w-html2docx 1.0.28 → 1.0.29

package/srcPython/htmlToDocx.py

@@ -1,385 +1,416 @@
-from win32com import client as wc  
-
-#使用win32com
-#win32com因更新問題會出現 ImportError: DLL load failed while importing win32api: 找不到指定的模組。
-#1. 安裝 pip install pywin32
-#2. 使用系統管理員權限開啟cmd
-#3. cd至安裝目錄 C:\ProgramData\Anaconda3\Scripts
-#4. 用python安裝腳本 python pywin32_postinstall.py -install
-
-#編譯
-#1. 安裝編譯套件 pip install pyinstaller
-#2. 編譯 pyinstaller -F htmlToDocx.py
-
-#win32com教學
-#https://zhuanlan.zhihu.com/p/67543981
-
-
-def getError():
-    import sys
-
-    #exc_info
-    type, message, traceback = sys.exc_info()
-
-    #es
-    es=[]
-    while traceback:
-        e={
-            'name':traceback.tb_frame.f_code.co_name,
-            'filename':traceback.tb_frame.f_code.co_filename,
-        }
-        es.append(e)
-        traceback = traceback.tb_next
-
-    #err
-    err={
-        'type':type,
-        'message':message,
-        'traceback':es,
-    }
-
-    return err
-
-
-def j2o(v):
-    #json轉物件
-    import json
-    return json.loads(v)
-
-
-def o2j(v):
-    #物件轉json
-    import json
-    return json.dumps(v, ensure_ascii=False)
-
-
-def str2b64(v):
-    #字串轉base64字串
-    import base64
-    v=base64.b64encode(v.encode('utf-8'))
-    return str(v,'utf-8')
-    
-
-def b642str(v):
-    #base64字串轉字串
-    import base64
-    return base64.b64decode(v).decode('utf-8')
-
-
-def readText(fn):
-    #讀取檔案fn內文字
-    import codecs
-    with codecs.open(fn,'r',encoding='utf8') as f:
-        return f.read()
-
-    
-def writeText(fn,str):
-    #寫出文字str至檔案fn
-    import codecs
-    with codecs.open(fn,'w',encoding='utf8') as f:
-        f.write(str)
-
-
-def htmlToDocx(fpInSrc, fpInTemp, fpOut, opt):
-
-    #Dispatch
-    app = wc.Dispatch('Word.Application')
-
-    #正式版須隱藏
-    app.Visible = False  
-
-    #不詢問使用者
-    app.DisplayAlerts = False 
-
-    #Open
-    docInTemp = app.Documents.Open(fpInTemp)
-    docInSrc = app.Documents.Open(fpInSrc)
-
-    # 取得來源文件全文 Range（含格式）
-    rngSrc = docInSrc.Range()
-
-    # 取得模板文件的主內容 Range
-    rngDst = docInTemp.Content  # 或 docInTemp.Range()
-
-    # 如果你想插在文件最後，可先收合到結尾
-    WdCollapseEnd = 0 #wdCollapseEnd, 收合到Range結尾
-    rngDst.Collapse(WdCollapseEnd)
-
-    # 直接複製格式化文字（等同貼上，但不經過剪貼簿也不靠 Selection）
-    rngDst.FormattedText = rngSrc.FormattedText
-
-    # #選擇src並複製全文
-    # docInSrc.Activate()
-    # app.Selection.WholeStory()
-    # app.Selection.Copy()
-
-    # #選擇模板並貼上全文
-    # docInTemp.Activate()
-    # app.Selection.Paste()
-    # # WdFormatOriginalFormatting = 16
-    # # app.Selection.PasteAndFormat(WdFormatOriginalFormatting)
-
-    #選擇模板調整字型
-    try:
-
-        fontFamilies=opt['fontFamilies']
-
-        if isinstance(fontFamilies, (list, tuple)) and len(fontFamilies) > 0:
-            #print(fontFamilies)
-
-            rng = docInTemp.Range()
-            for fn in fontFamilies:
-                # print(fn)
-                rng.Font.Name = fn
-
-    except:
-        err=getError()
-        print(err)
-
-    #選擇模板偵測各圖片寬度是否大於滿版(412), 並限制於最大值
-    try:
-
-        imgRatioWidthMax=opt['imgRatioWidthMax']
-        imgRatioHeightMax=opt['imgRatioHeightMax']
-
-        page_setup = docInTemp.PageSetup
-        pageW = page_setup.PageWidth #A4≈595 pt
-        pageH = page_setup.PageHeight #A4≈842 pt
-        leftM = page_setup.LeftMargin
-        rightM = page_setup.RightMargin
-        topM = page_setup.TopMargin
-        bottomM = page_setup.BottomMargin
-
-        #版心最大maxW, maxH
-        maxW = pageW - leftM - rightM
-        maxW = maxW * imgRatioWidthMax
-        maxH = (pageH - topM - bottomM) * 0.937043054427296 #高度須預留圖名空間
-        maxH = maxH * imgRatioHeightMax
-
-        # 使用 docInTemp.InlineShapes 直接取得整份文件所有圖片
-        for s in docInTemp.InlineShapes:
-            try:
-
-                #w, h
-                w = s.Width
-                h = s.Height
-
-                #scale, 計算縮放比例
-                scaleW = maxW / w if w > maxW else 1
-                scaleH = maxH / h if h > maxH else 1
-                scale = min(scaleW, scaleH)
-
-                if scale < 1:
-                    s.LockAspectRatio = True
-                    s.Width = w * scale   #只設width, LockAspectRatio=true會讓height依比例調整
-
-            except:
-                pass
-
-    except:
-        err=getError()
-        print(err)
-
-    #設定前後行距
-    try:
-
-        for p in docInTemp.Paragraphs:
-
-            try:
-                lf = p.Range.ListFormat
-            except:
-                #若可能沒有ListFormat須跳過
-                continue
-
-            listType = lf.ListType   # 0,1,2,3,4
-            listStr  = lf.ListString # '•', '1.', ''...
-            
-            # #測試顯示ListType
-            # try:
-            #     rng = p.Range.Duplicate
-            #     rng.End = rng.End - 1 #最末是換行符號, 不能插在換行符號後面
-            #     rng.InsertAfter(f"[{listType},{listStr}]")
-            # except:
-            #     err = getError()
-            #     print(err)
-            
-            WdListNoNumbering = 0       #普通段落
-            WdListSingleLevelNumber = 1 #單層編號清單
-            WdListMultiLevel = 2        #多層編號清單
-            WdListBullet = 3            #項目符號清單
-            WdListOutlineNumbering = 4  #大綱編號
-            is_list = (listType != WdListNoNumbering) or (listStr != "")
-
-            if is_list: 
-
-                fmt = p.Format
-
-                #關閉自動段前段後距離
-                try:
-                    fmt.SpaceBeforeAuto = False
-                    fmt.SpaceAfterAuto = False
-                except:
-                    pass
-
-                # 設定為 0.5 行
-                fmt.LineUnitBefore = 0.5
-                fmt.LineUnitAfter = 0.5
-
-    except:
-        err=getError()
-        print(err)
-
-    #設定本文段落(p)與清單項目(li)之水平對齊, 已為置中者(如表格標題列, 圖名)不更動, 其餘改為左右對齊(兩端對齊), 使右側邊界齊平較為美觀
-    try:
-
-        WdAlignParagraphCenter = 1   #置中對齊
-        WdAlignParagraphJustify = 3  #左右對齊
-        WdOutlineLevelBodyText = 10  #本文層級, 標題(h1-h6)匯入後為大綱層級1-9
-        WdWithInTable = 12           #Information之參數, 判斷是否位於表格內
-
-        for p in docInTemp.Paragraphs:
-            try:
-
-                rng = p.Range
-
-                #跳過標題段落
-                if p.OutlineLevel != WdOutlineLevelBodyText:
-                    continue
-
-                #跳過表格內段落
-                if rng.Information(WdWithInTable):
-                    continue
-
-                #跳過圖片段落
-                if rng.InlineShapes.Count > 0:
-                    continue
-
-                #跳過置中段落
-                if p.Alignment == WdAlignParagraphCenter:
-                    continue
-
-                p.Alignment = WdAlignParagraphJustify
-
-            except:
-                pass
-
-    except:
-        err=getError()
-        print(err)
-
-    #移除零寬空格佔位字元(U+200B): 來源為w-md2html於換行標記空div內插入之佔位字元, 用以撐過Word匯入不被當成空段落丟棄
-    #於此移除後該段落成為真正的空段落(乾淨段落標記, 顯示編輯標記時亦無可見字元)
-    try:
-
-        z = chr(0x200B) #U+200B 零寬空格
-
-        for p in docInTemp.Paragraphs:
-            rng = p.Range.Duplicate
-            rng.End = rng.End - 1 #排除段落結尾之段落標記, 避免破壞段落結構
-            t = rng.Text
-            if z in t:
-                rng.Text = t.replace(z, '') #移除佔位字元, 段落標記保留 => 乾淨空段落
-
-    except:
-        err=getError()
-        print(err)
-
-    #SaveAs
-    WdSaveFormat = 16 #wdFormatDocumentDefault=16(docx) #https://learn.microsoft.com/zh-tw/office/vba/api/word.wdsaveformat
-    docInTemp.SaveAs2(fpOut, WdSaveFormat)
-
-    # Close
-    WdDoNotSaveChanges = 0 #wdDoNotSaveChanges
-    docInTemp.Close(WdDoNotSaveChanges)
-    docInSrc.Close(WdDoNotSaveChanges)
-
-    #Quit
-    WdSaveOptions = 0 #wdDoNotSaveChanges=0(不儲存擱置中變更) #https://learn.microsoft.com/zh-tw/office/vba/api/word.wdsaveoptions
-    app.Quit(WdSaveOptions)
-
-
-def core(b64):
-    state=''
-
-    try:
-
-        #b642str
-        s=b642str(b64)
-
-        #j2o
-        o=j2o(s)
-
-        #params
-        fpInSrc=o['fpInSrc']
-        fpInTemp=o['fpInTemp']
-        fpOut=o['fpOut']
-
-        #opt
-        opt={}
-        opt['fontFamilies']=o['fontFamilies']
-        opt['imgRatioWidthMax']=o['imgRatioWidthMax']
-        opt['imgRatioHeightMax']=o['imgRatioHeightMax']
-
-        #htmlToDocx
-        htmlToDocx(fpInSrc, fpInTemp, fpOut, opt)
-
-        state='success'
-    except:
-        err=getError()
-        state='error: '+str(err["message"])
-
-    return state
-
-
-def run():
-    import sys
-
-    #由外部程序呼叫或直接給予檔案路徑
-    state=''
-    argv=sys.argv
-    #argv=['','']
-    if len(argv)==2:
-        
-        #b64
-        b64=sys.argv[1]
-        
-        #core
-        state=core(b64)
-        
-    else:
-        #print(sys.argv)
-        state='error: invalid length of argv'
-    
-    #print & flush
-    print(state)
-    sys.stdout.flush()
-
-
-if True:
-    #正式版
-    
-    #run
-    run()
-    
-    
-if False:
-    #產生測試輸入b64
-    
-    #inp
-    inp={
-        'fpInSrc':'./ztmp.html',
-        'fpInTemp':'./tmp.docx',
-        'fpOut':'./ztmp.docx',
-        'fontFamilies': ['標楷體','Times New Roman'], #word使用, 因由左往右設定, 故得先設定標楷體再設定Times New Roman
-    }
-    # print(o2j(inp))
-    
-    #str2b64
-    b64=str2b64(o2j(inp))
-    print(b64)
-
-    #core
-    state=core(b64)
-
-    print(state)
-
+from win32com import client as wc
+import sys
+
+#使用win32com
+#win32com因更新問題會出現 ImportError: DLL load failed while importing win32api: 找不到指定的模組。
+#1. 安裝 pip install pywin32
+#2. 使用系統管理員權限開啟cmd
+#3. cd至安裝目錄 C:\ProgramData\Anaconda3\Scripts
+#4. 用python安裝腳本 python pywin32_postinstall.py -install
+
+#編譯
+#1. 安裝編譯套件 pip install pyinstaller
+#2. 編譯 pyinstaller -F htmlToDocx.py
+
+#win32com教學
+#https://zhuanlan.zhihu.com/p/67543981
+
+
+def getError():
+    import sys
+
+    #exc_info
+    type, message, traceback = sys.exc_info()
+
+    #es
+    es=[]
+    while traceback:
+        e={
+            'name':traceback.tb_frame.f_code.co_name,
+            'filename':traceback.tb_frame.f_code.co_filename,
+        }
+        es.append(e)
+        traceback = traceback.tb_next
+
+    #err
+    err={
+        'type':type,
+        'message':message,
+        'traceback':es,
+    }
+
+    return err
+
+
+def j2o(v):
+    #json轉物件
+    import json
+    return json.loads(v)
+
+
+def o2j(v):
+    #物件轉json
+    import json
+    return json.dumps(v, ensure_ascii=False)
+
+
+def str2b64(v):
+    #字串轉base64字串
+    import base64
+    v=base64.b64encode(v.encode('utf-8'))
+    return str(v,'utf-8')
+
+
+def b642str(v):
+    #base64字串轉字串
+    import base64
+    return base64.b64decode(v).decode('utf-8')
+
+
+def readText(fn):
+    #讀取檔案fn內文字
+    import codecs
+    with codecs.open(fn,'r',encoding='utf8') as f:
+        return f.read()
+
+
+def writeText(fn,str):
+    #寫出文字str至檔案fn
+    import codecs
+    with codecs.open(fn,'w',encoding='utf8') as f:
+        f.write(str)
+
+
+def htmlToDocxCore(app, docs, fpInSrc, fpInTemp, fpOut, opt):
+
+    #Open, 開啟之文件推入docs, 供htmlToDocx於finally關閉
+    docInTemp = app.Documents.Open(fpInTemp)
+    docs.append(docInTemp)
+    docInSrc = app.Documents.Open(fpInSrc)
+    docs.append(docInSrc)
+
+    # 取得來源文件全文 Range（含格式）
+    rngSrc = docInSrc.Range()
+
+    # 取得模板文件的主內容 Range
+    rngDst = docInTemp.Content  # 或 docInTemp.Range()
+
+    # 如果你想插在文件最後，可先收合到結尾
+    WdCollapseEnd = 0 #wdCollapseEnd, 收合到Range結尾
+    rngDst.Collapse(WdCollapseEnd)
+
+    # 直接複製格式化文字（等同貼上，但不經過剪貼簿也不靠 Selection）
+    rngDst.FormattedText = rngSrc.FormattedText
+
+    # #選擇src並複製全文
+    # docInSrc.Activate()
+    # app.Selection.WholeStory()
+    # app.Selection.Copy()
+
+    # #選擇模板並貼上全文
+    # docInTemp.Activate()
+    # app.Selection.Paste()
+    # # WdFormatOriginalFormatting = 16
+    # # app.Selection.PasteAndFormat(WdFormatOriginalFormatting)
+
+    #選擇模板調整字型
+    try:
+
+        fontFamilies=opt['fontFamilies']
+
+        if isinstance(fontFamilies, (list, tuple)) and len(fontFamilies) > 0:
+            #print(fontFamilies)
+
+            rng = docInTemp.Range()
+            for fn in fontFamilies:
+                # print(fn)
+                rng.Font.Name = fn
+
+    except:
+        err=getError()
+        print(err, file=sys.stderr)
+
+    #選擇模板偵測各圖片寬度是否大於滿版(412), 並限制於最大值
+    try:
+
+        imgRatioWidthMax=opt['imgRatioWidthMax']
+        imgRatioHeightMax=opt['imgRatioHeightMax']
+
+        page_setup = docInTemp.PageSetup
+        pageW = page_setup.PageWidth #A4≈595 pt
+        pageH = page_setup.PageHeight #A4≈842 pt
+        leftM = page_setup.LeftMargin
+        rightM = page_setup.RightMargin
+        topM = page_setup.TopMargin
+        bottomM = page_setup.BottomMargin
+
+        #版心最大maxW, maxH
+        maxW = pageW - leftM - rightM
+        maxW = maxW * imgRatioWidthMax
+        maxH = (pageH - topM - bottomM) * 0.937043054427296 #高度須預留圖名空間
+        maxH = maxH * imgRatioHeightMax
+
+        # 使用 docInTemp.InlineShapes 直接取得整份文件所有圖片
+        for s in docInTemp.InlineShapes:
+            try:
+
+                #w, h
+                w = s.Width
+                h = s.Height
+
+                #scale, 計算縮放比例
+                scaleW = maxW / w if w > maxW else 1
+                scaleH = maxH / h if h > maxH else 1
+                scale = min(scaleW, scaleH)
+
+                if scale < 1:
+                    s.LockAspectRatio = True
+                    s.Width = w * scale   #只設width, LockAspectRatio=true會讓height依比例調整
+
+            except:
+                pass
+
+    except:
+        err=getError()
+        print(err, file=sys.stderr)
+
+    #設定前後行距
+    try:
+
+        for p in docInTemp.Paragraphs:
+
+            try:
+                lf = p.Range.ListFormat
+            except:
+                #若可能沒有ListFormat須跳過
+                continue
+
+            listType = lf.ListType   # 0,1,2,3,4
+            listStr  = lf.ListString # '•', '1.', ''...
+
+            # #測試顯示ListType
+            # try:
+            #     rng = p.Range.Duplicate
+            #     rng.End = rng.End - 1 #最末是換行符號, 不能插在換行符號後面
+            #     rng.InsertAfter(f"[{listType},{listStr}]")
+            # except:
+            #     err = getError()
+            #     print(err)
+
+            WdListNoNumbering = 0       #普通段落
+            WdListSingleLevelNumber = 1 #單層編號清單
+            WdListMultiLevel = 2        #多層編號清單
+            WdListBullet = 3            #項目符號清單
+            WdListOutlineNumbering = 4  #大綱編號
+            is_list = (listType != WdListNoNumbering) or (listStr != "")
+
+            if is_list:
+
+                fmt = p.Format
+
+                #關閉自動段前段後距離
+                try:
+                    fmt.SpaceBeforeAuto = False
+                    fmt.SpaceAfterAuto = False
+                except:
+                    pass
+
+                # 設定為 0.5 行
+                fmt.LineUnitBefore = 0.5
+                fmt.LineUnitAfter = 0.5
+
+    except:
+        err=getError()
+        print(err, file=sys.stderr)
+
+    #設定本文段落(p)與清單項目(li)之水平對齊, 已為置中者(如表格標題列, 圖名)不更動, 其餘改為左右對齊(兩端對齊), 使右側邊界齊平較為美觀
+    try:
+
+        WdAlignParagraphCenter = 1   #置中對齊
+        WdAlignParagraphJustify = 3  #左右對齊
+        WdOutlineLevelBodyText = 10  #本文層級, 標題(h1-h6)匯入後為大綱層級1-9
+        WdWithInTable = 12           #Information之參數, 判斷是否位於表格內
+
+        for p in docInTemp.Paragraphs:
+            try:
+
+                rng = p.Range
+
+                #跳過標題段落
+                if p.OutlineLevel != WdOutlineLevelBodyText:
+                    continue
+
+                #跳過表格內段落
+                if rng.Information(WdWithInTable):
+                    continue
+
+                #跳過圖片段落
+                if rng.InlineShapes.Count > 0:
+                    continue
+
+                #跳過置中段落
+                if p.Alignment == WdAlignParagraphCenter:
+                    continue
+
+                p.Alignment = WdAlignParagraphJustify
+
+            except:
+                pass
+
+    except:
+        err=getError()
+        print(err, file=sys.stderr)
+
+    #移除零寬空格佔位字元(U+200B): 來源為w-md2html於換行標記空div內插入之佔位字元, 用以撐過Word匯入不被當成空段落丟棄
+    #於此移除後該段落成為真正的空段落(乾淨段落標記, 顯示編輯標記時亦無可見字元)
+    try:
+
+        z = chr(0x200B) #U+200B 零寬空格
+
+        for p in docInTemp.Paragraphs:
+            rng = p.Range.Duplicate
+            rng.End = rng.End - 1 #排除段落結尾之段落標記, 避免破壞段落結構
+            t = rng.Text
+            if z in t:
+                rng.Text = t.replace(z, '') #移除佔位字元, 段落標記保留 => 乾淨空段落
+
+    except:
+        err=getError()
+        print(err, file=sys.stderr)
+
+    #SaveAs
+    WdSaveFormat = 16 #wdFormatDocumentDefault=16(docx) #https://learn.microsoft.com/zh-tw/office/vba/api/word.wdsaveformat
+    docInTemp.SaveAs2(fpOut, WdSaveFormat)
+
+
+def htmlToDocx(fpInSrc, fpInTemp, fpOut, opt):
+
+    #Dispatch
+    app = wc.Dispatch('Word.Application')
+
+    #正式版須隱藏
+    app.Visible = False
+
+    #不詢問使用者
+    app.DisplayAlerts = False
+
+    #docs, 已開啟之文件, 由htmlToDocxCore推入
+    docs = []
+
+    #任一步驟例外時亦須關閉已開啟之文件並結束Word, 否則殘留之WINWORD會鎖住模板與輸出檔, 使後續轉檔失敗
+    try:
+        htmlToDocxCore(app, docs, fpInSrc, fpInTemp, fpOut, opt)
+    finally:
+
+        # Close
+        WdDoNotSaveChanges = 0 #wdDoNotSaveChanges
+        for doc in reversed(docs):
+            try:
+                doc.Close(WdDoNotSaveChanges)
+            except:
+                pass
+
+        #Quit
+        WdSaveOptions = 0 #wdDoNotSaveChanges=0(不儲存擱置中變更) #https://learn.microsoft.com/zh-tw/office/vba/api/word.wdsaveoptions
+        try:
+            app.Quit(WdSaveOptions)
+        except:
+            pass
+
+
+def core(b64):
+    state=''
+
+    try:
+
+        #b642str
+        s=b642str(b64)
+
+        #j2o
+        o=j2o(s)
+
+        #params
+        fpInSrc=o['fpInSrc']
+        fpInTemp=o['fpInTemp']
+        fpOut=o['fpOut']
+
+        #opt
+        opt={}
+        opt['fontFamilies']=o['fontFamilies']
+        opt['imgRatioWidthMax']=o['imgRatioWidthMax']
+        opt['imgRatioHeightMax']=o['imgRatioHeightMax']
+
+        #htmlToDocx
+        htmlToDocx(fpInSrc, fpInTemp, fpOut, opt)
+
+        state='success'
+    except:
+        err=getError()
+        state='error: '+err['type'].__name__+': '+str(err['message'])
+
+    return state
+
+
+def run():
+
+    #stdout與stderr改為utf-8: Python寫pipe時預設用系統ANSI字碼頁(中文Windows為cp950), 呼叫端須猜編碼才能讀, 且訊息含cp950無法編碼之字元時print會拋UnicodeEncodeError;
+    #PyInstaller打包後不吃PYTHONIOENCODING與PYTHONUTF8, 只能於此設定
+    for f in (sys.stdout, sys.stderr):
+        try:
+            f.reconfigure(encoding='utf-8', errors='backslashreplace')
+        except:
+            pass
+
+    #由外部程序呼叫或直接給予檔案路徑
+    state=''
+    argv=sys.argv
+    #argv=['','']
+    if len(argv)==2:
+
+        #b64
+        b64=sys.argv[1]
+
+        #core
+        state=core(b64)
+
+    else:
+        #print(sys.argv)
+        state='error: invalid length of argv'
+
+    #print & flush
+    print(state)
+    sys.stdout.flush()
+
+    #失敗時以非0離開碼結束, 使呼叫端(如wsemi之execProcess)能直接以離開碼判定成敗, 不必解析輸出
+    if state != 'success':
+        sys.exit(1)
+
+
+if True:
+    #正式版
+
+    #run
+    run()
+
+
+if False:
+    #產生測試輸入b64
+
+    #inp
+    inp={
+        'fpInSrc':'./ztmp.html',
+        'fpInTemp':'./tmp.docx',
+        'fpOut':'./ztmp.docx',
+        'fontFamilies': ['標楷體','Times New Roman'], #word使用, 因由左往右設定, 故得先設定標楷體再設定Times New Roman
+    }
+    # print(o2j(inp))
+
+    #str2b64
+    b64=str2b64(o2j(inp))
+    print(b64)
+
+    #core
+    state=core(b64)
+
+    print(state)
+