Philippe Lhoste wrote:
Robert Roessler wrote:
You might look at the LexCaml source - specifically, the handling for
the SCE_CAML_IDENTIFIER case element.
Whenever it looks like we have hit a token which *could* be either an
identifier *or* a keyword, it is initially colored as an identifier.
Then, when we come to the end of the token, we check to see if it has
a match in any of the 3 keyword lists. If so, then we recolor it (by
changing the state/color) as a keyword of the appropriate color.
Thank you, but for the quick look I took, it seems to be the classical
sc.ChangeState() happening just after we quit the identifier state, ie.
on first non-word char found after this identifier.
ChangeState can only change the latest give state, on the current
segment, ie. from the latest SetState.
I need to recolorize a full segment, which can have various styles
already, often after meeting other styles as well:
MsgBox = value <--- We meet DEFAULT spaces before finding the = that
tell to recolorize MsgBox from COMMAND (back) to IDENTIFIER.
This+Is*A.Legal/Label!:: <-- When I see the double colon, I must
recolorize the whole segment as label...
Yes, that does look nasty... but does your recoloring need to
[possibly] go back over line boundaries, or will it always be confined
to a single line? If the single-line case, then it just can't be that
bad, since you have already typed the tokens you need to scan back
over - even if you will be recoloring them when you find the first one
in your sequence.
Robert Roessler
[EMAIL PROTECTED]
http://www.rftp.com
_______________________________________________
Scintilla-interest mailing list
[email protected]
http://mailman.lyra.org/mailman/listinfo/scintilla-interest
