Author Topic: Text File Corruption  (Read 3422 times)

Smuffle

  • Jr. Member
  • **
  • Posts: 55
    • View Profile
Text File Corruption
« on: January 28, 2006, 05:40:03 am »
I use the text editor (in Cacko) to store lists of Japanese vocab but every now and again when I open the file, the whole file contents has turned to mush (mojibake).

It only happens once every month or two but it's still a pain.  I've tried opening the file on another computer using various software but I always end up deleting the file.  It's a shame because it's made me a little paranoid about using it..

There doesn't seem to be rhyme or reason, just seems to happen under normal operation.

Anybody else have a similar problem?

rokugo

  • Full Member
  • ***
  • Posts: 130
    • View Profile
    • http://
Text File Corruption
« Reply #1 on: January 28, 2006, 11:38:23 am »
Assuming you installed the Japanese support package from the Cacko feed, its not 100% complete.
The Text Editor still doesn't support Japanese character encoding ie Shift-JIS etc.
My guess is you've previously opened the vocab lists in another SJIS-capable editor or in Windows (which defaults to SJIS for Japanese input) thus causing the mojibake.

Check out this post to enable SJIS support for the Text Editor:
https://www.oesf.org/forums/index.php?showtopic=16808

I wrote it for the C1000 but should work for the 3100 too.
« Last Edit: January 28, 2006, 11:43:57 am by rokugo »
currently:  SL-C860 Cacko 1.23 ROM converted to full Japanese
formerly: SL-C1000 Cacko 1.23 ROM converted to full Japanese
Sony WL100 wifi CF card (handed down from old Clie)
A-Data 60X 2GB SD card loaded with Super Daijirin, Green Goddess5, NJECD and Eijiro

Smuffle

  • Jr. Member
  • **
  • Posts: 55
    • View Profile
Text File Corruption
« Reply #2 on: January 29, 2006, 03:23:03 am »
Quote
Assuming you installed the Japanese support package from the Cacko feed, its not 100% complete.
The Text Editor still doesn't support Japanese character encoding ie Shift-JIS etc.
My guess is you've previously opened the vocab lists in another SJIS-capable editor or in Windows (which defaults to SJIS for Japanese input) thus causing the mojibake.

Check out this post to enable SJIS support for the Text Editor:
https://www.oesf.org/forums/index.php?showtopic=16808

I wrote it for the C1000 but should work for the 3100 too.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=112757\"][{POST_SNAPBACK}][/a][/div]

All my Japanese study is on the Z so I can't blame another source (like my desktop).  I've literally closed and opened the same file on the Z (straight away) and seen it open as mojibake; it's starting to make me paranoid.

The missing files you mentioned in your thread - if some were missing wouldn't all Japanese support fail ?  As it is, it works fine it just goes strange every month or two.

rokugo

  • Full Member
  • ***
  • Posts: 130
    • View Profile
    • http://
Text File Corruption
« Reply #3 on: January 30, 2006, 12:54:54 am »
Like I said, it isn't 100%.
Right now your Text Editor is using UTF-8 which obviously supports Japanese.

But since ShiftJIS is still the predominant Japanese encoding (I suspect some apps in the Zaurus use it as default), you're bound to run into mojibake sooner or later.
« Last Edit: January 30, 2006, 01:03:00 am by rokugo »
currently:  SL-C860 Cacko 1.23 ROM converted to full Japanese
formerly: SL-C1000 Cacko 1.23 ROM converted to full Japanese
Sony WL100 wifi CF card (handed down from old Clie)
A-Data 60X 2GB SD card loaded with Super Daijirin, Green Goddess5, NJECD and Eijiro

Smuffle

  • Jr. Member
  • **
  • Posts: 55
    • View Profile
Text File Corruption
« Reply #4 on: January 30, 2006, 02:52:40 am »
Quote
Like I said, it isn't 100%.
Right now your Text Editor is using UTF-8 which obviously supports Japanese.

But since ShiftJIS is still the predominant Japanese encoding (I suspect some apps in the Zaurus use it as default), you're bound to run into mojibake sooner or later.
[div align=\"right\"][a href=\"index.php?act=findpost&pid=112933\"][{POST_SNAPBACK}][/a][/div]

after reading that, it made me think of a possibility.  Sometimes I cut and paste from the built in dictionary into the text editor - maybe the dictionary uses a different encoding and it's confusing matters.