Skip to content
Fetching contributors…
Cannot retrieve contributors at this time
351 lines (290 sloc) 12.5 KB
from urllib import quote
from urllib import unquote_plus
import mimetypes
import re
from copy import copy
from django.contrib.gis.db import models
from django.core.urlresolvers import reverse
from django.template.defaultfilters import stringfilter
from django.core.exceptions import ValidationError
from django.utils.safestring import mark_safe
from import (
from ckeditor.models import HTML5FragmentField
from versionutils import diff
from versionutils import versioning
import exceptions
allowed_tags = ['p', 'br', 'a', 'em', 'strong', 'u', 'img', 'h1', 'h2', 'h3',
'h4', 'h5', 'h6', 'hr', 'ul', 'ol', 'li', 'pre', 'table',
'thead', 'tbody', 'tr', 'th', 'td', 'span', 'strike', 'sub',
'sup', 'tt', 'input']
allowed_attributes_map = {'p': ['class', 'style'],
'h1': ['style'],
'h2': ['style'],
'h3': ['style'],
'h4': ['style'],
'h5': ['style'],
'h6': ['style'],
'ul': ['class'],
'a': ['class', 'name', 'href', 'style'],
'img': ['class', 'src', 'alt', 'title', 'style'],
'span': ['class', 'style'],
'table': ['class', 'style'],
'th': ['class', 'colspan', 'rowspan', 'style'],
'td': ['class', 'colspan', 'rowspan', 'style'],
'input': ['class', 'type', 'value']
allowed_styles_map = {'p': ['text-align'],
'h1': ['text-align'],
'h2': ['text-align'],
'h3': ['text-align'],
'h4': ['text-align'],
'h5': ['text-align'],
'h6': ['text-align'],
'img': ['width', 'height'],
'span': ['width', 'height'],
'table': ['width', 'height'],
'th': ['text-align', 'background-color'],
'td': ['text-align', 'background-color', 'width',
'height', 'vertical-align'],
'a': ['width']
rename_elements = {'b': 'strong',
'i': 'em'
class Page(models.Model):
name = models.CharField(max_length=255)
slug = models.SlugField(max_length=255, editable=False, unique=True)
content = HTML5FragmentField(allowed_elements=allowed_tags,
def __unicode__(self):
def get_absolute_url(self):
return reverse('pages:show', args=[self.pretty_slug])
def save(self, *args, **kwargs):
self.slug = slugify(
super(Page, self).save(*args, **kwargs)
def clean(self): = clean_name(
if not slugify(
raise ValidationError('Page name is invalid.')
def exists(self):
True if the Page currently exists in the database.
if Page.objects.filter(slug=self.slug):
return True
return False
def pretty_slug(self):
if not
return self.slug
return name_to_url(
pretty_slug = property(pretty_slug)
def name_parts(self):
name_parts = property(name_parts)
def _get_slug_related_objs(self):
# Right now this is simply hard-coded.
# TODO: generalize this slug pattern, perhaps with some kind of
# AttachedSlugField or something.
return [
{'objs': PageFile.objects.filter(slug=self.slug),
'unique_together': ('name', 'slug')},
def rename_to(self, pagename):
Renames the page to `pagename`. Moves related objects around
def _get_slug_lookup(unique_together, obj, new_p):
d = {}
for field in unique_together:
d[field] = getattr(obj, field)
d['slug'] = new_p.slug
return d
from redirects.models import Redirect
from redirects.exceptions import RedirectToSelf
if Page.objects.filter(slug=slugify(pagename)):
if slugify(pagename) == self.slug:
# The slug is the same but we're changing the name.
old_name = = pagename'Renamed from "%s"' % old_name)
raise exceptions.PageExistsError(
"The page '%s' already exists!" % pagename)
# Copy the current page into the new page, zeroing out the
# primary key and setting a new name and slug.
new_p = copy(self) = None = pagename
new_p.slug = slugify(pagename)'Renamed from "%s"' %
# Get all related objects before the original page is deleted.
related_objs = []
for r in self._meta.get_all_related_objects():
rel_obj = getattr(self, r.get_accessor_name())
continue # No object for this relation.
# Is this a related /set/, e.g. redirect_set?
if isinstance(rel_obj, models.Manager):
# list() freezes the QuerySet, which we don't want to be
# fetched /after/ we delete the page.
(r.get_accessor_name(), list(rel_obj.all())))
related_objs.append((r.get_accessor_name(), rel_obj))
# Cache all ManyToMany values on related objects so we can restore them
# later--otherwise they will be lost when page is deleted.
for attname, rel_obj_list in related_objs:
if not isinstance(rel_obj_list, list):
rel_obj_list = [rel_obj_list]
for rel_obj in rel_obj_list:
rel_obj._m2m_values = dict(
(f.attname, list(getattr(rel_obj, f.attname).all()))
for f in rel_obj._meta.many_to_many)
# Create a redirect from the starting pagename to the new pagename.
redirect = Redirect(source=self.slug, destination=new_p)
# Creating the redirect causes the starting page to be deleted.
# Point each related object to the new page and save the object with a
# 'was renamed' comment.
for attname, rel_obj in related_objs:
if isinstance(rel_obj, list):
for obj in rel_obj: = None # Reset the primary key before saving.
getattr(new_p, attname).add(obj)"Parent page renamed")
# Restore any m2m fields now that we have a new pk
for name, value in obj._m2m_values.items():
setattr(obj, name, value)
except RedirectToSelf, s:
# We don't want to create a redirect to ourself.
# This happens during a rename -> rename-back
# cycle.
# This is an easy way to set obj to point to new_p.
setattr(new_p, attname, rel_obj) = None # Reset the primary key before saving."Parent page renamed")
# Restore any m2m fields now that we have a new pk
for name, value in rel_obj._m2m_values.items():
setattr(rel_obj, name, value)
# Do the same with related-via-slug objects.
for info in self._get_slug_related_objs():
unique_together = info['unique_together']
objs = info['objs']
for obj in objs:
# If we already have the same object with this slug then
# skip it. This happens when there's, say, a PageFile that's
# got the same name that's attached to the page -- which can
# happen during a page rename -> rename back cycle.
obj_lookup = _get_slug_lookup(unique_together, obj, new_p)
if obj.__class__.objects.filter(**obj_lookup):
obj.slug = new_p.slug = None # Reset the primary key before saving."Parent page renamed")
class PageDiff(diff.BaseModelDiff):
fields = ('name',
('content', diff.diffutils.HtmlFieldDiff),
diff.register(Page, PageDiff)
class PageFile(models.Model):
file = models.FileField(upload_to='pages/files/',
name = models.CharField(max_length=255)
slug = models.SlugField(max_length=255, editable=False)
_rough_type_map = [(r'^audio', 'audio'),
(r'^video', 'video'),
(r'^application/pdf', 'pdf'),
(r'^application/msword', 'word'),
(r'^text/html', 'html'),
(r'^text', 'text'),
(r'^image', 'image'),
(r'^application/', 'powerpoint'),
(r'^application/', 'excel')
def get_absolute_url(self):
return reverse('pages:file',
kwargs={'slug': self.slug, 'file':})
def attached_to_page(self):
p = Page.objects.get(slug=self.slug)
except Page.DoesNotExist:
p = Page(slug=self.slug, name=clean_name(self.slug))
return p
def rough_type(self):
mime = self.mime_type
if mime:
for regex, rough_type in self._rough_type_map:
if re.match(regex, mime):
return rough_type
return 'unknown'
def mime_type(self):
return mimetypes.guess_type([0]
def is_image(self):
return self.rough_type == 'image'
class Meta:
unique_together = ('slug', 'name')
ordering = ['-id']
def clean_name(name):
# underscores are used to namespace special URLs, so let's remove them
name = re.sub('_', ' ', name).strip()
# we allow / in page names so we want to strip each bit between slashes
name = '/'.join([part.strip()
for part in name.split('/') if slugify(part)])
return name
def slugify(value, keep=r"\-\.,'\"/!@$%&*()"):
Normalizes page name for db lookup
value: String or unicode object to normalize.
keep: Special non-word and non-space characters that should not get
stripped out and contribute to a slug's uniqueness. Defaults to
characters important to meaning.
Lowercase string with special characters removed.
value = url_to_name(value)
# normalize unicode
import unicodedata
value = unicodedata.normalize('NFKD', unicode(value))
# remove non-{word,space,keep} characters
misc_characters = re.compile(('[^\w\s%s]' % keep), re.UNICODE)
value = re.sub(misc_characters, '', value)
value = value.strip()
value = re.sub('[_\s]+', ' ', value)
return value.lower()
slugify = stringfilter(slugify)
def name_to_url(value):
"""Converts page name to its canonical URL path
# spaces to underscore
value = re.sub('[\s]', '_', value.strip())
# url-encode
value = quote(value.encode('utf-8'))
return mark_safe(value)
name_to_url.is_safe = True
name_to_url = stringfilter(name_to_url)
def url_to_name(value):
"""Converts URL to the intended page name
# decode URL-encoded chars
value = unquote_plus(value.encode('utf-8')).decode('utf-8')
return re.sub('_', ' ', value).strip()
url_to_name = stringfilter(url_to_name)
import feeds # To fire register() calls.
import signals # To fire signal calls.
Something went wrong with that request. Please try again.