librarian in submodule, HTML and TXT book previews, URL fixes
[redakcja.git] / apps / wiki / models.py
index 5faf1d3..66f8a28 100644 (file)
 # This file is part of FNP-Redakcja, licensed under GNU Affero GPLv3 or later.
 # Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information.
 #
-from django.db import models
+import itertools
 import re
-import os
-import vstorage
-from vstorage import DocumentNotFound
-from wiki import settings, constants
+
+from django.db import models
 from django.utils.translation import ugettext_lazy as _
 
-from django.http import Http404
+from dvcs import models as dvcs_models
+
 
 import logging
 logger = logging.getLogger("fnp.wiki")
 
 
-# _PCHARS_DICT = dict(zip((ord(x) for x in u"ĄĆĘŁŃÓŚŻŹąćęłńóśżź "), u"ACELNOSZZacelnoszz_"))
-_PCHARS_DICT = dict(zip((ord(x) for x in u" "), u"_"))
-
-# I know this is barbaric, but I didn't find a better solution ;(
-def split_name(name):
-    parts = name.translate(_PCHARS_DICT).split('__')
-    return parts
+RE_TRIM_BEGIN = re.compile("^<!-- TRIM_BEGIN -->$", re.M)
+RE_TRIM_END = re.compile("^<!-- TRIM_END -->$", re.M)
 
-def join_name(*parts, **kwargs):
-    name = u'__'.join(p.translate(_PCHARS_DICT) for p in parts)
-    logger.info("JOIN %r -> %r", parts, name)
-    return name
 
-def normalize_name(name):
-    """
-    >>> normalize_name("gąska".decode('utf-8'))
-    u'g\u0105ska'
-    """
-    return name.translate(_PCHARS_DICT).lower()
+class Book(models.Model):
+    """ A document edited on the wiki """
 
-STAGE_TAGS_RE = re.compile(r'^#stage-finished: (.*)$', re.MULTILINE)
+    title = models.CharField(_('title'), max_length=255)
+    slug = models.SlugField(_('slug'), max_length=128, unique=True)
+    gallery = models.CharField(_('scan gallery name'), max_length=255, blank=True)
 
+    parent = models.ForeignKey('self', null=True, blank=True, verbose_name=_('parent'), related_name="children")
+    parent_number = models.IntegerField(_('parent number'), null=True, blank=True, db_index=True)
 
-class DocumentStorage(object):
-    def __init__(self, path):
-        self.vstorage = vstorage.VersionedStorage(path)
-
-    def get(self, name, revision=None):
-        text, rev = self.vstorage.page_text(name, revision)
-        return Document(self, name=name, text=text, revision=rev)
+    class Meta:
+        ordering = ['parent_number', 'title']
+        verbose_name = _('book')
+        verbose_name_plural = _('books')
 
-    def get_by_tag(self, name, tag):
-        text, rev = self.vstorage.page_text_by_tag(name, tag)
-        return Document(self, name=name, text=text, revision=rev)
+    def __unicode__(self):
+        return self.title
+
+    @classmethod
+    def create(cls, creator=None, text=u'', *args, **kwargs):
+        """
+            >>> Book.create(slug='x', text='abc').materialize()
+            'abc'
+        """
+        instance = cls(*args, **kwargs)
+        instance.save()
+        instance.chunk_set.all()[0].doc.commit(author=creator, text=text)
+        return instance
+
+    @staticmethod
+    def trim(text, trim_begin=True, trim_end=True):
+        """ 
+            Cut off everything before RE_TRIM_BEGIN and after RE_TRIM_END, so
+            that eg. one big XML file can be compiled from many small XML files.
+        """
+        if trim_begin:
+            text = RE_TRIM_BEGIN.split(text, maxsplit=1)[-1]
+        if trim_end:
+            text = RE_TRIM_END.split(text, maxsplit=1)[0]
+        return text
+
+    def materialize(self):
+        """ 
+            Get full text of the document compiled from chunks.
+            Takes the current versions of all texts for now, but it should
+            be possible to specify a tag or a point in time for compiling.
+
+            First non-empty text's beginning isn't trimmed,
+            and last non-empty text's end isn't trimmed.
+        """
+        texts = []
+        trim_begin = False
+        text = ''
+        for chunk in self.chunk_set.all():
+            next_text = chunk.doc.materialize()
+            if not next_text:
+                continue
+            if text:
+                # trim the end, because there's more non-empty text
+                # don't trim beginning, if `text' is the first non-empty part
+                texts.append(self.trim(text, trim_begin=trim_begin))
+                trim_begin = True
+            text = next_text
+        # don't trim the end, because there's no more text coming after `text'
+        # only trim beginning if it's not still the first non-empty
+        texts.append(self.trim(text, trim_begin=trim_begin, trim_end=False))
+        return "".join(texts)
+
+    @staticmethod
+    def listener_create(sender, instance, created, **kwargs):
+        if created:
+            instance.chunk_set.create(number=1, slug='1')
+
+models.signals.post_save.connect(Book.listener_create, sender=Book)
+
+
+class Chunk(models.Model):
+    """ An editable chunk of text. Every Book text is divided into chunks. """
+
+    book = models.ForeignKey(Book)
+    number = models.IntegerField()
+    slug = models.SlugField()
+    comment = models.CharField(max_length=255)
+    doc = models.ForeignKey(dvcs_models.Document, editable=False, unique=True, null=True)
 
-    def revert(self, name, revision):
-        text, rev = self.vstorage.revert(name, revision)
-        return Document(self, name=name, text=text, revision=rev)
+    class Meta:
+        unique_together = [['book', 'number'], ['book', 'slug']]
+        ordering = ['number']
 
-    def get_or_404(self, *args, **kwargs):
-        try:
-            return self.get(*args, **kwargs)
-        except DocumentNotFound:
-            raise Http404
+    def __unicode__(self):
+        return "%d-%d: %s" % (self.book_id, self.number, self.comment)
 
-    def put(self, document, author, comment, parent=None):
-        self.vstorage.save_text(
-                title=document.name,
-                text=document.text,
-                author=author,
-                comment=comment,
-                parent=parent)
+    def save(self, *args, **kwargs):
+        if self.doc is None:
+            self.doc = dvcs_models.Document.objects.create()
+        super(Chunk, self).save(*args, **kwargs)
 
-        return document
+    @classmethod
+    def get(cls, slug, chunk=None):
+        if chunk is None:
+            return cls.objects.get(book__slug=slug, number=1)
+        else:
+            return cls.objects.get(book__slug=slug, slug=chunk)
 
-    def create_document(self, text, name):
-        title = u', '.join(p.title for p in split_name(name))
+    def pretty_name(self):
+        return "%s, %s (%d/%d)" % (self.book.title, self.comment, 
+                self.number, self.book.chunk_set.count())
 
-        if text is None:
-            text = u''
 
-        document = Document(self, name=name, text=text, title=title)
-        return self.put(document, u"<wiki>", u"Document created.")
 
-    def delete(self, name, author, comment):
-        self.vstorage.delete_page(name, author, comment)
 
-    def all(self):
-        return list(self.vstorage.all_pages())
+'''
+from wiki import settings, constants
+from slughifi import slughifi
 
-    def history(self, title):
-        def stage_desc(match):
-            stage = match.group(1)
-            return _("Finished stage: %s") % constants.DOCUMENT_STAGES_DICT[stage]
+from django.http import Http404
 
-        for changeset in self.vstorage.page_history(title):
-            changeset['description'] = STAGE_TAGS_RE.sub(stage_desc, changeset['description'])
-            yield changeset
 
 
 
 class Document(object):
-    META_REGEX = re.compile(r'\s*<!--\s(.*?)-->', re.DOTALL | re.MULTILINE)
-
-    def __init__(self, storage, **kwargs):
-        self.storage = storage
-        for attr, value in kwargs.iteritems():
-            setattr(self, attr, value)
 
     def add_tag(self, tag, revision, author):
         """ Add document specific tag """
@@ -127,7 +163,7 @@ class Document(object):
                 except ValueError:
                     continue
 
-        gallery = result.get('gallery', self.name.replace(' ', '_'))
+        gallery = result.get('gallery', slughifi(self.name.replace(' ', '_')))
 
         if gallery.startswith('/'):
             gallery = os.path.basename(gallery)
@@ -138,13 +174,9 @@ class Document(object):
     def info(self):
         return self.storage.vstorage.page_meta(self.name, self.revision)
 
-def getstorage():
-    return DocumentStorage(settings.REPOSITORY_PATH)
 
-#
-# Django models
-#
 
+'''
 class Theme(models.Model):
     name = models.CharField(_('name'), max_length=50, unique=True)