Some improvements in missing material handling.
[librarian.git] / librarian / __init__.py
index 52234fe..215cacb 100644 (file)
@@ -8,24 +8,34 @@ from __future__ import with_statement
 import os
 import re
 import shutil
+import urllib
 
-class ParseError(Exception):
+
+class UnicodeException(Exception):
     def __str__(self):
         """ Dirty workaround for Python Unicode handling problems. """
-        return self.message
+        return unicode(self).encode('utf-8')
 
     def __unicode__(self):
         """ Dirty workaround for Python Unicode handling problems. """
-        return self.message
+        args = self.args[0] if len(self.args) == 1 else self.args
+        try:
+            message = unicode(args)
+        except UnicodeDecodeError:
+            message = unicode(args, encoding='utf-8', errors='ignore')
+        return message
+
+class ParseError(UnicodeException):
+    pass
 
-class ValidationError(Exception):
+class ValidationError(UnicodeException):
     pass
 
 class NoDublinCore(ValidationError):
     """There's no DublinCore section, and it's required."""
     pass
 
-class NoProvider(Exception):
+class NoProvider(UnicodeException):
     """There's no DocProvider specified, and it's needed."""
     pass
 
@@ -66,60 +76,45 @@ WLNS = EmptyNamespace()
 
 
 class WLURI(object):
-    """Represents a WL URI. Extracts slug and language from it."""
-    DEFAULT_LANGUAGE = u'pol'
-
+    """Represents a WL URI. Extracts slug from it."""
     slug = None
-    language = None
 
-    example = 'http://wolnelektury.pl/katalog/lektura/template/'
-    _re_wl_uri = re.compile('http://wolnelektury.pl/katalog/lektura/'
-            '(?P<slug>[-a-z]+)(/(?P<lang>[a-z]{3})/?)?')
+    example = 'http://edukacjamedialna.edu.pl/'
+    _re_wl_uri = re.compile(r'http://(www\.)?edukacjamedialna.edu.pl/(lekcje/)?'
+            '(?P<slug>[-a-z0-9]+)/?$')
 
-    def __init__(self, uri=None):
-        if uri is not None:
-            self.uri = uri
-            match = self._re_wl_uri.match(uri)
-            assert match
-            self.slug = match.group('slug')
-            self.language = match.group('lang') or self.DEFAULT_LANGUAGE
+    def __init__(self, uri):
+        uri = unicode(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_and_lang(cls, slug, lang):
-        """Contructs an URI from slug and language code.
+    def from_slug(cls, slug):
+        """Contructs an URI from slug.
 
-        >>> WLURI.from_slug_and_lang('a-slug', WLURI.DEFAULT_LANGUAGE).uri
-        'http://wolnelektury.pl/katalog/lektura/a-slug/'
-        >>> WLURI.from_slug_and_lang('a-slug', 'deu').uri
-        'http://wolnelektury.pl/katalog/lektura/a-slug/deu/'
+        >>> WLURI.from_slug('a-slug').uri
+        u'http://wolnelektury.pl/katalog/lektura/a-slug/'
 
         """
-        if lang is None:
-            lang = self.DEFAULT_LANGUAGE
-        uri = 'http://wolnelektury.pl/katalog/lektura/%s/' % slug
-        if lang is not None and lang != cls.DEFAULT_LANGUAGE:
-            uri += lang + '/'
-        instance = cls()
-        instance.slug = slug
-        instance.language = lang
-        instance.uri = uri
-        return instance
+        uri = 'http://prawokultury.pl/publikacje/%s/' % slug
+        return cls(uri)
 
     def __unicode__(self):
         return self.uri
 
-    def __eq__(self, other):
-        return self.slug, self.language == other.slug, other.language
-
-    def filename_stem(self):
-        stem = self.slug
-        if self.language != self.DEFAULT_LANGUAGE:
-            stem += '_' + self.language
-        return stem
+    def __str__(self):
+        return self.uri
 
-    def validate_language(self, language):
-        if language != self.language:
-            raise ValidationError("Incorrect language definition in URI")
+    def __eq__(self, other):
+        return self.slug == other.slug
 
 
 class DocProvider(object):
@@ -128,18 +123,14 @@ class DocProvider(object):
     Used for generating joined files, like EPUBs.
     """
 
-    def by_slug_and_lang(self, slug, lang=None):
-        """Should return a file-like object with a WL document XML."""
-        raise NotImplementedError
-
     def by_slug(self, slug):
         """Should return a file-like object with a WL document XML."""
-        return self.by_slug_and_lang(slug)
+        raise NotImplementedError
 
-    def by_uri(self, uri):
+    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_and_lang(wluri.slug, wluri.language)
+        wluri = wluri(uri)
+        return self.by_slug(wluri.slug)
 
 
 class DirDocProvider(DocProvider):
@@ -148,11 +139,10 @@ class DirDocProvider(DocProvider):
     def __init__(self, dir_):
         self.dir = dir_
         self.files = {}
-        return super(DirDocProvider, self).__init__()
 
-    def by_slug_and_lang(self, slug, lang=None):
-        fname = WLURI.from_slug_and_lang(slug, lang).filename_stem() + '.xml'
-        return open(os.path.join(self.dir, fname))
+    def by_slug(self, slug):
+        fname = slug + '.xml'
+        return IOFile.from_filename(os.path.join(self.dir, fname))
 
 
 import lxml.etree as etree
@@ -160,13 +150,16 @@ import dcparser
 
 DEFAULT_BOOKINFO = dcparser.BookInfo(
         { RDFNS('about'): u'http://wiki.wolnepodreczniki.pl/Lektury:Template'},
-        { DCNS('creator'): [u'Some, Author'],
+        { 
+          DCNS('creator.expert'): [u'Some, Author'],
+          DCNS('creator.scenario'): [u'Some, Author'],
+          DCNS('creator.textbook'): [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'): [WLURI.DEFAULT_LANGUAGE],
+          DCNS('language'): [u'pol'],
           # DCNS('date'): [creation_date],
           DCNS('publisher'): [u"Fundacja Nowoczesna Polska"],
           DCNS('description'):
@@ -215,32 +208,35 @@ 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."""
-
+class IOFile(object):
+    """ Represents a file fed as input or returned as a result. """
     _string = None
     _filename = None
+    _filename_tmp = False
+
+    def __init__(self, attachments=None):
+        self.attachments = attachments or {}
 
     def __del__(self):
-        if self._filename:
+        if self._filename_tmp:
             os.unlink(self._filename)
 
     def __nonzero__(self):
         return self._string is not None or self._filename is not None
 
     @classmethod
-    def from_string(cls, string):
+    def from_string(cls, string, *args, **kwargs):
         """Converter returns contents of a file as a string."""
 
-        instance = cls()
+        instance = cls(*args, **kwargs)
         instance._string = string
         return instance
 
     @classmethod
-    def from_filename(cls, filename):
+    def from_filename(cls, filename, *args, **kwargs):
         """Converter returns contents of a file as a named file."""
 
-        instance = cls()
+        instance = cls(*args, **kwargs)
         instance._filename = filename
         return instance
 
@@ -273,6 +269,7 @@ class OutputFile(object):
             temp.write(self._string)
             temp.close()
             self._filename = temp.name
+            self._filename_tmp = True
             return self._filename
         else:
             return None
@@ -284,3 +281,25 @@ class OutputFile(object):
         if not os.path.isdir(dirname):
             os.makedirs(dirname)
         shutil.copy(self.get_filename(), path)
+
+    def dump_to(self, path, directory=None):
+        """ Path should be name for main file. """
+        self.save_as(path)
+        dirname = os.path.dirname(os.path.abspath(path))
+        for filename, attachment in self.attachments.items():
+            attachment.save_as(os.path.join(dirname, filename))
+
+
+class Format(object):
+    """ Generic format class. """
+    def __init__(self, wldoc, **kwargs):
+        self.wldoc = wldoc
+        self.customization = kwargs
+
+    def build(self):
+        raise NotImplementedError
+
+
+class URLOpener(urllib.FancyURLopener):
+    version = 'FNP Librarian (http://github.com/fnp/librarian)'
+urllib._urlopener = URLOpener()