Subject: Re: Commit: fix that spelling + images bug
From: WJCarpenter (bill-abisource@carpenter.ORG)
Date: Fri Mar 23 2001 - 18:29:18 CST
>> I don't think this is correct. If I insert an object, no matter
>> what kind, it's in my mind definitely two different words before
>> and after this object. To merge these into one word seems to
>> resemble stenography, and I'm not sure that's a job for
>> AbiWord. :-)
dom> I can agree with this. It'll be a little difficult to implement,
dom> but totally doable.
Wait a minute (he says without trolling through *all* the code)...
Spell-check decides word boundaries in part by calling
UT_isWordDelimiter(). Someone went to the trouble to decide that
UCS_ABI_OBJECT represents part of a word. If it can also represent
embedded graphics objects, then that clearly isn't valid (anymore).
Fragment from ut_misc.cpp UT_isWordDelimiter():
// This is for the unicode character used to represent AbiWord
// objects (which are used among other things for computed
// fields which should be considered words).
if (UCS_ABI_OBJECT == currentChar) return false;
Maybe the real problem is either that:
1. There should be different placeholders for textual things versus
non-textual things.
2. Or (if UT_isWordDelimiter() is used eleswhere [I suspect it's
not]), then at the point of calling it, spell-check knows it's
seeing a non-field thing and can stop its isolation algoritm.
-- bill@carpenter.ORG (WJCarpenter) PGP 0x91865119 38 95 1B 69 C9 C6 3D 25 73 46 32 04 69 D6 ED F3
This archive was generated by hypermail 2b25 : Fri Mar 23 2001 - 18:29:55 CST