use mapping char filter to split lucene-nnnn, etc., as a single token

Author: mikemccand

examples/jirasearch/handle.py

@@ -851,14 +851,21 @@ def toAgo(seconds):
   else:
     return addUnit('%.1f' % (seconds/24/3600/365), 'year')
 
-def renderJiraHits(w, groups, userDrillDowns):
+def renderJiraHits(w, text, groups, userDrillDowns):
 
   now = int(time.time() + getUTCOffset())
 
+  lowerText = text.strip().lower().replace('-', ' ')
+
   # Each group is a Jira issue; each (child) doc is a comment on that issue
   for group in groups:
     fields = group['fields']
     key = fields['key'].upper()
+
+    # hack alert!  really i should index a separate text+highlight field for this...
+    if key.lower().replace('-', ' ') == lowerText:
+      key = '<b>%s</b>' % key
+    
     if fields['status'] in ('Open', 'Reopened', 'In Progress'):
       skey = key
     else:
@@ -1931,9 +1938,9 @@ def handleQuery(path, isMike, environ):
     if groupBy is not None:
       for group in result['groups']:
         renderGroupHeader(w, group, groupDDField, hitsPerGroup)
-        renderJiraHits(w, group['hits'], userDrillDowns)
+        renderJiraHits(w, text, group['hits'], userDrillDowns)
     else:
-      renderJiraHits(w, result['groups'], userDrillDowns)
+      renderJiraHits(w, text, result['groups'], userDrillDowns)
 
     #w('</div>')
     w('</table>')

examples/jirasearch/indexJira.py

@@ -97,9 +97,12 @@ def createSchema(svr):
   svr.send('startIndex', {'indexName': 'jira'})
 
   analyzer = {
+    'charFilters': [{'class': 'Mapping',
+                     'mappingFileContents': '"LUCENE-" => "LUCENE"\n"lucene-" => "lucene"\n"SOLR-" => "SOLR"\n"solr-" => "solr"\n"INFRA-" => "INFRA"\n"infra-" => "infra"'}],
     'tokenizer': ICU_TOKENIZER_KEEP_ISSUES,
     'tokenFilters': ['EnglishPossessive',
                      {'class': 'WordDelimiter',
+                      'splitOnNumerics': 0,
                       'preserveOriginal': 1},
                      'LowerCase',
                      {'class': 'Synonym',
@@ -133,7 +136,7 @@ def createSchema(svr):
                      'EnglishMinimalStem']}
 
   #print('TOKENS')
-  #prettyPrintJSON(svr.send('analyze', {'analyzer': analyzer, 'indexName': 'jira', 'text': 'AnalyzingInfixSuggester'}))
+  #prettyPrintJSON(svr.send('analyze', {'analyzer': analyzer, 'indexName': 'jira', 'text': 'foo LUCENE-444 lucene-1123 AnalyzingInfixSuggester'}))
 
   fields = {'key': {'type': 'atom',
                     'sort': True,

examples/jirasearch/server.py

@@ -199,7 +199,7 @@ def main():
     else:
       port = int(sys.argv[idx+1])
 
-    httpd = wsgiref.simple_server.make_server('10.17.4.92', port, application)
+    httpd = wsgiref.simple_server.make_server('10.17.4.93', port, application)
 
     print('Ready on port %s' % port)
     httpd.serve_forever()

examples/jirasearch/static/about.html

@@ -23,6 +23,6 @@
 <br>
 &nbsp;&nbsp;<font size=+1>This site uses <a href="http://lucene.apache.org">Apache Lucene</a>, specifically <a href="https://github.com/mikemccand/luceneserver">this thin-wrapper HTTP server over Lucene</a>, to search all Lucene, Solr and Tika Jira issues.  See
 the <a href="http://blog.mikemccandless.com/2013/05/eating-dog-food-with-lucene.html">Eating
-dog food with Lucene</a> blog post for details.  The source code for indexing and searching Jira issues <a href="http://github.com/mikemccand/luceneserver/tree/master/examples/jirasearch">is here</a>.</font>
+dog food with Lucene</a> blog post and <a href="http://blog.mikemccandless.com/2016/10/jiraseseach-20-dog-food-using-lucene-to.html">Jirasearch 2.0 dog food</a> for details.  The source code for indexing and searching Jira issues <a href="http://github.com/mikemccand/luceneserver/tree/master/examples/jirasearch">is here</a>.</font>
 </body>
 </html>