[Other] Refactor js submodule (#415)

* Refactor js submodule

* Remove change-log

* Update ocr module

* Update ocr-detection module

* Update ocr-detection module

* Remove change-log

examples/application/js/package/packages/paddlejs-models/gesture/gulpfile.ts

@@ -0,0 +1,160 @@
+import path from 'path'
+import fse from 'fs-extra'
+import { series } from "gulp"
+import { paths, log } from "./build_package/util"
+import rollupConfig from './rollup.config'
+import { rollup } from 'rollup'
+import {
+  Extractor,
+  ExtractorConfig,
+  ExtractorResult,
+} from '@microsoft/api-extractor'
+/**
+ * 这里是由于 'conventional-changelog' 未提供类型文件
+ */
+// eslint-disable-next-line @typescript-eslint/ban-ts-comment
+// @ts-ignore
+import conventionalChangelog from 'conventional-changelog'
+
+interface TaskFunc {
+    // eslint-disable-next-line @typescript-eslint/ban-types
+    (cb: Function): void
+}
+
+const CHANGE_TRACE = ['paddlejs-models/gesture', 'paddle-js-models/gesture', 'paddlejs-models', 'paddle-js-models', 'all']
+
+/**
+ * 删除 lib 文件
+ * @param cb
+ * @returns {Promise<void>}
+ */
+const clearLibFile: TaskFunc = async (cb) => {
+    fse.removeSync(paths.lib)
+    log.progress('Deleted lib file')
+    cb()
+}
+
+/**
+ * rollup 打包
+ * @param cb
+ */
+const buildByRollup: TaskFunc = async (cb) => {
+    const inputOptions = {
+        input: rollupConfig.input,
+        external: rollupConfig.external,
+        plugins: rollupConfig.plugins,
+    }
+    const outOptions = rollupConfig.output
+    const bundle = await rollup(inputOptions)
+
+    // 写入需要遍历输出配置
+    if (Array.isArray(outOptions)) {
+        for (const outOption of outOptions) {
+            await bundle.write(outOption)
+        }
+        cb()
+        log.progress('Rollup built successfully')
+    }
+}
+
+/**
+ * api-extractor 整理 .d.ts 文件
+ * @param cb
+ */
+const apiExtractorGenerate: TaskFunc = async (cb) => {
+    const apiExtractorJsonPath: string = path.join(__dirname, './api-extractor.json')
+    // 加载并解析 api-extractor.json 文件
+    const extractorConfig: ExtractorConfig = await ExtractorConfig.loadFileAndPrepare(apiExtractorJsonPath)
+    // 判断是否存在 index.d.ts 文件，这里必须异步先访问一边，不然后面找不到会报错
+    const isdtxExist: boolean = await fse.pathExists(extractorConfig.mainEntryPointFilePath)
+    // 判断是否存在 etc 目录，api-extractor需要该目录存在
+    const isEtcExist: boolean = await fse.pathExists('./etc')
+
+    if (!isdtxExist) {
+        log.error('API Extractor not find index.d.ts')
+        return
+    }
+
+    if (!isEtcExist) {
+        fse.mkdirSync('etc');
+        log.progress('Create folder etc for API Extractor')
+    }
+
+    // 调用 API
+    const extractorResult: ExtractorResult = await Extractor.invoke(extractorConfig, {
+        localBuild: true,
+        // 在输出中显示信息
+        showVerboseMessages: true,
+    })
+
+    if (extractorResult.succeeded) {
+        // 删除多余的 .d.ts 文件
+        const libFiles: string[] = await fse.readdir(paths.lib)
+        for (const file of libFiles) {
+            if (file.endsWith('.d.ts') && !file.includes('index')) {
+                await fse.remove(path.join(paths.lib, file))
+            }
+        }
+        log.progress('API Extractor completed successfully')
+        // api-extractor 会生成 temp 文件夹，完成后进行删除
+        fse.ensureDirSync('temp')
+        fse.removeSync('temp')
+        cb()
+    } else {
+        log.error(`API Extractor completed with ${extractorResult.errorCount} errors`
+            + ` and ${extractorResult.warningCount} warnings`)
+    }
+}
+
+/**
+ * 完成
+ * @param cb
+ */
+const complete: TaskFunc = (cb) => {
+    log.progress('---- end ----')
+    cb()
+}
+
+/**
+ * 生成 CHANGELOG
+ * @param cb
+ */
+export const changelog: TaskFunc = async (cb) => {
+    const checkTrace = (chunk: string) => {
+        for (const keyWord of CHANGE_TRACE) {
+            if (chunk.includes(keyWord)) {
+                return true
+            }
+        }
+        return false
+    }
+    const changelogPath: string = path.join(paths.root, 'CHANGELOG.md')
+    // 对命令 conventional-changelog -p angular -i CHANGELOG.md -w -r 0
+    const changelogPipe = await conventionalChangelog({
+        preset: 'angular',
+        releaseCount: 0,
+    })
+    changelogPipe.setEncoding('utf8')
+
+    const resultArray = ['# 更新日志\n\n']
+    changelogPipe.on('data', (chunk) => {
+        // 原来的 commits 路径是进入提交列表
+        chunk = chunk.replace(/\/commits\//g, '/commit/')
+        /**
+         * title 或 指定跟踪 才会写入CHANGELOG
+         */
+        for (const log of chunk.split("\n")) {
+            if (log.includes('# ') || log.includes('### ') || checkTrace(log)) {
+                resultArray.push(log+"\n\n")
+            }
+        }
+    })
+    changelogPipe.on('end', async () => {
+        fse.createWriteStream(changelogPath).write(resultArray.join(''))
+        cb()
+        log.progress('CHANGELOG generation completed')
+    })
+}
+
+
+exports.build = series(clearLibFile, buildByRollup, apiExtractorGenerate, complete)