Epub: only attach images referenced in the text.

[librarian.git] / librarian / __init__.py

diff --git a/librarian/__init__.py b/librarian/__init__.py
index 5b6981d..119b6b1 100644 (file)
--- a/librarian/__init__.py
+++ b/librarian/__init__.py
@@ -3,21 +3,42 @@
 # 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.
 #

-import os
+from __future__ import print_function, unicode_literals

 
 

-class ParseError(Exception):
+import os
+import re
+import shutil
+from tempfile import NamedTemporaryFile
+import urllib
+from lxml import etree
+import six
+from six.moves.urllib.request import FancyURLopener
+from .util import makedirs
+
+
+@six.python_2_unicode_compatible
+class UnicodeException(Exception):

     def __str__(self):
         """ Dirty workaround for Python Unicode handling problems. """
     def __str__(self):
         """ Dirty workaround for Python Unicode handling problems. """

-        return self.message.message
-
-    def __unicode__(self):
-        """ Dirty workaround for Python Unicode handling problems. """
-        return self.message.message
+        args = self.args[0] if len(self.args) == 1 else self.args
+        try:
+            message = six.text_type(args)
+        except UnicodeDecodeError:
+            message = six.text_type(args, encoding='utf-8', errors='ignore')
+        return message
+
+class ParseError(UnicodeException):
+    pass

 
 

-class ValidationError(Exception):
+class ValidationError(UnicodeException):

     pass
 
 class NoDublinCore(ValidationError):
     pass
 
 class NoDublinCore(ValidationError):

+    """There's no DublinCore section, and it's required."""
+    pass
+
+class NoProvider(UnicodeException):
+    """There's no DocProvider specified, and it's needed."""

     pass
 
 class XMLNamespace(object):
     pass
 
 class XMLNamespace(object):


@@ -46,53 +67,97 @@ class EmptyNamespace(XMLNamespace):
         return tag
 
 # some common namespaces we use
         return tag
 
 # some common namespaces we use

+XMLNS = XMLNamespace('http://www.w3.org/XML/1998/namespace')

 RDFNS = XMLNamespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#')
 DCNS = XMLNamespace('http://purl.org/dc/elements/1.1/')
 XINS = XMLNamespace("http://www.w3.org/2001/XInclude")
 XHTMLNS = XMLNamespace("http://www.w3.org/1999/xhtml")
 NCXNS = XMLNamespace("http://www.daisy.org/z3986/2005/ncx/")
 OPFNS = XMLNamespace("http://www.idpf.org/2007/opf")
 RDFNS = XMLNamespace('http://www.w3.org/1999/02/22-rdf-syntax-ns#')
 DCNS = XMLNamespace('http://purl.org/dc/elements/1.1/')
 XINS = XMLNamespace("http://www.w3.org/2001/XInclude")
 XHTMLNS = XMLNamespace("http://www.w3.org/1999/xhtml")
 NCXNS = XMLNamespace("http://www.daisy.org/z3986/2005/ncx/")
 OPFNS = XMLNamespace("http://www.idpf.org/2007/opf")

+PLMETNS = XMLNamespace("http://dl.psnc.pl/schemas/plmet/")

 
 WLNS = EmptyNamespace()
 
 
 
 WLNS = EmptyNamespace()
 
 

+@six.python_2_unicode_compatible
+class WLURI(object):
+    """Represents a WL URI. Extracts slug from it."""
+    slug = None
+
+    example = 'http://wolnelektury.pl/katalog/lektura/template/'
+    _re_wl_uri = re.compile(r'http://(www\.)?wolnelektury.pl/katalog/lektur[ay]/'
+            '(?P<slug>[-a-z0-9]+)/?$')
+
+    def __init__(self, uri):
+        uri = six.text_type(uri)
+        self.uri = uri
+        self.slug = uri.rstrip('/').rsplit('/', 1)[-1]
+
+    @classmethod
+    def strict(cls, uri):
+        match = cls._re_wl_uri.match(uri)
+        if not match:
+            raise ValidationError(u'Invalid URI (%s). Should match: %s' % (
+                        uri, cls._re_wl_uri.pattern))
+        return cls(uri)
+
+    @classmethod
+    def from_slug(cls, slug):
+        """Contructs an URI from slug.
+
+        >>> print(WLURI.from_slug('a-slug').uri)
+        http://wolnelektury.pl/katalog/lektura/a-slug/
+
+        """
+        uri = 'http://wolnelektury.pl/katalog/lektura/%s/' % slug
+        return cls(uri)
+
+    def __str__(self):
+        return self.uri
+
+    def __eq__(self, other):
+        return self.slug == other.slug
+
+

 class DocProvider(object):
 class DocProvider(object):

-    """ Base class for a repository of XML files.
-        Used for generating joined files, like EPUBs
+    """Base class for a repository of XML files.
+
+    Used for generating joined files, like EPUBs.

     """
 
     def by_slug(self, slug):
     """
 
     def by_slug(self, slug):

-        raise NotImplemented
-
-    def __getitem__(self, slug):
-        return self.by_slug(slug)
+        """Should return a file-like object with a WL document XML."""
+        raise NotImplementedError

 
 

-    def by_uri(self, uri):
-        return self.by_slug(uri.rsplit('/', 1)[1])
+    def by_uri(self, uri, wluri=WLURI):
+        """Should return a file-like object with a WL document XML."""
+        wluri = wluri(uri)
+        return self.by_slug(wluri.slug)

 
 
 class DirDocProvider(DocProvider):
     """ Serve docs from a directory of files in form <slug>.xml """
 
 
 
 class DirDocProvider(DocProvider):
     """ Serve docs from a directory of files in form <slug>.xml """
 

-    def __init__(self, dir):
-        self.dir = dir
+    def __init__(self, dir_):
+        self.dir = dir_

         self.files = {}
 
     def by_slug(self, slug):
         self.files = {}
 
     def by_slug(self, slug):

-        return open(os.path.join(self.dir, '%s.xml' % slug))
+        fname = slug + '.xml'
+        return open(os.path.join(self.dir, fname), 'rb')

 
 
 
 

-import lxml.etree as etree
-import dcparser
+from . import dcparser

 
 DEFAULT_BOOKINFO = dcparser.BookInfo(
 
 DEFAULT_BOOKINFO = dcparser.BookInfo(

-        { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'}, \
+        { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'},

         { DCNS('creator'): [u'Some, Author'],
           DCNS('title'): [u'Some Title'],
           DCNS('subject.period'): [u'Unknown'],
           DCNS('subject.type'): [u'Unknown'],
           DCNS('subject.genre'): [u'Unknown'],
           DCNS('date'): ['1970-01-01'],
         { DCNS('creator'): [u'Some, Author'],
           DCNS('title'): [u'Some Title'],
           DCNS('subject.period'): [u'Unknown'],
           DCNS('subject.type'): [u'Unknown'],
           DCNS('subject.genre'): [u'Unknown'],
           DCNS('date'): ['1970-01-01'],

+          DCNS('language'): [u'pol'],

           # DCNS('date'): [creation_date],
           DCNS('publisher'): [u"Fundacja Nowoczesna Polska"],
           DCNS('description'):
           # DCNS('date'): [creation_date],
           DCNS('publisher'): [u"Fundacja Nowoczesna Polska"],
           DCNS('description'):


@@ -100,32 +165,32 @@ DEFAULT_BOOKINFO = dcparser.BookInfo(
              Wolne Lektury (http://wolnelektury.pl). Reprodukcja cyfrowa
              wykonana przez Bibliotekę Narodową z egzemplarza
              pochodzącego ze zbiorów BN."""],
              Wolne Lektury (http://wolnelektury.pl). Reprodukcja cyfrowa
              wykonana przez Bibliotekę Narodową z egzemplarza
              pochodzącego ze zbiorów BN."""],

-          DCNS('identifier.url'):
-            [u"http://wolnelektury.pl/katalog/lektura/template"],
+          DCNS('identifier.url'): [WLURI.example],

           DCNS('rights'):
             [u"Domena publiczna - zm. [OPIS STANU PRAWNEGO TEKSTU]"] })
 
 def xinclude_forURI(uri):
     e = etree.Element(XINS("include"))
     e.set("href", uri)
           DCNS('rights'):
             [u"Domena publiczna - zm. [OPIS STANU PRAWNEGO TEKSTU]"] })
 
 def xinclude_forURI(uri):
     e = etree.Element(XINS("include"))
     e.set("href", uri)

-    return etree.tostring(e, encoding=unicode)
+    return etree.tostring(e, encoding='unicode')

 
 def wrap_text(ocrtext, creation_date, bookinfo=DEFAULT_BOOKINFO):
     """Wrap the text within the minimal XML structure with a DC template."""
     bookinfo.created_at = creation_date
 
     dcstring = etree.tostring(bookinfo.to_etree(), \
 
 def wrap_text(ocrtext, creation_date, bookinfo=DEFAULT_BOOKINFO):
     """Wrap the text within the minimal XML structure with a DC template."""
     bookinfo.created_at = creation_date
 
     dcstring = etree.tostring(bookinfo.to_etree(), \

-        method='xml', encoding=unicode, pretty_print=True)
+        method='xml', encoding='unicode', pretty_print=True)

 
     return u'<utwor>\n' + dcstring + u'\n<plain-text>\n' + ocrtext + \
 
     return u'<utwor>\n' + dcstring + u'\n<plain-text>\n' + ocrtext + \

-        u'\n</plain-text>\n</utwor>';
+        u'\n</plain-text>\n</utwor>'

 
 
 def serialize_raw(element):
     b = u'' + (element.text or '')
 
     for child in element.iterchildren():
 
 
 def serialize_raw(element):
     b = u'' + (element.text or '')
 
     for child in element.iterchildren():

-        e = etree.tostring(child, method='xml', encoding=unicode, pretty_print=True)
+        e = etree.tostring(child, method='xml', encoding='unicode',
+                pretty_print=True)

         b += e
 
     return b
         b += e
 
     return b


@@ -140,3 +205,75 @@ def serialize_children(element, format='raw'):
 def get_resource(path):
     return os.path.join(os.path.dirname(__file__), path)
 
 def get_resource(path):
     return os.path.join(os.path.dirname(__file__), path)
 

+
+class OutputFile(object):
+    """Represents a file returned by one of the converters."""
+
+    _bytes = None
+    _filename = None
+
+    def __del__(self):
+        if self._filename:
+            os.unlink(self._filename)
+
+    def __nonzero__(self):
+        return self._bytes is not None or self._filename is not None
+
+    @classmethod
+    def from_bytes(cls, bytestring):
+        """Converter returns contents of a file as a string."""
+
+        instance = cls()
+        instance._bytes = bytestring
+        return instance
+
+    @classmethod
+    def from_filename(cls, filename):
+        """Converter returns contents of a file as a named file."""
+
+        instance = cls()
+        instance._filename = filename
+        return instance
+
+    def get_bytes(self):
+        """Get file's contents as a bytestring."""
+
+        if self._filename is not None:
+            with open(self._filename, 'rb') as f:
+                return f.read()
+        else:
+            return self._bytes
+
+    def get_file(self):
+        """Get file as a file-like object."""
+
+        if self._bytes is not None:
+            return six.BytesIO(self._bytes)
+        elif self._filename is not None:
+            return open(self._filename, 'rb')
+
+    def get_filename(self):
+        """Get file as a fs path."""
+
+        if self._filename is not None:
+            return self._filename
+        elif self._bytes is not None:
+            temp = NamedTemporaryFile(prefix='librarian-', delete=False)
+            temp.write(self._bytes)
+            temp.close()
+            self._filename = temp.name
+            return self._filename
+        else:
+            return None
+
+    def save_as(self, path):
+        """Save file to a path. Create directories, if necessary."""
+
+        dirname = os.path.dirname(os.path.abspath(path))
+        makedirs(dirname)
+        shutil.copy(self.get_filename(), path)
+
+
+class URLOpener(FancyURLopener):
+    version = 'FNP Librarian (http://github.com/fnp/librarian)'
+urllib._urlopener = URLOpener()