Changelog for perl-ElasticSearch-0.68-1.2.noarch.rpm :
Sat May 9 14:00:00 2015 cooloAATTsuse.com
- updated to 0.68
see /usr/share/doc/packages/perl-ElasticSearch/Changes
0.68 2013-03-05
Fixed the abstract to show DEPRECATED
0.67 2013-03-05
Deprecated the ElasticSearch namespace in favour of Search::Elasticsearch.
See https://github.com/elasticsearch/elasticsearch-perl/issues/20
0.66 2013-09-24
Deprecated in favour of the new Elasticsearch.pm
0.65 2013-05-08
Add support for ignore_missing to get_aliases()
0.64 2013-02-26
Default query for count()
0.63 2013-02-12
Removed unsupported params from mlt()
0.62 2013-01-03
BUG FIX
Don\'t destroy params passed in to new()
ENHANCEMENT
Added the \'order\' param to index templates
0.61 2012-11-09
ENHANCEMENT
Ignore empty queryb/filterb sections
0.60 2012-11-07
ENHANCEMENT:
Added a max_content_length attribute to ElasticSearch::Transport
to avoid posting too large a body to elasticsearch. The value
is auto-detected in versions 0.19.12 onwards.
Added doc param to update()
0.59 2012-10-31
BREAKING CHANGE:
Added the \'explain\' parameter to validate_query(), which now returns
a hash instead of true|false
BUG FIXES:
The queryb->query, filterb->filter SearchBuilder conversions no longer
change the original parameters
ENHANCEMENTS:
Added cluster_reroute()
Added type_exists()
Added warmer APIs
Added explain API
Added upsert to update()
Added ignore_indices param to search(), searchqs() and msearch()
Bulk error handlers now receive the element number of the error
to identify the original item in the array (required for Elastic::Model)
Errors now have string comparison overloaded
0.58 2012-08-09
Added a _method_name parameter to reindex() so that objects other
than ScrolledSearch can be passed as the $source. The scrolling
defaults to $source->next, but with $method_name will be called
as $source->$method_name.
0.57 2012-07-31
Minor changes to ElasticSearch::TestServer to make it work consistently
on Travis CI
0.56 2012-07-31
BUG FIXES
- Trying to bulk create a doc that already exists should also trigger
a conflict error, not a general error.
0.55 2012-07-06
ENHANCEMENTS
- Added on_conflict and on_error callbacks to bulk actions and reindex()
0.54 2012-06-30
NEW FEATURES
- Added exists()
ENHANCEMENTS
- An ElasticSearch::Error::Conflict object now contains the latest
version number
- delete_mapping() now throws an error when the index or type is
- get() now accepts a \'parent\' param
- searchqs() now accecpts \'lenient\', \'quote_analyzer\' and
\'quote_field_suffix\' params
- msearch() now accepts a \'search_type\' param
BUG FIXES
- Removed dependencies on Module::Build and base
missing
- Worked around bug https://github.com/elasticsearch/elasticsearch/issues/2072
in ES which causes the default HTTP backend to hang by returning
a response body to a HEAD request
- Fixed current_server_version() to work with async backends
0.53 2012-04-14
Bug fix:
- the timestamp param to index() should be a \'string\', not
a \'duration\'
0.52 2012-03-17
BREAKING CHANGE:
- get_alias() now uses the native GET alias request, and its output
format has changed.
Bug fix:
- put_mapping() incorrectly required a single index. Now it
can contain zero or more indices
0.51 2012-03-05
index() and create() were missing consistency and replication parameters
0.50 2012-03-02
- Added msearch()
- The short-circuit functions for bulk/mget weren\'t working correctly
either with as_json or async backends
- Fixed bug in TestServer where it wasn\'t shutting down ES correctly
if it failed to create a new ES.pm object
0.49 2012-02-17
Don\'t pass a custom port param through to ElasticSearch::TestServer
0.48 2012-02-17
BREAKING CHANGE:
Updated the bulk
* methods to support index, type, timestamp and ttl
args. As a result, the method signatures have changed.
ENHANCEMENTS:
- index(), create() and the index/create bulk actions now accept
a JSON encoded string as a value for the \'data\' param.
- added update()
- Added \'fields\' to clear_cache
- Added index_segments()
- Added validate_query()
- Added ClusterBlock exception (code 403)
- Added NotReady exception (code 503)
- index_stats() now supports the \'all\' flag
- Added params to nodes():
- http, jvm, network, os, process, thread_pool, transport
- Added params to nodes_stats():
- indices, clear, all, fs, http, jvm, network, os, process,
thread_pool, transport
- Added params to analyze():
- tokenizer, filters, field
- Added param \'partial_fields\' to search()
- ElasticSearch::TestServer will now die if it can\'t find the
PID of the started ES process
- Added drain_buffer(), refill_buffer() and facets() to
ElasticSearch::ScrolledSearch
BUG FIXES:
- Fixed support for as_json in ScrolledSearch
- Fixed implicit IO::Handle bug when trying to log in older Perls
- Fixed a bug in index_stats and addded \'get\' stats and groups.
- Corrected spelling of \'quorom\' to \'quorum\'
- create() now uses the index op_type instead of /_create to
fix a bug when creating a doc without an ID
- put_mapping wasn\'t respecting a type set with use_type()

Thu Oct 27 14:00:00 2011 pascal.bleserAATTopensuse.org
- update to 0.47:

* merged patch from AVAR, which ensures that the servers are used in random
order

Wed Sep 7 14:00:00 2011 pascal.bleserAATTopensuse.org
- update to 0.46

Sun May 1 14:00:00 2011 pascal.bleserAATTopensuse.org
- update to 0.37:

* BREAKING CHANGES:
+ ElasticSearch::ScrolledSearch
$scrolled_search->next() now returns 1 result by default, instead of
$size results, to make it easier to write:
while (my $result = $scroll->next) {}
+ ElasticSearch::TestServer is now an object which subclasses ElasticSearch itself
+ connect_to_server() has been removed
+ also, the port parameter can be used to specify a custom start port when
starting the test server, making it easier to run the test server while there
is an existing ES server running

Thu Apr 28 14:00:00 2011 pascal.bleserAATTopensuse.org
- update to 0.36:

* Updated ElasticSearch.pm to work with ElasticSearch version 0.16.0

* New features:
- no_refresh:
Disable the automatic discovery of live nodes, to work with
remote clusters or proxied clusters
- scrolled_search():
Convenience iterator to wrap search/scroll requests
- reindex():
Utility to reindex data from one index to another, or one
cluster to another
- use_index()/use_type():
Set default index/type parameters
- searchqs():
URI/query string search, (as opposed to search() which uses a
request body)
- New search_type\'s : \'scan\' and \'count\'
- index_settings():
Retrieve index settings
- changes from 0.35:

* added the \'dynamic_templates\' option to put_mapping

Sat Mar 19 13:00:00 2011 pascal.bleserAATTopensuse.org
- update to 0.33:

* added the ElasticSearch::Transport::HTTPTiny backend
- changes from 0.32:

* added as_json as an arg to all ES requests

* fixed a logic bug when setting the port manually in
ElasticSearch::TestServer
- changes from 0.30:

* added some missing requirements to Build.PL

* scroll() now accepts a \'scroll\' parameter and works (but sorting when
scrolling still broken on the server)

* reindex_es.pl now also preserves the _parent and _routing fields
- changes from 0.28:

* support added for
+ Percolators
+ Open / Close index
+ Index templates
+ Routing
+ Parent docs
+ Versioning
+ Analyzer
+ ElasticSearch::QueryParser
+ ElasticSearch::TestServer

* create_index() now supports adding mappings at the same time
+ added bulk_create(), bulk_index() and bulk_delete()
+ bulk actions now accept _id and id etc, which means that search results
can be passed directly to bulk without renaming these parameters
+ added river_status()

* in ElasticSearch::Transport
+ added max_requests() to force a refresh of known live nodes every
$max_requests
+ on the first request, the list of live nodes is automatically updated
+ made httplite recognise more error states as temporary Connection errors
+ added support for recognising 404 and 409 HTTP codes
+ upgraded Thrift to version 5.0
+ improvements to logging
- changes from 0.27:

* fixed an undefined bug in Transport::refresh_servers which would die if no
nodes were found
- changes from 0.26:

* added a missing dependency on parent.pm
- changes from 0.25:

* another ambiguous use of localtime in ElasticSearch::Transport
- changes from 0.24:

* corrected an ambiguous use of localtime in ElasticSearch::Transport
- changes from 0.22:

* support ElasticSearch version 0.12.0

* added the bulk() method, for bulk indexing, creating and deleting

* added create_river(), delete_river() and get_river()

* added delete_mapping()

* added error_trace()

* completely refactored ElasticSearch to enable pluggable backends:
+ the request parser methods now live in ElasticSearch/RequestParser.pm
although they are loaded into the ElasticSeach namespace
+ added ElasticSearch::Transport as a transport base class
+ added ElasticSearch::Transport::HTTP (default) as the LWP backend
+ added ElasticSearch::Transport::HTTPLite as the HTTP::Lite backend
+ added ElasticSearch::Transport::Thrift as the Thrift backend
+ added ElasticSearch::Error class for exception objects
+ added ElasticSearch::Util with some generally useful utilities
+ a number of methods have from ElasticSearch to ElasticSearch::transport
including current_server(), servers(), refresh_servers()
+ removed the debug() method - now use $ElasticSearch::DEBUG instead
+ added filter_keywords to ElasticSearch::Util
- changes from 0.21:

* supports ElasticSearch version 0.10.0

* added the _source parameter to put_mapping()

Wed Dec 1 13:00:00 2010 cooloAATTnovell.com
- switch to perl_requires macro

Wed Sep 8 14:00:00 2010 chrisAATTcomputersalat.de
- recreated by cpanspec 1.78
- noarch pkg
- fix deps
o no dependency on perl(common::sense) -> JSON::XS (line 104)

Sat Sep 4 14:00:00 2010 pascal.bleserAATTopensuse.org
- initial package (0.20)