[PATCH 2/2] contrib/nmbug/nmbug-status: combine thread messages

[Tomi Ollila <tomi.ollila@iki.fi>]

Newer patch email containing In-Reply-To: to an email sent some time ago
(i.e. to a "thread") was not visible in that "thread" in patch view when
another patch "thread" was submitted in between. This change collects
all messages in every (notmuch-created) thread together before printing
all these threads out in a patch view.

Thanks to Ethan Glasser-Camp for initial review and suggestions with
code examples.
---
Notes:

Thread class was added for convenience at late time to replace tuple
containing last dict & lines list; It could be more utilized -- my excuse
of not doing so is to minimize change...

OrderedDict could be used in place of threadlist but that requires 
Python 2.7 (I used python 2.6 where argparse.py was copied to cwd;
also nmbug site uses python 2.6...). 

The table row separator '\n<tr><td colspan="2"><br /></td></tr>\n'
is a hack; If someone knows better simple semantic alternative that
could be used later...

 contrib/nmbug/nmbug-status |   73 +++++++++++++++++++++++++++----------------
 1 files changed, 46 insertions(+), 27 deletions(-)

diff --git a/contrib/nmbug/nmbug-status b/contrib/nmbug/nmbug-status
index c663409..69b407c 100755
--- a/contrib/nmbug/nmbug-status
+++ b/contrib/nmbug/nmbug-status
@@ -51,12 +51,19 @@ if args.text:
 else:
     output_format = 'html'
 
-headers = ['date', 'from', 'subject']
-last = {}
+class Thread:
+    def __init__(self, last, lines):
+        self.last = last
+        self.lines = lines
+
+    def join_utf8_with_newlines(self):
+        return '\n'.join( (line.encode('utf-8') for line in self.lines) )
 
-def clear_last():
-    for header in headers:
-        last[header] = ''
+def output_with_separator(threadlist, sep):
+    outputs = (thread.join_utf8_with_newlines() for thread in threadlist)
+    print sep.join(outputs)
+
+headers = ['date', 'from', 'subject']
 
 def print_view(title, query, comment):
 
@@ -64,7 +71,12 @@ def print_view(title, query, comment):
     q_new = notmuch.Query(db, query_string)
     q_new.set_sort(notmuch.Query.SORT.OLDEST_FIRST)
 
-    last['thread_id'] = ''
+    last_thread_id = ''
+    threads = {}
+    threadlist = []
+    out = {}
+    last = None
+    lines = None
 
     if output_format == 'html':
         print '<h3><a name="%s" />%s</h3>' % (title, title)
@@ -77,11 +89,21 @@ def print_view(title, query, comment):
 
     for m in q_new.search_messages():
 
-        out = {}
-
         thread_id = m.get_thread_id()
-        if thread_id != last['thread_id']:
-            clear_last()
+
+        if thread_id != last_thread_id:
+            if threads.has_key(thread_id):
+                last = threads[thread_id].last
+                lines = threads[thread_id].lines
+            else:
+                last = {}
+                lines = []
+                thread = Thread(last, lines)
+                threads[thread_id] = thread
+                for h in headers:
+                    last[h] = ''
+                threadlist.append(thread)
+            last_thread_id = thread_id
 
         for header in headers:
             val = m.get_header(header)
@@ -94,38 +116,35 @@ def print_view(title, query, comment):
                 if val == '':
                     val = addr.split('@')[0]
 
-            if last[header] == val:
+            if header != 'subject' and last[header] == val:
                 out[header] = ''
             else:
-                out[header] = val.encode('utf-8')
+                out[header] = val
                 last[header] = val
 
         mid = m.get_message_id()
         out['id'] = 'id:"%s"' % mid
 
         if output_format == 'html':
-            # XXX using <br /> is a hack, but ... // 20111216 too
-            if thread_id != last['thread_id']:
-                br = '<br />'
-            else:
-                br = ''
 
             out['subject'] = '<a href="http://mid.gmane.org/%s">%s</a>' \
                 % (urllib.quote(mid), out['subject'])
 
-            print ' <tr><td>%s %s' % (br, out['date'])
-            print '</td><td>%s %s' % (br, out['id'])
-            print '</td></tr>'
-            print ' <tr><td>%s' % out['from']
-            print '</td><td>%s' % out['subject']
-            print '</td></tr>\n'
+            lines.append(' <tr><td>%s' % out['date'])
+            lines.append('</td><td>%s' % out['id'])
+            lines.append('</td></tr>')
+            lines.append(' <tr><td>%s' % out['from'])
+            lines.append('</td><td>%s' % out['subject'])
+            lines.append('</td></tr>')
         else:
-            print '%(date)-10.10s %(from)-20.20s %(subject)-40.40s\n%(id)72s\n' % out
-
-        last['thread_id'] = thread_id
+            lines.append('%(date)-10.10s %(from)-20.20s %(subject)-40.40s\n%(id)72s' % out)
 
     if output_format == 'html':
+        output_with_separator(threadlist,
+                              '\n<tr><td colspan="2"><br /></td></tr>\n')
         print '</table>'
+    else:
+        output_with_separator(threadlist, '\n\n')
 
 # main program
 
-- 
1.7.1