Problems with truncation of query strings

[Steven P. Venti]

I'm having trouble with Felix truncating query strings containing halfwidth kana.

For example, Felix has no trouble recognizing the entire string for 過給EGRバルブ選定,
in which the word バルブ comprises three fullwidth kana.

But in cases where the same word comprises five halfwidth kana (ﾊﾞﾙﾌﾞ),
the query string truncates at the end of the halfwidth kana and appears as 過
給EGRﾊﾞﾙﾌﾞ in the query line of the Felix Memory window, even though the
entire phrase is highlighted in the MS Office application.

And since the entire word is highlighted in the MS Office app, it is not
possible to extend or otherwise force recognition using shortcuts likes ctrl +
right arrow or alt + L.

I realize that it's a bit much to ask that Felix recognize either バルブ or ﾊ
ﾞﾙﾌﾞ as the same word, but is there at least a way to get it to stop
truncating the query string?

Using Felix v1.7 with MS Office 2010 on Win7.

------------------------------------------------------------------------
Steve Venti
spv...@bhk-limited.com

You want to squeeze my nose? Go ahead! I think we're all bozos on this bus.
-- The Firesign Theatre
------------------------------------------------------------------------

[Ginstrom IT Solutions]

> On Behalf Of Steven P. Venti

> But in cases where the same word comprises five halfwidth kana (ﾊﾞﾙﾌﾞ),
> the query string truncates at the end of the halfwidth kana and appears as

> 過給EGRﾊﾞﾙﾌﾞ in the query line of the Felix Memory window, even though the

> entire phrase is highlighted in the MS Office application.

This is definitely a bug; thanks for reporting it.

One workaround is to normalize width in the TM settings.

1. Go to Tools >> Preferences.
2. On the Memory tab, select Normalize >> Width.

This should recognize the half-width katakana correctly, and recognize the half and full-width variants as being the same.


Regards,
Ryan

=================================
Ryan Ginstrom
Felix Translation Memory Software
sup...@felix-cat.com
http://felix-cat.com/
+1-425-633-6039
=================================

[Steven P. Venti]

"Ginstrom IT Solutions" <sup...@felix-cat.com> wrote:
> One workaround is to normalize width in the TM settings.

Thanks for the fast response, Ryan.

I just checked and I have width normalized. In fact, as I work on this
document, I find that the behavior is very inconsistent, and that longer
query segments seem more robust.

For example, the query string ﾊﾞﾙﾌﾞ構造等複雑にしています does not
truncate. But if I insert a fullwidth space after 造, the query string truncates
to ﾊﾞﾙﾌﾞ. Interestingly enough, if I insert a fullwidth space after 等,
the query string truncates to ﾊﾞﾙﾌﾞ構.

Also, the query string ﾃﾞｨｾﾞﾙの際の比較からもそうです does not truncate,
until I place a fullwidth space after ﾙ, at which point it truncates to ﾃﾞｨｾ.

Anyway, thanks for your help.

Cheers,

[Ginstrom IT Solutions]

> On Behalf Of Steven P. Venti

> For example, the query string ﾊﾞﾙﾌﾞ構造等複雑にしています does not
> truncate. But if I insert a fullwidth space after 造, the query string
> truncates to ﾊﾞﾙﾌﾞ. Interestingly enough, if I insert a fullwidth space
> after 等,
> the query string truncates to ﾊﾞﾙﾌﾞ構.

Thanks for the added examples. I'm using Office 2010, and Felix is segmenting all of these cases correctly. Are you using a different version of Office?

[Steven P. Venti]

"Ginstrom IT Solutions" <sup...@felix-cat.com> wrote:

> I'm using Office 2010, and Felix is segmenting all of these cases correctly.
> Are you using a different version of Office?

No, I'm using Office 2010 as well, and although I get slightly different
behavior between Word and PowerPoint, the problem itself is essentially the
same.

It could be that the cause is in some screwed up setting in my OS, since the
entire system was in Japanese when I bought this machine new, but then
I switched everything to English, then later switched just the non-unicode
language back to Japanese.

お騒がせしました。