Initial filter support (no parameters)

filters/coral_cdn_filter.py

@@ -0,0 +1,17 @@
+"""
+Remap all images to take advantage of the Coral Content Distribution
+Network <http://www.coralcdn.org/>.
+"""
+
+import sys, urlparse, xml.dom.minidom
+
+entry = xml.dom.minidom.parse(sys.stdin).documentElement
+
+for node in entry.getElementsByTagName('img'):
+    if node.hasAttribute('src'):
+        component = list(urlparse.urlparse(node.getAttribute('src')))
+        if component[0]=='http' and component[1].find(':')<0:
+            component[1] += '.nyud.net:8080'
+            node.setAttribute('src', urlparse.urlunparse(component))
+
+print entry.toxml('utf-8')

planet/config.py

@@ -101,6 +101,8 @@ def __init__():
     define_planet_list('template_files')
     define_planet_list('bill_of_materials')
     define_planet_list('template_directories')
+    define_planet_list('filters')
+    define_planet_list('filter_directories')
     define_planet_list('reading_lists')
 
     # template options
@@ -151,6 +153,12 @@ def load(config_file):
         else:
             log.error('Unable to find theme %s', theme)
 
+    # Filter support
+    dirs = config.filter_directories()
+    filter_dir = os.path.join(sys.path[0],'filters')
+    if filter_dir not in dirs and os.path.exists(filter_dir):
+        parser.set('Planet', 'filter_directories', ' '.join(dirs+[filter_dir]))
+
     # Reading list support
     reading_lists = config.reading_lists()
     if reading_lists:
@@ -209,8 +217,8 @@ def feedtype():
 
 def subscriptions():
     """ list the feed subscriptions """
-    return filter(lambda feed: feed!='Planet' and feed not in template_files(),
-       parser.sections())
+    return filter(lambda feed: feed!='Planet' and 
+        feed not in template_files()+filters(), parser.sections())
 
 def planet_options():
     """ dictionary of planet wide options"""

planet/shell/__init__.py

@@ -2,16 +2,21 @@ import planet
 import os
 import sys
 
-def run(template_file, doc):
+def run(template_file, doc, mode='template'):
     """ select a template module based on file extension and execute it """
     log = planet.getLogger(planet.config.log_level())
 
+    if mode == 'template':
+        dirs = planet.config.template_directories()
+    else:
+        dirs = planet.config.filter_directories()
+ 
     # see if the template can be located
-    for template_dir in planet.config.template_directories():
+    for template_dir in dirs:
         template_resolved = os.path.join(template_dir, template_file)
         if os.path.exists(template_resolved): break
     else:
-        return log.error("Unable to locate template %s", template_file)
+        return log.error("Unable to locate %s %s", mode, template_file)
 
     # Add shell directory to the path, if not already there
     shellpath = os.path.join(sys.path[0],'planet','shell')
@@ -20,16 +25,22 @@ def run(template_file, doc):
 
     # Try loading module for processing this template, based on the extension
     base,ext = os.path.splitext(os.path.basename(template_resolved))
-    template_module_name = ext[1:]
+    module_name = ext[1:]
     try:
-        template_module = __import__(template_module_name)
+        module = __import__(module_name)
     except Exception, inst:
-        return log.error("Skipping template '%s' after failing to load '%s':" +
-            " %s", template_resolved, template_module_name, inst)
+        print module_name
+        return log.error("Skipping %s '%s' after failing to load '%s': %s", 
+            mode, template_resolved, module_name, inst)
 
     # Execute the shell module
-    log.info("Processing template %s using %s", template_resolved,
-        template_module_name)
-    output_dir = planet.config.output_dir()
-    output_file = os.path.join(output_dir, base)
-    template_module.run(template_resolved, doc, output_file)
+    if mode == 'filter':
+        log.debug("Processing filer %s using %s", template_resolved,
+            module_name)
+        return module.run(template_resolved, doc, None)
+    else:
+        log.info("Processing template %s using %s", template_resolved,
+            module_name)
+        output_dir = planet.config.output_dir()
+        output_file = os.path.join(output_dir, base)
+        module.run(template_resolved, doc, output_file)

planet/shell/py.py

@@ -0,0 +1,16 @@
+from subprocess import Popen, PIPE
+
+def run(script, doc, output_file=None):
+    """ process an Python script """
+
+    if output_file:
+        out = open(output_file, 'w')
+    else:
+        out = PIPE
+
+    proc = Popen(['python', script], stdin=PIPE, stdout=out, stderr=PIPE)
+    stdout, stderr = proc.communicate(doc)
+    if stderr:
+        print stderr
+
+    return stdout

planet/spider.py

@@ -7,7 +7,7 @@ and write each as a set of entries in a cache directory.
 import time, calendar, re, os
 from xml.dom import minidom
 # Planet modules
-import planet, config, feedparser, reconstitute
+import planet, config, feedparser, reconstitute, shell
 
 # Regular expressions to sanitise cache filenames
 re_url_scheme    = re.compile(r'^\w+:/*(\w+:|www\.)?')
@@ -39,9 +39,8 @@ def filename(directory, filename):
 def write(xdoc, out):
     """ write the document out to disk """
     file = open(out,'w')
-    file.write(xdoc.toxml('utf-8'))
+    file.write(xdoc)
     file.close()
-    xdoc.unlink()
 
 def spiderFeed(feed):
     """ Spider (fetch) a single feed """
@@ -116,30 +115,43 @@ def spiderFeed(feed):
     xdoc=minidom.parseString('''<feed xmlns:planet="%s"
       xmlns="http://www.w3.org/2005/Atom"/>\n''' % planet.xmlns)
     reconstitute.source(xdoc.documentElement, data.feed, data.bozo)
-    write(xdoc, filename(sources, feed))
+    write(xdoc.toxml('utf-8'), filename(sources, feed))
+    xdoc.unlink()
 
     # write each entry to the cache
     cache = config.cache_directory()
     for entry in data.entries:
+
+        # generate an id, if none is present
         if not entry.has_key('id') or not entry.id:
             entry['id'] = reconstitute.id(None, entry)
             if not entry['id']: continue
 
-        out = filename(cache, entry.id)
+        # compute cache file name based on the id
+        cache_file = filename(cache, entry.id)
 
+        # get updated-date either from the entry or the cache (default to now)
         mtime = None
         if entry.has_key('updated_parsed'):
             mtime = calendar.timegm(entry.updated_parsed)
             if mtime > time.time(): mtime = None
         if not mtime:
             try:
-                mtime = os.stat(out).st_mtime
+                mtime = os.stat(cache_file).st_mtime
             except:
                 mtime = time.time()
             entry['updated_parsed'] = time.gmtime(mtime)
 
-        write(reconstitute.reconstitute(data, entry), out) 
-        os.utime(out, (mtime, mtime))
+        # apply any filters
+        xdoc = reconstitute.reconstitute(data, entry)
+        output = xdoc.toxml('utf-8')
+        xdoc.unlink()
+        for filter in config.filters():
+            output = shell.run(filter, output, mode="filter")
+
+        # write out and timestamp the results
+        write(output, cache_file) 
+        os.utime(cache_file, (mtime, mtime))
 
 def spiderPlanet(configFile):
     """ Spider (fetch) an entire planet """

tests/data/filter/coral_cdn.xml

@@ -0,0 +1,7 @@
+<entry xmlns="http://www.w3.org/2005/xhtml">
+  <content>
+     <div xmlns="http://www.w3.org/1999/xhtml">
+       <img src="http://example.com/foo.png"/>
+     </div>
+  </content>
+</entry>

tests/test_filters.py

@@ -0,0 +1,15 @@
+#!/usr/bin/env python
+
+import unittest, xml.dom.minidom
+from planet import shell
+
+testfile = 'tests/data/filter/coral_cdn.xml'
+filter = 'coral_cdn_filter.py'
+
+class FilterTests(unittest.TestCase):
+
+    def test_coral_cdn(self):
+        output = shell.run(filter, open(testfile).read(), mode="filter")
+        dom = xml.dom.minidom.parseString(output)
+        imgsrc = dom.getElementsByTagName('img')[0].getAttribute('src')
+        self.assertEqual('http://example.com.nyud.net:8080/foo.png', imgsrc)