-- 25/02/2015, 16:43. Schema locations (with even amounts of uri's) are added as context parameters
to the application's web.xml

-- 25/02/2015.  The following fragment have been moved from the tomcat's web.xml to the 
applications' web.xml.
  <context-param> 
    <param-name>nl.mpi.cqlsearch2.imdiNamespaceURL</param-name>
    <param-value>http://www.mpi.nl/IMDI/Schema/IMDI</param-value>
  </context-param> 
  
  <context-param> 
    <param-name>nl.mpi.cqlsearch2.imdiLanguageMapURL</param-name>
    <param-value>http://www.mpi.nl/IMDI/Schema/LanguageCodeMap.xml</param-value>
  </context-param> 
  
  <context-param> 
    <param-name>nl.mpi.cqlsearch2.handleResolverURL</param-name>
    <param-value>http://hdl.handle.net</param-value>
  </context-param>
  
 

-- 16/02/2015. The bug with recordSchemaIdentifier is fixed. The bug is discovered after testing
lux17's deployment on http://weblicht.sfs.uni-tuebingen.de/Aggregator-testing. also, some
URL are moved from the source code to the web.xml as parameters.

-- 16/02/2015. schemaRecordURL corrected to http://clarin.eu/fcs/resource. The complete  set of 
permutations is generated now for clauses of an AND-request (using guava library). The mapping of 
obsolete language codes in imdi-corpora to the ISO-3 ones is now happens not inside the package
but using a corpus-management-generated mappoing at
http://www.mpi.nl/IMDI/Schema/LanguageCodeMap.xml.

-- 11/02/2015. 
1) The endpoint is adjusted for the new FCS specification, that is the responds for “explain” and 
“searchRetriwve” satisfy the new schema (it is done using JAXB translating schemata to java classes). 
2) The search termas with AND and OR are implemented. 
3) Languages of searchable corpora and their subcorpora are retrieved via retrieving languages 
of the corresponding sessions; cache is used to preserve retrieved languages; to initiate cache
(done in a separate thread) one needs to (re)start the application on the tomcat.

-- 29/11/2013. Resolved handle problem (discovered for talkbank and childes which do not have handles).
The new version delivers the link to the imd-file of a node if there is no handle.
Moreover, it recongnizes such links when they are given, 
instead of handles, as parameters for the search and scan operations. 

-- the landing-page URL refers now not to xml-file but to a user-friendly page.

November 2013 
-- the parent pom version is updated to 1.7
-- the mpi-libraries versions (including annot-search, corpusstructure- api) are updated to the latest ones mentioned on lux15
-- the scan operation now is able to recognize handles ending with
@format=cmdi, which is necessary froVLO-aggregator connection

-- June 2013. CQLConfig.xml is removed, because all necessary parameters (incl. the list of searchable corpora)
are placed in web.xml


-- 24-25/01: README.txt, INSTALL.txt, UPDATE.txt, CHANGES.txt are added.

January 2013:
-- testing
-- redistributing parameters in the project's context.xml, web.xml, CQLConfig.xml so that they
satisfy criteria of the reasonable architecture and requirements of the external sru-library. 


December 2012:
-- small update of the code according to be compatible with the new version of the search library
(done by Eric)