bc976f1bb6
New node_info API
2014-01-08 20:23:24 +01:00
80dcf406b7
Add support for flat_settings flag to all REST APIs that output settings
...
See https://github.com/elasticsearch/elasticsearch/issues/4140
2014-01-08 19:44:56 +01:00
08b2e94333
Add a not on versioning and branches
2014-01-07 16:48:30 +01:00
21e7c6e5d8
Have our master track the master (1.0) branch of ES
2014-01-07 16:40:41 +01:00
6d3a4e238d
Adding description of force and full options to indices.flush
2014-01-06 18:20:30 +01:00
aaeab13f8b
Handle regexp-based catch statements in the yaml test suite
2014-01-03 18:05:25 +01:00
2ced9697be
Use new yaml tests
2014-01-02 20:07:44 +01:00
a49371b553
Reflect new ES api for get aliases.
...
see https://github.com/elasticsearch/elasticsearch/issues/4539 for more
details
2014-01-02 20:00:59 +01:00
fb14082f89
Adding new parameters to indices APIs
...
See https://github.com/elasticsearch/elasticsearch/issues/4436 for more details
2014-01-02 19:51:08 +01:00
f7d7931a2d
Rest-api-spec repo has been merged with elasticsearch proper
...
Changing the git submodule and moving it top level to avoid traversing
the es source code during test discovery
2013-12-28 17:40:00 +01:00
d1e031ac03
Version bump to 0.4.4
2013-12-23 16:35:34 +01:00
0f440508ce
Added configuration for the wheel package format
2013-12-23 16:30:09 +01:00
9b60521b34
Tweaking the changelog a bit to prepare for release
2013-12-23 16:27:27 +01:00
06efe2c15b
Update to latest tests
2013-12-23 16:03:04 +01:00
18f2ac1231
#29 - corrected target_client assignment
2013-12-21 01:18:17 +05:30
6f7349eec4
make Elasticsearch.scroll POST the scroll ID
...
Without this, if you have a large number of indices the `scroll_id` will be very long, and ElasticSearch will return an invalid response. (A urllib3 `ConnectionError` to be precise.) This changes the behavior to POST to the scroll endpoint, which gets around the URL length restriction.
2013-12-16 16:33:32 +01:00
05c4eefbf3
Enable deserialization of different content types coming from the server.
2013-12-14 18:15:15 +01:00
d1ccc2430d
Mark 0.4.4 as development at leats in Changelog
2013-12-13 19:38:59 +01:00
e2ab7ea80d
Adding a third return value from connection.perform_request - headers
...
This will be used to implement deserializing based on mimetype
2013-12-13 19:36:21 +01:00
be9acfad90
Make sure we handle quotes correctly in the trace logger
2013-12-11 14:41:42 +01:00
1345b36948
Make sure actions aren't modified when passed into bulk
2013-12-11 14:30:59 +01:00
79c59d6f80
Use fully qualified names for the generated tests
2013-12-09 22:24:54 +01:00
a19595e803
Report the name of the generated test case when running the tests with -v
2013-12-09 22:17:36 +01:00
99eb8936a3
Another example query, this time with an actual query
2013-12-09 17:48:54 +01:00
3e262779bb
Example code using the client
2013-12-09 03:17:32 +01:00
6947b01275
Removed extra reference to the id parameter for create API
2013-12-09 02:53:07 +01:00
0ab904c4de
Use dict.popitem when we know the dict only has one key
2013-12-08 19:27:44 +01:00
e9eaefd999
Add a note to changelog about the unicode fix
2013-12-04 15:10:31 +01:00
95fba05e02
Fix unicode body handling.
...
Only body should be encoded as bytes, not anything going through the
serializer (mostly because of bulk)
2013-12-04 15:07:56 +01:00
4420296636
For string body make sure we serialize unicode properly
2013-12-04 13:22:33 +01:00
19f9eaab95
Make sure bulk requests are logged correctly
2013-12-03 02:04:11 +01:00
c5c98dcaac
Add logging of body as INFO on connection error
2013-12-03 01:55:32 +01:00
cf03f9831a
Ass send_get_body_as parameter for situations where GET cannot have a body
2013-12-02 22:57:47 +01:00
69b8023c15
Doc fix: bulk_index -> bulk rename
2013-12-02 17:17:06 +01:00
6f6c864bc8
Bulk supports timeout parameter in new ES
2013-11-25 11:57:21 +01:00
10ae841db3
Add a not on the optional test client override to CONTRIBUTING.md
2013-11-23 19:34:54 +01:00
911817d514
Have git ignore local overrides in tests
2013-11-23 19:22:38 +01:00
82fa7b03a4
Enable users to override the client used in tests.
...
Fixes #7 , thanks nkvoll!
2013-11-23 19:20:44 +01:00
5ce42094c6
Skip test for different op types before 0.90.1
2013-11-22 22:28:29 +01:00
5f8c5322b0
Have bulk tests work for 0.90.0 too
2013-11-22 16:05:51 +01:00
8e8981d9c9
New bulk docs updated
2013-11-22 14:38:38 +01:00
3ae44db90c
Bulk helpers overhaul
...
streaming_bulk for iterative consumption of the input
bulk_index -> bulk
callback expand_action to transform input into action/data pairs
support for non-index actions through _op_type key
2013-11-22 14:29:08 +01:00
a4e068c505
Fix the helpers tests.
...
ES doesn't guarantee that refresh will wait for cluster state
propagation and thus the type might not exist right after the index
being created during the bulk load. Just check the documents instead.
2013-11-19 22:33:53 +01:00
1c279d5aad
By default allow 10 connections to be held to the server
2013-11-19 22:31:25 +01:00
eed4698c2c
Merge branch 'urllib3_pool_maxsize' of github.com:rboulton/elasticsearch-py
2013-11-19 22:29:08 +01:00
7c84613abc
Document the format of the body parameter for bulk operations, and correct a couple of typos in the documentation.
2013-11-19 17:03:32 +00:00
e8439372d6
Allow the maxsize parameter to be passed to the urllib3 connection pool
2013-11-19 16:45:39 +00:00
5557fbff1e
Make sure jenkins output contains logs
2013-11-19 04:10:34 +01:00
daad6eacca
Spell mistake corrected
2013-11-18 02:17:17 +05:30
4bce8a464c
Added more detailed error handling
...
namely being able to detect conflicts and bad requests
2013-11-14 01:12:55 +01:00
Honza Král