xml: change from amara to lxml

[treecutter.git] / treecutter / sitemap.py

diff --git a/treecutter/sitemap.py b/treecutter/sitemap.py
index e7a2e4b3b99a0c8e7cee21f8e2c11f53073f4601..a8612f23b9ba9b1c1955bd370af771b59a008b9f 100644 (file)
--- a/treecutter/sitemap.py
+++ b/treecutter/sitemap.py
@@ -4,8 +4,10 @@ import re
 import shutil
 import gettext
 import tempfile
-from amara import bindery
+from lxml import etree
+from lxml.builder import ElementMaker
 from time import time
+from treecutter import constants as const
 from treecutter.trie import Trie
 from treecutter.link import Link
 from treecutter.tools import ssh_cmd, publish, mkdir_p
@@ -16,10 +18,14 @@ class Sitemap():
         self._file = 'sitemap.txt'
         self._tree = Trie()
         self._sitelang = set()
-        self._isocode = bindery.parse('/usr/share/xml/iso-codes/iso_639_3.xml')
+        self._isocode = etree.parse('/usr/share/xml/iso-codes/iso_639_3.xml')
         self._tranlang = {}
         self._tmptarget = tempfile.mkdtemp()+'/'
 
+    # The sitemap uses a trie structure to keep track of links
+    # A link represents the path to the document and the link
+    # representing the text on the site.
+    # A link can have several pages in different languages.
     def add_link(self, link):
         tokens = filter(None,re.split(r'(^/[\w-]*/|[\w-]*/)',link))
         self._tree.add(tokens,Link(link))
@@ -39,9 +45,13 @@ class Sitemap():
         except IOError, what_error:
             print 'INFO: Could not read sitemap.txt - one will be created'
 
+    # Create a set of the current tree for comparison with the
+    # directory scan
     def set(self):
         return set(link.link() for link in self._tree)
 
+    # Main driver in the application processing the documents
+    # in the collected sitemap
     def process(self, style):
         t1 = time()
         print "Prepareing the input"
@@ -67,7 +77,7 @@ class Sitemap():
         print "Template [%5.2f s]" % (round(t5-t4,2))
         t6 = time()
         res = set()
-        cwd = os.getcwd()
+        # Collect all files used by the documents
         for link in self._tree:
             res = res.union(link.resources())
         for f in res:
@@ -75,6 +85,8 @@ class Sitemap():
             mkdir_p(os.path.dirname(outfile))
             shutil.copyfile(f,outfile)
         print "Resources[%5.2f s]" % (round(t6-t5,2))
+        # TODO: Improve the sitemap, it is a page that is generated from
+        #       the ground up and added a bit adhoc.
         sitmaplink = Link('/sitemap')
         for l in self._sitelang:
             sitmaplink.add_page((l,'/sitemap.'+l+'.xml'))
@@ -91,19 +103,20 @@ class Sitemap():
         return self._tree.menu(lang,page,cssclass)
 
     def lang_menu(self,lang,link):
-        html = "<ul>"
+        html = ElementMaker(namespace=const.HTML_NS)
+        menu = html.ul()
         for l in link.languages():
             isoxml = u"//iso_639_3_entry[@*='"+l+"']"
-            ln = self._isocode.xml_select(isoxml)[0].name
+            ln = self._isocode.xpath(isoxml)[0].get('name')
             if lang != 'en':
                 ln = self._tranlang[lang].gettext(ln)
             p = link.link()
             if p[-1] == '/':
                 p = p +'index'
             p = p+'.'+l
-            html += '<li><a href="%s" hreflang="%s">%s</a></li>' % (p, l, ln)
-        html += "</ul>"
-        return html
+            li = html.li(html.a(ln,href=p,hreflang=l))
+            menu.append(li)
+        return etree.tostring(menu,encoding='UTF-8',pretty_print=False)
 
     def publish(self,output,style):
         ssh_cmd(output,"mkdir -p")