From 2d601445941b51ad14d6dc851ead1c0fbde3a8bf Mon Sep 17 00:00:00 2001
From: francis <francis>
Date: Fri, 7 Mar 2008 10:13:57 +0000
Subject: Index word docs and PDFs

---
 app/models/incoming_message.rb | 32 ++++++++++++++++++++++++++++++--
 todo.txt                       |  7 +++++++
 2 files changed, 37 insertions(+), 2 deletions(-)

diff --git a/app/models/incoming_message.rb b/app/models/incoming_message.rb
index 49b24b694..a22385347 100644
--- a/app/models/incoming_message.rb
+++ b/app/models/incoming_message.rb
@@ -18,7 +18,7 @@
 # Copyright (c) 2007 UK Citizens Online Democracy. All rights reserved.
 # Email: francis@mysociety.org; WWW: http://www.mysociety.org/
 #
-# $Id: incoming_message.rb,v 1.54 2008-03-06 23:17:28 francis Exp $
+# $Id: incoming_message.rb,v 1.55 2008-03-07 10:13:57 francis Exp $
 
 
 # TODO
@@ -314,9 +314,37 @@ class IncomingMessage < ActiveRecord::Base
         text = IncomingMessage.remove_quoted_sections(text, "")
     end
 
+    # Returns text version of attachment text
+    def get_attachment_text
+        text = ''
+        attachments = self.get_attachments_for_display
+        for attachment in attachments
+            if attachment.content_type == 'text/plain'
+                text += attachment.body
+            elsif attachment.content_type == 'application/msword'
+                tempfile = Tempfile.new('foipdf')
+                tempfile.print attachment.body
+                tempfile.flush
+                system("/usr/bin/wvText " + tempfile.path + " " + tempfile.path + ".txt")
+                text += File.read(tempfile.path + ".txt")
+                File.unlink(tempfile.path + ".txt")
+                tempfile.close
+            elsif attachment.content_type == 'application/pdf'
+                tempfile = Tempfile.new('foipdf')
+                tempfile.print attachment.body
+                tempfile.flush
+                IO.popen("/usr/bin/pdftotext " + tempfile.path + " -", "r") do |child|
+                    text += child.read()
+                end
+                tempfile.close
+            end
+        end
+        return text
+    end
+
     # Returns text for indexing
     def get_text_for_indexing
-        return get_body_for_quoting()
+        return get_body_for_quoting + get_attachment_text
     end
 
     # Returns the name of the person the incoming message is from, or nil if there isn't one
diff --git a/todo.txt b/todo.txt
index acc31b894..3bb7ea53a 100644
--- a/todo.txt
+++ b/todo.txt
@@ -1,8 +1,11 @@
 Search:
 
+Don't show same request so many times
 Add indexing of PDFs and DOCs etc.
 Date ranges and types and stuff
 
+Search for users
+Search for public bodies
 
 FOI requests to use to test it
 ==============================
@@ -57,6 +60,10 @@ eived from server "/data/vhost/foi.mysociety.org/docs/dispatch.fcgi"
 
 "Government" in about page
 
+Send email to remind people to classify
+Send email to tell admins something isn't classified
+Send email to remind people to clarify
+
 Later
 =====
 
-- 
cgit v1.2.3