Commit Graph
842 Commits
Author SHA1 Message Date
Nick LangandHonza Král 790c80fedb bumping requests to 2.0.0+ (#587)
requests < 2.0.0 causes the following attribute error:
    AttributeError: type object 'Session' has no attribute 'prepare_request'
2017-05-15 18:36:51 +02:00
Honza Král 3b41c3bca7 Add develop extra to allow for easy dependency install 2017-05-10 16:52:58 +02:00
Patryk KrawaczyńskiandHonza Král 6ccc08520c doc_type = type to sample (#572)
elasticsearch.exceptions.RequestError: TransportError(400, 'illegal_argument_exception', 'request [/_cluster/nodes/hotthreads] contains unrecognized parameter: [doc_type]')

https://www.elastic.co/guide/en/elasticsearch/reference/current/cluster-nodes-hot-threads.html
2017-05-10 16:33:38 +02:00
Honza Král 5c658fcfb8 Allow pipeline as a parameter in bulk actions
Fixes #564
2017-04-17 22:12:41 +02:00
Honza Král c318f7fa63 Make sure environment settings are merged in for requests
Fixes #557
2017-04-17 21:59:03 +02:00
Honza Král 5e63ea4f93 hardcoded ES version in travis.yml 2017-04-17 21:27:26 +02:00
Honza Král 618b69ae7a Merge branch 'master' of github.com:elasticsearch/elasticsearch-py 2017-04-13 23:51:25 +02:00
Patrick MézardandHonza Král 7f6d38273e Unquote user:password in hosts (#569)
Before turning them into an authentication header, the user or password
supplied in hosts must be URL decoded.

The RFC https://tools.ietf.org/html/rfc3986 describes userinfo like:

  userinfo    = *( unreserved / pct-encoded / sub-delims / ":" )

and python documentation about urlparse says:

  The components are not broken up in smaller parts (for example, the
  network location is a single string), and % escapes are not expanded.

It is the caller job to unquote them before using them.

For instance, a character like "]" cannot be supplied in a host or
urllib will incorrectly interpret it using ipv6 syntax.

Fix #568
2017-04-13 15:10:40 +02:00
Honza Král 2f62f9e3d4 version bump for 5.3 2017-03-30 22:32:38 -07:00
Honza Král fe897ebe0d Wait_for_completion defaults to true 2017-03-26 14:11:01 +02:00
Honza Král c32e3e52f7 new API compatibility 2017-03-21 13:49:17 +01:00
allias888andHonza Král 05ad5ef96b wait for completion is on by default (#550) 2017-03-12 14:51:40 -07:00
Honza Král 597bf7ade9 Log proper headers in trace logs 2017-03-05 11:19:01 -08:00
Husain ZafarandHonza Král ed5a1cce8e This commit modifies the elasticsearch/connection_pool.py file to solve the list index out of range error. (#539)
Fixes #371
2017-02-19 16:04:32 +05:30
Honza Král 66cc0a82b4 allow fielddata for files field to allow aggregations 2017-02-18 15:49:13 +05:30
Honza Král f638c3f625 bump version to 2.0 2017-02-12 16:26:02 +01:00
Honza Král fc25175130 Added API docs for ingest and tasks
Fixes #526
2017-02-12 16:23:49 +01:00
Glen SmithandGitHub 1d3e7022b5 Merge pull request #532 from keith-pinto/fix_doc
fix ordering of arguments in documentation for client's get method
2017-02-11 10:54:26 -07:00
keith pinto 52f8fd86ac fix order of arguments in documentation for client's get method 2017-02-09 04:57:42 +05:30
Honza Král 1a8896e207 setup.py typo 2017-02-08 21:30:23 +01:00
Honza Král a180459d69 Update API parameters to 5.x 2017-02-08 21:07:29 +01:00
Honza Král a96f5f7284 Sending raw text as body is no longer supported, wrap scroll_id 2017-02-08 20:51:13 +01:00
Honza Král d862c8a799 Add content-type default headers with application/json 2017-02-08 20:50:50 +01:00
Alex ChanandHonza Král c7441e9dc0 Fix a few spelling mistakes in the docs (#522) 2017-01-21 16:44:07 +01:00
Aggelos AvgerinosandHonza Král 7fa7deab1e exceptions-docs: Remove duplicate entry for ConnectionError (#523) 2017-01-18 18:07:31 +01:00
Honza Král f38fa51fea When user is passing params we need to make a copy
Fixes #517
2017-01-17 03:10:50 +01:00
Honza Král 5ecfc70a16 bunp version to 5.1.0 for release 2017-01-11 15:45:55 +01:00
Honza Král 0bf58d7717 Updated API 2017-01-11 15:42:33 +01:00
Honza Král a870b4bf40 test es 5.1.1 2017-01-11 14:26:56 +01:00
ihipopandHonza Král fa0ee04b8f Add elasticsearch 5.x intrusion info in index.rst (#514)
for issue https://github.com/elastic/elasticsearch-py/issues/492
2017-01-10 09:50:18 +01:00
Yotam TanayandHonza Král a644af0e07 Fix error where address is null in _get_host_info (#506)
* Fix error where address is null in _get_host_info

* Bump CI
2017-01-03 15:12:15 +01:00
Honza Král 8107bbcf93 update headers in lowercase to allow overriding of auth headers 2017-01-03 15:07:28 +01:00
Honza Král 75bc46f506 add unauthorized catch code for common test runner 2017-01-03 15:07:28 +01:00
Honza Král b9ba12977f Refactored connection error handling 2016-12-15 16:46:05 +11:00
Honza Král c2fc5aa18c Be more explicit with stats_only behavior 2016-12-15 10:21:38 +11:00
Honza Král ef3820c4fe use text/keyword instead of string in load example 2016-12-12 14:10:06 +11:00
Javier AguirreandHonza Král 8e69b98844 Fix UnboundLocalError response (#476) (#481) 2016-12-06 18:10:44 +01:00
Honza Král 4812980479 Merge PR #497, thanks veatch! 2016-12-06 12:33:35 +01:00
Honza Král fe53ab588e Fix sniffing for 5.0
Fixes #477
2016-12-06 12:17:57 +01:00
Jason Veatch 7fbcfc2c71 Cleaned up indentation in script block 2016-12-01 16:29:05 -05:00
Jason Veatch d4748c98c3 Restore fixed update example in example/load.py
Earlier I'd misread the Scripting and Security page about inline scripts
- turns out they are enabled by default. Earlier failures I saw were due
  to old groovy syntax. Things are working after changing to painless.
2016-12-01 16:03:46 -05:00
Jason VeatchandJason Veatch b46f7e74e6 Querying with parent id has changed in ES v5 2016-12-01 14:17:44 -05:00
Jason VeatchandJason Veatch 3401b21f1c Updated example queries for ES v5 2016-12-01 13:47:08 -05:00
Jason VeatchandJason Veatch cdd1fbf6a0 Scripted updates are disabled by default in ES v5 2016-12-01 13:45:21 -05:00
Jason VeatchandJason Veatch 9ac70f3dcb Fix UnicodeDecodeError 2016-12-01 12:08:39 -05:00
Glen SmithandGitHub 594f194bff Merge pull request #478 from ShaneHarvey/test-elastic-5-java-8
Travis: install Java 8 and test against Elasticsearch 5.0.0
2016-12-01 09:29:57 -07:00
Jason VeatchandJason Veatch 5f199882d8 multi_field type is removed from 5, move to new format
Instead of ignoring all 400s when creating the index, only ignore
index_already_exists_exception and raise other exceptions.
2016-12-01 10:51:04 -05:00
Noah WoodandHonza Král 2bc407e633 add '_mapping' to exists_type path (#473)
This adds ‘_mapping’ to the path for exists_type to avoid the
deprecation warning consistent with:
https://github.com/elastic/elasticsearch/commit/a4ea7e72234b480f89e8c59b
b12061be11172db1
2016-11-30 12:53:56 +01:00
Honza Král 1780214a19 explain localhost:9200 in the trace log 2016-11-13 05:52:59 +01:00
Shane Harvey 91787fcf66 Test Elasticsearch-5.0.0 on Travis 2016-11-02 20:31:09 -07:00