* harbour/contrib/libct/ctc.c
* harbour/contrib/libct/files.c
* harbour/contrib/libnf/ftattr.c
* harbour/contrib/libnf/n2color.c
* cleaned some BCC warnings
* harbour/include/hbapi.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbsetup.ch
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* harbour/source/common/hbver.c
* harbour/source/common/hbverdsp.c
* harbour/source/compiler/hbcomp.c
* harbour/source/vm/arrays.c
* harbour/source/vm/arrayshb.c
* harbour/source/vm/cmdarg.c
* harbour/source/vm/hvm.c
! fixed Harbour support for accessing string characters with array
index operators []
+ added support for #define HB_COMPAT_XHB - it's disabled by default.
When set it enables some of xHarbour features like negative array
indexes calculated from tail, accessing string characters with []
operators (warning it's not compatible with default Harbour version
enabled by -ks during compilation and //flags:s at runtime - in
xHarbour compatibility mode "ABC"[2] returns "B" when in Harbour 66)
and some other minor features.
* harbour/source/compiler/cmdcheck.c
* added support for optional setting C output type in -go switch
Now it accepts optional digit [0123] after -go in the same way
as -gc, f.e.: -go3
* harbour/source/compiler/harbour.c
* call hb_compExprLstDealloc() only when errors appeared during
compilation - it disable automatic freeing all allocated expressions
in clean compilation. I make it intentionally for easier detecting
problems one some expressions are not freed. So far the final cleanup
code always deallocated expressions what can hide some real bugs.
* harbour/source/vm/hvm.c
! fixed very serious bug in HVM closing. The s_uiActionRequest flag
was not cleared before hb_rddShutDown() so when RDD buffers were
hot and flushing required some PCODE evaluation (f.e. index KEY
or FOR expression) then they were not properly evaluated.
* harbour/source/vm/macro.c
* minor improvement
* harbour/utils/hbtest/rt_misc.prg
! fixed one of expected result, it was neither Harbour nor Clipper
result (I've just made some test with Clipper), I do not know why
it existed
* harbour/contrib/rdd_ads/adsfunc.c
* harbour/source/rdd/workarea.c
* formatting
* harbour/include/hbapi.h
* harbour/source/vm/extend.c
+ added missing hb_storptrGC()
* harbour/include/hbapicdp.h
* harbour/source/rtl/cdpapi.c
+ added parameter to unicode translation function to protect control
code translation
* formatting
* harbour/source/rtl/gtsln/kbsln.c
* harbour/source/rtl/gtxwc/gtxwc.c
* use new unicode translation function format. It also fixes problem
with space translation in GTLSN used in UTF8 mode
* harbour/include/hbapiitm.h
* harbour/source/vm/itemapi.c
+ added hb_itemCopyToRef(), hb_itemMoveToRef(), hb_itemMoveRef(),
hb_itemUnRefWrite()
* harbour/include/hbexprb.c
! use HB_P_DUPLUNREF instead of HB_P_DUPLICATE when assign is used in
push context - Clipper compatible behavior, see:
x:=10
f(y:=iif(.t.,@x,))
? x, y
func f(x)
x+=100
return nil
* harbour/include/hbver.h
* harbour/source/common/hbver.c
* changed status to "devel"
* harbour/source/common/expropt2.c
* disabled setting NIL when iif(.t.,,sth()) is optimized to avoid
warning when iif() is used as statement - to fix number of parameters
passed to function in code like f(iif(.t.,,sth())) we will need a
little bit deeper modification
* harbour/source/compiler/complex.c
* added hack for potential problems caused by WITH OBJECT syntax, f.e.
in code like:
proc p(obj,x)
case:=myCaseClass():new()
retu:=myCaseClass():new()
WITH OBJECT obj
do case
case x==nil
case:value()
case:action()
retu:sth()
end
END
case:action()
retu:sth()
return
case:action(), case:value(), retu:sth() can means two different things
and compiler cannot guess which one should use. It cannot be fixed
because it's caused by bad WITH OBJECT syntax. In xHarbour constractions
like case:value() or retu:sth() are always used as in WITH OBJECT
context what is not Clipper compatible. I do not want to introduce
such limitations. We have two choices: change WITH OBJECT syntax or
add some hack to "hide" the problem a little bit. Now I intorduce
ugly hack which check number of spaces between ':' so
case :action()
is compiled as case condition using WITH OBJECT message and
case:action() as sending message to object in case variable
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
+ added support for passing base value to FOR EACH by reference.
* harbour/source/vm/arrays.c
* use hb_itemMoveRef() in AINS()/ADEL() to avoid creation of cyclic
reference chains
* harbour/source/vm/debug.c
* harbour/source/vm/memvars.c
* use hb_itemCopyToRef() to avoid creation of cyclic reference chains
* harbour/source/vm/classes.c
* harbour/source/vm/hvm.c
* use hb_itemMoveToRef() to avoid creation of cyclic reference chains
+ added support for for writable string enumerators, f.e.:
proc main()
local s, c
s:="abcdefghijklmnopqrst"
for each c in @s
if c$"aeio"
c:="*"
endif
next
? s // "*bcd*fgh*jklmn*pqrst"
return
* harbour/source/compiler/hbusage.c
* update the year in hb_compPrintLogo() from 2006 to 2007
* harbour/source/vm/garbage.c
! fixed potentially very serious bug - when I added array items
references I forgot to update GC. Fortunately it was caught
by code to detect bugs in object destructors so only RT error
was reported instead of internal HVM memory corruption
* harbour/tests/hbinline.prg
* updated comments
* harbour/include/hbexpra.c
! fixed GPF which can appear on syntax error during compilation
one of enumerator message
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* accept the folowing syntax: (@var):msg
With this modification FOR EACH overloading is completed.
* harbour/source/vm/hvm.c
* small modification to protect against creating long references chains
which can hardly reduce the performance in some cases
* harbour/include/hbcomp.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbexprop.h
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* harbour/source/compiler/Makefile
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/harbour.yyh
* harbour/source/compiler/hbcomp.c
* harbour/source/compiler/hbusage.c
* harbour/source/compiler/ppcomp.c
* harbour/source/macro/Makefile
* harbour/source/macro/macro.y
* harbour/source/macro/macro.yyc
* harbour/source/macro/macro.yyh
* harbour/source/vm/macro.c
* harbour/source/vm/memvars.c
* harbour/source/vm/memvclip.c
* keep prefix definition for bison public symbols in *.y files instead
of passing them as bison arguments to avoid possible mistakes during
*.yy[hc] file generation
! fixed iif() reduction when used as single function parameter
+ add -kM compiler switch - turn off macrotext substitution
* updated <string> + <string> optimization and other macrotext usage
to keep more Clipper compatible behavior
! fixed typo in expression type checking in IIF() reduction - it was
working because wrongly used macros has the same values as the valid
ones
- removed compiler switches and macros for old HB_INLINE() syntax - it
was not supported after last months modifications. Only:
HB_INLINE([params,...]) { C code }
is supported.
! fixed passing parameters by reference in IIF() statement and macro
compiler. The following valid Clipper code:
&("myfunc(<exp>, @var1, @var2)")
was not supported
! do not allow to use references or empty expressions in expression list
AFAIK it's Clipper compatible. If I'm wrong then please fix me but
Clipper allows to use references only as function arguments, 2-nd and
3-rd IIF() argument and as directly created array items ( {...@var...} )
DO ... WITH ... is a special case of function call.
! Changed grammar rules to accept only Clipper compatible parameters.
! fixed typo in macro name in hb_memvarNewParameter() function.
* use HB_FUNC_EXEC(...) instead of HB_FUNCNAME(...)()
* harbour/include/hbexprb.c
* make function call optimizations only when hb_compFunCallCheck()
do not report errors
* harbour/source/common/reserved.c
+ added VALTYPE to list of reserved words and shortcuts (Clipper
compatible)
* harbour/source/compiler/hbfunchk.c
* checking for number of parameters in FILE() function call is Harbour
extension - do not enable it when HB_C52_STRICT macro is set
* harbour/source/rtl/filehb.c
* formatting
* harbour/include/hbcompdf.h
+ added comment
* harbour/include/hbexpra.c
* change expression type for <exp> = <exp> from HB_EO_EQUAL
to HB_EO_ASSIGN in hb_compExprGenStatement() function so
later we can safely make all optimization in hb_compExprUseEqual()
during reduce process - optimization only when PCODE is generated
is too late and causes that expressions like:
iif( 1 = 1, <exp>, <exp> )
are not reduces (Clipper reduce them)
* do not execute hb_compFunCallCheck() in hb_compExprNewFunCall()
to avoid repeating the same error message in hb_compExprUseFunCall()
* harbour/include/hbexprb.c
* enabled reduction for hb_compExprUseEqual() and generate errors
when it's used as statement
* harbour/source/common/expropt2.c
* added some missing reductions for expressions like: NIL == NIL
* harbour/source/common/reserved.c
% very basic optimization: use binary search instead of linear
scanning for reserved words and shortcuts - it gives noticeable
speed improvement in macro compiler
* harbour/source/compiler/complex.c
* changed the order of special keywords to keep alphabetic sorting
* harbour/source/compiler/harbour.c
% some minor optimizations in aliased expressions
! added protection in hb_compIsJump() to avoid possible problems
when this function is executed after replacing some code with
jumps by series of NOOPs in optimizations process
* harbour/source/compiler/hbfix.c
* better optimization of PCODE generated for logical expressions
* harbour/source/compiler/hbfunchk.c
% use binary search instead of linear scan
* harbour/source/rtl/minmax.c
! fixed GPF when MIN() or MAX() function is called with less then
two parameters
* harbour/source/vm/itemapi.c
* minor optimization and formatting
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/source/common/expropt1.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/hbgenerr.c
* harbour/source/macro/macro.l
* harbour/source/macro/macro.y
* harbour/source/macro/macro.yyc
* harbour/source/macro/macro.yyh
* harbour/source/macro/macrolex.c
* harbour/source/vm/macro.c
* final grammar cleanup: removed unused and/or unnecessary grammar
expressions in *.y files, fixed some differences between compiler
and macrocompiler, fixed some other bugs, typos, clipper
incompatibilities, etc.
It's end of year, it's late and sorry but I'd like to begin making
something absolutely differ then creating detail description ;-)
If you will have any questions then I'll answer in new year.
* harbour/source/rdd/dbffpt/Makefile
* small hack for XCC compilation
best regards and happy new year,
Przemek
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbpcode.h
* harbour/include/hbpp.h
* harbour/include/hbxvm.h
* harbour/source/compiler/genc.c
* harbour/source/compiler/gencc.c
* harbour/source/compiler/gencli.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/hbgenerr.c
* harbour/source/compiler/hbpcode.c
* harbour/source/vm/hvm.c
* harbour/source/vm/macro.c
* changed HB_P_MACROPUSHINDEX PCODE behavior - now it does not evaluated
macro expression
* eliminated HB_ET_MACRO_INDEX - HB_ET_MACRO_LIST can be used instead
* accept ... in codeblocks only when codeblocks parameters are defined
as variables, f.e: {|a,b,...|qout(a+b,...)}
! fixed error in a[&s]:=100
+ added support for a[...] - now ... can be used in the same context as
xbase++ like multivalue macro (a[&s])
* disabled ++, --, <op>= optimizations when -kc (strict Clipper
compatibility mode) compiler switch is used
* harbour/include/hbcompdf.h
* removed not longer used pFunCall from asMacro expression.
This item was used only in a very old s:="1,2";f(&s) implementation.
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprop.h
* harbour/source/common/expropt1.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* a little bit more simple implementation of passing optional arguments
and simplified some code. Now they can be used in any place in argument
list (also repeated). It's also possible to create arrays using "..."
as array element(s), f.e.: func f(...); return { "{", ..., "}" }
If you think it will be usable then I can also add support for accessing
array items with optional parameters, f.e.: ? a[ ... ]
so it can be used in the same context as xbase++ like macro lists
but for this I will need new PCODE so I'd like to hear your opinion
ASAP - I'm closing any extensions today.
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hberrors.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbpcode.h
* harbour/include/hbxvm.h
* harbour/source/common/expropt1.c
* harbour/source/compiler/genc.c
* harbour/source/compiler/gencc.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/hbdead.c
* harbour/source/compiler/hbfix.c
* harbour/source/compiler/hbgenerr.c
* harbour/source/compiler/hblbl.c
* harbour/source/compiler/hbpcode.c
* harbour/source/compiler/hbstripl.c
* harbour/source/vm/hvm.c
+ added support for passing optional parameters to called function, f.e.:
proc p( a, b, ... )
qout( a, b, "X", ... )
* forbid creating function pointer when function is used with parameters,
f.e.: x:=@f1(1,2,3)
* harbour/source/compiler/complex.c
* cleaned all restrictions on reserved words I've found - if I missed
sth then please inform me. The previous behavior and some error
messages are used with -kc compiler switch
* harbour/source/rtl/symbol.prg
* harbour/source/rtl/tobject.prg
! eliminated limitations in number of constructor parameters
(new syntax with passing optional parameters used)
* harbour/include/hbcompdf.h
* harbour/include/hbexprb.c
* harbour/source/common/expropt1.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/hbcomp.c
- removed not longer used fExternal from compiler structure
* allow to pass array items by reference ( func(@a[1]) ) and some
cleanups in reference rules. Probably we should also block using:
@func(param1 [,param2,[,...paramN]])
and accept only @func()
! fixed using statements in codeblocks used to initialize static
variables and some other cleanups (f.e. allow to use as static
initializers any functions which are later eliminated by expression
optimizer.
* harbour/include/hbexprb.c
* do not generate HB_P_FUNCPTR - it's not longer necessary
+ added optimization for <exp> + 1, <exp> - 1 - disabled by default
because it changes error messages in hbtest, code like NIL + 1
generates 'argument error ++' instead of 'argument error +'. I'd like
you will decide what you prefer - faster code or strict Clipper error
messaged.
* harbour/source/common/expropt2.c
+ added optimizations for <num> + <date>
+ added ooptimizations for <exp> + 0, 0 + <exp>, <exp> - 0
This is also sth what may interact with operators overloading in object
system. When we will have strong typing then we should think about an
option which will optionally disable some optimizations - someone may
want to define arithmetic where <obj> + 0 gives differ then <obj>
results.
* harbour/include/hbcomp.h
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/harbour.yyh
+ added hb_compStatmentStart()
! restrict MEMVAR and FIELD usage - now they have to be located before
executable statements like in Clipper.
! generate error when PARAMETERS is used as file wide declaration.
! generate errors when different executable statements are used before
first procedure - now such code was simply ignored without any errors.
! generate valid error messages when some declarations are used in
wrong places
! fixed setting begin of executable statement flag in different .prg
constructions.
* harbour/contrib/rdd_ads/adsfunc.c
* do not use hb_param( pItem, HB_IT_BYREF ) but ISBYREF() macro
* harbour/include/hbdefs.h
* formatting
* harbour/include/hbexprb.c
! fixed code generated for logical expressions when shortcutting
is enabled to be Clipper compatible
* harbour/source/compiler/cmdcheck.c
! added missing break in -z compiler switch decoding
* harbour/source/compiler/hbfix.c
% added optimizations for code generated for logical expressions,
it strips a chain of HB_P_DUPLICATE -> HB_P_JUMP[TRUFALES] ->
HB_P_DUPLICATE -> HB_P_JUMP[TRUFALES] -> ...
In fact it should not be done here but because we do not have
any metacode level where such optimizations should be done and
PCODE is generated online then it will have to be here. Maybe
in the future we will add metacode.
* harbour/source/rdd/dbcmd.c
* make "M" alias Clipper compatible. After some thoughts I decided
that in some cases it can be usable so now "M" will set WA number
to 65535 like in Clipper.
* harbour/source/rdd/dbf1.c
* allow to open VFP DBFs with _NullFlags system command
* harbour/source/rdd/delim1.c
* harbour/source/rdd/sdf1.c
! updated field size calculation for some some extended types.
* harbour/source/rdd/workarea.c
* generate RT error when CREATEFIELD() method does not accept some
parameters to avoid silent DBCREATE() failing.
* harbour/source/vm/macro.c
* removed hb_compMemvarCheck() - this function was not dooing
anything - only slows the TYPE() by calling hb_dynsymFind()
one or two times.
+ added TODO note: memvars create by TYPE() have PUBLIC scope
in Clipper. Clipper also always tries to evaluate valid expressions
inside TYPE even if the contain UDF. UDF only stops evaluation
so:
TYPE( "(val:=1) + f() + (val2:=2)" )
creates in clipper public variable VAL but not VAL2.
* harbour/contrib/rdd_ads/adsfunc.c
* do not use hb_param( pItem, HB_IT_BYREF ) but ISBYREF() macro
* harbour/include/hbdefs.h
* formatting
* harbour/include/hbexprb.c
! fixed code generated for logical expressions when shortcutting
is enabled to be Clipper compatible
* harbour/source/compiler/cmdcheck.c
! added missing break in -z compiler switch decoding
* harbour/source/compiler/hbfix.c
% added optimizations for code generated for logical expressions,
it strips a chain of HB_P_DUPLICATE -> HB_P_JUMP[TRUFALES] ->
HB_P_DUPLICATE -> HB_P_JUMP[TRUFALES] -> ...
In fact it should not be done here but because we do not have
any metacode level where such optimizations should be done and
PCODE is generated online then it will have to be here. Maybe
in the future we will add metacode.
* harbour/source/rdd/dbcmd.c
* make "M" alias Clipper compatible. After some thoughts I decided
that in some cases it can be usable so now "M" will set WA number
to 65535 like in Clipper.
* harbour/source/rdd/dbf1.c
* allow to open VFP DBFs with _NullFlags system command
* harbour/source/rdd/delim1.c
* harbour/source/rdd/sdf1.c
! updated field size calculation for some some extended types.
* harbour/source/rdd/workarea.c
* generate RT error when CREATEFIELD() method does not accept some
parameters to avoid silent DBCREATE() failing.
* harbour/source/vm/macro.c
* removed hb_compMemvarCheck() - this function was not dooing
anything - only slows the TYPE() by calling hb_dynsymFind()
one or two times.
+ added TODO note: memvars create by TYPE() have PUBLIC scope
in Clipper. Clipper also always tries to evaluate valid expressions
inside TYPE even if the contain UDF. UDF only stops evaluation
so:
TYPE( "(val:=1) + f() + (val2:=2)" )
creates in clipper public variable VAL but not VAL2.
* harbour/include/hbexprop.h
* harbour/source/common/expropt1.c
+ added hb_compExprParamListLen()
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* use hb_compExprParamListLen()
* harbour/include/hbexprc.c
* generate error message for syntax like:
o:var()++ or o:var()+=10
only:
o:var++ and o:var+=10
is supported, just like for @o:var
* harbour/include/hbapi.h
* harbour/source/common/hbstr.c
+ hb_numDecConv() used to convert double value to/from binary integer
fields with fixed number of decimal places
* harbour/common.mak
* harbour/include/hbapicdp.h
* harbour/source/codepage/Makefile
+ harbour/source/codepage/cptrdos.c
+ harbour/source/codepage/cptrwin.c
* harbour/source/lang/Makefile
+ harbour/source/lang/msgtrdos.c
+ harbour/source/lang/msgtrwin.c
+ added Turkish CPs and MSGs created by Bicahi Esgici
+ harbour/source/codepage/uc1254.c
+ harbour/source/codepage/uc857.c
+ added unicode tables for Turkish CPs
* harbour/source/rdd/dbf1.c
* harbour/source/rdd/workarea.c
+ added support for decimal places in binary integer fields ("I")
! fixed currency field decoding ("Y") - it's integer binary value
not IEEE758 double. Warning: on write the value is rounded not
truncated like in VFP.
* harbour/config/w32/bcc32.cf
* cleanup
+ harbour/config/w32/xcc.cf
+ added XCC support
* harbour/contrib/btree/hb_btree.c
* use hb_vmAtInit()/hb_vmAtExit() instead of INIT/EXIT functions defined
from C code
* harbour/contrib/libct/tab.c
* casting
* harbour/contrib/odbc/odbc.c
! removed #include <malloc.h> - it should not be used with new C
compilers
I think that we should replace this library with hbodbc from xHarbour.
* harbour/contrib/ole/ole2.c
* updated for XCC
* harbour/include/hbdefs.h
* include stdint.h if available
* harbour/source/compiler/complex.c
! fixed yet another stupid mistake in WITHOBJECT token
* harbour/include/hbpcode.h
* harbour/include/hbxvm.h
* harbour/source/compiler/genc.c
* harbour/source/compiler/gencc.c
* harbour/source/compiler/hbdead.c
* harbour/source/compiler/hbfix.c
* harbour/source/compiler/hblbl.c
* harbour/source/compiler/hbpcode.c
* harbour/source/compiler/hbstripl.c
* harbour/source/vm/hvm.c
+ added HB_P_SWAP <n> PCODE
* harbour/include/hbcompdf.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbexprop.h
! fixed GPF during compilation of @:var, @:&var, @o:&var
+ added support for @o:&var and @:var, @:&var inside WITH OBJECT
statement
* changed PCODE generated of ++,--,+=,-=,...
Now left side expression is evaluated _ONLY_ once and when object
messages are used it's guarantied that exactly the same object
variable will be used. It also fixes some problems which exists
in Clipper. This optimization is enabled by -kh flag (by default)
and can be disabled with -kc.
% optimize ald macros in ++,--,+=,-=,... operations
* add automatically "_" prefix when macro message is used in assignment
context, f.e.:
s:="osCode"
o:=errorNew()
? o:&s
o:&s := 100
? ++o:&s
? o:&s *= 5
* harbour/include/hblang.ch
- removed #xtranslate - this file is included by C code and some C
compiler do not like unknown directives
* harbour/utils/hbtest/rt_math.prg
+ added test code for <op>assign and (pre/post)(inc/dec)rementation,
macro messages and WITH OBJECT
* harbour/source/compiler/harbour.c
* harbour/source/rdd/dbf1.c
* harbour/source/rtl/errorapi.c
* harbour/source/rtl/hbgtcore.c
* harbour/source/rtl/gtdos/gtdos.c
* harbour/source/rtl/gtos2/gtos2.c
* harbour/source/rtl/gtpca/gtpca.c
* harbour/source/rtl/gtsln/gtsln.c
* harbour/source/rtl/gtstd/gtstd.c
* harbour/source/rtl/gtwin/gtwin.c
* casting and warning cleanup
* harbour/utils/Makefile
+ added $(HB_UTILS)
- harbour/utils/hbpp/hbpp.h
+ harbour/utils/hbpp/hbppdef.h
* harbour/utils/hbpp/hbpp.c
* harbour/utils/hbpp/hbppcomp.c
* harbour/utils/hbpp/hbppcore.c
* harbour/utils/hbpp/hbpplib.c
* harbour/utils/hbpp/hbpptbl.c
* harbour/utils/hbpp/pragma.c
* renamed hbpp.h to hbppdef.h to avoid possible conflict with hbpp.h
in include directory
* casting and warning cleanup
* harbour/include/hbapi.h
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hbdefs.h
* harbour/include/hberrors.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbpcode.h
* harbour/include/hbxvm.h
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* harbour/source/common/hbstr.c
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/genc.c
* harbour/source/compiler/gencc.c
* harbour/source/compiler/gencobj.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/hbdead.c
* harbour/source/compiler/hbfix.c
* harbour/source/compiler/hbgenerr.c
* harbour/source/compiler/hblbl.c
* harbour/source/compiler/hbpcode.c
* harbour/source/compiler/hbstripl.c
* harbour/source/macro/macro.y
* harbour/source/macro/macro.yyc
* harbour/source/rtl/console.c
* harbour/source/rtl/isprint.c
* harbour/source/rtl/left.c
* harbour/source/rtl/right.c
* harbour/source/rtl/strtran.c
* harbour/source/vm/codebloc.c
* harbour/source/vm/hvm.c
* harbour/source/vm/macro.c
* general PCODE cleanup and address most of TODO/TOFIX notes in
source code:
! fixed GPF traps when too long string or codeblock is generatd
+ added support for 16MB codeblocks and strings
! removed macrocompiler limitation for jumps range
! fixed GPF when more then 255 local variables is used and added
support for 2^15 locals
! removed all strtok() functions
% added optimization for all +=, -=, *=, %=, ^=, **= operations
when left side of expression is variable or array item
% added optimization for all +=, -=, *=, %=, ^=, **= operations
when left side of expression is object method and updated ++, --
for new code. It's still disabled until we will not add support
for late evaluated reference items to HVM
! fixed a[++i]++ and similar operations (a[++i]*=2, ...). Now ++i is
executed only once. It's not Clipper compatible but it was in
TODO note in source code. It can be disabled by -kc option
* finished support to xHarbour like #pragma TEXTHIDDEN(1)
! fixed local add int optimization when PARAMETERS used after
optimization changed local variable number over 255
! fixed GPF trap when in HB_P_<op>EQ PCODEs when executed for
direct values
* others
! fixed problems reported by Chen
* optimized strtran(), left(), right() to not create new string copy
when the same value is returned
* harbour/source/rtl/console.c
* pacify false CG error messages - Chen Kedem modifications (sizeof(int)
in #if replaced by hardcoded 4 for compilers which does not allow to
use sizeof() in #if directives)
* harbour/TODO
- removed console.c cleanup note from TOFIX
* harbour/include/hbcomp.h
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/source/common/expropt1.c
* minor code cleanup
* harbour/include/hbstack.h
! fixed typo in function name, should be: hb_stackLocalVariable()
not hb_stacklocalVariable()
+ added hb_stackId()
* hide hb_stack declaration when HB_STACK_MACROS is not set
* harbour/source/vm/dynlibhb.c
* use hb_stackId()
* harbour/source/vm/estack.c
+ added hb_stackId()
! fixed some condition used to check stack access - they were
reporting internal errors for valid code
* harbour/source/vm/hvm.c
! fixed code which executed hb_stackLocalVariable() with pointer
to SHORT not int
* harbour/source/rtl/memoline.c
! added fix sent to me by Randy Portnoff - it's a little bit differ
then the one Randy sent to list but this one seems to be correct.
I never used this function in Clipper and made only very basic
tests. I believe that Randy make better tests and this fix will
not break some other constructions.
* harbour/config/bsd/gcc.cf
* harbour/config/darwin/gcc.cf
* harbour/config/dos/bcc16.cf
* harbour/config/dos/djgpp.cf
* harbour/config/dos/owatcom.cf
* harbour/config/dos/rsx32.cf
* harbour/config/hpux/gcc.cf
* harbour/config/linux/gcc.cf
* harbour/config/linux/owatcom.cf
* harbour/config/os2/gcc.cf
* harbour/config/os2/icc.cf
* harbour/config/sunos/gcc.cf
* harbour/config/w32/bcc32.cf
* harbour/config/w32/gcc.cf
* harbour/config/w32/mingw32.cf
* harbour/config/w32/msvc.cf
* harbour/config/w32/rsxnt.cf
* harbour/config/w32/watcom.cf
* added respecting HB_GT_DEFAULT, though it's quite possible that I'll
remove it at all soon.
* some other cleanup
* harbour/source/compiler/complex.c
! fixed wrongly replicated from FLEX lexer condition used to DECLARE.
Now they should be the same as in FLEX lexer but IMHO they are not
Clipper compatible.
* harbour/source/compiler/gencli.c
* harbour/source/compiler/gencobj.c
* harbour/source/compiler/genhrb.c
* harbour/source/compiler/genjava.c
* harbour/source/compiler/genobj32.c
* harbour/source/compiler/harbour.c
* harbour/source/rtl/hbffind.c
* harbour/source/pp/ppcore.c
* harbour/source/rdd/workarea.c
! fixed sizes of some C stack buffers and strncpy() parameters
* use hb_strn*() instead of strn*() in few places to be sure that 0
is always stored in destination buffer
* harbour/source/rtl/gtcgi/gtcgi.c
* change reported GT name from "Standard stream console" to
"Raw stream console" to make it differ then GTSTD one.
* harbour/bin/hb-func.sh
* harbour/bin/hb-mkslib.sh
* harbour/bin/postinst.sh
* harbour/config/bsd/gcc.cf
* harbour/config/darwin/gcc.cf
* harbour/config/dos/owatcom.cf
* harbour/config/hpux/gcc.cf
* harbour/config/linux/gcc.cf
* harbour/config/linux/owatcom.cf
* harbour/config/os2/gcc.cf
* harbour/config/sunos/gcc.cf
* harbour/config/w32/watcom.cf
* harbour/contrib/rdd_ads/ads1.c
* harbour/contrib/rdd_ads/adsfunc.c
* harbour/include/hbapifs.h
* harbour/include/hbexprc.c
* harbour/include/hbsetup.h
* harbour/source/common/expropt2.c
* harbour/source/common/hbdate.c
* harbour/source/common/hbgete.c
* harbour/source/common/hbstr.c
* harbour/source/common/hbver.c
* harbour/source/common/hbverdsp.c
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/fixflex.c
* harbour/source/compiler/gencobj.c
* harbour/source/compiler/genobj32.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.l
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/hbfix.c
* harbour/source/compiler/hbfunchk.c
* harbour/source/pp/ppcore.c
* harbour/source/pp/ppgen.c
* harbour/source/pp/pplib.c
* harbour/source/rdd/dbcmd.c
* harbour/source/rdd/workarea.c
* harbour/source/rdd/dbfcdx/dbfcdx1.c
* harbour/source/rdd/dbffpt/dbffpt1.c
* harbour/source/rdd/hsx/hsx.c
* harbour/source/rtl/accept.c
* harbour/source/rtl/ampm.c
* harbour/source/rtl/console.c
* harbour/source/rtl/defpath.c
* harbour/source/rtl/errorint.c
* harbour/source/rtl/filesys.c
* harbour/source/rtl/fkmax.c
* harbour/source/rtl/fstemp.c
* harbour/source/rtl/hbffind.c
* harbour/source/rtl/hbgtcore.c
* harbour/source/rtl/langapi.c
* harbour/source/rtl/seconds.c
* harbour/source/rtl/set.c
* harbour/source/rtl/tobject.prg
* harbour/source/rtl/transfrm.c
* harbour/source/rtl/gtcrs/chrmap.c
* harbour/source/rtl/gtcrs/gtcrs.c
* harbour/source/rtl/gtcrs/gtcrs.h
* harbour/source/rtl/gtpca/gtpca.c
* harbour/source/rtl/gtwvt/gtwvt.c
* harbour/source/rtl/gtxwc/gtxwc.c
* harbour/source/vm/classes.c
* harbour/source/vm/cmdarg.c
* harbour/source/vm/debug.c
* harbour/source/vm/estack.c
* harbour/source/vm/fm.c
* harbour/source/vm/hvm.c
* harbour/source/vm/macro.c
* harbour/source/vm/proc.c
* harbour/source/vm/runner.c
* harbour/utils/hbmake/hbmlang.c
* harbour/utils/hbpp/hbpp.c
* harbour/utils/hbpp/hbppcomp.c
* harbour/utils/hbpp/hbppcore.c
* harbour/utils/hbpp/hbpplib.c
* harbour/utils/hbpp/pragma.c
* harbour/utils/hbver/hbverfix.c
* general code cleanup: all strcpy() and strcat() replaced hb strn*()
and hb_strn*(), sprintf() by snprintf(), etc.
It fixed some possible buffer overflow but it's also possible that
it will exploit some some hidden so far problems but I strongly
prefer to know about them ASAP to fix them before final 1.0 release.
! fixed bug I introduce in codeblock falgs which disabled early macro
evaluation in codeblocks.
! added missing protection against execution by bison destructor for
used CBSTART token on syntax error.
% use PP stringify logik added for FLEX support to remove all redundant
spaces in stringified for early macro evaluation codeblocks and add
spaces between operators if after preprocessing they were removed.
Now we are supporting also some code which cannot be compiled by
Clipper.
% cleaned and improved the speed of printer redirecting in console code.
Chen please check if CG still reports problems and if not ubdate
TODO/TOFIX or inform me about them.
* some cleanups in building process for different *nixes: MacOSX, SunOS,
*BSD, etc.
* removed not longer necessary (I hope) compiler flag to force default
'char' type as signed. Now Harbour should work with signed and unsigned
char without any problems. To other developers: please remember that
when you are using 'char' as 8 bytes signed integer then it's a bug.
Always explicitly use: 'signed char' in such case. On some platforms
it's not possible to set default type for 'char' so if you hardcode
that 'char' is signed or unsigned somewhere then code will not be
portable. Chen if you will find a while to also rebuild Harbour with
BCC and -K switch to make some tests with CG then I'll be thankful.
* harbour/source/macro/macro.y
* harbour/source/macro/macro.yyc
* harbour/source/macro/macro.yyh
* harbour/source/macro/macrolex.c
* added support for extended string (e"...") to macro compiler and
strings with embedded ASCII NUL character (chr(0))
* harbour/source/rtl/valtostr.c
* added HB_STRTOEXP() function which converts string item to valid
expression which can be compiled by macro compiler. String may
contain any characters.
* harbour/source/rtl/persist.prg
* use HB_STRTOEXP() for sting conversion and 0d....... for date
constants
* harbour/source/rtl/readvar.prg
* harbour/source/rtl/xsavescr.c
* minor cleanup
* harbour/bin/pack_src.sh
+ added packing *.yy[ch] files
* harbour/makefile.bc
* harbour/makefile.vc
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbexprop.h
* harbour/include/hbpp.h
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* harbour/source/compiler/Makefile
* harbour/source/compiler/complex.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.l
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/harbour.yyh
* harbour/source/compiler/hbgenerr.c
* harbour/source/macro/macro.y
* harbour/source/macro/macro.yyc
* harbour/source/pp/ppcore.c
+ added hb_comp prefix to grammar/lexer compiler public functions to
reduce possible conflict with 3-rd party code which may use default
yy prefix.
! do not use bison destructors for expressions. Internal bison logic
cannot properly detect if expression was used or not in some of our
grammar rules and it's possible that some expressions will not be freed
and some other freed twice.
! added protection against multiple destructors execution for CBSTART
and LITERAL tokens
* added small garbage collector for deallocating expressions which were
not freed (such situation can happen in syntax errors)
% some optimizations in used structures to reduce their sizes
+ added protection against execution PCODE optimizations for functions
which were not cleanly compiled.
* harbour/source/rtl/idle.c
* use const in nanosec() timeout declaration
* harbour/include/hbcompdf.h
* harbour/source/compiler/harbour.l
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* removed declaration and all references to global variable hb_comp_data
it's not longer necessary after last modifications
* harbour/source/compiler/harbour.yyc
* synced compiler grammar parser
* harbour/source/macro/macro.yyc
! regenerated macro compiler grammar parser with -p hb_macro - I forgot
about this prefix before last commit
* harbour/include/Makefile
+ added hbcompdf.h and hbstdgen.ch
* harbour/source/compiler/harbour.y
* harbour/source/macro/macro.y
+ added #define __STDC__ as workaround for __BORLANDC__ which seems
to not properly understand ANSI C declarations. It should pacify
some of BCC warning messages about undeclared functions
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/harbour.yyh
* harbour/source/macro/macro.yyc
* harbour/source/macro/macro.yyh
* updated generated grammar parser files for above modification
in parser definition files (.y)
* harbour/source/compiler/harbour.y
* use HB_COMP_PARAM macro in expression destructor instead of
direct accessing hb_comp_data global variable to make compiler
code MT safe
+ harbour/source/compiler/harbour.yyc
+ harbour/source/compiler/harbour.yyh
+ harbour/source/macro/macro.yyc
+ harbour/source/macro/macro.yyh
+ added generated by bison 2.3 grammar parsers for compiler and
macro compiler
* harbour/config/dos/global.cf
* harbour/config/dos/install.cf
* harbour/config/w32/global.cf
* changed CP definition for COMMAND.COM so it can work as standalone
command and added DIRSEP macro
* harbour/config/c.cf
+ added support for HB_REBUILD_PARSER=no environment variable.
When it's set then during build process grammars parsers
will not be generated from grammar definitions files (.y)
but instead the predefined ones (.yyc) used.
Tested and works well in Linux and DOS
* harbour/makefile.bc
+ added support for HB_REBUILD_PARSER=no environment variable.
Not tested. It's even possible that I've just broken the BCC
compilation with this modification. Please check it and fix
me if necessary.
* harbour/source/compiler/harbour.c
! fixed typo in last modification which caused memory leak
* harbour/source/vm/codebloc.c
+ initialize freed codeblock structure with pointer to static PCODE
which returns NIL for buggy .prg destructors.
* harbour/ChangeLog
! fixed date of my last commit
* harbour/tests/codebl.prg
* harbour/tests/langmsg.prg
* harbour/utils/hbpptest/pp_test.prg
* harbour/utils/hbpptest/pretest.prg
* harbour/samples/pe/license.txt
* removed carriage return (\r) characters
* harbour/include/hbcomp.h
* harbour/include/hbapi.h
* harbour/include/hbcompdf.h
* harbour/include/hbexprb.c
* harbour/include/hbexprop.h
* harbour/source/vm/macro.c
* harbour/source/compiler/hbcomp.c
* harbour/source/compiler/complex.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.l
* harbour/source/compiler/harbour.y
* harbour/source/compiler/genjava.c
* harbour/source/compiler/hbcomp.c
* harbour/source/compiler/hbident.c
* harbour/source/compiler/hbpcode.c
* harbour/source/compiler/ppcomp.c
* separated HB_COMP and HB_MACRO definitions and added HB_COMMON
structure used when HB_COMMON_SUPPORT macro is set
* hide HB_MACRO internal definitions in hbapi.h when HB_MACRO_SUPPORT
is not set
These modifications should help in future compiler and macro compiler
integration into single binary.
+ added clipper compatible error message in unclosed structures
In general we should cleanup rules to make error reporting more
user friendly and Clipper compatible
% eliminated unnecessary allocations
+ added int mode member to common part of HB_COMP and HB_MACRO
structure. It's initialized to HB_MODE_MACRO or HB_MODE_COMPILER
+ add TOFIX note about memory leaks in cParamTypes and pParamClasses
members of COMDECLARED structure. It's a part of unfinished strong
typing code. I do not know what Ryszard plan to do with it so I'm
leaving it as is. The part of strong typing which were operating
on emulated HVM stack has been removed.
* removed static variables from genjava.c. Also gencli.c and genobj32.c
should be fixed but this will not make this code working so I haven't
touched it.
* changed 3-rd parameter of hb_compIdentifierNew() from BOOL to int.
Now it can have the following values: HB_IDENT_STATIC, HB_IDENT_FREE,
HB_IDENT_COPY. This modification will allow in the future using common
for static and dynamic symbol hash table without additional source code
changes. I also plan to use identifier hash table in PP the reduce
number of memory allocations and speed-up preprocessing by using
second level hashing for hash keys.
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* changed HB_MACRO_SUPPORT to HB_COMMON_SUPPORT to automatically detect
possible conflicts between compiler and macro compiler
! removed first detected conflict which existed for long time
+ added hb_compExprMacroAsAlias() function to convert HB_ET_VARIABLE
expression to HB_ET_ALIAS.
* harbour/source/macro/macro.y
! fixed using prefix¯o-> and ¯o.sufix-> when macro cannot be
substituted by compiler, f.e.:
M->v := NIL
x := "&v.1->fld"
M->v := "v"
M->v1:= "data"
? &x
! fixed memory leak in macro substituted expression compilation
* harbour/source/vm/macro.c
! fixed initialization of some HB_MACRO members
! fixed value returned for type("&V->F") when alias does not exist
- removed not longer used hb_comp_bShortCuts global variable
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbexprop.h
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* harbour/source/compiler/complex.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.y
* harbour/source/macro/macro.y
+ added support for strings with with ASCII NUL character
Strings with ASCII NUL character are not stored with hash table
but dynamically allocated/deallocated
! fixed compilation of strings with ASCII NUL character which
can appear in escaped strings e"..." or after compiler chr(0)
optimization
% removed some not longer necessary memory allocations
* harbour/source/vm/macro.c
! fixed GPF caused by HB_MACRO structure double freeing
- removed some code which was added as workaround for the old
behavior of HVM QUIT and false FM stat memory leak reports.
It's not longer necessary in current HVM.
* harbour/source/common/hbstr.c
! fixed final string size calculation in extended strings
* harbour/source/compiler/gencli.c
* cleaned BCC warning
* harbour/config/bsd/gcc.cf
* harbour/config/darwin/gcc.cf
* harbour/config/dos/djgpp.cf
* harbour/config/dos/owatcom.cf
* harbour/config/hpux/gcc.cf
* harbour/config/linux/gcc.cf
* harbour/config/linux/owatcom.cf
* harbour/config/os2/gcc.cf
* harbour/config/sunos/gcc.cf
* harbour/config/w32/bcc32.cf
* harbour/config/w32/gcc.cf
* harbour/config/w32/mingw32.cf
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hberrors.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbver.h
* harbour/source/common/hbstr.c
* harbour/source/common/hbver.c
* harbour/source/compiler/Makefile
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/complex.c
* harbour/source/compiler/gencli.c
* harbour/source/compiler/genjava.c
* harbour/source/compiler/genobj32.c
* harbour/source/compiler/harbour.c
- harbour/source/compiler/harbour.simple
- harbour/source/compiler/harbour.slx
- harbour/source/compiler/harbour.sly
* harbour/source/compiler/harbour.y
* harbour/source/compiler/hbgenerr.c
- harbour/source/compiler/hbslex.c
- harbour/source/compiler/hbstrong.c
* harbour/source/compiler/hbusage.c
- harbour/source/compiler/simplex.c
* harbour/source/macro/Makefile
- harbour/source/macro/macro.slx
- harbour/source/macro/macroslx.c
* harbour/source/pp/ppcore.c
* harbour/utils/hbpp/hbppcomp.c
- removed SIMPLEX references
! fixed memory leak in request _GET_
* use hb_compIdentifiersNew() in INLINE and AUTOOPEN structures
members
* eliminated all internal exit() calls. Now hb_compGenError() with
fatal errors does not interrupt application but only sets
HB_COMP_PARAM->fExit = TRUE;
What should cause silent clean application close.
Please keep it in mind making future modifications and do not
introduce and exit() function call.
+ added support for \x<hex> and \<oct> quoting in extended strings
(just like in C). Warning !!!. Embedded 0 bytes are supported by
PP, lexer and expression optimizer but not by bison. To make them
working it will be necessary to change Literal terminal symbol
holder and add length field in bison rules and also change the hash
table used to hold identifiers, literals and macros to store and use
real string length. If someone will try to use strings with embedded
0 now then they will be cut but nothing worser should happen.
* harbour/source/compiler/harbour.y
* fixed some memory leaks which cannot be addressed by destructor
because they are inside our own code which were not freeing
properly passed expressions, f.e.:
x:=1->f
or:
x:=field->(1)
This should close the memory leaks problem in grammar parser.
* indenting
* harbour/include/hbapi.h
* harbour/include/hbcomp.h
+ harbour/include/hbcompdf.h
* harbour/include/hberrors.h
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/include/hbexprop.h
* harbour/include/hbmacro.h
* harbour/include/hbpp.h
* harbour/include/hbvm.h
* harbour/source/common/expropt1.c
* harbour/source/common/expropt2.c
* harbour/source/compiler/Makefile
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/complex.c
* harbour/source/compiler/genc.c
* harbour/source/compiler/gencc.c
* harbour/source/compiler/gencli.c
* harbour/source/compiler/gencobj.c
* harbour/source/compiler/genhrb.c
* harbour/source/compiler/genjava.c
* harbour/source/compiler/genobj32.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.l
* harbour/source/compiler/harbour.y
* harbour/source/compiler/hbcomp.c
* harbour/source/compiler/hbdead.c
* harbour/source/compiler/hbfix.c
* harbour/source/compiler/hbfunchk.c
* harbour/source/compiler/hbgenerr.c
* harbour/source/compiler/hbident.c
* harbour/source/compiler/hbpcode.c
* harbour/source/compiler/ppcomp.c
* harbour/source/macro/macro.y
* harbour/source/pp/ppcore.c
* harbour/source/pp/ppgen.c
* harbour/source/pp/pplib.c
* harbour/source/vm/cmdarg.c
* harbour/source/vm/estack.c
* harbour/source/vm/fm.c
* harbour/source/vm/macro.c
* harbour/utils/hbpp/hbpp.c
* harbour/utils/hbpp/hbpp.h
* harbour/utils/hbpp/hbppcomp.c
* harbour/utils/hbpp/hbppcore.c
* harbour/utils/hbpp/hbpplib.c
* harbour/utils/hbpp/pragma.c
* changed the internal compiler API to be MT safe.
All global and static non constant variables replaced by
HB_COMP structure which have all compiler context settings.
It's possible to allocate simultaneously many compiler contexts
and compile code. Only constant/read only variables are shared.
In macro compiler HB_COMP is replaced by HB_MACRO.
In source code I everywhere used to macros: HB_COMP_DECL and
HB_COMP_PARAM which are equivalents of old HB_MACRO_DECL and
HB_MACRO_PARAM definitions.
We will only have to change compiler FATAL errors support to
not execute exit() for non batch systems.
* make macro compiler fully MT safe
* removed not longer used definitions and variables
! add direct accessing to PP line number information - it fixes missing
line numbers reported recently
! added generating .ppo files when tokens are teken directly by compiler
- it fixes empty .ppo file problem
! clear hb_stack internall variables after removing hb_stack to avoid
possible GPF if application still works and try to access unexsiting
hb_stack
! do not allocate new memory block when final FM statistic report is
generated - it fixes problem with GPF when application compiled with
FM statistic exits and CLIPPER envvar is set.
! fixed memory leak hb_cmdargCheck()
! added to harbour.y symbol destructors - it should fixes memory leaks
in syntax errors but it uses quite new bison feature which is not
fully supported yet. Unfortunately it's also not MT safe and the
destructors implementation in 1.875c does not respect %parse-param.
Bison documentation says that it should so I hope it will be fixed
soon (or maybe even already is in the newest bison versions) before
I'll add multi context compilation support. If not then I will have
to add some workaround. I can pass compiler context pointer inside
YYSTYPE using hack in a lexer but it will force really huge number
modifications in existing bison rules so probably it will be much
easier to fix bison or write a small tool to update generated parser.
* other modifications - the total size of patch is nearly 1MB and
detail description will have to take me few days.
* harbour/common.mak
* harbour/source/compiler/Makefile
* harbour/include/hbcomp.h
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/genc.c
* harbour/source/compiler/harbour.c
* harbour/source/compiler/harbour.l
* harbour/source/compiler/harbour.slx
* harbour/source/compiler/harbour.y
* harbour/source/compiler/hbgenerr.c
* harbour/source/compiler/hbident.c
* harbour/source/compiler/ppcomp.c
+ harbour/source/compiler/complex.c
+ added new PP based compiler lexer - it's smaller, MT safe and a
little bit faster then then the FLEX version.
+ added HB_COMP structure to hold compiler data in future MT version
+ added global variable HB_COMP_PTR hb_comp_data to make conversion
to MT easier - now it holds only PP and lexer data.
* update PP related code in compiler to be MT safe
+ added %pure-parser, %parse-param and %lex-param for bison to generate
MT safe grammar parser.
* updated FLEX to work with recent compiler modifications and pure-parser
bison API
* harbour/makefile.bc
* harbour/makefile.vc
* harbour/source/macro/Makefile
* harbour/source/macro/macro.l
* harbour/source/macro/macro.y
* harbour/source/macro/macrolex.c
* use hb_macro prefix instead of hb_comp in bison/flex parser/lexer
used in macro compiler to avoid possible conflicts in the future
* separated lexer data
* harbour/include/hbapi.h
* harbour/include/hbpp.h
* harbour/source/pp/ppcore.c
* harbour/source/pp/ppgen.c
* harbour/source/pp/pplib.c
* harbour/source/vm/macro.c
* removed not used members from HB_MACRO structure to make it
cleaner before creating common to compiler and macro compiler
structure
+ added new token HB_PP_TOKEN_EPSILON
+ added void * cargo parameters passed to executed user functions
+ hb_pp_tokenGet(), hb_pp_tokenToString(), hb_pp_tokenBlockString()
functions for new PP based compiler lexer
* harbour/utils/hbpp/hbpp.c
* harbour/utils/hbpp/hbpp.h
* harbour/utils/hbpp/hbppcomp.c
* harbour/utils/hbpp/hbppcore.c
* harbour/utils/hbpp/hbpplib.c
* harbour/utils/hbpp/pragma.c
* updated to compile with recent compiler header file modifications
PP, new lexer and most of grammar parser should be MT safe. Now we should
update all compiler functions to pass pointer to HB_COMP data structure
where we should all current global variables. This structure as first
member should have HB_CMPCOMMON structure which will hold common to
compiler and macro compiler data. Ryszard I think you are the best person
to define this structure.
We have new lexer which is MT safe but please note that it has to be
extensively tested so I would like to ask everybody to compile as much
as possible different code and check if the final programs work as
expected. Working on new code I removed some limitations existing in
FLEX though not all. At the beginning I tried to replicate the exact
FLEX behavior but I've found that in few places it does not work as
it should so I begin to encode rules in a way which remove some
limitations. In fact now it's much easier to control some things.
I kept the FLEX code working and made all necessary modifications
so it still can be used but keeping FLEX working cost us IMHO too
much. It's not possible to introduce some improvements to grammar
parser. All identifiers, keyword and macros returned by new lexer
are converted to upper letters, do not have to be freed by hb_xfree()
and is guarantied that will be always accessible. So from grammar file
we can remove all hb_compIdentifierNew( hb_strupr($1), TRUE ) what
should give noticeable speed improvement but will break the FLEX code.
Ryszard and other you will have to decide if we will support FLEX in
the future. We can also clean the code and remove most of other
redundant hb_strupr() and hb_strdup() used in many places. BTW only
one terminal symbol can be returned with lower letters: DOIDENT
and I make it intentionally so it's possible to use:
DO prog1 WITH "sth"
on case sensitive file systems so this symbol should be cloned in
upper cases as function symbol but used without modification as
file name. It's current behavior but I'm not sure you will want
to keep it. Maybe compiler switch to always convert file names
created from
DO <id> [WITH <params,...>]
to lower cases will be better. Please think about it.
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* harbour/include/hbexprc.c
* harbour/source/macro/macro.y
* harbour/source/macro/macrolex.c
* do not allocate separate memory in macro compiler for terminal
symbols which needs string representation - it fixes some set
of memory leaks and increase a little bit macro compilation
! fixed accessing freed memory in:
private cMacro := "cEarly", cEarly := {"Early"}
@ 12,10 GET &cMacro[1]
! fixed GPF in:
p:=@left()
and similar expressions
* harbour/source/common/expropt2.c
% minor modification
* harbour/source/compiler/harbour.c
! fixed HB_TRACE message
* harbour/source/pp/ppcore.c
! fixed casting - it resolves problem reported by Alexander with
preprocessing chr(255) from static buffer (PP used by .prg code)
* harbour/source/macro/macro.y
+ added some comments
* harbour/source/macro/Makefile
* commented out compilation of FLEX based lexer for macro compiler and
added compilation of new small lexer written in pure .c
* harbour/common.mak
* replaced macrol.obj by macrolex.obj
+ harbour/source/macro/macrolex.c
+ new small lexer written in pure .c. The code is much smaller,
MT safe and more friendly for grammar parser. I think it should
be also faster but I haven't made any tests yet. When we update
grammar parser to not free string representation of terminal
symbols then we should expect some farther speed improvement,
it will also resolve some set of memory leaks like in:
type("user input")
and will allow to make some parts of grammar parser common to
compiler and macro compiler by removing #if[n]def HB_MACRO_SUPPORT
* harbour/source/common/hbstr.c
! fixed decoding strings without trailing ASCII NUL character in
hb_strRemEscSeq()
* harbour/source/pp/ppcore.c
+ added decoding C escaped strings (e"...") to PP so they can be
preprocessed like a normal strings, it's not enabled when
HB_C52_STRICT macro is set
* harbour/include/hbapi.h
* harbour/source/common/hbstr.c
* harbour/source/compiler/harbour.l
* harbour/source/macro/macro.l
* harbour/source/macro/macro.slx
* harbour/source/compiler/harbour.slx
* changed parameters in function: hb_compStrToNum() now the string
size is passed explicitly so it can work with strings which does
not have trailing ASCII NUL character.
* harbour/source/macro/macro.y
* indenting
* harbour/include/hbmacro.h
* harbour/source/common/hbstr.c
* harbour/source/macro/macro.l
* harbour/source/macro/macro.y
* added two new functions for macro compiler: hb_macroLexNew() and
hb_macroLexDelete() to clearly separate lexer from grammar parser.
* harbour/source/compiler/harbour.l
* harbour/source/common/hbstr.c
* harbour/source/pp/ppcore.c
* added error generation for wrong e"..." strings and fixed \" quoting
* removed \q quoting
* harbour/source/pp/ppcore.c
* harbour/source/pp/ppgen.c
* added small description in headers
* harbour/common.mak
* harbour/source/pp/Makefile
* harbour/source/pp/pplib.c
+ harbour/source/pp/pplib2.c
* harbour/include/hbextern.ch
+ added auto destructor for allocated PP context, it's not longer
necessary to execute __PP_FREE() so this function was moved from
the standard .prg API to file with backward compatible functions
* changed __PREPROCESS() to __PP_PROCESS()
* moved setting std rules to separate file so now if user uses PP
with only his own rules we are not forcing linking static tables
which are quite huge (over 100kb)
To include stdandard rules in static binaries user should add
to his code:
REQUEST __PP_STDRULES
+ harbour/source/pp/pplib3.c
* old PP functions for backward compatibility:
__PPADDRULE(), __PREPROCESS(), __PP_FREE()
Please note that using this function is not MT safe. They should
work like with old PP code. Using any of this function automatically
forces linking __PP_STDRULES
* harbour/include/hbapi.h
+ added
void * pLex;
member to HB_MACRO structure - it will be necessary for MT safe
macro parser
* harbour/include/set.ch
* harbour/include/hbexprb.c
* harbour/source/macro/macro.l
* harbour/source/macro/macro.y
* indenting, formatting and converting tabs to spaces
* harbour/include/hbpp.h
* harbour/source/compiler/ppcomp.c
* harbour/source/pp/ppcore.c
* harbour/source/pp/ppgen.c
* harbour/source/pp/pplib.c
* make HB_PP_TOKEN structure public for other code and changed
public PP API to not force including static rules by programs
which does not need them.
+ added basic API for using PP as lexer
* divided HB_PP_TOKEN_MACRO to HB_PP_TOKEN_MACROVAR and
HB_PP_TOKEN_MACROTEXT and mark
[<keyword>][&<keyword>[.[<nextidchars>]]]+
as HB_PP_TOKEN_MACROTEXT not pair HB_PP_TOKEN_KEYWORD and
HB_PP_TOKEN_MACRO
It allows easier integration PP as (macro)compiler lexer and
fixes preprocessing [<keyword>][&<keyword>[.[<nextidchars>]]]+
in some patterns.
* harbour/utils/hbtest/Makefile
+ added PP library to linked library list
* harbour/source/macro/macro.y
+ added seaport to use PP as lexer. It's disabled by default and
can be enabled using by defining HB_PP_MACROLEX. It eliminates
using FLEX or SIMPLEX lexer in macro compiler. Now it emulates
the behavior of FLEX lexer because I wanted to minimize the
modification but in fact because parsed token are freed when
macro compiler finish its job then we can resolve RT memory
leaks which apear during parsing wrong expressions. It will be
enough to drop allocating memory for new terminal symbols and
using them as is like in compiler. It will also quite nice
simplify include/hbexpr*.c code and will allow to move some
functions to COMMON library. I want to make also pure compiler
code MT safe so I will have to add passing pointer to a structure
with compiler context just like in macro compiler. It means that
if we define a structure which will be common to compiler and macro
compiler with local (not common) data stored in other structure for
which will will keep only a pointer then we will be able to move most
of include/hbexpr*.c code to common library. But I do not want to
make farther modification without consultation without Ryszard.
Also to make PP default lexer some farther modifications should
be done in core PP code and include/hbexpr*.c files to not reduce
the speed. Now using HB_PP_MACROLEX for sure will have to cause
speed reduction in macro compiler. But if we make necessary
modifications then we should keep at least the same speed or
probably reach noticeable better results.
