("identifier.url"), DCNS("rights.license")]
                 re.compile("(?<=[^-\s])-(?=[^-\s])"),
                 "dywiz",
                 exclude=[DCNS("identifier.url"), DCNS("rights.license")]


@@ -75,7 +88,7 @@ def substitute_hyphens(doc):
 
 
 def fix_hanging(doc):
 
 
 def fix_hanging(doc):

-    insert_tags(doc, 
+    insert_tags(doc,

                 re.compile("(?<=\s\w)\s+"),
                 "nbsp",
                 exclude=[DCNS("identifier.url"), DCNS("rights.license")]
                 re.compile("(?<=\s\w)\s+"),
                 "nbsp",
                 exclude=[DCNS("identifier.url"), DCNS("rights.license")]


@@ -128,13 +141,22 @@ def hack_motifs(doc):
 
 
 def parse_creator(doc):
 
 
 def parse_creator(doc):

-    """ find all dc:creator tags and add dc:creator_parsed with forenames first """
-    for creator in doc.findall('//'+DCNS('creator')):
-        p = Person.from_text(creator.text)
-        creator_parsed = deepcopy(creator)
-        creator_parsed.tag = DCNS('creator_parsed')
-        creator_parsed.text = ' '.join(p.first_names + (p.last_name,))
-        creator.getparent().insert(0, creator_parsed)
+    """Generates readable versions of creator and translator tags.
+
+    Finds all dc:creator and dc.contributor.translator tags
+    and adds *_parsed versions with forenames first.
+    """
+    for person in doc.xpath("|".join('//dc:'+(tag) for tag in (
+                    'creator', 'contributor.translator')),
+                    namespaces = {'dc': str(DCNS)})[::-1]:
+        if not person.text:
+            continue
+        p = Person.from_text(person.text)
+        person_parsed = deepcopy(person)
+        person_parsed.tag = person.tag + '_parsed'
+        person_parsed.set('sortkey', person.text)
+        person_parsed.text = p.readable()
+        person.getparent().insert(0, person_parsed)

 
 
 def get_stylesheet(name):
 
 
 def get_stylesheet(name):


@@ -142,12 +164,12 @@ def get_stylesheet(name):
 
 
 def package_available(package, args='', verbose=False):
 
 
 def package_available(package, args='', verbose=False):

-    """ check if a verion of a latex package accepting given args is available """  
+    """ check if a verion of a latex package accepting given args is available """

     tempdir = mkdtemp('-wl2pdf-test')
     fpath = os.path.join(tempdir, 'test.tex')
     f = open(fpath, 'w')
     f.write(r"""
     tempdir = mkdtemp('-wl2pdf-test')
     fpath = os.path.join(tempdir, 'test.tex')
     f = open(fpath, 'w')
     f.write(r"""

-        \documentclass{book}
+        \documentclass{wl}

         \usepackage[%s]{%s}
         \begin{document}
         \end{document}
         \usepackage[%s]{%s}
         \begin{document}
         \end{document}


@@ -161,37 +183,58 @@ def package_available(package, args='', verbose=False):
     return p == 0
 
 
     return p == 0
 
 

-def transform(provider, slug=None, file_path=None, 
-              output_file=None, output_dir=None, make_dir=False, verbose=False, save_tex=None, morefloats=None):
+def transform(wldoc, verbose=False, save_tex=None, morefloats=None,
+              cover=None, flags=None, customizations=None):

     """ produces a PDF file with XeLaTeX
 
     """ produces a PDF file with XeLaTeX
 

-    provider: a DocProvider
-    slug: slug of file to process, available by provider
-    file_path can be provided instead of a slug
-    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>.pdf istead of <output_dir>/<slug>.pdf
+    wldoc: a WLDocument

     verbose: prints all output from LaTeX
     save_tex: path to save the intermediary LaTeX file to
     morefloats (old/new/none): force specific morefloats
     verbose: prints all output from LaTeX
     save_tex: path to save the intermediary LaTeX file to
     morefloats (old/new/none): force specific morefloats

+    cover: a cover.Cover factory or True for default
+    flags: less-advertising,
+    customizations: user requested customizations regarding various formatting parameters (passed to wl LaTeX class)

     """
 
     # Parse XSLT
     try:
     """
 
     # Parse XSLT
     try:

-        if file_path:
-            if slug:
-                raise ValueError('slug or file_path should be specified, not both')
-            document = load_including_children(provider, file_path=file_path)
-        else:
-            if not slug:
-                raise ValueError('either slug or file_path should be specified')
-            document = load_including_children(provider, slug=slug)
+        book_info = wldoc.book_info
+        document = load_including_children(wldoc)
+        root = document.edoc.getroot()
+
+        if cover:
+            if cover is True:
+                cover = DefaultEbookCover
+            bound_cover = cover(book_info)
+            root.set('data-cover-width', str(bound_cover.width))
+            root.set('data-cover-height', str(bound_cover.height))
+            if bound_cover.uses_dc_cover:
+                if book_info.cover_by:
+                    root.set('data-cover-by', book_info.cover_by)
+                if book_info.cover_source:
+                    root.set('data-cover-source',
+                            book_info.cover_source)
+        if flags:
+            for flag in flags:
+                root.set('flag-' + flag, 'yes')

 
         # check for LaTeX packages
         if morefloats:
 
         # check for LaTeX packages
         if morefloats:

-            document.edoc.getroot().set('morefloats', morefloats.lower())
+            root.set('morefloats', morefloats.lower())

         elif package_available('morefloats', 'maxfloats=19'):
         elif package_available('morefloats', 'maxfloats=19'):

-            document.edoc.getroot().set('morefloats', 'new')
+            root.set('morefloats', 'new')
+
+        # add customizations
+        if customizations is not None:
+            root.set('customizations', u','.join(customizations))
+
+        # add editors info
+        root.set('editors', u', '.join(sorted(
+            editor.readable() for editor in document.editors())))
+        if document.book_info.funders:
+            root.set('funders', u', '.join(document.book_info.funders))
+        if document.book_info.thanks:
+            root.set('thanks', document.book_info.thanks)

 
         # hack the tree
         move_motifs_inside(document.edoc)
 
         # hack the tree
         move_motifs_inside(document.edoc)


@@ -200,19 +243,21 @@ def transform(provider, slug=None, file_path=None,
         substitute_hyphens(document.edoc)
         fix_hanging(document.edoc)
 
         substitute_hyphens(document.edoc)
         fix_hanging(document.edoc)
 

-        # find output dir
-        if make_dir and output_dir is not None:
-            author = unicode(document.book_info.author)
-            output_dir = os.path.join(output_dir, author)
-

         # wl -> TeXML
         style_filename = get_stylesheet("wl2tex")
         style = etree.parse(style_filename)
         # wl -> TeXML
         style_filename = get_stylesheet("wl2tex")
         style = etree.parse(style_filename)

+

         texml = document.transform(style)
         texml = document.transform(style)

-        del document # no longer needed large object :)

 
         # TeXML -> LaTeX
         temp = mkdtemp('-wl2pdf')
 
         # TeXML -> LaTeX
         temp = mkdtemp('-wl2pdf')

+
+        if cover:
+            with open(os.path.join(temp, 'cover.png'), 'w') as f:
+                bound_cover.save(f)
+
+        del document # no longer needed large object :)
+

         tex_path = os.path.join(temp, 'doc.tex')
         fout = open(tex_path, 'w')
         process(StringIO(texml), fout, 'utf-8')
         tex_path = os.path.join(temp, 'doc.tex')
         fout = open(tex_path, 'w')
         process(StringIO(texml), fout, 'utf-8')


@@ -223,10 +268,13 @@ def transform(provider, slug=None, file_path=None,
             shutil.copy(tex_path, save_tex)
 
         # LaTeX -> PDF
             shutil.copy(tex_path, save_tex)
 
         # LaTeX -> PDF

-        shutil.copy(get_resource('pdf/wl.sty'), temp)
+        shutil.copy(get_resource('pdf/wl.cls'), temp)

         shutil.copy(get_resource('res/wl-logo.png'), temp)
 
         shutil.copy(get_resource('res/wl-logo.png'), temp)
 

-        cwd = os.getcwd()
+        try:
+            cwd = os.getcwd()
+        except OSError:
+            cwd = None

         os.chdir(temp)
 
         if verbose:
         os.chdir(temp)
 
         if verbose:


@@ -236,59 +284,42 @@ def transform(provider, slug=None, file_path=None,
         if p:
             raise ParseError("Error parsing .tex file")
 
         if p:
             raise ParseError("Error parsing .tex file")
 

-        os.chdir(cwd)
+        if cwd is not None:
+            os.chdir(cwd)

 
 

-        # save the PDF
+        output_file = NamedTemporaryFile(prefix='librarian', suffix='.pdf', delete=False)

         pdf_path = os.path.join(temp, 'doc.pdf')
         pdf_path = os.path.join(temp, 'doc.pdf')

-        if output_dir is not None:
-            try:
-                os.makedirs(output_dir)
-            except OSError:
-                pass
-            if slug:
-                output_path = os.path.join(output_dir, '%s.pdf' % slug)
-            else:
-                output_path = os.path.join(output_dir, os.path.splitext(os.path.basename(file_path))[0] + '.pdf')
-            shutil.move(pdf_path, output_path)
-        else:
-            if hasattr(output_file, 'write'):
-                # file-like object
-                with open(pdf_path) as f:
-                    output_file.write(f.read())
-                output_file.close()
-            else:
-                # path to output file
-                shutil.copy(pdf_path, output_file)
+        shutil.move(pdf_path, output_file.name)

         shutil.rmtree(temp)
         shutil.rmtree(temp)

+        return OutputFile.from_filename(output_file.name)

 
     except (XMLSyntaxError, XSLTApplyError), e:
         raise ParseError(e)
 
 
 
     except (XMLSyntaxError, XSLTApplyError), e:
         raise ParseError(e)
 
 

-def load_including_children(provider, slug=None, uri=None, file_path=None):
-    """ makes one big xml file with children inserted at end 
-    either slug or uri must be provided
+def load_including_children(wldoc=None, provider=None, uri=None):
+    """ Makes one big xml file with children inserted at end.
+    
+    Either wldoc or provider and URI must be provided.

     """
 
     """
 

-    if uri:
+    if uri and provider:

         f = provider.by_uri(uri)
         f = provider.by_uri(uri)

-    elif slug:
-        f = provider[slug]
-    elif file_path:
-        f = open(file_path, 'r')
+        text = f.read().decode('utf-8')
+        f.close()
+    elif wldoc is not None:
+        text = etree.tostring(wldoc.edoc, encoding=unicode)
+        provider = wldoc.provider

     else:
     else:

-        raise ValueError('Neither slug, URI nor file path provided for a book.')
+        raise ValueError('Neither a WLDocument, nor provider and URI were provided.')

 
 

-    text = f.read().decode('utf-8')

     text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
 
     text = re.sub(ur"([\u0400-\u04ff]+)", ur"<alien>\1</alien>", text)
 

-    document = WLDocument.from_string(text, True,
-        parse_dublincore=True)
-
-    f.close()
+    document = WLDocument.from_string(text,
+                parse_dublincore=True, provider=provider)
+    document.swap_endlines()

 
     for child_uri in document.book_info.parts:
 
     for child_uri in document.book_info.parts:

-        child = load_including_children(provider, uri=child_uri)
+        child = load_including_children(provider=provider, uri=child_uri)

         document.edoc.getroot().append(child.edoc.getroot())
         document.edoc.getroot().append(child.edoc.getroot())

-

     return document
     return document