<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to patches</title><link href="https://sourceforge.net/p/spambayes/patches/" rel="alternate"/><link href="https://sourceforge.net/p/spambayes/patches/feed.atom" rel="self"/><id>https://sourceforge.net/p/spambayes/patches/</id><updated>2015-05-18T15:32:00.001000Z</updated><subtitle>Recent changes to patches</subtitle><entry><title>#96 Patch for bug when header_score_logarithm option is set</title><link href="https://sourceforge.net/p/spambayes/patches/96/?limit=25#baee" rel="alternate"/><published>2015-05-18T15:32:00.001000Z</published><updated>2015-05-18T15:32:00.001000Z</updated><author><name>Christian Hudon</name><uri>https://sourceforge.net/u/chrish42/</uri></author><id>https://sourceforge.net07362ca9817a081c1ddbc7c626cb8cf0c7f4ced5</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Well, I guess it didn't when I ran into that bug. That was a long time ago, now. This ticket can be closed, then.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>#96 Patch for bug when header_score_logarithm option is set</title><link href="https://sourceforge.net/p/spambayes/patches/96/?limit=25#9b9a" rel="alternate"/><published>2015-05-17T08:51:34.164000Z</published><updated>2015-05-17T08:51:34.164000Z</updated><author><name/><uri>https://sourceforge.net</uri></author><id>https://sourceforge.net7cc437d0b1e11d821d69b99bc4ab00bb438b3890</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;How is that possible, as the hammie.py module imports "math" as its first statement?&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Patch for bug when header_score_logarithm option is set</title><link href="https://sourceforge.net/p/spambayes/patches/96/" rel="alternate"/><published>2014-02-28T16:26:12.889000Z</published><updated>2014-02-28T16:26:12.889000Z</updated><author><name>Christian Hudon</name><uri>https://sourceforge.net/u/chrish42/</uri></author><id>https://sourceforge.net19cfccb21ab3ed9d4ece1cfef7a821ae71e27c83</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;There is a bug in the code when the header_score_logarithm option is set and the first message processed has a high value for the variable "prob". The "math" module is then undefined, because it is only imported in the "low-probability" branch.&lt;/p&gt;
&lt;p&gt;The following simple patch fixes this:&lt;/p&gt;
&lt;p&gt;--- hammie.py~  2008-12-06 18:33:39.000000000 -0500&lt;br /&gt;
+++ hammie.py   2012-10-28 00:38:36.000000000 -0400&lt;br /&gt;
@@ -116,8 +116,8 @@&lt;br /&gt;
         basic_disp = disp&lt;br /&gt;
         disp += "; %.*f" % (options&lt;span&gt;["Headers", "header_score_digits"]&lt;/span&gt;, prob)&lt;br /&gt;
         if options&lt;span&gt;["Headers", "header_score_logarithm"]&lt;/span&gt;:&lt;br /&gt;
+            import math&lt;br /&gt;
             if prob &amp;lt;= 0.005 and prob &amp;gt; 0.0:&lt;br /&gt;
-                import math&lt;br /&gt;
                 x = -math.log10(prob)&lt;br /&gt;
                 disp += " (%d)" % x&lt;br /&gt;
             if prob &amp;gt;= 0.995 and prob &amp;lt; 1.0:&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Bad regexp in sb_imapfilter.py causing DB corruption</title><link href="https://sourceforge.net/p/spambayes/patches/95/" rel="alternate"/><published>2011-04-29T07:06:39Z</published><updated>2011-04-29T07:06:39Z</updated><author><name>AJ Ostergaard</name><uri>https://sourceforge.net/u/ajostergaard/</uri></author><id>https://sourceforge.net65272c42b6a3067e7d2113ed08bbfbb76851981c</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Broken MessageId regexp caused useless IDs like '000' and '4', this in turn caused messages to be confused and resulted in lost messages and corrupted DB.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>tte.py patch for python backward compatibility</title><link href="https://sourceforge.net/p/spambayes/patches/94/" rel="alternate"/><published>2009-06-19T14:00:30Z</published><updated>2009-06-19T14:00:30Z</updated><author><name>Dave Abrahams</name><uri>https://sourceforge.net/u/userid-52572/</uri></author><id>https://sourceforge.net9fc23816e5b6982c1802a6ba3fded41a9ff88fac</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;reversed() is newer than Python 2.3, I think?&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Add Date header to msgs created by XMLRPCPlugin.form_to_mime</title><link href="https://sourceforge.net/p/spambayes/patches/93/" rel="alternate"/><published>2008-10-20T23:07:34Z</published><updated>2008-10-20T23:07:34Z</updated><author><name>Chris Arndt</name><uri>https://sourceforge.net/u/strogon14/</uri></author><id>https://sourceforge.net4327457f0e50e1e22b2c51f2de7618b1a04e73a3</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Here's a tiny patch that adds a Date hader with the current date/time to the MIME message created by the "form_to_mime" function of the XMLRPCPlugin module. This enables one to set the "display_headers" option in the "html_ui" section of the configuration and then see the date of submissions through the XMLRPC-interface in the "Review" page of the core server web interface for training messages.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Update german localization message files</title><link href="https://sourceforge.net/p/spambayes/patches/92/" rel="alternate"/><published>2008-04-02T14:29:26Z</published><updated>2008-04-02T14:29:26Z</updated><author><name>Marko von Oppen</name><uri>https://sourceforge.net/u/igs/</uri></author><id>https://sourceforge.net651a42d01ccaba235eeb0d6dd19b0002abda78d7</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Here are updated German message files.&lt;/p&gt;
&lt;p&gt;So far I can see the file messages.po is a binary file which was checked into CVS initial as a text file. This has changed now in SVN but this action destroyed the file.&lt;/p&gt;
&lt;p&gt;In the supplied ZIP file there are newer versions of German message files.&lt;/p&gt;
&lt;p&gt;Files where built against updated messages.pot.&lt;br /&gt;
Tested on German Vista x86 + Office 2007 + Spambayes SVN trunk version.&lt;/p&gt;
&lt;p&gt;Remark: The Poedit application was improved really in the meantime. It is really easy and comfortable now to localize the message files. Just open the messages.mo file, import the catalog file, make changes, save the file, restart Outlook, see the result!&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Make tte ratio well-behaved when balance is good</title><link href="https://sourceforge.net/p/spambayes/patches/91/" rel="alternate"/><published>2007-09-26T23:50:46Z</published><updated>2007-09-26T23:50:46Z</updated><author><name>Dave Abrahams</name><uri>https://sourceforge.net/u/userid-52572/</uri></author><id>https://sourceforge.netc14d4e3c4b3e255112928a53516624b0df12a0d0</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;Right now, if you pass a spam/ham ratio to tte and the training set is actually more balanced than indicated by the ratio, training will skip some messages in order to achieve that more-unbalanced ratio.  Nobody wants that.  The enclosed patch fixes it.&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Protection for dnscache corruption</title><link href="https://sourceforge.net/p/spambayes/patches/90/" rel="alternate"/><published>2007-09-26T23:43:47Z</published><updated>2007-09-26T23:43:47Z</updated><author><name>Dave Abrahams</name><uri>https://sourceforge.net/u/userid-52572/</uri></author><id>https://sourceforge.net76137b2c1a92cec6b6f1f16ae314903fd65c503f</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;My spam filtering regularly starts failing because the dnscache is corrupted.  I don't have any idea how it gets that way, but a broken cache should never scuttle the process, so I propose the enclosed patch.&lt;/p&gt;
&lt;p&gt;procmail: Executing "sb_filter.py"&lt;br /&gt;
Traceback (most recent call last):&lt;br /&gt;
File "/usr/local/bin/sb_filter.py", line 83, in ?&lt;br /&gt;
from spambayes import hammie, Options, mboxutils, storage&lt;br /&gt;
File "/usr/local/lib/python2.4/site-packages/spambayes/hammie.py", line 7, in ?&lt;br /&gt;
from spambayes.tokenizer import tokenize&lt;br /&gt;
File "/usr/local/lib/python2.4/site-packages/spambayes/tokenizer.py", line 44, in ?&lt;br /&gt;
cache = dnscache.cache(cachefile=options["Tokenizer", "lookup_ip_cache"])&lt;br /&gt;
File "/usr/local/lib/python2.4/site-packages/spambayes/dnscache.py", line 98, in __init__&lt;br /&gt;
self.caches = pickle.load(open(self.cachefile, "rb"))&lt;br /&gt;
cPickle.UnpicklingError: invalid load key, '&lt;br /&gt;
'.&lt;br /&gt;
procmail: Program failure (1) of "sb_filter.py"&lt;/p&gt;&lt;/div&gt;</summary></entry><entry><title>Fix for bsddb._db.DBAccessError: (13, 'Permission denied -- </title><link href="https://sourceforge.net/p/spambayes/patches/89/" rel="alternate"/><published>2007-09-26T23:36:58Z</published><updated>2007-09-26T23:36:58Z</updated><author><name>Dave Abrahams</name><uri>https://sourceforge.net/u/userid-52572/</uri></author><id>https://sourceforge.net74bc1fc23a7c52f030cb14942a31ff903f7d33c9</id><summary type="html">&lt;div class="markdown_content"&gt;&lt;p&gt;&lt;a href="http://article.gmane.org/gmane.mail.spam.spambayes.devel/3953" rel="nofollow"&gt;http://article.gmane.org/gmane.mail.spam.spambayes.devel/3953&lt;/a&gt; describes a problem and offers an analysis for which I am submitting a patch.  This is the most conservative change I could think of.&lt;/p&gt;&lt;/div&gt;</summary></entry></feed>