User Tool Import Fail

[Outis]

Hello all,

I just made a pretty simple user tool (at least I tried to) and the result was pretty ugly (see attached screenshot).

I haven't had problems like this before with importing user tools from .html. So, my question is this:

Am I the only having this problem or are there others out there who are having the same problem?

Thanks.

[Joel Brown]

Someone else has reported this problem, and I believe it is due to the html coming in as UTF-8 (unicode). This definitely seems to be a bug in 9.3 that we are looking into. Until we fix it for 9.3.2, if you open the html file in a text editor, like TextWrangler, and resave it as MacRoman encoding, it should import fine.

[mikes]

Someone else has reported this problem, and I believe it is due to the html coming in as UTF-8 (unicode). This definitely seems to be a bug in 9.3 that we are looking into. Until we fix it for 9.3.2, if you open the html file in a text editor, like TextWrangler, and resave it as MacRoman encoding, it should import fine.

While trying to help another user (I'm guessing the same one?) I ran into the same problem, except re-encoding into MacRoman scrambles the text... is there a better way to do this until the bug is fixed?

 

Here's the error I get when trying to save as MacRoman using textmate:

The document “Test” cannot be saved using the MacRoman encoding. Would you like to save it using UTF-8 instead?

[Joel Brown]

If you are scrambling the text by reencoding, that is reinterpreting, not reencoding. By definition, reencoding changes the backend, but visually the file is the same. I'm not familiar with TextMate, but it works fine in TextWrangler. If you can't reencode it as MacRoman, that means there are some unicode only characters that need to be manually changed or removed. If TextMate doesn't do that, again, I recommend TextWrangler. Its free, fantastic, and extremely helpful.

[Outis]

If you can't reencode it as MacRoman, that means there are some unicode only characters that need to be manually changed or removed. If TextMate doesn't do that, again, I recommend TextWrangler. Its free, fantastic, and extremely helpful.

 

Thanks Joel. By default, I have the "export unicode" box checked in Accordance. Thankfully, in this user tool there were only a couple greek references. So I deleted them. Then I saved the .html document as "Western Mac OS Roman" in BBEdit. It imported just fine. I did have to retype in the Greek words. But it imported ok.

 

Thanks for the 'work-around.' And I look forward to the bug being fixed (-:

[mikes]

If you are scrambling the text by reencoding, that is reinterpreting, not reencoding. By definition, reencoding changes the backend, but visually the file is the same. I'm not familiar with TextMate, but it works fine in TextWrangler. If you can't reencode it as MacRoman, that means there are some unicode only characters that need to be manually changed or removed. If TextMate doesn't do that, again, I recommend TextWrangler. Its free, fantastic, and extremely helpful.

I use textmate for development, but have text wrangler as well.

 

Same problem: UTF-8 Greek doesn't fit on the MacRoman map using either tool when using the "Reopen using encoding". If you have a particular methodology, please let me know what you're doing.

[Joel Brown]

Yes, the high-unicode characters, such as greek unicode, won't encode to MacRoman so you need to delete them or convert them manually to something acceptable in ASCII.

 

'Reopen in encoding' is the reinterpretation that I was talking about *not* doing. That is reading the same data in the file, but pretending it is macroman not utf8. Instead, open the file normally as a utf8 file, but then change that to MacRoman using the selector at the bottom. When you attempt to save if there are still bad characters you need to remove them (can do it in the warning dialog), but once removed it should save fine and then import into Accordance.

 

Or just wait a week for the fix.

[mikes]

Yes, the high-unicode characters, such as greek unicode, won't encode to MacRoman so you need to delete them or convert them manually to something acceptable in ASCII.

 

'Reopen in encoding' is the reinterpretation that I was talking about *not* doing. That is reading the same data in the file, but pretending it is macroman not utf8. Instead, open the file normally as a utf8 file, but then change that to MacRoman using the selector at the bottom. When you attempt to save if there are still bad characters you need to remove them (can do it in the warning dialog), but once removed it should save fine and then import into Accordance.

 

Or just wait a week for the fix.

Thanks for your help: being that the document is 99% Greek, I'll wait for the fix.