>>>>> "Jean-Marc" == Jean-Marc Lasgouttes <[EMAIL PROTECTED]> writes:
Jean-Marc> An alternative way of doing it would be not to use
Jean-Marc> cursorLeftOneWord at all, but do the move to the left 'by
Jean-Marc> hand'. Would you prefer that?
OK, I just did that and it turns out to be much simpler. Additionally,
the following patch does the following:
- use Paragraph::isWord, like cursorLeftOneWord does; the old code was
not consistent, since the criterion for finding the beginning of the
word was not the same than for finding the right part
- change Paragraph::isWord to report true for insets that have the
isLetter property. This is the case for ligature breaks and
hyphenation points. Now these characters are not a problem anymore
for index entries, for example.
I like this patch so much that I am just going to apply it :)
Note that there are still two ways to define what a word is:
- IsLetter lists all the characters that can be part of a word.
This is used by the spellchecker
- isWord is true except if the character is a known word separator.
This is used by the word-related cursor functions.
Of course having two ways of defining the same thing is not very
healthy... We'll fix it later, I guess.
JMarc
Index: src/ChangeLog
===================================================================
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/ChangeLog,v
retrieving revision 1.1928
diff -u -p -r1.1928 ChangeLog
--- src/ChangeLog 27 Jun 2004 13:14:07 -0000 1.1928
+++ src/ChangeLog 29 Jun 2004 15:25:21 -0000
@@ -1,3 +1,11 @@
+2004-06-29 Jean-Marc Lasgouttes <[EMAIL PROTECTED]>
+
+ * paragraph.C (isWord): return true on insets that report
+ isLetter().
+
+ * text.C (getWord): use Paragraph::isWord to decide what is in a
+ word and what is not; fix bug 1609.
+
2004-06-27 Jürgen Spitzmüller <[EMAIL PROTECTED]>
* tex-strings.C: add "none" to string_paperpackages[], fixes
Index: src/text.C
===================================================================
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/text.C,v
retrieving revision 1.564
diff -u -p -r1.564 text.C
--- src/text.C 17 May 2004 11:28:27 -0000 1.564
+++ src/text.C 29 Jun 2004 15:25:21 -0000
@@ -1842,46 +1842,36 @@ void LyXText::getWord(CursorSlice & from
switch (loc) {
case lyx::WHOLE_WORD_STRICT:
if (from.pos() == 0 || from.pos() == from_par.size()
- || from_par.isSeparator(from.pos())
- || from_par.isKomma(from.pos())
- || from_par.isNewline(from.pos())
- || from_par.isSeparator(from.pos() - 1)
- || from_par.isKomma(from.pos() - 1)
- || from_par.isNewline(from.pos() - 1)) {
+ || !from_par.isWord(from.pos())
+ || !from_par.isWord(from.pos() - 1)) {
to = from;
return;
}
// no break here, we go to the next
case lyx::WHOLE_WORD:
- // Move cursor to the beginning, when not already there.
- if (from.pos() && !from_par.isSeparator(from.pos() - 1)
- && !(from_par.isKomma(from.pos() - 1)
- || from_par.isNewline(from.pos() - 1)))
- cursorLeftOneWord(bv()->cursor());
- break;
+ // If we are already at the beginning of a word, do nothing
+ if (!from.pos() || !from_par.isWord(from.pos() - 1))
+ break;
+ // no break here, we go to the next
+
case lyx::PREVIOUS_WORD:
// always move the cursor to the beginning of previous word
- cursorLeftOneWord(bv()->cursor());
+ while (from.pos() && from_par.isWord(from.pos() - 1))
+ --from.pos();
break;
case lyx::NEXT_WORD:
lyxerr << "LyXText::getWord: NEXT_WORD not implemented yet"
<< endl;
break;
case lyx::PARTIAL_WORD:
+ // no need to move the 'from' cursor
break;
}
to = from;
Paragraph & to_par = pars_[to.par()];
- while (to.pos() < to_par.size()
- && !to_par.isSeparator(to.pos())
- && !to_par.isKomma(to.pos())
- && !to_par.isNewline(to.pos())
- && !to_par.isHfill(to.pos())
- && !to_par.isInset(to.pos()))
- {
+ while (to.pos() < to_par.size() && to_par.isWord(to.pos()))
++to.pos();
- }
}
Index: src/paragraph.C
===================================================================
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/paragraph.C,v
retrieving revision 1.368
diff -u -p -r1.368 paragraph.C
--- src/paragraph.C 17 May 2004 11:28:26 -0000 1.368
+++ src/paragraph.C 29 Jun 2004 15:25:21 -0000
@@ -1524,7 +1524,9 @@ bool Paragraph::isLetter(pos_type pos) c
bool Paragraph::isWord(pos_type pos) const
{
- unsigned char const c = getChar(pos);
+ if (isInset(pos))
+ return getInset(pos)->isLetter();
+ value_type const c = getChar(pos);
return !(IsSeparatorChar(c)
|| IsKommaChar(c)
|| IsInsetChar(c));
