change to create a PR

.gitignore

@@ -127,3 +127,7 @@ build
 .vscode
 
 wasm
+
+# benchmark results
+results.ndjson
+*.lock_core

.gitlab-ci.yml

@@ -0,0 +1,6 @@
+# Add a new stage called “benchmarks”
+stages:
+  - benchmarks
+
+  # … and include a file that specifies benchmarks CI jobs
+include: ".gitlab/benchmarks.yml"

.gitlab/benchmarks.yml

@@ -0,0 +1,33 @@
+variables:
+  BASE_CI_IMAGE: 486234852809.dkr.ecr.us-east-1.amazonaws.com/ci/benchmarking-platform:dd-wasm-js-rewriter
+
+benchmarks:
+  stage: benchmarks
+  when: on_success
+  tags: ["runner:apm-k8s-tweaked-metal"]
+  image: $BASE_CI_IMAGE
+  interruptible: true
+  timeout: 1h
+  script:
+    - export ARTIFACTS_DIR="$(pwd)/artifacts" && (mkdir "${ARTIFACTS_DIR}" || :)
+    - git config --global url."https://gitlab-ci-token:${CI_JOB_TOKEN}@gitlab.ddbuild.io/DataDog/".insteadOf "https://github.com/DataDog/"
+    - git clone --branch dd-wasm-js-rewriter https://github.com/DataDog/benchmarking-platform platform && cd platform
+    - ./steps/capture-hardware-software-info.sh
+    - ./steps/run-benchmarks.sh
+    - ./steps/analyze-results.sh
+    - "./steps/upload-results-to-s3.sh || :"
+    - "./steps/post-pr-comment.sh || :"
+  artifacts:
+    name: "artifacts"
+    when: always
+    paths:
+      - artifacts/
+    expire_in: 3 months
+  variables:
+    KUBERNETES_SERVICE_ACCOUNT_OVERWRITE: dd-wasm-js-rewriter
+    FF_USE_LEGACY_KUBERNETES_EXECUTION_STRATEGY: "true"
+    UPSTREAM_BRANCH: $CI_COMMIT_REF_NAME # The branch or tag name for which project is built.
+    UPSTREAM_COMMIT_SHA: $CI_COMMIT_SHA # The commit revision the project is built for.
+    UPSTREAM_PROJECT_ID: $CI_PROJECT_ID # The ID of the current project. This ID is unique across all projects on the GitLab instance.
+    UPSTREAM_PROJECT_NAME: $CI_PROJECT_NAME # "dd-wasm-js-rewriter"
+    MAJOR_VERSION: 22

benchmark/sirun/get-variants.js

@@ -0,0 +1,9 @@
+#!/usr/bin/env node
+
+'use strict'
+
+const path = require('path')
+const metaJson = require(path.join(process.cwd(), 'meta.json'))
+const variants = (metaJson.variants && Object.keys(metaJson.variants)) || []
+
+process.stdout.write(variants.join(' '))

benchmark/sirun/initialize/init.js

@@ -0,0 +1,172 @@
+const configs = [{
+  csiMethods: [
+    { src: 'concat' },
+    { src: 'join' },
+    { src: 'parse' },
+    { src: 'plusOperator', operator: true },
+    { src: 'random' },
+    { src: 'replace' },
+    { src: 'slice' },
+    { src: 'substr' },
+    { src: 'substring' },
+    { src: 'toLowerCase', dst: 'stringCase' },
+    { src: 'toUpperCase', dst: 'stringCase' },
+    { src: 'tplOperator', operator: true },
+    { src: 'trim' },
+    { src: 'trimEnd' },
+    { src: 'trimStart', dst: 'trim' },
+    { src: 'eval', allowedWithoutCallee: true }
+  ],
+  orchestrion: `version: 1
+dc_module: dc-polyfill
+instrumentations:
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/runnables/base.js
+    function_query:
+      name: invoke
+      type: method
+      kind: async
+      class: RunnableSequence
+    operator: tracePromise
+    channel_name: "RunnableSequence_invoke"
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/runnables/base.js
+    function_query:
+      name: batch
+      type: method
+      kind: async
+      class: RunnableSequence
+    operator: tracePromise
+    channel_name: "RunnableSequence_batch"
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/language_models/chat_models.js
+    function_query:
+      name: generate
+      type: method
+      kind: async
+      class: BaseChatModel
+    operator: tracePromise
+    channel_name: "BaseChatModel_generate"
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/language_models/llms.js
+    function_query:
+      name: generate
+      type: method
+      kind: async
+    operator: tracePromise
+    channel_name: "BaseLLM_generate"
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/embeddings.js
+    function_query:
+      name: constructor
+      type: method
+      kind: sync
+      class: Embeddings
+    operator: traceSync
+    channel_name: "Embeddings_constructor"`
+}, {
+  csiMethods: [
+    { src: 'concat' },
+    { src: 'join' },
+    { src: 'parse' },
+    { src: 'plusOperator', operator: true },
+    { src: 'random' }
+  ],
+  orchestrion: `version: 1
+dc_module: dc-polyfill
+instrumentations:
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/runnables/base.js
+    function_query:
+      name: invoke
+      type: method
+      kind: async
+      class: RunnableSequence
+    operator: tracePromise
+    channel_name: "RunnableSequence_invoke"
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/language_models/chat_models.js
+    function_query:
+      name: generate
+      type: method
+      kind: async
+      class: BaseChatModel
+    operator: tracePromise
+    channel_name: "BaseChatModel_generate"`
+}, {
+  csiMethods: [
+    { src: 'slice' },
+    { src: 'substr' },
+    { src: 'substring' },
+    { src: 'toLowerCase', dst: 'stringCase' },
+    { src: 'toUpperCase', dst: 'stringCase' }
+  ],
+  orchestrion: `version: 1
+dc_module: dc-polyfill
+instrumentations:
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/language_models/llms.js
+    function_query:
+      name: generate
+      type: method
+      kind: async
+    operator: tracePromise
+    channel_name: "BaseLLM_generate"
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/embeddings.js
+    function_query:
+      name: constructor
+      type: method
+      kind: sync
+      class: Embeddings
+    operator: traceSync
+    channel_name: "Embeddings_constructor"`
+}, {
+  csiMethods: [
+    { src: 'trim' },
+    { src: 'trimEnd' },
+    { src: 'trimStart', dst: 'trim' },
+    { src: 'eval', allowedWithoutCallee: true }
+  ],
+  orchestrion: `version: 1
+dc_module: dc-polyfill
+instrumentations:
+  - module_name: "@langchain/core"
+    version_range: ">=0.1.0"
+    file_path: dist/runnables/base.js
+    function_query:
+      name: batch
+      type: method
+      kind: async
+      class: RunnableSequence
+    operator: tracePromise
+    channel_name: "RunnableSequence_batch"`
+}]
+
+let count = 0
+
+const { Rewriter } = require('../../../main')
+
+for (let i = 0; i < 1e4; i++) {
+  for (const config of configs) {
+    const instance = new Rewriter({
+      csiMethods: config.csiMethods,
+      orchestrion: config.orchestrion
+    })
+    count += instance.orchestrionModules ? 1 : 0
+  }
+}
+
+if (count === 0) {
+  // eslint-disable-next-line no-console
+  console.log('This cannot happen, but count is used to avoid optimizations')
+}

benchmark/sirun/initialize/meta.json

@@ -0,0 +1,13 @@
+{
+  "name": "initialize",
+  "cachegrind": false,
+  "instructions": true,
+  "iterations": 100,
+  "timeout": 15,
+  "variants": {
+    "main": {
+      "run": "node ./init.js",
+      "run_with_affinity": "bash -c \"taskset -c $CPU_AFFINITY node ./init.js\""
+    }
+  }
+}

benchmark/sirun/rewrite/.gitignore

@@ -0,0 +1 @@
+!node_modules

benchmark/sirun/rewrite/meta.json

@@ -0,0 +1,13 @@
+{
+  "name": "rewrite",
+  "cachegrind": false,
+  "instructions": true,
+  "iterations": 100,
+  "timeout": 15,
+  "variants": {
+    "iast+orchestrion": {
+      "run": "node rewrite.js",
+      "run_with_affinity": "bash -c \"taskset -c $CPU_AFFINITY node rewrite.js\""
+    }
+  }
+}

benchmark/sirun/rewrite/orchestrion.yml

@@ -0,0 +1,12 @@
+version: 1
+dc_module: dc-polyfill
+instrumentations:
+  - module_name: "test"
+    version_range: ">=0.1.0"
+    file_path: file.js
+    function_query:
+      name: methodToBeRewritten
+      type: method
+      kind: async
+    operator: tracePromise
+    channel_name: "methodToBeRewritten_invoke"

benchmark/sirun/rewrite/rewrite.js

@@ -0,0 +1,30 @@
+'use strict'
+
+const fs = require('fs')
+const path = require('path')
+
+const { Rewriter } = require('../../../main')
+
+const csiMethods = [
+  { src: 'concat' },
+  { src: 'join' },
+  { src: 'parse' },
+  { src: 'plusOperator', operator: true }
+]
+const filename = path.join(__dirname, 'node_modules', 'test', 'file.js')
+const code = fs.readFileSync(filename, 'utf8').toString()
+const orchestrion = fs.readFileSync(path.join(__dirname, 'orchestrion.yml'), 'utf8').toString()
+
+const rewriter = new Rewriter({ csiMethods, orchestrion })
+
+let count = 0
+
+for (let i = 0; i < 1e4; i++) {
+  const rewritten = rewriter.rewrite(code, filename, ['iast', 'orchestrion'])
+  count += rewritten.content ? 1 : 0
+}
+
+if (count === 0) {
+  // eslint-disable-next-line no-console
+  console.log('This cannot happen, but count is used to avoid optimizations')
+}

benchmark/sirun/run-one-variant.js

@@ -0,0 +1,9 @@
+#!/usr/bin/env node
+
+'use strict'
+
+const { exec, stdio } = require('./run-util')
+
+const env = Object.assign({}, process.env)
+
+exec('sirun', ['meta-temp.json'], { env, stdio })

benchmark/sirun/run-util.js

@@ -0,0 +1,42 @@
+'use strict'
+
+const childProcess = require('child_process')
+const readline = require('readline')
+
+function exec (...args) {
+  return new Promise((resolve, reject) => {
+    const proc = childProcess.spawn(...args)
+    streamAddVersion(proc.stdout)
+    proc.on('error', (e) => {
+      reject(e)
+    })
+    proc.on('exit', (code) => {
+      if (code === 0) {
+        resolve()
+      } else {
+        reject(new Error('Process exited with non-zero code. - ' + code))
+      }
+    })
+  })
+}
+
+function streamAddVersion (input) {
+  input.rl = readline.createInterface({ input })
+  input.rl.on('line', function (line) {
+    try {
+      const json = JSON.parse(line.toString())
+      json.nodeVersion = process.versions.node
+      // eslint-disable-next-line no-console
+      console.log(JSON.stringify(json))
+    } catch (e) {
+      // eslint-disable-next-line no-console
+      console.log(line)
+    }
+  })
+}
+
+module.exports = {
+  exec,
+  stdio: ['inherit', 'pipe', 'inherit'],
+  streamAddVersion
+}