treecutter/sitemap.py

   1 #!/usr/bin/python
   2 import os
   3 import codecs
   4 import re
   5 import shutil
   6 import sys
   7 import gettext
   8 import tempfile
   9 from lxml import etree
  10 from lxml.builder import ElementMaker
  11 from time import time
  12 from treecutter import constants as const
  13 from treecutter.trie import Trie
  14 from treecutter.link import Link
  15 from treecutter.tools import ssh_cmd, publish, mkdir_p,get_folder_size,sizeof_fmt
  16
  17
  18 class Sitemap():
  19     """Class keeping the internal site structure"""
  20     def __init__(self,args):
  21         self._output = args.output
  22         self._style = args.style
  23         self._subdir = args.subdir
  24         self._file = 'sitemap.txt'
  25         self._tree = Trie()
  26         self._sitelang = set()
  27         self._isocode = etree.parse('/usr/share/xml/iso-codes/iso_639_3.xml')
  28         self._tranlang = {}
  29         self._tmptarget = tempfile.mkdtemp()+'/'
  30
  31     # The sitemap uses a trie structure to keep track of links
  32     # A link represents the path to the document and the link
  33     # representing the text on the site.
  34     # A link can have several pages in different languages.
  35     def add_link(self, link):
  36         tokens = filter(None,re.split(r'(^/[\w\.:-]*$|^/[\w\.:-]*/|[\w\.:-]*/)',link,flags=re.UNICODE))
  37         self._tree.add(tokens,Link(link))
  38
  39     def write_map(self):
  40         f = codecs.open(self._file,'w','utf-8')
  41         s = '\n'.join(link.link() for link in self._tree)
  42         f.write(s)
  43         f.close()
  44
  45     def read_map(self):
  46         try:
  47             f = codecs.open(self._file, 'r', 'utf-8')
  48             sml = f.read().split()
  49             f.close()
  50             for line in sml:
  51                 self.add_link(line)
  52         except IOError, what_error:
  53             print 'INFO: Could not read sitemap.txt - one will be created'
  54
  55     # Create a set of the current tree for comparison with the
  56     # directory scan
  57     def set(self):
  58         return set(link.link() for link in self._tree)
  59
  60     # Main driver in the application processing the documents
  61     # in the collected sitemap
  62     def process(self):
  63         t1 = time()
  64         print "Prepareing the input"
  65         for link in self._tree:
  66             link.prepare()
  67         t2 = time()
  68         print "Prepare  [%5.2f s]" % (round(t2-t1,2))
  69         for link in self._tree:
  70             self._sitelang = self._sitelang.union(set(link.languages()))
  71         for tran in self._sitelang:
  72             if tran != 'en':
  73                 self._tranlang[tran] = gettext.translation('iso_639_3',
  74                                                            languages=[tran])
  75         t3 = time()
  76         print "Language [%5.2f s]" % (round(t3-t2,2))
  77         transform = {}
  78         transform['xhtml5'] = etree.XSLT(etree.parse(self._style+"docbook.xhtml5.xsl"))
  79         for link in self._tree:
  80             link.render(transform)
  81         t4 = time()
  82         print "Render   [%5.2f s]" % (round(t4-t3,2))
  83         for link in self._tree:
  84             link.template(self, self._style, self._tmptarget,self._subdir)
  85         t5 = time()
  86         print "Template [%5.2f s]" % (round(t5-t4,2))
  87         t6 = time()
  88         res = set()
  89         # Collect all files used by the documents
  90         for link in self._tree:
  91             res = res.union(link.resources())
  92         for f in res:
  93             outfile = self._tmptarget+f
  94             mkdir_p(os.path.dirname(outfile))
  95             shutil.copyfile(f,outfile)
  96         print "Resources[%5.2f s]" % (round(t6-t5,2))
  97         # TODO: Improve the sitemap, it is a page that is generated from
  98         #       the ground up and added a bit adhoc.
  99         sitmaplink = Link('/sitemap')
 100         for l in self._sitelang:
 101             sitmaplink.add_page((l,'/sitemap.'+l+'.xml'))
 102         for l in self._sitelang:
 103             sitmaplink.page(l).set_article(self.gen_menu(l,None,"tree sitemap"))
 104             sitmaplink.page(l).template(self,self._style,self._tmptarget,self._subdir)
 105         t7 = time()
 106         print "Sitemap  [%5.2f s]" % (round(t7-t6,2))
 107
 108     def graph(self):
 109         self._tree.graph()
 110
 111     def gen_menu(self,lang,page,cssclass):
 112         return self._tree.menu(lang,page,cssclass,self._subdir)
 113
 114     def lang_menu(self,lang,link):
 115         html = ElementMaker()
 116         menu = html.ul()
 117         for l in link.languages():
 118             isoxml = u"//iso_639_3_entry[@*='"+l+"']"
 119             ln = self._isocode.xpath(isoxml)[0].get('name')
 120             if lang != 'en':
 121                 ln = self._tranlang[lang].gettext(ln)
 122             p = unicode(link.link())
 123             if p[-1] == u'/':
 124                 p = p +u'index'
 125             p = p+u'.'+l
 126             li = html.li(html.a(ln.decode('utf-8'),
 127                                 href=self._subdir+p,hreflang=l))
 128             menu.append(li)
 129         return etree.tostring(menu,encoding='UTF-8',pretty_print=False)
 130
 131     def publish(self):
 132         print "Size [ %7s ]" % (sizeof_fmt(get_folder_size(self._tmptarget)))
 133         ssh_cmd(self._output,"mkdir -p")
 134         publish(self._tmptarget, self._output)
 135         for res in ["stylesheets","images","js","fonts","favicon.ico"]:
 136             if (os.path.exists(self._style+res)):
 137                 publish(self._style+res, self._output)
 138         ssh_cmd(self._output,"chmod a+rx")