Applescript: Save Word documents as plaintext while retaining accents

1
votes

I'm trying to save Word documents as plain text docs. Currently, some times the accents turn into other symbols (usually the same ones, for example: é turns into a theta). Other times it works fine. How do I prevent this?

Currently using the line:

save as active document file name FullDocPath file format format Unicode text

When I encounter this error, I can save the document using the dialog (selecting Western Mac OS Roman encoding...that fixes the problem.

The applescript Word dictionary mentions:

[text encoding unsigned integer] : Text encoding to use when saving out as text file

I have no idea if this is the piece I'm missing or how to utilize it (is there a set integer that designates Western Mac OS Roman encoding?)

Anyone have any ideas?

2
ms-wordapplescript
FWIW the "save as" keywords for encoding that work in VBA do not appear to work in applescript. For example, for UTF-8 encoding, Word VBA records FileFormat:=wdFormatText, Encoding:=65001 (the value for Western (Mac Os Roman) is 10000). But the equivalent applescript "file format format text text encoding 65001" does not work. Nor does "file format format Unicode text" etc. Also, when Word saves as UTF-8 via VBA or the UI, it saves a byte order mark ("BOM") at the beginning that it does not recognise when opening the file. It does recognise the BOM for the 16-bit Unicode formats. - user1379931
Thanks Stefan and bibadiak for the insights. Using the textutil shell script, I can now get the accents to save properly (they appear when I open the file in textedit) however, when I open the text file via MS Word, there are still issues (ie. the é turns to a normal e and now additional characters are being switched, ' turns to Äô) - I'm new to encoding but I would assume it's trying to open it as a different type. - user2575047

2 Answers

2
votes

Try:

set wordDoc to choose file
do shell script "textutil -convert txt " & quoted form of POSIX path of (wordDoc as text)

Check out StefanK's solution using textutil

0
votes

This is in response to your comment beginning "Thanks Stefan and bibadiak"

With .txt file formats is that there is no universally used way to specify the encoding of a file inside the file, so either the application has to guess, or you have to know the encoding and the application has to let you tell it.

AFAIK if you do not specify an output encoding when you use textutil to convert from .doc or .docx format to text, you get UTF-8. But Mac Word just does not seem to recognise that when you try to open it, either programmatically or in the UI.

So I think you need to do some mix of the following:

a. save in, and work with, a format that uses 16-bit Unicode encoding. Word should recognise that, certainly if the BOM is preserved

b. save to UTF and work with UTF elsewhere, but use textutil to do the conversion back to (say) .docx before you re-open the document in Mac Word

c. if all your characters can be encoded using Mac OS Roman, use e.g. 

textutil -convert txt -encoding 30

to save, ensure you work only with that character set, and re-open with Word. (30 is the value of the APple NSString value NSMacOSRomanStringEncoding). I think textutil will fail to convert documents that contain characters outside the MacOS Roman set.