WDocx2html.mjs

import path from 'path'
import process from 'process'
import get from 'lodash-es/get.js'
import isestr from 'wsemi/src/isestr.mjs'
import isp0int from 'wsemi/src/isp0int.mjs'
import str2b64 from 'wsemi/src/str2b64.mjs'
import execProcess from 'wsemi/src/execProcess.mjs'
import getPathParent from 'wsemi/src/getPathParent.mjs'
import fsIsFile from 'wsemi/src/fsIsFile.mjs'
import fsIsFolder from 'wsemi/src/fsIsFolder.mjs'
import fsCreateFolder from 'wsemi/src/fsCreateFolder.mjs'
import ensureBins from './ensureBins.mjs'
import judgeConvertOutput, { judgeConvertFile } from './judgeConvertOutput.mjs'


//fnExe, docx2html執行檔名稱(須與bin/manifest.json之name一致)
let fnExe = 'docx2html.exe'


function isWindows() {
    return process.platform === 'win32'
}


/**
 * 以Microsoft Word將Docx檔轉為Html檔
 *
 * @param {String} fpIn 輸入來源Docx檔位置字串
 * @param {String} fpOut 輸入轉出Html檔位置字串
 * @param {Object} [opt={}] 輸入設定物件，預設{}
 * @param {Integer} [opt.execFontGrow=0] 輸入全文件放大字型次數整數，預設0
 * @returns {Promise} 回傳Promise，resolve回傳成功訊息'ok'，reject回傳錯誤訊息。reject值有下列型態：參數檢測失敗為說明字串；取不到執行檔為`[code] 訊息`字串(code取自ensureBins，如`[E_OFFLINE_MISSING] ...`)；docx2html.exe無法執行或離開碼非0為execProcess之訊息字串(如`code[N]:\n...`)；docx2html.exe執行後自報失敗(如輸入非docx、docx損毀、Word無法開啟或無法另存)為`[E_CONVERT_FAILED] 原因`，原因為exe輸出之診斷(已依系統字碼頁解碼，不回顯輸入與輸出路徑，但Word之診斷可能自行含檔名)；exe報成功而fpOut未產生或為空檔(如fpOut無副檔名，Word另存為`<fpOut>.htm`)亦為`[E_CONVERT_FAILED] output file was not generated or is empty`。docx2html.exe不論成敗離開碼皆為0，故以其輸出判定成敗：須有`success`行且無`error:`行才resolve，空輸出或未知輸出亦視為失敗；失敗(reject)時fpOut之內容不可採用，本套件不刪除fpOut既有之舊檔
 * @example
 *
 * import w from 'wsemi'
 * import WDocx2html from './src/WDocx2html.mjs'
 * //import WDocx2html from 'w-docx2html/src/WDocx2html.mjs'
 * //import WDocx2html from 'w-docx2html'
 *
 * async function test() {
 *
 *     let fpIn = `./test/docin.docx`
 *     let fpOut = `./test/docout.html`
 *     let opt = {
 *         execFontGrow: 1,
 *     }
 *
 *     let r = await WDocx2html(fpIn, fpOut, opt)
 *     console.log(r)
 *     // => ok
 *
 *     w.fsDeleteFile(fpOut)
 *
 * }
 * test()
 *     .catch((err) => {
 *         console.log('catch', err)
 *     })
 *
 */
async function WDocx2html(fpIn, fpOut, opt = {}) {
    let errTemp = null

    //isWindows
    if (!isWindows()) {
        return Promise.reject('operating system is not windows')
    }

    //check
    if (!fsIsFile(fpIn)) {
        return Promise.reject(`fpIn[${fpIn}] does not exist`)
    }

    //check
    if (!isestr(fpOut)) {
        return Promise.reject(`fpOut[${fpOut}] is not an effective string`)
    }

    //execFontGrow
    let execFontGrow = get(opt, 'execFontGrow')
    if (!isp0int(execFontGrow)) {
        execFontGrow = 0
    }

    //轉絕對路徑
    fpIn = path.resolve(fpIn)
    fpOut = path.resolve(fpOut)

    //fdOut
    let fdOut = getPathParent(fpOut)
    if (!fsIsFolder(fdOut)) {
        fsCreateFolder(fdOut)
    }

    //prog, 經ensureBins取得docx2html.exe: 依bin/manifest.json校驗, 首次使用時下載(或自鏡像取得), 之後自快取取得, 見README〈外部執行檔〉
    //本函數以字串reject(同w-docx-split、w-html2docx), 故將ensureBins帶code之Error轉為`[code] message`
    let prog = ''
    let errBin = null
    await ensureBins({ names: fnExe })
        .then((r) => {
            prog = r[fnExe]
        })
        .catch((err) => {
            errBin = `${isestr(get(err, 'code')) ? `[${err.code}] ` : ''}${isestr(get(err, 'message')) ? err.message : String(err)}`
        })
    if (errBin) {
        return Promise.reject(errBin)
    }
    // console.log('prog', prog)

    //inp
    let inp = {
        fpIn,
        fpOut,
        execFontGrow,
    }
    // console.log('inp', inp)

    //input to b64
    let cInput = JSON.stringify(inp)
    let b64Input = str2b64(cInput)
    // console.log('b64Input', b64Input)

    //execProcess, docx2html.exe以ANSI字碼頁輸出(中文系統為big5), 故codeCmd給auto使失敗訊息不亂碼
    let output = ''
    await execProcess(prog, b64Input, { codeCmd: 'auto' })
        .then((r) => {
            output = r
        })
        .catch((err) => {
            console.log('execProcess catch', err)
            errTemp = err.toString()
        })

    //check
    if (errTemp) {
        return Promise.reject(errTemp)
    }

    //check, docx2html.exe內部失敗時只於stdout印出`error: ...`且離開碼為0, execProcess不會reject, 故須依輸出判定
    let j = judgeConvertOutput(output)
    if (!j.ok) {
        return Promise.reject(j.msg)
    }

    //check, exe報成功後另驗fpOut存在且非空(例: fpOut無副檔名時Word另存為<fpOut>.htm, exe仍報success)
    let jf = judgeConvertFile(fpOut)
    if (!jf.ok) {
        return Promise.reject(jf.msg)
    }

    return 'ok'
}


//ensureBins, 供部署前預先取得外部執行檔(以靜態屬性匯出, 不加具名匯出以免改變UMD主檔之形狀)
WDocx2html.ensureBins = ensureBins


export default WDocx2html