Hi again, Przemek
----- Original Message -----
From: "Przemyslaw Czerpak" <[EMAIL PROTECTED]>
To: "Harbour Project Main Developer List." <harbour@harbour-project.org>
Sent: Thursday, May 08, 2008 3:07 AM
Subject: Re: [Harbour] Re: .ppo selectively skips #line ... directives
In current PP <*id*> should work like in Clipper allocating data
to the end of processed line. AFAIR in the old PP ; was additional
stop condition. In some cases it may be useful to have some stop
tokens for wild markers. I left note about it in ppcore.c[3596]
ca-clipper's wild match marker grabing beyond ; tokens makes sense since it
doesn't support multiple statement pp directive continuations except for 1
level, but for harbour that's sth to consider; to avoid introducing more
than one new marker f.i. <;marker;> could be useful for mentioned additional
stop condition, but imo it's more important to have wild match marker
variant which acts like ca-clipper but leaving multiple statement continued
pp directives untouched
#xCOM FOO <#wild#> => ok()
#xCOM TEST => FOO command1 ; command2 ; #include
"header.ch"
proc main
TEST
should show in .ppo as
ok() ; #include "header.ch"
The second rules is used for %% and ^^ says that expression cannot
contain some combinations of operators. There are much more similar
combinations, f.e. %*, +*, /+, etc. f.e.:
#trans a( <x> ) => qout( "translated", #<x> )
// not translated
a( %% )
[> snip ]
You do not have to use :: %% or ^^ in the above example. You can use
anything what will cause that in:
a( <x> )
<x> will not be single expression for CA-Clipper. F.e. add to above
code:
#xTRAN __ =>
#xTRAN 1_ => // these are two separated tokesns 1 and _ for PP
and then test:
a( __ b( .T. ))
a( 1_ b( .F. ))
however, when using repeating clause, there's still this difference
// in this test, none of %% ^^ :: are preliminary translated by
// #xTRAN %% =>
// #xTRAN :: =>
// #xTRAN ^^ =>
#xTRAN foo( [<x>] ) => [<x>] OK
#xTRAN test( .T. ) => IS()
-------------------- // ca-clipper // harbour
foo( !! test( .T. )) // !! IS() OK // ! ! IS() OK
foo( ! ! test( .T. )) // ! ! IS() OK // ! ! IS() OK
foo( %% test( .T. )) // foo( %% IS()) // % % IS() OK
foo( % % test( .T. )) // foo( % % IS()) // % % IS() OK
foo( ^^ test( .T. )) // foo( ^^ IS()) // ^ ^ IS() OK
foo( ^ ^ test( .T. )) // foo( ^ ^ IS()) // ^ ^ IS() OK
foo( :: test( .T. )) // foo( :: IS()) // :: IS() OK
foo( : : test( .T. )) // foo( : : IS()) // : : IS() OK
foo( *+ test( .T. )) // foo( *+ IS()) // * + IS() OK
foo( * + test( .T. )) // foo( * + IS()) // * + IS() OK
foo( ++ test( .T. )) // ++ IS() OK // ++ IS() OK
foo( + + test( .T. )) // foo( + + IS()) // + + IS() OK
foo( -- test( .T. )) // -- IS() OK // -- IS() OK
foo( - - test( .T. )) // - - IS() OK // - - IS() OK
foo( test( .T. )) // IS() OK // IS() OK
foo( . test( .T. )) // . IS() OK // . IS() OK
meanwhile is clear that ^^ is a valid harbour operator, so i'll leave that
one aside from here on
#xTRAN a( [<x>] ) => qout( "translated" [,#<x>] )
--------------- ca-clipper ----- harbour -----
a( %% x y ) // a( %% x y ) // qout( "translated","%% x","y" )
a( % % x y ) // a( % % x y ) // qout( "translated","% % x","y" )
imo, %% as forcer or delayer appears to still be a good option, but it
doesn't work in harbour as in ca-clipper and is thus not usable as forcer in
harbour
frank
No virus found in this outgoing message.
Checked by AVG.
Version: 8.0.100 / Virus Database: 269.23.10/1421 - Release Date: 7/05/2008
17:23
_______________________________________________
Harbour mailing list
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listinfo/harbour
