From 01f9b10c275d741255a6a40376b8817f1982b856 Mon Sep 17 00:00:00 2001
From: Samuel Lai <sam@edgylogic.com>
Date: Thu, 27 Feb 2014 17:57:34 +1100
Subject: [PATCH] Fixed #7. Turns out post IDs are not unique across sites.

This change will require re-indexing of all sites unfortunately. On the upside, more questions to browse!
---
 java/solr/server/solr/stackdump/conf/schema.xml | 6 +++++-
 python/src/stackdump/commands/import_site.py    | 3 +++
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/java/solr/server/solr/stackdump/conf/schema.xml b/java/solr/server/solr/stackdump/conf/schema.xml
index ff4de8c..5de8dc2 100644
--- a/java/solr/server/solr/stackdump/conf/schema.xml
+++ b/java/solr/server/solr/stackdump/conf/schema.xml
@@ -110,6 +110,10 @@
 
    <!-- we'll get the values out of the JSON, so most fields are not stored -->
    <!-- fields are listed here so searches can be performed against them -->
+   <!-- this is used by Lucene to uniquely identify a post across all sites.
+        It is of the form "siteKey-id" and is necessary because post IDs are
+        reused across sites. -->
+   <field name="documentId" type="string" indexed="true" stored="true" required="true" />
    <!-- the ID field needs to be a string for the QueryElevationComponent -->
    <field name="id" type="string" indexed="true" stored="true" required="true" />
    <field name="siteKey" type="string" indexed="true" stored="true" required="true" />
@@ -196,7 +200,7 @@
  <!-- Field to use to determine and enforce document uniqueness. 
       Unless this field is marked with required="false", it will be a required field
    -->
- <uniqueKey>id</uniqueKey>
+ <uniqueKey>documentId</uniqueKey>
 
  <!-- DEPRECATED: The defaultSearchField is consulted by various query parsers when
   parsing a query string that isn't explicit about the field.  Machine (non-user)
diff --git a/python/src/stackdump/commands/import_site.py b/python/src/stackdump/commands/import_site.py
index fc2c296..f2b71db 100644
--- a/python/src/stackdump/commands/import_site.py
+++ b/python/src/stackdump/commands/import_site.py
@@ -459,6 +459,9 @@ class PostContentHandler(xml.sax.ContentHandler):
         doc['answers-json'] = [ json.dumps(a, default=self.json_default_handler) for a in q['answers'] ]
         
         # map other fields to search index doc
+        # this is the ID for Solr to uniquely identify this question across all
+        # sites
+        doc['documentId'] = self.site.key + '-' + str(q['id'])
         doc['id'] = str(q['id'])
         doc['siteKey'] = self.site.key
         doc['creationDate'] = q['creationDate']