actions/opensearch-pyd
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

Merge PR #497, thanks veatch!

Browse Source
This commit is contained in:
Honza Král
2016-12-06 12:33:35 +01:00
parent fe53ab588e 7fbcfc2c71
commit 4812980479
2 changed files with 78 additions and 84 deletions
Download Patch File Download Diff File Expand all files Collapse all files
example/load.py
+67 -58
View File
@@ -10,6 +10,7 @@ import logging
import git
from elasticsearch import Elasticsearch
from elasticsearch.exceptions import TransportError
from elasticsearch.helpers import bulk, streaming_bulk
def create_git_index(client, index):
@@ -17,69 +18,75 @@ def create_git_index(client, index):
user_mapping = {
'properties': {
'name': {
'type': 'multi_field',
'type': 'string',
'fields': {
'raw': {'type' : 'string', 'index' : 'not_analyzed'},
'name': {'type' : 'string'}
}
}
}
}
create_index_body = {
'settings': {
# just one shard, no replicas for testing
'number_of_shards': 1,
'number_of_replicas': 0,
# custom analyzer for analyzing file paths
'analysis': {
'analyzer': {
'file_path': {
'type': 'custom',
'tokenizer': 'path_hierarchy',
'filter': ['lowercase']
}
}
}
},
'mappings': {
'commits': {
'_parent': {
'type': 'repos'
},
'properties': {
'author': user_mapping,
'authored_date': {'type': 'date'},
'committer': user_mapping,
'committed_date': {'type': 'date'},
'parent_shas': {'type': 'string', 'index' : 'not_analyzed'},
'description': {'type': 'string', 'analyzer': 'snowball'},
'files': {'type': 'string', 'analyzer': 'file_path'}
}
},
'repos': {
'properties': {
'owner': user_mapping,
'created_at': {'type': 'date'},
'description': {
'type': 'string',
'analyzer': 'snowball',
},
'tags': {
'type': 'string',
'index': 'not_analyzed'
}
}
}
}
}
# create empty index
client.indices.create(
index=index,
body={
'settings': {
# just one shard, no replicas for testing
'number_of_shards': 1,
'number_of_replicas': 0,
# custom analyzer for analyzing file paths
'analysis': {
'analyzer': {
'file_path': {
'type': 'custom',
'tokenizer': 'path_hierarchy',
'filter': ['lowercase']
}
}
}
},
'mappings': {
'commits': {
'_parent': {
'type': 'repos'
},
'properties': {
'author': user_mapping,
'authored_date': {'type': 'date'},
'committer': user_mapping,
'committed_date': {'type': 'date'},
'parent_shas': {'type': 'string', 'index' : 'not_analyzed'},
'description': {'type': 'string', 'analyzer': 'snowball'},
'files': {'type': 'string', 'analyzer': 'file_path'}
}
},
'repos': {
'properties': {
'owner': user_mapping,
'created_at': {'type': 'date'},
'description': {
'type': 'string',
'analyzer': 'snowball',
},
'tags': {
'type': 'string',
'index': 'not_analyzed'
}
}
}
}
},
try:
client.indices.create(
index=index,
body=create_index_body,
)
except TransportError as e:
# ignore already existing index
ignore=400
)
if e.error == 'index_already_exists_exception':
pass
else:
raise
def parse_commits(head, name):
"""
@@ -157,7 +164,7 @@ REPO_ACTIONS = [
},
{'_type': 'repos', '_id': 'elasticsearch-py', '_op_type': 'update', 'doc': {
'owner': {'name': 'Honza Král', 'email': '[email protected]'},
'owner': {'name': u'Honza Král', 'email': '[email protected]'},
'created_at': datetime(2013, 5, 1, 16, 37, 32),
'tags': ['elasticsearch', 'search', 'python', 'client'],
'description': 'For searching snakes.'}
@@ -190,9 +197,11 @@ if __name__ == '__main__':
doc_type='repos',
id='elasticsearch',
body={
"script" : "ctx._source.tags += tag",
"params" : {
"tag" : "java"
"script": {
"inline" : "ctx._source.tags.add(params.tag)",
"params" : {
"tag" : "java"
}
}
}
)