Source Code for Module apesmit

  1  #!/usr/bin/env python 
  2  #-*- coding: utf-8 -*- 
  3   
  4  ##     ApeSmit - A simple Python module to create XML sitemaps 
  5  ##                     <http://www.florian-diesch.de/software/apesmit/> 
  6  ##     Copyright (C) 2008  Florian Diesch <devel@florian-diesch.de> 
  7   
  8  ##     This program is free software; you can redistribute it and/or modify 
  9  ##     it under the terms of the GNU General Public License as published by 
 10  ##     the Free Software Foundation; either version 2 of the License, or 
 11  ##     (at your option) any later version. 
 12   
 13  ##     This program is distributed in the hope that it will be useful, 
 14  ##     but WITHOUT ANY WARRANTY; without even the implied warranty of 
 15  ##     MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the 
 16  ##     GNU General Public License for more details. 
 17   
 18  ##     You should have received a copy of the GNU General Public License along 
 19  ##     with this program; if not, write to the Free Software Foundation, Inc., 
 20  ##     51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. 
 21   
 22   
 23  import datetime, codecs 
 24   
 25  FREQ=set((None, 'always', 'hourly', 'daily', 'weekly', 'monthly', 
 26            'yearly', 'never'))  #: values for changefreq 
 27   
 28        
 29 -class Url(object): 
 30      """ 
 31      Class to handle a URL in `Sitemap` 
 32      """ 
 33 -    def __init__(self, loc, lastmod, changefreq, priority, escape=True): 
 34          """ 
 35          Constructor 
 36   
 37          :Parameters: 
 38            loc : string 
 39              Location (URL). See http://www.sitemaps.org/protocol.php#locdef 
 40            lastmod : ``datetime.date`` or ``string`` 
 41              Date of last modification. 
 42              See http://www.sitemaps.org/protocol.php#lastmoddef 
 43              The ``today`` is replaced by today's date 
 44            changefreq : One of the values in `FREQ` 
 45              Expected frequency for changes. 
 46              See http://www.sitemaps.org/protocol.php#changefreqdef 
 47            priority : ``float`` or ``string`` 
 48              Priority of this URL relative to other URLs on your site. 
 49              See http://www.sitemaps.org/protocol.php#prioritydef 
 50            escape 
 51              True if escaping for XML special characters should be done. 
 52              See http://www.sitemaps.org/protocol.php#escaping         
 53          """ 
 54          if escape:             
 55              self.loc=self.escape(loc) 
 56          else: 
 57              self.loc=loc 
 58          if lastmod=='today': 
 59              lastmod=datetime.date.today().isoformat() 
 60          if lastmod is not None: 
 61              self.lastmod=unicode(lastmod) 
 62          else: 
 63              self.lastmod=None 
 64          if changefreq not in FREQ: 
 65              raise ValueError("Invalid changefreq value: '%s'"%changefreq) 
 66          if changefreq is not None: 
 67              self.changefreq=unicode(changefreq) 
 68          else: 
 69              self.changefreq=None 
 70          if priority is not None: 
 71              self.priority=unicode(priority) 
 72          else: 
 73              self.priority=None 
 74          self.urls=[] 
 75   
 76 -    def escape(self, s): 
 77          """ 
 78          Escaping XML special chracters 
 79   
 80          :Parameters: 
 81            s 
 82              String to escape 
 83          :return: Escaped string 
 84          """ 
 85          s=s.replace('&', '&amp;') 
 86          s=s.replace("'", '&apos;') 
 87          s=s.replace('"', '&quod;') 
 88          s=s.replace('>', '&gt;') 
 89          s=s.replace('<', '&lt;') 
 90          return s 
 91       
 92 -class Sitemap(object): 
 93      """ 
 94      Class to manage a sitemap 
 95      """ 
 96 -    def __init__(self, lastmod=None, changefreq=None, priority=None): 
 97          """ 
 98          Constructor 
 99   
100          :Parameters: 
101            lastmod 
102               Default value for `lastmod`. See `Url.__init__()`. 
103            changefreq 
104               Default value for `changefreq`. See `Url.__init__()`. 
105            priority 
106               Default value for `priority`. See `Url.__init__()`. 
107          """ 
108           
109          self.lastmod=lastmod 
110          self.changefreq=changefreq 
111          self.priority=priority 
112          self.urls=[] 
113   
114   
115 -    def add(self, loc, lastmod=None, changefreq=None, priority=None, escape=True): 
116          """ 
117          Add a new URl. Parameters are the same as in  `Url.__init__()`. 
118          If ``lastmod``, ``changefreq`` or ``priority`` is ``None`` the default 
119          value is used (see `__init__()`) 
120          """ 
121           
122          if lastmod is None: 
123              lastmod=self.lastmod 
124          if changefreq is None: 
125              changefreq=self.changefreq 
126          if priority is None: 
127              priority=self.priority 
128          self.urls.append(Url(loc, lastmod, changefreq, priority, escape)) 
129   
130               
131 -    def write(self, out): 
132          """ 
133          Write sitemap to ``out`` 
134   
135          :Parameters: 
136             out 
137               file name or anything with a ``write()`` method   
138          """ 
139   
140          if isinstance(out, basestring): 
141              try: 
142                  output=codecs.open(out, 'w', 'utf-8') 
143              except Exception, e: 
144                  print "Can't open file '%s': %s"%(path, str(e)) 
145                  return 
146          else: 
147              output=out 
148          output.write("<?xml version='1.0' encoding='UTF-8'?>\n" 
149          '<urlset xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"\n' 
150          '        xsi:schemaLocation="http://www.sitemaps.org/schemas/sitemap/0.9\n' 
151          '        http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd"\n' 
152          '        xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">\n') 
153            
154   
155          for url in self.urls: 
156              lastmod=changefreq=priority='' 
157              if url.lastmod is not None: 
158                  lastmod='  <lastmod>%s</lastmod>\n'%url.lastmod 
159              if url.changefreq is not None: 
160                  changefreq='  <changefreq>%s</changefreq>\n'%url.changefreq 
161              if url.priority is not None: 
162                  priority='  <priority>%s</priority>\n'%url.priority 
163              output.write(" <url>\n" 
164                           "  <loc>%s</loc>\n%s%s%s" 
165                           " </url>\n"%(url.loc.decode('utf-8'), 
166                                        lastmod.decode('utf-8'), 
167                                        changefreq.decode('utf-8'), 
168                                        priority.decode('utf-8'))) 
169          output.write('</urlset>\n') 
170          if output is not out: 
171              output.close() 
172