Gentics Content.Connector Changelog Summary

Gentics Content.Connector 2.4.0


Manual Change

Some dependencies have been updated:

  • log4j was updated to 2.17.0. This requires the log configuration file conf/gentics/nodelog.properties to be migrated
    to the configuration file conf/gentics/nodelog.yml. See Migrating from Log4j 1.x for details.
  • node-lib was updated to 2.4.2
  • mysql connector was updated to 8.0.27
  • commons-compress was updated to 1.21
  • commons-io was updated to 2.11.0
  • httpclient was updated to 4.5.13
  • jsoup was updated to 1.14.2
  • pdfbox was updated to 2.0.24
  • jackson was updated to 2.12.6
  • tika was updated to 1.27

Gentics Content.Connector 2.3.1


Bugfix SUP-8960

Updated internal utility dependency.


Bugfix SUP-8056

Updated dependency to node-lib.


Gentics Content.Connector 2.3.0


Manual Change

Java 8 is now required. If you are using a Java 7, please update your JRE/JDK to Java 8.


Bugfix SUP-7992

Updated the dependency Apache Tika to 1.20 in order to fix security vulnerabilities.


Gentics Content.Connector 2.2.10


Bugfix SUP-7992

Updated some dependencies in order to fix security vulnerabilities:

  • pdfbox to 2.0.15
  • c3p0 to 0.9.5.4
  • httpclient to 4.5.9
  • xalan to 2.7.2
  • velocity-tools to 1.4
  • node to 2.3.8

Gentics Content.Connector 2.2.9


Bugfix SUP-7578

Updated dependency to node-lib.


Gentics Content.Connector 2.2.8


Gentics Content.Connector 2.2.7


Bugfix SUP-6797

Updated Apache PDFBox version to 2.0.11 to prevent OutOfMemory errors.


Gentics Content.Connector 2.2.6


Bugfix SUP-4077

Updated dependency to node-lib.
This fixes a problem with background sync checking of datasource using persistent caches.


Bugfix SUP-6718

Update the Apache Tika, Apache POI, Apache Pdfbox and Apache Commons Compress dependencies to the newest versions.


Bugfix SUP-6717

Fixed missing log output for Java-Erros (such as “NoClassDefFoundError”) thrown inside of IndexJobs.


Bugfix SUP-6561

Updated dependency to node-lib.
This solves an issue where images did not get the correct filename when being resized by Gentics Image Store.


Gentics Content.Connector 2.2.5


Bugfix

Isolate the Velocity-runtime for the contentconnector. The contentconnector will now use
the new-able VelocityEngine class internally instead of the Velocity class singleton.
This ensures that the configuration of the Velocity engine in the contentconnector will
not interfere with that of other libraries or frameworks which use (and configure) the
singleton model.


Gentics Content.Connector 2.2.4


Enhancement SUP-3060

Add new LDAPRequestProcessor. This RequestProcessor can be used to access LDAP entries
via the Gentics ContentConnector. For usage in the Lucene indexer see the
deact_indexer.ldap.properties file in the config-sample.


Bugfix

The node version has been updated to 2.3.1. This node version includes a fix for multithreaded access of
caches for contentobjects.
Also the version of the trove library has been updated to 2.1.0.


Bugfix

The NavigationCacheRequestProcessor has been fixed to get the navigation synchronously,
if it is not found in the cache, instead of returning null.


Gentics Content.Connector 2.2.3


Bugfix

The dependencies have been cleaned again to avoid some more errors
due to conflicting libraries.


Gentics Content.Connector 2.2.2


Bugfix

The dependencies have been cleaned to avoid various errors
due to conflicting libraries.


Gentics Content.Connector 2.2.1


Bugfix

The node version has been increased to 2.1.2. This node version does not use quartz for scheduling of
datasource related background jobs. Due to a version conflict of quartz, sometimes the background jobs were
not scheduled (without any error message), and clearing the caches due to changes in the contentrepository did not work.


Gentics Content.Connector 2.2.0


Security

Updated the dependencies org.apache.poi and org.apache.tika to their latest versions in order to fix the security vulnerabilities CVE-2014–3529 and CVE-2016–5000


Enhancement

The delimiter used in the transformer SimpleMVString to join values can now be configured.
The previous static delimiter, a whitespace, is now the default delimiter.
The use of stopwords can now be toggled in CustomPatternAnalyzer. Stopwords are enabled by default.


Gentics Content.Connector 2.1.1


Bugfix

The request processor accessing an autocomplete index sometimes started to
return incorrect results due to index readers that were not reopened after the
index was modified.
This has been fixed now.


Bugfix

When beans returned from a Request Processor were missing the idattribute,
a NullPointerException could occur. This has been fixed now, a CRException with an
explicit error message will now be thrown.


Gentics Content.Connector 2.1.0


Enhancement

The new maven module contentconnector-navcache provides the NavigationCacheRequestProcessor,
which will cache responses to navigation requests and will automatically refresh the caches before the
cached navigations will become invalid.
See the configuration file navcache.properties in the module contentconnector-config-sample for an example
configuration.


Gentics Content.Connector 2.0.12


Bugfix

Updated the dependency node to 2.0.45 in order to fix the initialization of the JSC caches not being synchronized. This caused exceptions like a NPE.


Gentics Content.Connector 2.0.11


Enhancement

The contentconnector-csv module was added to the contentconnector-servlet sample.


Gentics Content.Connector 2.0.10


Enhancement

Implemented a CSV request processor that reads a CSV file (with file path, delimiter and escape characters defined in a properties file)


Gentics Content.Connector 2.0.9


Bugfix

When using a collector in the result of a query with result type JSON, a ClassCastException was thrown, which has been fixed now.


Gentics Content.Connector 2.0.8


Bugfix

Due to a race condition in the initializtion of the JCS cache in the content-connector, the PortalCache for the node-lib was intialized before the cache in the content-connector. Due to this bug the JCS cache was not intialized correctly for some request-processors. This has been fixed: the content-connector now uses the node-lib‘s JCS-wrapper (the PortalCache) instead of JCS directly. Switching to PortalCache ensurse that the same cache is used for all objects and that JCS is initialized only once.


Gentics Content.Connector 2.0.7


Bugfix

Fixed rarely occurring bug when using CRRecencyBoostingQuery and having elements with big timestamps


Bugfix

Fixed bug when using autocomplete feature with stopwords.
Now uppercase searchTerms get automatically converted to lowercase
and if the searchTerm is an stopword the result will be empty.


Bugfix

Changed the loglevel of the "Data could not be found" message from
ERROR to INFO since a request for non-existent data is not a serious
error in content connector itself.


Gentics Content.Connector 2.0.6


Bugfix

When fetching search results containing facets and using the XmlContentRepository (type=xml) an error was thrown
because the facets results where not parsed correctly to the XML-Response. This has been fixed.


Gentics Content.Connector 2.0.5


Bugfix

The version of the node-lib dependency has been raised to 2.0.31 (for various bugfixes).


Bugfix

Add relevant facetconfig mapping to search.properties config-sample to ensure working facet search with sample-config. Print warn message if facet search is enabled and no mapping is configured.


Gentics Content.Connector 2.0.4


Bugfix

Fixed a Java 7 compatibility issue which first occurred in version 2.0.3.


Gentics Content.Connector 2.0.3


Enhancement

Add possibility to define multiple database handles for one datasource.
A Round-Robin algorithm is used to choose a handle.
If one handle is not available it will be removed temporarily from pool
For configuration see: https://github.com/gtxcontentconnector/contentconnector/wiki/Configuration


Bugfix

Search queries containing the characters ‘&‘, ‘|‘ or ‘~‘ returned unexpected results when
using the CRQueryParser and inherited classes because those characters where always
replaced with ‘AND‘, ‘OR‘ and ‘NOT‘. This has been fixed: if those characters are
escaped (using backslash ‘\‘) they won‘t be replaced anymore.


Gentics Content.Connector 2.0.2


Bugfix

Sometimes an indexing job failed with “org.apache.lucene.store.AlreadyClosedException: this IndexReader is closed”.


Gentics Content.Connector 2.0.1


Bugfix

CustomPatternAnalyzer would stop working after processing the first string.


Gentics Content.Connector 2.0.0


Enhancement

Dependency management: The portalnode-lib dependency was removed from the Gentics Content.Connector

  • The ITemplateManager interface was moved to node-lib and will be available starting from version 2.0.13.
  • The ITemplateManager interface in node-lib was updated. The render method will no longer throw CRException. Instead a normal Exception will be thrown.
  • The PortalNodeTemplateManager class was moved to portalnode-lib and can still be used when manually adding the portalnode-lib dependency.
  • The PluggableActionCaller class was moved to portalnode-lib.
  • The PortalNodeInteractor class is no longer interfacing directly with portalnode. Instead the PortalConnectorFactory class is used for datasource creation.
  • Various portalnode-lib dependencies were removed in various pom.xml files.

Gentics Content.Connector 1.17.1


Bugfix

CustomPatternAnalyzer would stop working after processing the first string.


Gentics Content.Connector 1.17.0


Enhancement

This change implements a new RequestProcessor (HTTPClientAllParamsRequestProcessor) which will forward all custom parameters (but not the default parameters) set in the CRRequest.


Enhancement

-Added a custom collector class LanguageFallbackSortingCollector that is able to do a language fallback while searching and sort for string fields.


Enhancement

-Added a custom Transformer (DOCParagraphContentTransformer) that will transform word documents using the paragraph API of the Apache POI package and therefore leave out meta attributes such as DOCPROPERTY


Enhancement

A new highlighter was added. The CharacterCountPhraseBolder creates its fragments based on character count.


Bugfix

In some cases the new lucene LanguageFallbackSortingTopDocsCollector ignored the fallbackpriority parameter. This has been fixed, the request parameter “languagefallbackpriority” will now be used regardless if the request was created via the Java API or the REST API.


Bugfix

Fixed a bug in KeywordListAnalyzer where it would not produce any tokens.


Gentics Content.Connector 1.16.2


Bugfix RT58553

When searching the autocompleter for terms with special characters, the query
was not parsed correctly and no results where shown for many words containing
special characters like umlaute.


Gentics Content.Connector 1.16.1


Bugfix RT58458

Fixed class cast exception in Autocompleter Index creation.


Bugfix RT58446

When using the CRLuceneIndexJob to index objects from a Content Repository in a lucene index,
attributes were not removed from the index if they had been removed from the objects.
This has been fixed now.


Gentics Content.Connector 1.16.0


Manual Change

JDK compliance update

Due to the updated node and apache lucene dependency it is now mandatory to use JDK 1.7.


Manual Change

The com.gentics.cr.XMLRequestProcessor was removed.

The class was removed in order to simplify the inherited dependencies. You may add your own implementation which you can base upon this revision


Enhancement

Updated Apache Lucene to 4.9.0. In order to use the new Lucene version, it is recommended to re-index every configured index with the new version.
Please check your analyzer.properties configuration, as the package names of some analyzers may have changed in the lucene packages.


Gentics Content.Connector 1.15.2


Bugfix RT58553

When searching the autocompleter for terms with special characters, the query
was not parsed correctly and no results where shown for many words containing
special characters like umlaute.


Gentics Content.Connector 1.15.1


Bugfix RT58458

Fixed class cast exception in Autocompleter Index creation.


Bugfix RT58446

When using the CRLuceneIndexJob to index objects from a Content Repository in a lucene index,
attributes were not removed from the index if they had been removed from the objects.
This has been fixed now.


Gentics Content.Connector 1.15.0


The dependency on com.gentics:node-lib has been updated to 1.21.21


This is the last feature release to support Java 6. It will get bugfix releases until February 1st, 2015.


Gentics Content.Connector 1.14.0 (Unreleased)


Enhancement

Updated Apache Lucene to 4.9.0. In order to use the new Lucene version, it is recommended to re-index every configured index with the new version.
Please check your analyzer.properties configuration, as the package names of some analyzers may have changed in the lucene packages.


Gentics Content.Connector 1.13.3 (Unreleased)


In this release, only dependencies have been updated, namely node-lib to 2.0.4 and portalnode-lib to 4.6.0


Gentics Content.Connector 1.13.2


Bugfix

Removed a ClassCastException when a language not existing in the CR is requested using CMSPageLanguageFallbackRequestProcessor


Gentics Content.Connector 1.13.1


Bugfix

Lucene Index Update Checker now exits gracefully on server shutdown.


Bugfix

A bug in the velocity template caching was fixed. Previously the template caching didn‘t cache the evaluated templates correctly.