Dynastie/generators/index.py

427 lines
15 KiB
Python

import os
import datetime
import xml
from xml.parsers.expat import *
from xml.dom.minidom import parse, parseString
from dynastie.generators.generator import DynastieGenerator, StrictUTF8Writer
from django.db import models
class Index(DynastieGenerator):
cur_page = 0
nb_pages = 0
cur_post = 0
cur_post_obj = None
posts_per_page = 0
filename = 'index'
dirname = ''
blog = None
def __init__(self, hash_posts=None, hash_posts_content=None):
DynastieGenerator.__init__(self, hash_posts, hash_posts_content)
self.hooks = {'posts' : self.createPosts,
'title' : self.createTitle,
'navigation' : self.createNavigation,
'recents' : self.createRecents,
'tags' : self.createTags,
'replace' : self.createReplace}
self.first_try = True
def createReplace(self, posts, dom, root, replace_elem):
if not replace_elem.hasAttribute('div_name'):
self.addError('No attribute div_name for a replace tag')
return
div_element = replace_elem.cloneNode(True)
div_element.tagName = replace_elem.getAttribute('div_name')
div_element.removeAttribute('div_name')
for key,value in replace_elem.attributes.items():
if key == 'div_name': continue
value = value.replace('dyn:blog_id', str(self.blog.id))
div_element.setAttribute(key, value)
root.replaceChild(div_element, replace_elem)
return div_element
def createNavigation(self, posts, dom, root, node):
if self.nb_pages == 0 or self.nb_pages == 1:
return None
if self.dirname != '':
if self.dirname.startswith('/'):
href = '<a href="' + self.dirname + '/' + self.filename
else:
href = '<a href="/' + self.dirname + '/' + self.filename
else:
href = '<a href="/' + self.filename
nav = ''
if self.cur_page != 0:
nav = nav + href + '.html">&lt;&lt; First</a> '
if self.cur_page == 1:
nav = nav + href + '.html">&lt; Prev</a> '
else:
nav = nav + href + str(self.cur_page-1) + '.html">&lt; Prev</a> '
start = self.cur_page-5
if start < 0:
start = 0
end = start + 10
if end > self.nb_pages:
end = self.nb_pages
for i in range(start, end):
if i == self.cur_page:
nav = nav + str(i+1) + ' '
else:
if i == 0:
nav = nav + href + '.html">1</a> '
else:
nav = nav + href + str(i) + '.html">' + str(i+1) + '</a> '
if self.cur_page < self.nb_pages-1:
nav = nav + href + str(self.cur_page+1) + '.html">Next &gt;</a> '
nav = nav + href + str(self.nb_pages-1) + '.html">Last &gt;&gt;</a>'
new_dom = parseString('<div class="navigation">' + nav + '</div>')
new_node = new_dom.getElementsByTagName('div')[0]
res = new_node.cloneNode(True)
root.replaceChild(res, node)
return res
def createTitle(self, posts, dom, root, title_elem):
create_link = (title_elem.getAttribute('link') == '1')
post = self.cur_post_obj
if create_link == True:
node = self.createElement(dom, 'title')
node.appendChild(self.createLinkElem(dom, post.getPath(), post.title))
else:
node = self.createElement(dom, 'title', post.title)
root.replaceChild(node, title_elem)
def createPost(self, posts, dom, post_elem, root):
post = self.cur_post_obj
if post.id in self.hash_posts and not self.first_try:
node = self.hash_posts[post.id]
return node.cloneNode(0)
values = {}
values['author'] = post.author.first_name + ' ' + post.author.last_name
values['date'] = post.creation_date.strftime('%A, %d %B %Y %H:%m')
values['post_content'] = ''
blog = post.blog
blog.create_paths()
filename = blog.src_path + '/_post/' + str(post.id)
if not os.path.exists(filename):
self.addError('File does not exists ' + filename)
return None
if not filename in self.hash_posts_content:
f = open(filename, 'rb')
post_content = f.read()
f.close()
self.hash_posts_content[filename] = post_content
else:
post_content = self.hash_posts_content[filename]
while True:
start = post_content.find('<dyn:code')
if start == -1: break
end = post_content.find('</dyn:code>')
if end < start:
self.addError('Invalid <dyn:code> tags in ' + filename)
break
try:
dom2 = parseString(post_content[start:end+11])
except xml.dom.DOMException as e:
self.addError('Error parsing ' + filename)
break
res = self.createCode(dom2, dom2.firstChild)
post_content = post_content.replace(post_content[start:end+11], res)
self.simpleTransform(values, dom, post_elem, root)
content_nodes = post_elem.getElementsByTagName('div')
post_transform = ['post_content']
for content_node in content_nodes:
the_class = content_node.getAttribute('class')
if not the_class in post_transform:
continue
new_node = dom.createTextNode(post_content)
content_node.appendChild(new_node)
if post.id in self.hash_posts:
import hashlib
# Here, we are in first_try, check that computed
# post has the same result than the one in cache
self.first_try = False
writer = StrictUTF8Writer()
node = self.hash_posts[post.id]
node.writexml(writer)
content1 = writer.getvalue().encode('utf-8')
writer = StrictUTF8Writer()
post_elem.writexml(writer)
content2 = writer.getvalue().encode('utf-8')
md5_1 = hashlib.md5()
md5_1.update(content1)
md5_2 = hashlib.md5()
md5_2.update(content2)
# If not, clear cache
if md5_1.digest() != md5_2.digest():
self.hash_posts = {}
self.hash_posts[post.id] = post_elem.cloneNode(0)
return post_elem
def createPosts(self, posts, dom, root, node):
posts_elem = self.createElement(dom, 'posts')
create_link = (node.getAttribute('link') == '1')
for i in range(0, self.posts_per_page):
post_elem = self.createElement(dom, 'post')
if len(posts) > self.cur_post:
self.cur_post_obj = posts[self.cur_post]
post_elem = self.createPost(posts, dom, post_elem, node)
else:
post_elem = self.createElement(dom, '', '<b>No posts yet</b>')
self.cur_post_obj = None
if not post_elem is None:
posts_elem.appendChild(post_elem)
# Parse inner HTML
self._parse(self.hooks, posts, dom, post_elem)
self.cur_post = self.cur_post + 1
if self.cur_post == len(posts):
break
root.replaceChild(posts_elem, node)
return posts_elem
def createRecents(self, posts, dom, root, node):
if self.cur_post == len(posts):
root.removeChild(node)
return
if node.hasAttribute("limit"):
nb_recents = int(node.getAttribute("limit"))
else:
nb_recents = 5
recents_elem = self.createElement(dom, 'recents')
for child in node.childNodes:
recents_elem.appendChild(child.cloneNode(True))
list_elem = dom.createElement('ul')
for i in range(0, nb_recents):
post_elem = dom.createElement('li')
if self.cur_post+i < len(posts):
post = posts[self.cur_post+i]
link_elem = self.createLinkElem(dom, post.getPath(), post.title)
post_elem.appendChild(link_elem)
else:
break
list_elem.appendChild(post_elem)
recents_elem.appendChild(list_elem)
root.replaceChild(recents_elem, node)
return recents_elem
def createTags(self, posts, dom, root, node):
from dynastie.models import Post
tags_elem = self.createElement(dom, 'tags')
create_link = (node.getAttribute('link') == '1')
if type(posts) == models.query.QuerySet:
if len(posts) > self.cur_post:
cur_post = posts[self.cur_post]
else:
cur_post = None
elif type(posts) == Post:
cur_post = posts
else:
cur_post = None
if not cur_post is None:
for tag in cur_post.tags.all():
if create_link:
tag_elem = self.createElement(dom, 'tag')
link_elem = self.createLinkElem(dom, '/tag/' + tag.name_slug, '#' + tag.name)
tag_elem.appendChild(link_elem)
else:
tag_elem = self.createElement(dom, 'tag', '#' + tag.name)
tags_elem.appendChild(tag_elem)
if len(cur_post.tags.all()) == 0:
root.removeChild(node)
return None
else:
root.replaceChild(tags_elem, node)
else:
root.removeChild(node)
return None
return tags_elem
def createCode(self, dom, node):
try:
from pygments import highlight
from pygments.util import ClassNotFound
from pygments.lexers import get_lexer_by_name
from pygments.formatters import get_formatter_by_name
except ImportError:
self.addError('Pygments package is missing, please install it in order to use <dyn:code>')
return
language = node.getAttribute('language')
if language is None:
self.addWarning('No language defined for <dyn:code> assuming C language')
language = "c"
else:
language = language.lower()
lexer_options = {}
try:
lexer = get_lexer_by_name(language, **lexer_options)
except ClassNotFound:
self.addWarning('Language ' + language + ' not supported by current version of pygments')
lexer = get_lexer_by_name('c', **lexer_options)
formatter_options = {'classprefix' : 'color_emacs_', 'style' : 'emacs'}
for k in node.attributes.keys():
attr = node.attributes[k]
if attr.prefix != '': continue
if attr.name == 'language': continue
name = attr.name
value = attr.value
if name == 'colouring':
name = 'style'
formatter_options['classprefix'] = 'color_' + value + '_'
formatter_options[name] = value
formatter = get_formatter_by_name('html', **formatter_options)
lexer.encoding = 'utf-8'
formatter.encoding = 'utf-8'
writer = StrictUTF8Writer()
node.firstChild.writexml(writer)
code = writer.getvalue().encode('utf-8')
r,w = os.pipe()
r,w=os.fdopen(r,'r',0), os.fdopen(w,'w',0)
highlight(code, lexer, formatter, w)
w.close()
code = r.read()
r.close()
# Remove <pre> after <div class="highlight">
code = code[28:-13]
code = '<div class="highlight">' + code + '</div>'
return code
def parseTemplate(self, blog, src, output, name, directory=None):
self.blog = blog
if not os.path.exists(src + '/_%s.html' % name):
self.addError('No _%s.html found, exiting' % name)
return None
try:
dom = parse(src + '/_%s.html' % name)
except xml.dom.DOMException as e:
self.addError('Error parsing _%s.html : ' + e)
return None
if not directory is None and not os.path.exists(output + '/' + directory):
os.mkdir(output + '/' + directory)
post_nodes = dom.getElementsByTagNameNS(self.URI, "posts")
if not post_nodes is None:
if post_nodes[0].hasAttribute("limit"):
self.posts_per_page = int(post_nodes[0].getAttribute("limit"))
else:
self.posts_per_page = 5
else:
self.addError('No tag dyn:posts found')
return dom
def generatePages(self, dom, posts, src, output, name, directory=None):
if len(posts) > self.posts_per_page:
self.nb_pages = self.computeNbPages(len(posts), self.posts_per_page)
if not directory is None and not os.path.exists(output + self.dirname):
os.mkdir(output + self.dirname)
filename = self.dirname + '/' + self.filename + '.html'
impl = xml.dom.getDOMImplementation()
while self.cur_page <= self.nb_pages:
#print 'Generate ' + filename
dom_ = impl.createDocument('', 'xml', None)
dom_.replaceChild(dom.firstChild.cloneNode(0), dom_.firstChild)
nodes = dom.getElementsByTagName("*")
nodes[0] = self.parse(src, self.hooks, posts, dom_, nodes[0])
self.writeIfNotTheSame(output + filename, nodes[0])
self.cur_page = self.cur_page + 1
filename = self.dirname + '/' + self.filename + str(self.cur_page) + '.html'
filename = output + filename
while os.path.exists(filename):
self.addReport('Removing unused ' + filename)
os.unlink(filename)
filename = filename + '.gz'
if os.path.exists(filename):
self.addReport('Removing unused ' + filename)
os.unlink(filename)
self.cur_page = self.cur_page + 1
filename = self.dirname + '/' + self.filename + str(self.cur_page) + '.html'
filename = output + filename
def generate(self, blog, src, output):
from dynastie.models import Post, Blog
dom = self.parseTemplate(blog, src, output, 'index')
if dom is None: return self.report
now = datetime.datetime.now()
cur_year = now.year
posts = Post.objects.filter(creation_date__year=cur_year, published=True, front_page=True).order_by('-creation_date')
if posts.count() < self.posts_per_page:
posts = Post.objects.filter(published=True, front_page=True).order_by('-creation_date')[:self.posts_per_page]
self.dirname = ''
self.generatePages(dom, posts, src, output, 'index')
if not self.somethingWrote:
self.addReport('Nothing changed')
return self.report