|
| 1 | +/*! |
| 2 | + * Copyright Amazon.com, Inc. or its affiliates. All Rights Reserved. |
| 3 | + * SPDX-License-Identifier: Apache-2.0 |
| 4 | + */ |
| 5 | +import * as vscode from 'vscode' |
| 6 | +import { existsSync, statSync, readdirSync } from 'fs' |
| 7 | +import { getLogger } from '../../../shared/logger' |
| 8 | +import * as CodeWhispererConstants from '../../models/constants' |
| 9 | +import { readFileAsString } from '../../../shared/filesystemUtilities' |
| 10 | +import { sleep } from '../../../shared/utilities/timeoutUtils' |
| 11 | +import { DependencyGraphConstants, DependencyGraph, Truncation } from './dependencyGraph' |
| 12 | +import path = require('path') |
| 13 | + |
| 14 | +export const importRegex = /(require|require_relative|load|include|extend)\s+('[^']+'|"[^"]+"|\w+)(\s+as\s+(\w+))?/gm |
| 15 | + |
| 16 | +export class RubyDependencyGraph extends DependencyGraph { |
| 17 | + // Payload Size for Ruby: 1MB |
| 18 | + getPayloadSizeLimitInBytes(): number { |
| 19 | + return CodeWhispererConstants.codeScanRubyPayloadSizeLimitBytes |
| 20 | + } |
| 21 | + override updateSysPaths(uri: vscode.Uri) { |
| 22 | + this.getDirPaths(uri).forEach(dirPath => { |
| 23 | + this._sysPaths.add(dirPath) |
| 24 | + }) |
| 25 | + } |
| 26 | + |
| 27 | + private generateFilePath(modulePath: string, dirPath: string) { |
| 28 | + const filePath = path.join(dirPath, modulePath + DependencyGraphConstants.rubyExt) |
| 29 | + return existsSync(filePath) ? filePath : '' |
| 30 | + } |
| 31 | + |
| 32 | + //For Generating File Paths |
| 33 | + private generateFilePaths(modulePaths: string[], dirPaths: string[]) { |
| 34 | + return modulePaths |
| 35 | + .flatMap(modulePath => dirPaths.map(dirPath => this.generateFilePath(modulePath, dirPath))) |
| 36 | + .filter(filePath => filePath !== '') |
| 37 | + } |
| 38 | + |
| 39 | + //Generate the combinations for module paths |
| 40 | + private generateModulePaths(inputPath: string): string[] { |
| 41 | + const positionOfExt = inputPath.indexOf(DependencyGraphConstants.rubyExt) //To remove imports having .rb |
| 42 | + if (positionOfExt !== -1) { |
| 43 | + inputPath = inputPath.substring(0, positionOfExt).trim() |
| 44 | + } |
| 45 | + |
| 46 | + const inputPaths = inputPath.split('/') |
| 47 | + let outputPath = '' |
| 48 | + return inputPaths.map(pathSegment => { |
| 49 | + outputPath += (outputPath ? '/' : '') + pathSegment |
| 50 | + return path.join(...outputPath.split('/')) |
| 51 | + }) |
| 52 | + } |
| 53 | + |
| 54 | + private getModulePath(modulePathStr: string) { |
| 55 | + const pos = modulePathStr.indexOf(DependencyGraphConstants.as) |
| 56 | + if (pos !== -1) { |
| 57 | + modulePathStr = modulePathStr.substring(0, pos) |
| 58 | + } |
| 59 | + |
| 60 | + return this.generateModulePaths(modulePathStr.replace(/[",'\s()]/g, '').trim()) |
| 61 | + } |
| 62 | + |
| 63 | + private extractModulePaths(importStr: string) { |
| 64 | + let modulePaths: string[] = [] |
| 65 | + const { |
| 66 | + require: requireKeyword, |
| 67 | + require_relative: requireRelativeKeyword, |
| 68 | + include: includeKeyword, |
| 69 | + extend: extendKeyword, |
| 70 | + load: loadKeyword, |
| 71 | + } = DependencyGraphConstants |
| 72 | + |
| 73 | + let keyword: string | undefined |
| 74 | + let indexOfKeyword: number | undefined |
| 75 | + |
| 76 | + switch (true) { |
| 77 | + case importStr.startsWith(requireRelativeKeyword): |
| 78 | + keyword = requireRelativeKeyword |
| 79 | + indexOfKeyword = importStr.indexOf(requireRelativeKeyword) |
| 80 | + break |
| 81 | + case importStr.startsWith(requireKeyword): |
| 82 | + keyword = requireKeyword |
| 83 | + indexOfKeyword = importStr.indexOf(requireKeyword) |
| 84 | + break |
| 85 | + case importStr.startsWith(includeKeyword): |
| 86 | + keyword = includeKeyword |
| 87 | + indexOfKeyword = importStr.indexOf(includeKeyword) |
| 88 | + break |
| 89 | + case importStr.startsWith(extendKeyword): |
| 90 | + keyword = extendKeyword |
| 91 | + indexOfKeyword = importStr.indexOf(extendKeyword) |
| 92 | + break |
| 93 | + case importStr.startsWith(loadKeyword): |
| 94 | + keyword = loadKeyword |
| 95 | + indexOfKeyword = importStr.indexOf(loadKeyword) |
| 96 | + break |
| 97 | + default: |
| 98 | + break |
| 99 | + } |
| 100 | + |
| 101 | + if (keyword && indexOfKeyword !== -1 && keyword !== undefined && indexOfKeyword !== undefined) { |
| 102 | + const modulePathStr = importStr |
| 103 | + .substring(indexOfKeyword + keyword.length) |
| 104 | + .trim() |
| 105 | + .replace(/\s+/g, '') |
| 106 | + modulePaths = this.getModulePath(modulePathStr) |
| 107 | + } |
| 108 | + |
| 109 | + return modulePaths |
| 110 | + } |
| 111 | + |
| 112 | + override parseImport(importStr: string, dirPaths: string[]) { |
| 113 | + if (this._parsedStatements.has(importStr)) { |
| 114 | + return [] |
| 115 | + } |
| 116 | + |
| 117 | + this._parsedStatements.add(importStr) |
| 118 | + const modulePaths = this.extractModulePaths(importStr) |
| 119 | + const dependencies = this.generateFilePaths(modulePaths, dirPaths) |
| 120 | + return dependencies |
| 121 | + } |
| 122 | + |
| 123 | + override getDependencies(uri: vscode.Uri, imports: string[]) { |
| 124 | + const dependencies: string[] = [] |
| 125 | + imports.forEach(importStr => { |
| 126 | + this.updateSysPaths(uri) |
| 127 | + const importString = importStr.replace(';', '') |
| 128 | + const findings = this.parseImport(importString, Array.from(this._sysPaths.values())) |
| 129 | + const validSourceFiles = findings.filter(finding => !this._pickedSourceFiles.has(finding)) |
| 130 | + validSourceFiles.forEach(file => { |
| 131 | + if (existsSync(file) && !this.willReachSizeLimit(this._totalSize, statSync(file).size)) { |
| 132 | + dependencies.push(file) |
| 133 | + } |
| 134 | + }) |
| 135 | + }) |
| 136 | + return dependencies |
| 137 | + } |
| 138 | + private async readImports(content: string) { |
| 139 | + this._totalLines += content.split(DependencyGraphConstants.newlineRegex).length |
| 140 | + const regExp = new RegExp(importRegex) |
| 141 | + return content.match(regExp) ?? [] |
| 142 | + } |
| 143 | + |
| 144 | + override async searchDependency(uri: vscode.Uri): Promise<Set<string>> { |
| 145 | + const filePath = uri.fsPath |
| 146 | + const q: string[] = [] |
| 147 | + q.push(filePath) |
| 148 | + while (q.length > 0) { |
| 149 | + let count: number = q.length |
| 150 | + while (count > 0) { |
| 151 | + if (this.reachSizeLimit(this._totalSize)) { |
| 152 | + return this._pickedSourceFiles |
| 153 | + } |
| 154 | + count -= 1 |
| 155 | + const currentFilePath = q.shift() |
| 156 | + if (currentFilePath === undefined) { |
| 157 | + throw new Error('"undefined" is invalid for queued file.') |
| 158 | + } |
| 159 | + this._pickedSourceFiles.add(currentFilePath) |
| 160 | + this._totalSize += statSync(currentFilePath).size |
| 161 | + const uri = vscode.Uri.file(currentFilePath) |
| 162 | + const content: string = await readFileAsString(uri.fsPath) |
| 163 | + const imports = await this.readImports(content) |
| 164 | + const dependencies = this.getDependencies(uri, imports) |
| 165 | + dependencies.forEach(dependency => { |
| 166 | + q.push(dependency) |
| 167 | + }) |
| 168 | + } |
| 169 | + } |
| 170 | + return this._pickedSourceFiles |
| 171 | + } |
| 172 | + |
| 173 | + override async traverseDir(dirPath: string) { |
| 174 | + if (this.reachSizeLimit(this._totalSize)) { |
| 175 | + return |
| 176 | + } |
| 177 | + readdirSync(dirPath, { withFileTypes: true }).forEach(async file => { |
| 178 | + const absPath = path.join(dirPath, file.name) |
| 179 | + if (file.name.charAt(0) === '.' || !existsSync(absPath)) { |
| 180 | + return |
| 181 | + } |
| 182 | + if (file.isDirectory()) { |
| 183 | + await this.traverseDir(absPath) |
| 184 | + } else if (file.isFile()) { |
| 185 | + if ( |
| 186 | + file.name.endsWith(DependencyGraphConstants.rubyExt) && |
| 187 | + !this.reachSizeLimit(this._totalSize) && |
| 188 | + !this.willReachSizeLimit(this._totalSize, statSync(absPath).size) && |
| 189 | + !this._pickedSourceFiles.has(absPath) |
| 190 | + ) { |
| 191 | + await this.searchDependency(vscode.Uri.file(absPath)) |
| 192 | + } |
| 193 | + } |
| 194 | + }) |
| 195 | + } |
| 196 | + |
| 197 | + override async generateTruncation(uri: vscode.Uri): Promise<Truncation> { |
| 198 | + try { |
| 199 | + const workspaceFolder = vscode.workspace.getWorkspaceFolder(uri) |
| 200 | + if (workspaceFolder === undefined) { |
| 201 | + this._pickedSourceFiles.add(uri.fsPath) |
| 202 | + } else { |
| 203 | + await this.searchDependency(uri) |
| 204 | + await this.traverseDir(this.getProjectPath(uri)) |
| 205 | + } |
| 206 | + await sleep(1000) |
| 207 | + const truncDirPath = this.getTruncDirPath(uri) |
| 208 | + this.copyFilesToTmpDir(this._pickedSourceFiles, truncDirPath) |
| 209 | + const zipFilePath = this.zipDir(truncDirPath, CodeWhispererConstants.codeScanZipExt) |
| 210 | + const zipFileSize = statSync(zipFilePath).size |
| 211 | + return { |
| 212 | + rootDir: truncDirPath, |
| 213 | + zipFilePath: zipFilePath, |
| 214 | + scannedFiles: new Set(this._pickedSourceFiles), |
| 215 | + srcPayloadSizeInBytes: this._totalSize, |
| 216 | + zipFileSizeInBytes: zipFileSize, |
| 217 | + buildPayloadSizeInBytes: 0, |
| 218 | + lines: this._totalLines, |
| 219 | + } |
| 220 | + } catch (error) { |
| 221 | + getLogger().error(`${this._languageId} dependency graph error caused by:`, error) |
| 222 | + throw new Error(`${this._languageId} context processing failed.`) |
| 223 | + } |
| 224 | + } |
| 225 | + |
| 226 | + async getSourceDependencies(uri: vscode.Uri, content: string): Promise<string[]> { |
| 227 | + return [] |
| 228 | + } |
| 229 | + async getSamePackageFiles(uri: vscode.Uri, projectPath: string): Promise<string[]> { |
| 230 | + return [] |
| 231 | + } |
| 232 | + async isTestFile(content: string): Promise<boolean> { |
| 233 | + return false |
| 234 | + } |
| 235 | +} |
0 commit comments