This is the release notes for XWiki Commons, XWiki Rendering, XWiki Platform and XWiki Enterprise. They share the same release notes as they are released together and have the same version.

This is mostly a stabilization release leading to XWiki 5.3 final and it brings a few small improvements to the Solr search UI and the ability to import large XARs using the WikiStream framework. The Solr index has been modified to support the search for documents matching specific values in multiple object properties. As a consequence you can now facet and sort the search results based on property values. Besides this the new release includes 38 bug fixes and 14 small improvements.

New and Noteworthy (since XWiki 5.3 Milestone 2)

Full list of issues fixed and Dashboard for 5.3.

Solr Search Improvements

We made a few small improvements to the search facets:

• We added a hint to explain the 'No language' value in the language search facet

    

• We improved the way a custom interval is specified in the date search facet

    

• We improved the 'Object Type' facet

    

We also improved the way Static List properties are indexed. We now index the raw value (what is saved in the database) as verbatim (without being analysed, for exact matching) and the display value (what is specified in the XClass) as localized text (using the locale of the document that holds the XObject). 

See the Solr Search Application for more details.

Import large XARs

It's possible to enable the use of new WikiStream framework in standard import action (used in the standard import UI for example). Among other things it allows importing a wiki XAR of pretty much any size.

WikiStream is too young to be enabled by default and will require heavy testing. In the meantime you can try it by uncommenting #xwiki.action.import.xar.usewikistream=1 at the end of xwiki.cfg file.

Miscellaneous

• Added new configuration parameter for the Link Checker: rendering.transformation.linkchecker.excludedReferencePatterns (See Reference documentation).
• If you want to change the wiki owner or the prettyname of the wiki, you cannot go to the "Wiki" section of the Wiki Administration anymore since it's been removed for the moment due to some technical issue. Instead, go to the Wiki Index and then click on the "edit" action corresponding to the wiki for which to update the settings.
• When creating an empty wiki you now get the Distribution Wizard even when ALLINONE upgrade is selected (this used to disable the Distribution Wizard on subwikis)

See the full list of JIRA issues fixed in this release.

For Developers

Search for Documents Matching Specific Values in Multiple XObject Properties

Unlike a relational database, the Solr index has a single 'table'. So normally you index just one type of entity. Still, in XWiki you can search for documents, attachments, objects and object properties (see the 'Result Type' advanced search option). This is possible because we use a 'type' column in the index. So basically we index multiple types of entities, and we use the 'type' column to determine the type of entity an index row corresponds to. The entities are related though (the documents have objects and attachments, the objects have properties) and so you often need to put constraints on different entity types when performing a search (e.g. search for documents that have an object of a specific type with a property that matches a given value). In a relational database when you have the information spread in multiple places (tables) you do joins. Solr supports joins but the query syntax is complex and not as powerful as the SQL version. Plus, each join means index X index where X is the cartesian product, which can be expensive if the index is very large.

In this version we removed the need to perform joins when searching for documents matching specific values in multiple XObject properties. We achieved this by indexing XObject data on the document rows also. Of course, this means the index will have redundant information (XObject data is duplicated on document and object rows), but we believe this is not a problem and the increase in the Solr index size pays off when you take into the accout the search speed and the simplified query syntax.

Here's a short summary of what we implemented:

• We're using an encoding scheme similar to the URL-encoding to support special characters in the Solr index field names. We didn't use directly the URL-encoding because '+' (plus) and '%' (percent) have special meaning in Solr query syntax. Also, we didn't want to encode Unicode letters.
  E.g. "Somé Spâce.Bob's Claß" is encoded as "Somé$20Spâce.Bob$27s$20Claß"
• We wanted to be able to extract the class and property reference from a field name in order to display the location where the search text has been found. We couldn't use the default class / property reference serialization syntax because '\' and '^' have special meaning in the Solr query syntax. So we implemented a simple serialization syntax that uses only '.' as entity separator and the dot is escaped by repeating it.
  E.g. "wiki:Some\.Space.My\.Class^color" is serialized as "wiki.Some..Space.My..Class.color"
• We added the following fields to a document's index:
  • object : all types of objects found on the indexed document, e.g. ['Blog.BlogPostClass', 'XWiki.TagClass']
  • object.Space.Class : collects the values from all the properties of Space.Class objects found on the indexed document
  • property.Space.Class.propName : indexes the values of Space.Class^propName (multiple values if there are multiple objects of type Space.Class)
• object.* and property.* are multilingual fields so they are indexed in multiple languages. We added support for dynamic aliases (for dynamic fields) so we can write:
  object:Blog.BlogPostClass AND property.Blog.BlogPostClass.title:text AND object.XWiki.TagClass:news

  and it will be expanded into

  object:Blog.BlogPostClass AND
  (property.Blog.BlogPostClass.title_en:text OR property.Blog.BlogPostClass.title_fr:text OR ...) AND
  (object.XWiki.TagClass_en:news OR object.XWiki.TagClass_fr:news OR ...)

  Note that Solr doesn't support dynamic fields as default fields, i.e. as fields that are matched when you search for free text (without field:value in the query). This is not a problem for the search results, as dynamic fields like object.* and property.* are copied and aggregated in objcontent which is a default field. The issue is that we can't know what is exactly the XClass property that was matched by a free text search, we just know that the text was found inside an object.

• We also added support for indexing non-string XObject properties based on their type. This means Boolean, Number (int, long, float, double) and Date properties are now correctly indexed (on document rows) and we can write something like:
  property.Blog.BlogPostClass.publishDate:[NOW-1MONTH TO NOW]

Faceting and Sorting on Object Properties

We added support for faceting and sorting on XObject properties. See the Solr Search Query API for details.

See also the Solr schema design for the full list of fields that can be used in search queries.

Miscellaneous

• You can now easily run functional tests using PhantomJS by using -Dbrowser=*phantomjs.

Translations

The following translations have been updated: 

• German
• French
• Latvian
• Dutch
• Russian

Tested Browsers & Databases

Here is the list of browsers we support and how they have been tested for this release:

	Browser	Status
	Google Chrome 31	Jira Tickets Marked as Fixed in the Release Notes
	Mozilla Firefox 25	Not Tested
	Internet Explorer 8	Not Tested
	Internet Explorer 9	Not Tested

Here is the list of databases we support and how they have been tested for this release:

	Database	Status
	HyperSQL 2.3.0	Jira Tickets Marked as Fixed in the Release Notes
	MySQL 5.6.10	Not Tested
	Oracle 11.2	Not Tested
	PostgreSQL 9.2.4	Not Tested

Known issues

• Bugs we know about

Backward Compatibility and Migration Notes

General Notes

When upgrading make sure you compare your xwiki.cfg, xwiki.properties and web.xml files with the newest version since some configuration parameters may have been modified or added. Note that you should add xwiki.store.migration=1 so that XWiki will attempt to automatically migrate your current database to the new schema. Make sure you backup your Database before doing anything.

Solr Search Index

API Breakages

The following APIs were modified since XWiki 5.2.1:

• Requesting a dependency without indicating the namespace does not really make sense. This interface is not really supposed to be implemented anyway.

• Needed to filter search by namespace. This interface is not really supposed to be implemented anyway.

• Young API. Not supposed to be implemented from outside.

• Young API. Not supposed to be implemented from outside.

• New configuration property rendering.transformation.linkchecker.excludedReferencePatterns

• Young API that has been refactored to introduce the new Resource module.

• XWikiURL (now Resource) should only be in the Execution Context and not in the Request.

• Lots of refactoring in WikiStream.