Python win32com - Automating Word - How to replace

问题:

I'm trying to automate word to replace text in a word document using Python. (I'm on word 2003 if that matters and Python 2.4)

The first part of my replace method below works on everything except text in text boxes. The text just doesn't get selected. I notice when I go into Word manually and hit ctrl-A all of the text gets selected except for the text box.

Here's my code so far:

class Word:
    def __init__(self,visible=0,screenupdating=0):
        pythoncom.CoInitialize()
        self.app=gencache.EnsureDispatch(WORD)
        self.app.Visible = visible
        self.app.DisplayAlerts = 0
        self.app.ScreenUpdating = screenupdating
        print 'Starting word'
    def open(self,doc):
        self.opendoc=os.path.basename(doc)
        self.app.Documents.Open(FileName=doc)
    def replace(self,source,target):
        if target=='':target=' '
        alltext=self.app.Documents(self.opendoc).Range(Start=0,End=self.app.Documents(self.opendoc).Characters.Count) #select all
        alltext.Find.Text = source
        alltext.Find.Replacement.Text = target
        alltext.Find.Execute(Replace=1,Forward=True)
        #Special handling to do replace in text boxes
        #http://word.tips.net/Pages/T003879_Updating_a_Field_in_a_Text_Box.html
        for shp in self.app.Documents(self.opendoc).Shapes:
            if shp.TextFrame.HasText:
                shp.TextFrame.TextRange.Find.Text = source
                shp.TextFrame.TextRange.Find.Replacement.Text = target
                shp.TextFrame.TextRange.Find.Execute(Replace=1,Forward=True)

#My Usage
word=Word(visible=1,screenupdating=1)
word.open(r'C:\Invoice Automation\testTB.doc')
word.replace('[PGN]','1')

The for shp in self.app .. section is my attempt to hit the text boxes. It seems to find the text box, but it doesn't replace anything.

回答1:

When I add text boxes to a word document, they are added inside a drawing canvas. Therefore the top level shape is the canvas, and the text boxes are contained within the canvas. You should use the CanvasItems method to access the objects in the canvas, ie the text boxes

The following example works for me. I created a word document with a single text box.

import win32com.client

word = win32com.client.Dispatch("Word.Application")
canvas = word.ActiveDocument.Shapes[0]
for item in canvas.CanvasItems:
    print item.TextFrame.TextRange.Text

---

Update: answering OP's comment.

I think the problem with your code is that each line of code with Find creates a new Find object. You have to create and bind a Find object to a name, then modify its attributes and execute it. So in your code you should have:

find = shp.TextFrame.TextRange.Find
find.Text = source
find.Replacement.Text = target
find.Execute(Replace=1, Forward=True)

Or a single line:

shp.TextFrame.TextRange.Find.Execute(FindText=source, ReplaceWith=target, Replace=1, Forward=True)

Both of these methods work in my test code.