From 7cc47ab92bb9a38d31821ab1847442396af19287 Mon Sep 17 00:00:00 2001
From: Laurent Bachelier <laurent@bachelier.name>
Date: Wed, 3 Oct 2012 21:54:57 +0200
Subject: [PATCH] Add support for external pre-filtering

With manpage documentation including example script.
Also add the full e-mail to the available metadata (only git for now).

Signed-off-by: Eric S. Raymond <esr@thyrsus.com>
---
 irkerhook.py  | 26 +++++++++++++++++++++-----
 irkerhook.xml | 27 +++++++++++++++++++++++++++
 2 files changed, 48 insertions(+), 5 deletions(-)

diff --git a/irkerhook.py b/irkerhook.py
index ee9b671..4619310 100755
--- a/irkerhook.py
+++ b/irkerhook.py
@@ -51,6 +51,7 @@ class Commit:
         self.commit = commit
         self.branch = None
         self.rev = None
+        self.mail = None
         self.author = None
         self.files = None
         self.logmsg = None
@@ -93,6 +94,7 @@ class GenericExtractor:
         self.template = None
         self.urlprefix = None
         self.host = socket.getfqdn()
+        self.filtercmd = None
         # Color highlighting is disabled by default.
         self.color = None
         self.bold = self.green = self.blue = ""
@@ -190,6 +192,7 @@ class GitExtractor(GenericExtractor):
         self.template = '%(bold)s%(project)s:%(reset)s %(green)s%(author)s%(reset)s %(repo)s:%(yellow)s%(branch)s%(reset)s * %(bold)s%(rev)s%(reset)s / %(bold)s%(files)s%(reset)s: %(logmsg)s %(brown)s%(url)s%(reset)s'
         self.color = do("git config --get irker.color")
         self.urlprefix = do("git config --get irker.urlprefix") or "gitweb"
+        self.filtercmd = do("git config --get irker.filtercmd")
         # These are git-specific
         self.refname = do("git symbolic-ref HEAD 2>/dev/null")
         self.revformat = do("git config --get irker.revformat")
@@ -235,11 +238,12 @@ class GitExtractor(GenericExtractor):
         # other VCSes a different choice may be appropriate.
         metainfo = do("git log -1 '--pretty=format:%an <%ae>|%s' " + shellquote(commit.commit))
         (commit.author, commit.logmsg) = metainfo.split("|")
+        commit.mail = commit.author.split()[-1].strip("<>")
         # This discards the part of the author's address after @.
         # Might be be nice to ship the full email address, if not
         # for spammers' address harvesters - getting this wrong
         # would make the freenode #commits channel into harvester heaven.
-        commit.author = commit.author.replace("<", "").split("@")[0].split()[-1]
+        commit.author = commit.mail.split("@")[0]
         return commit
 
 class SvnExtractor(GenericExtractor):
@@ -359,6 +363,12 @@ extractors = [GitExtractor, HgExtractor, SvnExtractor]
 def ship(extractor, commit, debug):
     "Ship a notification for the specified commit."
     metadata = extractor.commit_factory(commit)
+
+    # Run through an external filter if required.
+    channels = extractor.channels.split(",")
+    if extractor.filtercmd:
+        channels = filterc(extractor.filtercmd, channels, metadata) or []
+
     # Message reduction.  The assumption here is that IRC can't handle
     # lines more than 510 characters long. If we exceed that length, we
     # try knocking out the file list, on the theory that for notification
@@ -371,15 +381,14 @@ def ship(extractor, commit, debug):
         privmsg = str(metadata)
 
     # Anti-spamming guard.
-    channel_list = extractor.channels.split(",")
     if extractor.maxchannels != 0:
-        channel_list = channel_list[:extractor.maxchannels]
+        channels = channels[:extractor.maxchannels]
 
     # Ready to ship.
-    message = json.dumps({"to":channel_list, "privmsg":privmsg})
+    message = json.dumps({"to": channels, "privmsg": privmsg})
     if debug:
         print message
-    else:
+    elif channels:
         try:
             if extractor.tcp:
                 try:
@@ -397,6 +406,13 @@ def ship(extractor, commit, debug):
         except socket.error, e:
             sys.stderr.write("%s\n" % e)
 
+def filterc(cmd, channels, commit):
+    channels, update = json.loads(do('%s %s %s' % (cmd,
+        shellquote(json.dumps(channels)),
+        shellquote(json.dumps(commit.__dict__)))))
+    commit.__dict__.update(update)
+    return channels
+
 if __name__ == "__main__":
     notify = True
     repository = os.getcwd()
diff --git a/irkerhook.xml b/irkerhook.xml
index e10d5b8..a44c4f2 100644
--- a/irkerhook.xml
+++ b/irkerhook.xml
@@ -296,6 +296,33 @@ exists.</para>
 
 </refsect2>
 
+<refsect2 id="filter"><title>Filtering</title>
+
+<para>It is possible to filter commits before sending them to <application>irkerd</application>.</para>
+
+<para>You have to specify the <option>filtercmd</option> option, which will be the command <application>irkerhook.py</application> will run. This command should accept two arguments: the first one being the commit metadata, the second one being the channel list. The command should then return a tuple, with the new channel list then an optional list of commit metadata updates.</para>
+
+<para>Below is an example filter:</para>
+
+<programlisting>
+#!/usr/bin/env python
+import sys, json
+channels = json.loads(sys.argv[1])
+commit = json.loads(sys.argv[2])
+update = {}
+# we have two bobs in the project
+if commit.get("mail", "") == "bob@foo":
+    update["author"] = "bobm"
+if commit.get("mail", "") == "bob@bar":
+    update["author"] = "bobp"
+# ignore some commits
+if 'automated commit' in commit['logmsg']:
+    channels = []
+print json.dumps([channels, update])
+</programlisting>
+
+</refsect2>
+
 </refsect1>
 
 <refsect1 id='options'><title>OPTIONS</title>
-- 
2.26.2