Allow filters to "tee"

docs/filters.html

@@ -77,8 +77,18 @@ python. <code>.xslt</code> involkes XSLT.  <code>.sed</code> and
 perl or ruby or class/jar (java), aren't supported at the moment, but these
 would be easy to add.</li>
 
-<li>Templates written using htmltmpl currently only have access to a fixed set
+<li>If the filter name contains a redirection character (<code>&gt;</code>),
-of fields, whereas XSLT templates have access to everything.</li>
+then the output stream is
+<a href="http://en.wikipedia.org/wiki/Tee_(Unix)">tee</a>d; one branch flows
+through the specified filter and the output is planced into the named file; the
+other unmodified branch continues onto the next filter, if any.
+One use case for this function is to use
+<a href="../filters/xhtml2html.py">xhtml2html</a> to produce both an XHTML and
+an HTML output stream from one source.</li>
+
+<li>Templates written using htmltmpl or django currently only have access to a
+fixed set of fields, whereas XSLT and genshi templates have access to
+everything.</li>
 </ul>
 </body>
 </html>

planet/splice.py

@@ -122,10 +122,22 @@ def apply(doc):
             output = open(output_file).read()
             for filter in config.filters(template_file):
                 if filter in planet_filters: continue
-                output = shell.run(filter, output, mode="filter")
+                if filter.find('>')>0:
-                if not output:
+                    # tee'd output
-                    os.unlink(output_file)
+                    filter,dest = filter.split('>',1)
-                    break
+                    tee = shell.run(filter.strip(), output, mode="filter")
+                    if tee:
+                        output_dir = planet.config.output_dir()
+                        dest_file = os.path.join(output_dir, dest.strip())
+                        dest_file = open(dest_file,'w')
+                        dest_file.write(tee)
+                        dest_file.close()
+                else:
+                    # pipe'd output
+                    output = shell.run(filter, output, mode="filter")
+                    if not output:
+                        os.unlink(output_file)
+                        break
             else:
                 handle = open(output_file,'w')
                 handle.write(output)

tests/data/apply/config-html.ini

@@ -0,0 +1,25 @@
+[Planet]
+output_theme = genshi_fancy
+output_dir = tests/work/apply
+name = test planet
+cache_directory = tests/work/spider/cache
+
+bill_of_materials:
+  images/#{face}
+
+[index.html.genshi]
+filters:
+   xhtml2html.py>index.html4
+
+[tests/data/spider/testfeed0.atom]
+name = not found
+
+[tests/data/spider/testfeed1b.atom]
+name = one
+face = jdub.png
+
+[tests/data/spider/testfeed2.atom]
+name = two
+
+[tests/data/spider/testfeed3.rss]
+name = three

tests/test_apply.py

@@ -55,6 +55,16 @@ class ApplyTest(unittest.TestCase):
         config.load(configfile % 'genshi')
         self.apply_fancy()
 
+    def test_apply_filter_html(self):
+        config.load(configfile % 'html')
+        self.apply_fancy()
+
+        output = open(os.path.join(workdir, 'index.html')).read()
+        self.assertTrue(output.find('/>')>=0)
+
+        output = open(os.path.join(workdir, 'index.html4')).read()
+        self.assertTrue(output.find('/>')<0)
+
     def apply_fancy(self):
         splice.apply(self.feeddata)
 
@@ -106,3 +116,4 @@ for method in dir(test_filter_genshi.GenshiFilterTests):
     if method.startswith('test_'): break
 else:
     delattr(ApplyTest,'test_apply_genshi_fancy')
+    delattr(ApplyTest,'test_apply_filter_html')