The dictionary/lexicon window has a similar problem although it never caused exception errors. It does not remember the last word looked up if it uses higher UTF-8 characters.
ak
Troy A. Griffitts wrote:
Adrian,
Thanks for the detailed bug report. It's probably a simple fix and a blockheaded error on my part. I would bet I'm not even trying to do a utf8towide conversion on read of the .conf file but I am doing the conversion on save (or vice versa) I'll try to get a fixed version for you quickly.
-Troy.
Adrian Korten wrote:
Good day,
We have run into a new and somewhat irregular error with Thai texts. On exiting the program, the following window would appear. "Sword caused an invalid page fault in module <unknown> at 0000:08cb8e0". Once this message occured, it would keep occuring with each exit of the program. This happened with both the v1.5.4 and v1.5.6 programs. At first, I thought it had something to do with the Thai text modules (zipped and encrypted). It was erratic and would show up after an install on some machines but not others. Uninstalling and re-installing the programs would fix the problem.
I finally had the problem on a local machine and the time to dig into it more. Neither uninstalling the Thai modules, turning off all switches (Strongs, headings, hebrew, etc.) nor turning off 'Autosave layout' and 'Autosave Personal bookmarks' helped. But I began to notice it was not saving layout changes that had been previously made. Examining the layout.conf file, it had a some lines that looked like Thai characters as untranslated utf-8 characters. After deleting these lines, the problem disappears and there is no error on exit.
Looking at this problem a bit more when the user does a search and looks up a previous search, the Thai searches are all unreadable. Rather the searches from the current session are displayed okay but searches from previous program sessions are unreadable. I assume the previous searches are being read from the layout.conf file. The layout.conf file is a non-UTF-8 format and cannot handle non-roman characters properly.
It also seems like the lookup and search texts that are saved for Thai
are quite long. They seem to be longer than what would have been entered
and then saved as ascii (3x characters). After doing a before program use and after, these strings are being repeated with each running of the program. The first shutdown messes up the utf-8 characters and afterwards it string is duplicated each time.
I'm thinking of setting the 'LookupSaveCount' and 'SearchSaveCount' to zero. Or does anyone have a better suggestion?
Adrian
------------------------------------------------------------------------
_______________________________________________ sword-devel mailing list [EMAIL PROTECTED] http://www.crosswire.org/mailman/listinfo/sword-devel
