feat(schema-compiler): Move transpiling to worker threads (under the …

.github/workflows/publish.yml

@@ -191,7 +191,7 @@ jobs:
           components: rustfmt
           target: ${{ matrix.target }}
       - name: Install Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         if: (matrix.python-version != 'fallback')
         with:
           python-version: ${{ matrix.python-version }}
@@ -265,7 +265,7 @@ jobs:
           rustflags: ""
           components: rustfmt
       - name: Install Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         if: (matrix.python-version != 'fallback')
         with:
           python-version: ${{ matrix.python-version }}
@@ -807,7 +807,7 @@ jobs:
       - name: Checkout
         uses: actions/checkout@v4
       - name: Trigger runtime
-        uses: actions/github-script@v6
+        uses: actions/github-script@v7
         with:
           github-token: ${{ secrets.GH_TRIGGER_TOKEN }}
           script: |

.github/workflows/push.yml

@@ -42,8 +42,11 @@ jobs:
       matrix:
         # Current docker version + next LTS
         node-version: [20.x, 22.x]
+        transpile-worker-threads: [false, true]
       fail-fast: false
 
+    env:
+      CUBEJS_TRANSPILATION_WORKER_THREADS: ${{ matrix.transpile-worker-threads }}
     steps:
       - id: get-tag-out
         run: echo "$OUT"

.github/workflows/rust-cubesql.yml

@@ -245,7 +245,7 @@ jobs:
           components: rustfmt
           target: ${{ matrix.target }}
       - name: Install Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         if: (matrix.python-version != 'fallback')
         with:
           python-version: ${{ matrix.python-version }}
@@ -327,7 +327,7 @@ jobs:
           rustflags: ""
           components: rustfmt
       - name: Install Python
-        uses: actions/setup-python@v4
+        uses: actions/setup-python@v5
         if: (matrix.python-version != 'fallback')
         with:
           python-version: ${{ matrix.python-version }}

.github/workflows/rust-cubestore-master.yml

@@ -149,7 +149,7 @@ jobs:
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3
       - name: Push to Docker Hub
-        uses: docker/build-push-action@v3
+        uses: docker/build-push-action@v6
         with:
           context: ./rust/cubestore
           file: ./rust/cubestore/Dockerfile

.github/workflows/rust-cubestore.yml

@@ -97,7 +97,7 @@ jobs:
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3
       - name: Build only
-        uses: docker/build-push-action@v3
+        uses: docker/build-push-action@v6
         with:
           context: ./rust/cubestore/
           file: ./rust/cubestore/Dockerfile

packages/cubejs-backend-shared/src/env.ts

@@ -193,6 +193,12 @@ const variables: Record<string, (...args: any) => any> = {
     .default('1')
     .asInt(),
   nativeSqlPlanner: () => get('CUBEJS_TESSERACT_SQL_PLANNER').asBool(),
+  transpilationWorkerThreads: () => get('CUBEJS_TRANSPILATION_WORKER_THREADS')
+    .default('false')
+    .asBoolStrict(),
+  transpilationWorkerThreadsCount: () => get('CUBEJS_TRANSPILATION_WORKER_THREADS_COUNT')
+    .default('0')
+    .asInt(),
 
   /** ****************************************************************
    * Common db options                                               *

packages/cubejs-schema-compiler/package.json

@@ -54,7 +54,8 @@
     "node-dijkstra": "^2.5.0",
     "ramda": "^0.27.2",
     "syntax-error": "^1.3.0",
-    "uuid": "^8.3.2"
+    "uuid": "^8.3.2",
+    "workerpool": "^9.2.0"
   },
   "devDependencies": {
     "@cubejs-backend/apla-clickhouse": "^1.7.0",

packages/cubejs-schema-compiler/src/compiler/CubeSymbols.js

@@ -8,7 +8,7 @@ import { camelizeCube } from './utils';
 import { BaseQuery } from '../adapter';
 
 const FunctionRegex = /function\s+\w+\(([A-Za-z0-9_,]*)|\(([\s\S]*?)\)\s*=>|\(?(\w+)\)?\s*=>/;
-const CONTEXT_SYMBOLS = {
+export const CONTEXT_SYMBOLS = {
   SECURITY_CONTEXT: 'securityContext',
   // SECURITY_CONTEXT has been deprecated, however security_context (lowecase)
   // is allowed in RBAC policies for query-time attribute matching
@@ -19,7 +19,7 @@ const CONTEXT_SYMBOLS = {
   SQL_UTILS: 'sqlUtils'
 };
 
-const CURRENT_CUBE_CONSTANTS = ['CUBE', 'TABLE'];
+export const CURRENT_CUBE_CONSTANTS = ['CUBE', 'TABLE'];
 
 export class CubeSymbols {
   constructor(evaluateViews) {

packages/cubejs-schema-compiler/src/compiler/DataSchemaCompiler.js

@@ -6,8 +6,9 @@ import { parse } from '@babel/parser';
 import babelGenerator from '@babel/generator';
 import babelTraverse from '@babel/traverse';
 import R from 'ramda';
+import workerpool from 'workerpool';
 
-import { isNativeSupported } from '@cubejs-backend/shared';
+import { getEnv, isNativeSupported } from '@cubejs-backend/shared';
 import { UserError } from './UserError';
 import { ErrorReporter } from './ErrorReporter';
 
@@ -26,6 +27,8 @@ export class DataSchemaCompiler {
     this.preTranspileCubeCompilers = options.preTranspileCubeCompilers || [];
     this.cubeNameCompilers = options.cubeNameCompilers || [];
     this.extensions = options.extensions || {};
+    this.cubeDictionary = options.cubeDictionary;
+    this.cubeSymbols = options.cubeSymbols;
     this.cubeFactory = options.cubeFactory;
     this.filesToCompile = options.filesToCompile;
     this.omitErrors = options.omitErrors;
@@ -38,6 +41,7 @@ export class DataSchemaCompiler {
     this.yamlCompiler = options.yamlCompiler;
     this.yamlCompiler.dataSchemaCompiler = this;
     this.pythonContext = null;
+    this.workerPool = null;
   }
 
   compileObjects(compileServices, objects, errorsReport) {
@@ -87,17 +91,52 @@ export class DataSchemaCompiler {
     const errorsReport = new ErrorReporter(null, [], this.errorReport);
     this.errorsReport = errorsReport;
 
-    // TODO: required in order to get pre transpile compilation work
-    const transpile = () => toCompile.map(f => this.transpileFile(f, errorsReport)).filter(f => !!f);
+    if (getEnv('transpilationWorkerThreads')) {
+      const wc = getEnv('transpilationWorkerThreadsCount');
+      this.workerPool = workerpool.pool(
+        path.join(__dirname, 'transpilers/transpiler_worker'),
+        wc > 0 ? { maxWorkers: wc } : undefined,
+      );
+    }
+
+    const transpile = async () => {
+      let cubeNames;
+      let cubeSymbolsNames;
+
+      if (getEnv('transpilationWorkerThreads')) {
+        cubeNames = Object.keys(this.cubeDictionary.byId);
+        // We need only cubes and all its member names for transpiling.
+        // Cubes doesn't change during transpiling, but are changed during compilation phase,
+        // so we can prepare them once for every phase.
+        // Communication between main and worker threads uses
+        // The structured clone algorithm (@see https://developer.mozilla.org/en-US/docs/Web/API/Web_Workers_API/Structured_clone_algorithm)
+        // which doesn't allow passing any function objects, so we need to sanitize the symbols.
+        cubeSymbolsNames = Object.fromEntries(
+          Object.entries(this.cubeSymbols.symbols)
+            .map(
+              ([key, value]) => [key, Object.fromEntries(
+                Object.keys(value).map((k) => [k, true]),
+              )],
+            ),
+        );
+      }
+      const results = await Promise.all(toCompile.map(f => this.transpileFile(f, errorsReport, { cubeNames, cubeSymbolsNames })));
+      return results.filter(f => !!f);
+    };
 
-    const compilePhase = (compilers) => this.compileCubeFiles(compilers, transpile(), errorsReport);
+    const compilePhase = async (compilers) => this.compileCubeFiles(compilers, await transpile(), errorsReport);
 
     return compilePhase({ cubeCompilers: this.cubeNameCompilers })
       .then(() => compilePhase({ cubeCompilers: this.preTranspileCubeCompilers }))
       .then(() => compilePhase({
         cubeCompilers: this.cubeCompilers,
         contextCompilers: this.contextCompilers,
-      }));
+      }))
+      .then(() => {
+        if (this.workerPool) {
+          this.workerPool.terminate();
+        }
+      });
   }
 
   compile() {
@@ -113,7 +152,7 @@ export class DataSchemaCompiler {
     return this.compilePromise;
   }
 
-  transpileFile(file, errorsReport) {
+  async transpileFile(file, errorsReport, options) {
     if (R.endsWith('.jinja', file.fileName) ||
       (R.endsWith('.yml', file.fileName) || R.endsWith('.yaml', file.fileName))
       // TODO do Jinja syntax check with jinja compiler
@@ -132,31 +171,47 @@ export class DataSchemaCompiler {
     } else if (R.endsWith('.yml', file.fileName) || R.endsWith('.yaml', file.fileName)) {
       return file;
     } else if (R.endsWith('.js', file.fileName)) {
-      return this.transpileJsFile(file, errorsReport);
+      return this.transpileJsFile(file, errorsReport, options);
     } else {
       return file;
     }
   }
 
-  transpileJsFile(file, errorsReport) {
+  async transpileJsFile(file, errorsReport, { cubeNames, cubeSymbolsNames }) {
     try {
-      const ast = parse(
-        file.content,
-        {
-          sourceFilename: file.fileName,
-          sourceType: 'module',
-          plugins: ['objectRestSpread']
-        },
-      );
+      if (getEnv('transpilationWorkerThreads')) {
+        const data = {
+          fileName: file.fileName,
+          content: file.content,
+          transpilers: this.transpilers.map(t => t.constructor.name),
+          cubeNames,
+          cubeSymbolsNames,
+        };
+
+        const res = await this.workerPool.exec('transpile', [data]);
+        errorsReport.addErrors(res.errors);
+        errorsReport.addWarnings(res.warnings);
+
+        return Object.assign({}, file, { content: res.content });
+      } else {
+        const ast = parse(
+          file.content,
+          {
+            sourceFilename: file.fileName,
+            sourceType: 'module',
+            plugins: ['objectRestSpread'],
+          },
+        );
 
-      this.transpilers.forEach((t) => {
-        errorsReport.inFile(file);
-        babelTraverse(ast, t.traverseObject(errorsReport));
-        errorsReport.exitFile();
-      });
+        this.transpilers.forEach((t) => {
+          errorsReport.inFile(file);
+          babelTraverse(ast, t.traverseObject(errorsReport));
+          errorsReport.exitFile();
+        });
 
-      const content = babelGenerator(ast, {}, file.content).code;
-      return Object.assign({}, file, { content });
+        const content = babelGenerator(ast, {}, file.content).code;
+        return Object.assign({}, file, { content });
+      }
     } catch (e) {
       if (e.toString().indexOf('SyntaxError') !== -1) {
         const line = file.content.split('\n')[e.loc.line - 1];

packages/cubejs-schema-compiler/src/compiler/ErrorReporter.ts

@@ -137,6 +137,22 @@ export class ErrorReporter {
     }
   }
 
+  public getErrors() {
+    return this.rootReporter().errors;
+  }
+
+  public addErrors(errors: CompilerErrorInterface[]) {
+    this.rootReporter().errors.push(...errors);
+  }
+
+  public getWarnings() {
+    return this.rootReporter().warnings;
+  }
+
+  public addWarnings(warnings: SyntaxErrorInterface[]) {
+    this.rootReporter().warnings.push(...warnings);
+  }
+
   protected rootReporter(): ErrorReporter {
     return this.parent ? this.parent.rootReporter() : this;
   }

packages/cubejs-schema-compiler/src/compiler/PrepareCompiler.ts

@@ -64,6 +64,8 @@ export const prepareCompiler = (repo: SchemaFileRepository, options: PrepareComp
     contextCompilers: [contextEvaluator],
     cubeFactory: cubeSymbols.createCube.bind(cubeSymbols),
     compilerCache,
+    cubeDictionary,
+    cubeSymbols,
     extensions: {
       Funnels,
       RefreshKeys,

packages/cubejs-schema-compiler/src/compiler/transpilers/CubePropContextTranspiler.ts

@@ -2,7 +2,12 @@ import * as t from '@babel/types';
 import R from 'ramda';
 
 import type { NodePath } from '@babel/traverse';
-import type { TranspilerInterface, TraverseObject } from './transpiler.interface';
+import {
+  TranspilerCubeResolver,
+  TranspilerInterface,
+  TranspilerSymbolResolver,
+  TraverseObject
+} from './transpiler.interface';
 import type { CubeSymbols } from '../CubeSymbols';
 import type { CubeDictionary } from '../CubeDictionary';
 
@@ -39,8 +44,8 @@ transpiledFieldsPatterns?.forEach((r) => {
 
 export class CubePropContextTranspiler implements TranspilerInterface {
   public constructor(
-    protected readonly cubeSymbols: CubeSymbols,
-    protected readonly cubeDictionary: CubeDictionary,
+    protected readonly cubeSymbols: TranspilerSymbolResolver,
+    protected readonly cubeDictionary: TranspilerCubeResolver,
   ) {
   }
 

packages/cubejs-schema-compiler/src/compiler/transpilers/LightweightNodeCubeDictionary.ts

@@ -0,0 +1,14 @@
+import { TranspilerCubeResolver } from './transpiler.interface';
+
+export class LightweightNodeCubeDictionary implements TranspilerCubeResolver {
+  public constructor(private cubeNames: string[] = []) {
+  }
+
+  public resolveCube(name: string): boolean {
+    return this.cubeNames.includes(name);
+  }
+
+  public setCubeNames(cubeNames: string[]): void {
+    this.cubeNames = cubeNames;
+  }
+}

packages/cubejs-schema-compiler/src/compiler/transpilers/LightweightSymbolResolver.ts

@@ -0,0 +1,30 @@
+import { TranspilerSymbolResolver } from './transpiler.interface';
+import { CONTEXT_SYMBOLS, CURRENT_CUBE_CONSTANTS } from '../CubeSymbols';
+
+type CubeSymbols = Record<string, Record<string, boolean>>;
+
+export class LightweightSymbolResolver implements TranspilerSymbolResolver {
+  public constructor(private symbols: CubeSymbols = {}) {
+  }
+
+  public setSymbols(symbols: CubeSymbols) {
+    this.symbols = symbols;
+  }
+
+  public isCurrentCube(name): boolean {
+    return CURRENT_CUBE_CONSTANTS.indexOf(name) >= 0;
+  }
+
+  public resolveSymbol(cubeName, name): any {
+    if (name === 'USER_CONTEXT') {
+      throw new Error('Support for USER_CONTEXT was removed, please migrate to SECURITY_CONTEXT.');
+    }
+
+    if (CONTEXT_SYMBOLS[name]) {
+      return true;
+    }
+
+    const cube = this.symbols[this.isCurrentCube(name) ? cubeName : name];
+    return cube || (this.symbols[cubeName] && this.symbols[cubeName][name]);
+  }
+}

packages/cubejs-schema-compiler/src/compiler/transpilers/transpiler.interface.ts

@@ -6,3 +6,12 @@ export type TraverseObject = TraverseOptions;
 export interface TranspilerInterface {
   traverseObject(reporter: ErrorReporter): TraverseObject;
 }
+
+export interface TranspilerSymbolResolver {
+  resolveSymbol(cubeName, name): any;
+  isCurrentCube(name): boolean;
+}
+
+export interface TranspilerCubeResolver {
+  resolveCube(name): boolean;
+}