From c147dc4b5b8cad935131b44cb78fd1181b036e56 Mon Sep 17 00:00:00 2001
From: Jeremy Stanley <fungi@yuggoth.org>
Date: Fri, 11 Sep 2020 00:39:10 +0000
Subject: [PATCH] Publish structured data listing our ML archives

In order to collect historical statistics on usage of our mailing
lists, we need an index of not only the current lists (which we
could get from Mailman) but also retired lists (which could only be
found by knowing the URL to their archives). Ultimately we should
publish hyperlinks to these so they'll continue to be indexed by
search engines, but for now start with structured YAML, which we
could later use to build that too.

Because the only way to determine the names of retired lists is from
the listserv's filesystem, we'll run a simple script once daily to
refresh the index and keep it in the Web root alongside the
robots.txt file. In the future, this could be triggered instead by
addition of new mailing lists, though while we're still managing
them with Puppet it's not clear how to go about doing that.

Of course restrict this to only indexing public list archives, as
privately-archived lists won't be accessible to the general public
by design.

Change-Id: Ibe3175a56831b7a43698d6fe454d70e93fcd0bc7
---
 .../files/mailman/mk-archives-index           |  9 +++++++++
 modules/openstack_project/manifests/lists.pp  | 20 +++++++++++++++++++
 2 files changed, 29 insertions(+)
 create mode 100755 modules/openstack_project/files/mailman/mk-archives-index

diff --git a/modules/openstack_project/files/mailman/mk-archives-index b/modules/openstack_project/files/mailman/mk-archives-index
new file mode 100755
index 0000000000..a814ad8a0c
--- /dev/null
+++ b/modules/openstack_project/files/mailman/mk-archives-index
@@ -0,0 +1,9 @@
+#!/bin/sh
+
+for site in `cut -d: -f1 /etc/mailman/sites | sort`; do
+    echo "$site:"
+    dir=`grep ^${site}: /etc/mailman/sites | cut -d' ' -f2`
+    for ml in `ls ${dir}/archives/public/ | sort`; do
+        echo "  - $ml"
+    done
+done
diff --git a/modules/openstack_project/manifests/lists.pp b/modules/openstack_project/manifests/lists.pp
index 2bf246600a..7a126741bd 100644
--- a/modules/openstack_project/manifests/lists.pp
+++ b/modules/openstack_project/manifests/lists.pp
@@ -17,6 +17,26 @@ class openstack_project::lists(
     mode    => '0444',
   }
 
+  file { '/usr/local/sbin/mk-archives-index':
+    ensure  => present,
+    source  => 'puppet:///modules/openstack_project/mailman/mk-archives-index',
+    owner   => 'root',
+    group   => 'root',
+    replace => true,
+    mode    => '0744',
+  }
+
+  cron { 'mk-archives-index':
+    user        => root,
+    hour        => '0',
+    minute      => '0',
+    command     => 'mk-archives-index > /var/www/archives.yaml',
+    environment => 'PATH=/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin',
+    require     => File['/usr/local/sbin/mk-archives-index',
+    ],
+  }
+
+
   # Disable inactive admins
   user::virtual::disable { 'oubiwann': }
   user::virtual::disable { 'rockstar': }