Client helpers (#1107)

* Added client helpers

* Updated test

* The search helper should return only the documents

* Added code comments

* Fixed bug

* Updated test

* Removed bulkSize and added flushBytes

* Updated test

* Added concurrency

* Updated test

* Added support for 429 handling in the scroll search helper

* Updated test

* Updated stats count

* Updated test

* Fix test

* Use client maxRetries as default

* Updated type definitions

* Refactored bulk helper to be more consistent with the client api

* Updated test

* Improved error handling, added refreshOnCompletion option and forward additinal options to the bulk api

* Updated type definitions

* Updated test

* Fixed test on Node v8

* Updated test

* Added TODO

* Updated docs

* Added Node v8 note

* Updated scripts

* Removed useless files

* Added helpers to integration test

* Fix cli argument position

* Moar fixes

* Test run elasticsearch in github actions

* Use master action version

* Add vm.max_map_count step

* Test new action setup

* Added Configure sysctl limits step

* Updated action to latest version

* Don't run helpers integration test in jenkins

* Run helpers integratino test also with Node v10

* Updated docs

* Updated docs

* Updated helpers type definitions

* Added test for helpers type definitions

* Added license header
This commit is contained in:
Tomas Della Vedova
2020-03-23 17:43:10 +01:00
committed by GitHub
parent 6c82a4967e
commit d7836a16af
24 changed files with 7654 additions and 11 deletions

View File

@ -0,0 +1,189 @@
// Licensed to Elasticsearch B.V under one or more agreements.
// Elasticsearch B.V licenses this file to you under the Apache 2.0 License.
// See the LICENSE file in the project root for more information
'use strict'
const { createReadStream } = require('fs')
const { join } = require('path')
const split = require('split2')
const { test, beforeEach, afterEach } = require('tap')
const { waitCluster } = require('../../utils')
const { Client } = require('../../../')
const datasetPath = join(__dirname, '..', '..', 'fixtures', 'stackoverflow.ndjson')
const INDEX = `test-helpers-${process.pid}`
const client = new Client({
node: process.env.TEST_ES_SERVER || 'http://localhost:9200'
})
beforeEach(async () => {
await waitCluster(client)
await client.indices.create({ index: INDEX })
})
afterEach(async () => {
await client.indices.delete({ index: INDEX }, { ignore: 404 })
})
test('bulk index', async t => {
const stream = createReadStream(datasetPath)
const result = await client.helpers.bulk({
datasource: stream.pipe(split()),
refreshOnCompletion: INDEX,
onDrop (doc) {
t.fail('It should not drop any document')
},
onDocument (doc) {
return {
index: { _index: INDEX }
}
}
})
t.type(result.time, 'number')
t.type(result.bytes, 'number')
t.match(result, {
total: 5000,
successful: 5000,
retry: 0,
failed: 0,
aborted: false
})
const { body } = await client.count({ index: INDEX })
t.match(body, { count: 5000 })
})
test('bulk index with custom id', async t => {
const stream = createReadStream(datasetPath)
const result = await client.helpers.bulk({
datasource: stream.pipe(split(JSON.parse)),
onDrop (doc) {
t.fail('It should not drop any document')
},
onDocument (doc) {
return {
index: {
_index: INDEX,
_id: doc.id
}
}
}
})
t.type(result.time, 'number')
t.type(result.bytes, 'number')
t.match(result, {
total: 5000,
successful: 5000,
retry: 0,
failed: 0,
aborted: false
})
const { body } = await client.get({
index: INDEX,
id: '19273860' // id of document n° 4242
})
t.strictEqual(body._index, INDEX)
t.strictEqual(body._id, '19273860')
t.strictEqual(body._source.id, '19273860')
})
test('abort the operation on document drop', async t => {
const stream = createReadStream(datasetPath)
const b = client.helpers.bulk({
datasource: stream.pipe(split(JSON.parse)),
concurrency: 1,
onDrop (doc) {
t.strictEqual(doc.status, 400)
t.strictEqual(doc.error.type, 'mapper_parsing_exception')
t.strictEqual(doc.document.id, '45924372')
b.abort()
},
onDocument (doc) {
if (doc.id === '45924372') { // id of document n° 500
// this will break the mapping
doc.title = { foo: 'bar' }
}
return {
index: {
_index: INDEX,
_id: doc.id
}
}
}
})
const result = await b
t.type(result.time, 'number')
t.type(result.bytes, 'number')
t.strictEqual(result.total - 1, result.successful)
t.match(result, {
retry: 0,
failed: 1,
aborted: true
})
})
test('bulk delete', async t => {
const indexResult = await client.helpers.bulk({
datasource: createReadStream(datasetPath).pipe(split(JSON.parse)),
refreshOnCompletion: true,
onDrop (doc) {
t.fail('It should not drop any document')
},
onDocument (doc) {
return {
index: {
_index: INDEX,
_id: doc.id
}
}
}
})
t.type(indexResult.time, 'number')
t.type(indexResult.bytes, 'number')
t.match(indexResult, {
total: 5000,
successful: 5000,
retry: 0,
failed: 0,
aborted: false
})
const { body: afterIndex } = await client.count({ index: INDEX })
t.match(afterIndex, { count: 5000 })
const deleteResult = await client.helpers.bulk({
datasource: createReadStream(datasetPath).pipe(split(JSON.parse)),
refreshOnCompletion: true,
onDrop (doc) {
t.fail('It should not drop any document')
},
onDocument (doc) {
return {
delete: {
_index: INDEX,
_id: doc.id
}
}
}
})
t.type(deleteResult.time, 'number')
t.type(deleteResult.bytes, 'number')
t.match(deleteResult, {
total: 5000,
successful: 5000,
retry: 0,
failed: 0,
aborted: false
})
const { body: afterDelete } = await client.count({ index: INDEX })
t.match(afterDelete, { count: 0 })
})

View File

@ -0,0 +1,103 @@
// Licensed to Elasticsearch B.V under one or more agreements.
// Elasticsearch B.V licenses this file to you under the Apache 2.0 License.
// See the LICENSE file in the project root for more information
'use strict'
const { createReadStream } = require('fs')
const { join } = require('path')
const split = require('split2')
const { test, beforeEach, afterEach } = require('tap')
const { waitCluster } = require('../../utils')
const { Client } = require('../../../')
const INDEX = `test-helpers-${process.pid}`
const client = new Client({
node: process.env.TEST_ES_SERVER || 'http://localhost:9200'
})
beforeEach(async () => {
await waitCluster(client)
await client.indices.create({ index: INDEX })
const stream = createReadStream(join(__dirname, '..', '..', 'fixtures', 'stackoverflow.ndjson'))
const result = await client.helpers.bulk({
datasource: stream.pipe(split()),
refreshOnCompletion: true,
onDocument (doc) {
return {
index: { _index: INDEX }
}
}
})
if (result.failed > 0) {
throw new Error('Failed bulk indexing docs')
}
})
afterEach(async () => {
await client.indices.delete({ index: INDEX }, { ignore: 404 })
})
test('search helper', async t => {
const scrollSearch = client.helpers.scrollSearch({
index: INDEX,
body: {
query: {
match: {
title: 'javascript'
}
}
}
})
var count = 0
for await (const search of scrollSearch) {
count += 1
for (const doc of search.documents) {
t.true(doc.title.toLowerCase().includes('javascript'))
}
}
t.strictEqual(count, 11)
})
test('clear a scroll search', async t => {
const scrollSearch = client.helpers.scrollSearch({
index: INDEX,
body: {
query: {
match: {
title: 'javascript'
}
}
}
})
var count = 0
for await (const search of scrollSearch) {
count += 1
if (count === 2) {
search.clear()
}
}
t.strictEqual(count, 2)
})
test('scroll documents', async t => {
const scrollSearch = client.helpers.scrollDocuments({
index: INDEX,
body: {
query: {
match: {
title: 'javascript'
}
}
}
})
var count = 0
for await (const doc of scrollSearch) {
count += 1
t.true(doc.title.toLowerCase().includes('javascript'))
}
t.strictEqual(count, 106)
})

View File

@ -0,0 +1,56 @@
// Licensed to Elasticsearch B.V under one or more agreements.
// Elasticsearch B.V licenses this file to you under the Apache 2.0 License.
// See the LICENSE file in the project root for more information
'use strict'
const { createReadStream } = require('fs')
const { join } = require('path')
const split = require('split2')
const { test, beforeEach, afterEach } = require('tap')
const { waitCluster } = require('../../utils')
const { Client } = require('../../../')
const INDEX = `test-helpers-${process.pid}`
const client = new Client({
node: process.env.TEST_ES_SERVER || 'http://localhost:9200'
})
beforeEach(async () => {
await waitCluster(client)
await client.indices.create({ index: INDEX })
const stream = createReadStream(join(__dirname, '..', '..', 'fixtures', 'stackoverflow.ndjson'))
const result = await client.helpers.bulk({
datasource: stream.pipe(split()),
refreshOnCompletion: true,
onDocument (doc) {
return {
index: { _index: INDEX }
}
}
})
if (result.failed > 0) {
throw new Error('Failed bulk indexing docs')
}
})
afterEach(async () => {
await client.indices.delete({ index: INDEX }, { ignore: 404 })
})
test('search helper', async t => {
const results = await client.helpers.search({
index: INDEX,
body: {
query: {
match: {
title: 'javascript'
}
}
}
})
t.strictEqual(results.length, 10)
for (const result of results) {
t.true(result.title.toLowerCase().includes('javascript'))
}
})