* harbour/include/hbdefs.h
* moved HB_WCHAR definition from hbapicdp.h to hbdefs.h
* harbour/include/hbapicdp.h
* harbour/include/hbcdpreg.h
* harbour/src/rtl/cdpapi.c
* harbour/src/rtl/cdpapihb.c
+ added support for custom sorting redirected from HVM
+ added support for custom character indexes in strings
+ added support for custom character flags (upper, lower, alpha, digit)
+ added support for custom upper/lower conversions
+ added support for CPs using unicode character values instead of
ASCII ones
+ added new CP functions which respects custom CP settings:
hb_cdpUpperWC(), hb_cdpTextLen(), hb_cdpTextPos(),
hb_cdpTextPosEx(), hb_cdpTextGetU16(), hb_cdpTextPutU16(),
hb_cdpCharEq(), hb_cdpCharCaseEq()
+ added new conversion functions:
hb_cdpGetUC(), hb_cdpGetWC(), hb_cdpGetU16Ctrl()
+ added macros to detect codepage parameters:
HB_CDP_ISBINSORT() - codepage uses simple binary sorting
HB_CDP_ISCUSTOM() - codepage uses custom string decoding
HB_CDP_ISCHARIDX() - codepage use character indexes instead
of bytes ones
HB_CDP_ISCHARUNI() - CHR(), ASC() and similar functions operates
on Unicode values instead of bytes
HB_CDP_ISUTF8() - codepage uses UTF-8 encoding
* harbour/include/inkey.ch
- removed HB_INKEY_EXTENDED - it was not used in Harbour
+ added new flag HB_INKEY_EXT - it allows to used different
event encoding system with automatic translation to
standard Clipper values. It will be used in the future
as base for low level GTs code. Now it's possible to
selectively switch to the new system. New codes are
unique and do not interacts with Clipper.
+ added HB_INKEY_ALL macro - it's similar to INKEY_ALL but
enables also GTEVENT keys.
* harbour/include/hbgtcore.h
* harbour/src/rtl/hbgtcore.c
* changed character values in internal screen buffer to unicode.
Warning: this modification interacts with SAVESCREE()/RESTSCREEN()
data. Now all GTs uses extended definition. If somene
has code which needs VGA like screen buffers returned
by SAVESCREE() then it can be forced by
hb_gtInfo( HB_GTI_COMPATBUFFER, .t. )
Anyhow it will for to use only one CP in output.
+ added new GT methods which operated on HB_WCHAR values
* modified existing methods using HB_WCHAR/HB_USHORT parameters
to operate on HB_WCHAR values
+ added codepage conversion code directly to core code so it can
be eliminated from low level GT drivers which do not need some
special operations on it.
+ added support for extended inkey codes, they allow to encoded
unicode values, mouse events with flags, keyboard events with
modifier flags (shift,alt,ctrl,...), etc.
Now in low level GT code only unicode key values are used but
I plan to updated all GTs and switch to new code - it should
resolve many small problems inheriting with Clipper inkey codes.
* harbour/include/hbapifs.h
* moved hb_fsNameConvU16() definition from hbapicdp.h to hbapifs.h
* harbour/include/hbxvm.h
- removed comment copied by mistake from GT header file
* harbour/include/hbwinuni.h
* changed HB_CHARDUP() and HB_CHARDUPN() macros - now they
operate on functions which can be used without active HVM.
In such case ANSI CP is used as source encoding.
+ added new macros: HB_OSSTRDUP() and HB_OSSTRDUP2().
They make conversions from OS encoding to HVM one.
They can be used without active HVM and in such case ANSI CP
is used as destination encoding.
* harbour/include/hbapi.h
* harbour/src/vm/cmdarg.c
+ added new function hb_cmdargProgName()
It returns application name with path or NULL if not set,
caller must free returned value with hb_xfree() if not NULL.
The string is in HVM encoding so it does not need any additional
encodings.
* modified hb_cmdargString() to return strings in HVM encoding.
% some small code optimizations
; NOTE: parameters passed to MS-WINDOWS GUI programs which use
WinMain() instead of main() as startup entry are translated
to ANSICP before they can be accessed by application.
This can be eliminated though not for console programs
using where C compiler makes such translation to main()
parameters.
in all systems parameters are passed to application
startup and init functions before programmer can set OS CP.
This can be resolved by adding:
INIT PROC CLIPINIT()
SET( _SET_CODEPAGE, <userCP> )
SET( _SET_OSCODEPAGE, <systemCP> )
RETURN
to linked code. INIT procedures called CLIPINIT() are
executed before any other ones.
* harbour/include/hbapi.h
* harbour/src/vm/set.c
+ added new functions for CP converisons which can be used with
and without active HVM:
hb_osStrEncode(), hb_osStrEncodeN(),
hb_osStrDecode(), hb_osStrDecode2(),
hb_osStrU16Encode(), hb_osStrU16EncodeN(),
hb_osStrU16Decode(), hb_osStrU16Decode2(),
* harbour/include/hbvm.h
* harbour/src/vm/hvm.c
+ added new function hb_vmIsReady() - it's similar to hb_vmActive()
but it also checks if current thread has active HVM stack.
* moved hb_cdpReleaseAll() to the end of HVM cleanup code so
CP conversions are longer active.
* harbour/include/hbcomp.h
* harbour/src/compiler/cmdcheck.c
* harbour/src/compiler/hbusage.c
* harbour/src/common/expropt2.c
+ added new compiler switch:
-ku - strings in user encoding
Now it informs compiler that strings use custom encoding so some
optimizations which are byte oriented cannot be used.
It's possible that in the future we will change above definition
to sth like: "strings in UTF8 encoding" but now I would like to
keep more general.
* harbour/src/vm/macro.c
* inform macrocompiler about custom CPs using own character indexes
to disable byte oriented optimizations.
* harbour/include/hbapigt.h
* harbour/src/rtl/gtapi.c
* harbour/src/rtl/inkeyapi.c
+ added HB_B_*_W macros with unicode box character definitions
+ added HB_MBUTTON_* macros
+ added new GT functions: hb_gtHostCP() and hb_gtBoxCP() which
allows to extract CPs used in translations by GTs.
+ added new function hb_inkeyKeyString() - it converts inkey value
to corresponding string
+ added new function hb_inkeyKeyStd() - it converts new extended
key value to standard Clipper one.
* harbour/include/hbapifs.h
* harbour/src/common/hbffind.c
* moved OS codepage translations fully to hb_fsFind*() functions.
It fixes few problems which existed before, i.e. double CP
conversions in MS-Windows builds and simplifies upper level code.
* harbour/src/pp/hbpp.c
* harbour/src/rtl/direct.c
* harbour/src/rtl/fssize.c
* harbour/src/rtl/file.c
* harbour/contrib/hbct/files.c
* eliminated not longer necessary CP conversions in code calling
hb_fsFind*() functions.
* harbour/src/common/hbgete.c
* moved OS codepage translations to hb_getenv(), hb_getenv_buffer()
and hb_setenv() functions. It fixes few problems which existed
before, i.e. double CP conversions in MS-Windows builds and
simplifies upper level code.
* harbour/src/rtl/net.c
* harbour/src/rtl/gete.c
* eliminated not longer necessary CP conversions in code calling
hb_getenv()/hb_setenv() functions.
; NOTE: additional parameters in HB_GETENV() and HB_SETENV() which
disabled CP conversions are not longer supported.
They were strictly platform dependent and ignored in chosen
cases (i.e. in MS-Windows UNICODE builds we always have to
convert strings transferred between HVM and OS. If someone
needs old functionality for other platform then he should
temporary disable _SET_OSCODEPAGE.
* harbour/src/common/hbver.c
* harbour/src/common/hbfsapi.c
* harbour/src/rtl/fstemp.c
* harbour/src/rtl/fslink.c
* eliminated HB_TCHAR_*() macros
* harbour/src/common/strwild.c
+ added supprot for custom CPs using own character indexes in:
hb_strMatchWild(),
hb_strMatchWildExact(),
hb_strMatchCaseWildExact()
* harbour/src/nortl/nortl.c
+ added new dummy function replacement for binaries which are not
linked with HVM.
* harbour/src/rtl/filesys.c
! fixed double CP conversions in MS-Windows builds of hb_fsCurDirBuff()
* use hb_vmIsReady() instead of hb_stackId() in file name conversions.
* use hb_cmdargProgName() in hb_fsBaseDirBuff()
* harbour/src/rtl/philes.c
* use hb_cmdargProgName() in HB_PROGNAME() function.
* harbour/src/rtl/gtcgi/gtcgi.c
* harbour/src/rtl/gtstd/gtstd.c
* harbour/src/rtl/gtpca/gtpca.c
* harbour/src/rtl/gtdos/gtdos.c
* harbour/src/rtl/gtos2/gtos2.c
* harbour/src/rtl/gtwin/gtwin.c
* harbour/src/rtl/gtwvt/gtwvt.h
* harbour/src/rtl/gtwvt/gtwvt.c
* harbour/src/rtl/gttrm/gttrm.c
* harbour/src/rtl/gtcrs/gtcrs.c
* harbour/src/rtl/gtsln/gtsln.c
* harbour/src/rtl/gtsln/kbsln.c
* harbour/src/rtl/gtsln/gtsln.h
* harbour/src/rtl/gtxwc/gtxwc.h
* harbour/src/rtl/gtxwc/gtxwc.c
* harbour/contrib/gtwvg/gtwvg.c
* harbour/contrib/gtwvg/gtwvg.h
* harbour/contrib/gtalleg/gtalleg.c
* harbour/contrib/hbqt/gtqtc/gtqtc.cpp
* harbour/contrib/hbqt/gtqtc/gtqtc.h
* updated to work with new unicode GT API
please make tests with different GTs - I'm not able to test
all of them, i.e. I do not have any OS2 machine.
* harbour/src/rtl/box.c
* harbour/src/rtl/oldbox.c
* harbour/src/rtl/scroll.c
* harbour/src/rtl/console.c
* updated to work with new unicode GT API and
CPs using custom character indexes
* harbour/src/rtl/at.c
* harbour/src/rtl/ati.c
* harbour/src/rtl/rat.c
* harbour/src/rtl/len.c
* harbour/src/rtl/transfrm.c
* harbour/src/rtl/left.c
* harbour/src/rtl/right.c
* harbour/src/rtl/substr.c
* harbour/src/rtl/stuff.c
* harbour/src/rtl/padc.c
* harbour/src/rtl/padl.c
* harbour/src/rtl/padr.c
* updated to work with CPs using custom character indexes
* harbour/src/rtl/chrasc.c
+ added support for HB_CDP_ISCHARUNI() CPs.
* harbour/src/rtl/mlcfunc.c
* rewritten from scratch to work with CPs using custom character
indexes
* harbour/src/rtl/accept.c
* updated to work with unicode inkey values and
CPs using custom character indexes
* harbour/src/rtl/strmatch.c
% small optimization
* harbour/src/rtl/Makefile
+ harbour/src/rtl/chruni.c
+ added new PRG functions which allows to make byte/binary and
unicode/character operations on strings:
HB_UCHAR( <nCode> ) -> <cText>
return string with U+nCode character in HVM CP encoding
HB_BCHAR( <nCode> ) -> <cText>
return 1 byte string with <nCode> value
HB_UCODE( <cText> ) -> <nCode>
return unicode value of 1-st character (not byte) in given string
HB_BCODE( <cText> ) -> <nCode>
return value of 1-st byte in given string
HB_ULEN( <cText> ) -> <nChars>
return string length in characters
HB_BLEN( <cText> ) -> <nBytes>
return string length in bytes
HB_UPEEK( <cText>, <n> ) -> <nCode>
return unicode value of <n>-th character in given string
HB_BPEEK( <cText>, <n> ) -> <nCode>
return value of <n>-th byte in given string
HB_UPOKE( [@]<cText>, <n>, <nVal> ) -> <cText>
change <n>-th character in given string to unicode <nVal> one and
return modified text
HB_BPOKE( [@]<cText>, <n>, <nVal> ) -> <cText>
change <n>-th byte in given string to <nVal> and return modified
text
* harbour/src/rtl/hbdoc.prg
* harbour/src/rtl/memvarhb.prg
* use HB_BCHAR() for binary string definitions
* harbour/src/rtl/hbi18n2.prg
* use hb_utf8CHR( 0xFEFF ) instead of hardcoded binary sting and
HB_BLEN() instead of LEN()
* harbour/src/rtl/inkey.c
+ added new functions:
HB_KEYCHAR( <nKey> ) -> <cChar>
HB_KEYSTD( <nExtKey> ) -> <nClipKey>
! use HB_INKEY_ALL instead of INKEY_ALL in LASTKEY()
* harbour/src/rtl/achoice.prg
* harbour/src/rtl/browse.prg
* harbour/src/rtl/menuto.prg
* harbour/src/rtl/tgetlist.prg
* harbour/src/rtl/teditor.prg
* harbour/src/rtl/tlabel.prg
* harbour/src/rtl/tpopup.prg
* harbour/src/rtl/radiobtn.prg
* harbour/src/rtl/radiogrp.prg
* harbour/src/rtl/wait.prg
* updated to work with unicode inkey() values
* harbour/src/rtl/listbox.prg
* use box.ch macros instead of explicit CHR(...) definitions
* harbour/src/rtl/ttopbar.prg
* updated to work with different type of CPs
* harbour/src/rtl/scrollbr.prg
* formatting
* harbour/src/rtl/mouse53.c
* use HB_MBUTTON_* macros instead of local ones
* harbour/src/codepage/cp_utf8.c
+ harbour/src/codepage/uc16def.c
+ harbour/src/codepage/utf8sort.c
* replaced UTF8ASC with new CP: UTF8EX
This CP uses character indexes instead of bytes one
and operates on unicode characters flags.
Tables for upper/lower conversions and upper/lower/alpha/digit
flags were generated automatically from
http://www.unicode.org/Public/UNIDATA/UnicodeData.txt
It also uses custom collation rules. It's very simple one
level sorting based on UTF8 C collation.
If someone needs some advanced sorting rules, then it's enough
to create copy of this cp with user custom version of UTF8_cmp()
and UTF8_cmpi() functions, i.e. they can be redirected to some
external library like ICU (icu-project.org).
* harbour/contrib/hbct/ctwin.c
* harbour/contrib/hbct/ctwin.h
* harbour/contrib/hbct/ctwfunc.c
* added support for new unicode GT API
* harbour/contrib/xhb/xhbfunc.c
* redirected HB_CMDARGARGV() to HB_PROGNAME()
* harbour/contrib/hbnf/origin.c
* redirected FT_ORIGIN() to HB_PROGNAME()
* harbour/contrib/hbnf/getenvrn.c
! windows version of FT_GETE() fully rewritten - it should
fix well known problems reported to the devel list. Please test.
% small optimization for other systems
* harbour/contrib/hbfship/exec.c
* use hb_cmdargProgName() in EXECNAME()
+ harbour/tests/uc16_gen.prg
+ added code which generates tables with unicode character flags
from http://www.unicode.org/Public/UNIDATA/UnicodeData.txt.
harbour/src/codepage/uc16def.c was generated by this code.
* harbour/tests/inkeytst.prg
* harbour/tests/wvtext.prg
* harbour/tests/gtkeys.prg
* use HB_INKEY_ALL
* harbour/include/harbour.hbx
* harbour/include/hbcpage.hbx
* harbour/include/hblang.hbx
* regenerated
; It was quite big peace of modifications and for sure not everything
is well tested so please make test and report problems you will find.
; This is basic version which introduce to HVM CPs with custom encodings.
Some contrib code has to be updated to work correctly with it.
I hope that developers interesting in will make necessary updates.
I haven't touched GTWVW code at all - sorry but it needs very serious
work to make it production ready and fix all existing problems.
; Special thanks to OTC - this firm sponsored adding basic UTF8 support
to HVM.
* src/3rd/pcre/Makefile
+ enabled UTF8 support in PCRE (binary size overhead ~150KB)
+ enabled unicode properties support in PCRE (binary size overhead ~50KB)
* src/rtl/hbregex.c
+ added commented logic to detect UTF8 support in PCRE
and enable UTF8 mode when the HVM CP is also UTF8. Latter
logic is missing yet and subject to review.
* contrib/hbnetio/utils/hbnetio/netiosrv.prg
* contrib/hbnetio/utils/hbnetio/netiocon.prg
* contrib/hbnetio/utils/hbnetio/netiomgm.prg
* contrib/hbrun/hbrun.prg
* tests/ac_test.prg
* tests/ac_test2.prg
* tests/tstmacro.prg
* tests/tstalias.prg
* formatted using automatic tools
* src/rtl/run.c
* src/rtl/hbrunfun.c
* experimental: use hb_processRun() on all win targets
(was wince targets) instead of system() from C RTL.
QUESTION: do the same for all platforms?
* contrib/hbwin/win_prn1.c
* contrib/hbwin/hbwin.ch
* contrib/hbwin/hbwin.hbx
+ WIN_ENUMFONTS() modified to work also when no HDC is passed.
+ WIN_ENUMFONTFAMILIES( [ <nCharset> ][, <cName> ] ) -> <aFonts>
* contrib/gtwvg/wvgcuig.c
* contrib/gtwvg/wvgcore.c
! fixed possible GPF in some functions if font name was not passed
* INSTALL
+ extended TROUBLESHOOTING about generic common-sense
information about not overdoing custom configuration,
especially for C compiler. I didn't specifically include
anything about bcc, but let me here note, that hbmk2
and Harbour core build system doesn't require bcc32.cfg
and ilink32.cfg to be _present at all_ since quite long,
so please remove them for best results.
* utils/hbmk2/hbmk2.prg
* do not consider ".C" (uppercase c) as C++ source file.
it might have caused confusion when bringing .c source
files from legacy filesystems like MS-DOS, it may also
be an advantage when doing multiplatform development
between MS-DOS and newer systems.
* contrib/gtwvg/gtwvg.c
! HB_GTINFOEX(): fixed using object after freeing it
! renamed HB_GTINFOEX() -> WVG_GTINFOEX() to avoid occupying
Harbour core namespace in non-core component. It also uses
raw pointer to access internals.
(if such functionality is to be ever implemented in core,
it should probably be done by extending existing HB_GTINFO())
; INCOMPATIBLE. Update your function calls according to above.
* src/rtl/fscopy.c
+ using hb_file*() API instead of hb_fs*() API.
Completing TODO originating from:
https://groups.google.com/d/msg/harbour-devel/0QY0SJ8HBFU/rWbUvJXygWYJ
; review me. I had to add an ugly-looking hack for win, because
error 38 was returned when reaching the end of file successfully.
with hb_fs*() API this didn't happen.
* utils/hbmk2/hbmk2.prg
* bin/hb3rdpat.hbs
! fixed minor formatting problems mainly using custom automatic script
* include/harbour.hbx
* contrib/hbsms/hbsms.hbx
+ applied std casing to function names. it enhances
readability, plus makes possible to use .hbx file for
automatic formatting purposes.
* utils/hbmk2/hbmk2.prg
* config/postinst.hbs
+ preserve casing of function names in .hbx files
! fixed non-fatal typo in regex expressions
* contrib/hbhpdf/3rd/libhpdf/libhpdf.hbp
* contrib/hbhpdf/3rd/libhpdf/libhpdf.dif
+ contrib/hbhpdf/3rd/libhpdf/hpdf3dme.h
+ contrib/hbhpdf/3rd/libhpdf/hpdf3dme.c
+ contrib/hbhpdf/3rd/libhpdf/hpdfencu.c
+ contrib/hbhpdf/3rd/libhpdf/hpdfexda.h
+ contrib/hbhpdf/3rd/libhpdf/hpdfexda.c
+ contrib/hbhpdf/3rd/libhpdf/hpdfimac.c
+ contrib/hbhpdf/3rd/libhpdf/t4.h
* contrib/hbhpdf/3rd/libhpdf/*
+ libharu 2.2.1 -> 2.3.0RC2
updated to latest RC2 because I may also need it and
the project is pretty much dead these days, so it might
take a while to decide on a release. (plus I submitted
pull request for warnings and report about more warnings)
beware of bugs and if you find any, report to libharu
developers. most important new stuff is partial unicode
(1-2 bytes UTF8) and PDF/A support. it also incorporates
our png patch, so local patching is minimal now.
; TODO: add hbhpdf wrappers for new functions not yet covered.
* harbour/include/hbapifs.h
* harbour/src/common/hbffind.c
! fixed potential freeing uninitialized handlers on some platforms
* hide private members
* harbour/src/codepage/Makefile
* harbour/include/hbcpage.hbx
+ harbour/src/codepage/cp950.c
+ harbour/src/codepage/cp_950.c
+ added new Harbour codepage "CP950". It's similar to BIG5 and
used by MS-Windows. Please test and inform me if you can replace
BIG5 in your applications with CP950 without any problems.
* harbour/contrib/hbnf/hbnf.hbx
+ added FT_IDLE()
* harbour/src/rtl/hbsocket.c
! fixed problems with missing error setting in select() executed
for asynchronous connect() (windows builds) and overwritten error
codes in connect() and accept() (all builds). Many thanks for
Mindaugas for locating the problem and patch.
* src/compiler/hbopt.c
! fixed -w3 warning 'Variable ... is assigned but not used' in case of
BEGIN SEQUENCE/END SEQUENCE sentence. Ex.:
BEGIN SEQUENCE
nI := 1
BREAK(NIL)
RECOVER
? nI
END SEQUENCE
* contrib/make.hbs
* contrib/hbpost.hbm
+ use HB_LIB3RD hbmk2 macro instead of rolling custom logic
to detect and enable unicows lib.
* utils/hbmk2/hbmk2.prg
+ added HB_LIB3RD internal variable (usable as filter and as macro)
it's filled when '<hbroot>/lib/3rd/<plat>/<comp>' directory is present
+ added EXPERIMENTAL pseudo-function for filters to detect presence
of file or directory: {hb_ispath='<file or directory>'}. if the
value is not an absolute path, it will be meant relative to source
hbmk2 file.
* package/harb_win.mft
* package/harb_win.rc
* package/mpkg_win.nsi
* package/mpkg_src_nightly.sh
* package/winuni/mpkg_win_uni.nsi
* "Harbour Project" -> "Harbour"
* "hbrun" shortcut -> "Harbour (Interactive shell)"
* harbour/include/hbapicdp.h
* harbour/src/rtl/cdpapi.c
+ added new C functions:
hb_cdpDupn(), hb_cdpStrDupU16(), hb_cdpnStrDupU16()
* harbour/include/hbwinuni.h
+ added new macros: HB_CHARDUP() and HB_CHARDUPN()
Warning: this macros uses HVM functions so cannot be used in
code which is executed without active HVM i.e. in
pure harbour compiler code (common library)
* harbour/src/rtl/filesys.c
* use hb_cdpStrDupU16()
* harbour/src/rtl/hbproces.c
! fixed double OS codepage conversion in hb_fsProcessRun()
* use HB_CHARDUP*() macros to respect _SET_CODEPAGE and _SET_OSCODEPAGE
in Windows builds
* src/common/hbverdsp.c
* do not show '(WINANSI)' in the list of version flags.
* utils/hbmk2/hbmk2.prg
+ add -inc to xbuild and xmate converted projects, this being
the default mode of operation (AFAI guess) of these tools.
* contrib/hbct/disk.c
* contrib/hbnf/kspeed.c
* contrib/hbnf/peek.c
* contrib/hbnf/proper.c
* contrib/hbnf/ftidle.c
* contrib/hbnf/mouse.c
* contrib/hbnf/stod.c
* contrib/hbnf/getvid.c
* contrib/hbnf/getver.c
* contrib/hbnf/getenvrn.c
* contrib/hbnf/mkdir.c
* contrib/hbnf/rmdir.c
! formatting
* config/win/mingw.mk
* removed undocumented way (committed in prev rev) to disable
separate compile pass for mingw Harbour-*.dll. if you're bothered,
use existing and documented: HB_BUILD_DYN=no
; NOTE: separate compile pass has another important benefit:
it will finally remove the implicit .dll exports, IOW now
only explicitly exported symbols will be public, in sync
harbour-*.dlls created with all the other C compilers.
* config/win/mingw.mk
+ solved mingw harbour dll and unicows problem, so now
-shared Harbour executables built using mingw will be able
to run on Win9x, "out of the box".
(has a downside that now separate compile pass is needed
to build objects with -DHB_DYNLIB (like with all the other
C compilers), but I left it enabled anyways to give uniform
support for unicows by default, including the nightly and
official builds. Look into the diff for an undocumented
way to disable this)
* config/win/watcom.mk
* config/win/xcc.mk
* config/win/icc.mk
* config/win/pocc.mk
* config/win/bcc.mk
* config/win/msvc.mk
* restored to link 3rd libs (currently unicows, if available)
to harbour-*.dll with above compilers. These are currently
noops (except for bcc, where it just doesn't work),
they are fundemantally correct, but may require further
C compiler specific mods.
; TODO: problem with bcc harbour-*.dll remains. unicows is present,
specified, but ignored when linking, even though it's not ignored
when contrib dlls are built using similar method. I leave
resolving this problem to bcc users.
* config/win/watcom.mk
* config/win/xcc.mk
* config/win/icc.mk
* config/win/mingw.mk
* config/win/pocc.mk
* config/win/bcc.mk
* config/win/msvc.mk
! do not link unicows to harbour dll. it seems a noop
for some compilers, and breaks some others (mingw).
* utils/hbmk2/hbmk2.prg
+ will now add lib/3rd dir to the link commands, so
unicows lib will be found in uninstalled Harbour SVN
source tree.
* contrib/make.hbs
* contrib/hbpost.hbm
+ automatically enable unicows support for contribs, if available.
- deleted support for undocumented __HB_BUILD_WINUNI envvar
* config/global.mk
* config/rules.mk
- deleted support for undocumented __HB_BUILD_WINUNI envvar
* package/winuni/mpkg_win_uni.bat
* package/mpkg_win_nightly.bat
- deleted special support for unicows libs for nightly
and official releases, along with HB_DIR_UNICOWS envvar.
Now these libs are included in all builds.
* config/win/global.mk
+ automatically enable unicows support for core, if available.
* config/win/xcc.mk
* config/win/icc.mk
* config/win/mingw.mk
* config/win/pocc.mk
* config/win/bcc.mk
* config/win/msvc.mk
+ added support for SYSLIBPATHS. (used by unicows support)
! fixed bcc dynlib link command missing libpaths.
; NOTE: watcom unicows support is missing, as it needs
special tricks due to missing separate libpath option
in this compiler/linker.
* config/wce/mingwarm.mk
* config/wce/poccarm.mk
* config/wce/msvcarm.mk
* synced with config/win files
+ config/libbin.mk
+ added rule to install pre-built binary libs.
(for unicows libs)
* Makefile
+ lib/3rd
+ lib/3rd/win
+ lib/3rd/win/mingw
+ lib/3rd/win/mingw/Makefile
+ lib/3rd/win/mingw/libunicows.a
+ lib/3rd/win/mingw/libunicows_license.txt
+ lib/3rd/win/bcc
+ lib/3rd/win/bcc/unicows_license.txt
+ lib/3rd/win/bcc/unicows.lib
+ lib/3rd/win/bcc/Makefile
+ added locally hosted unicows runtime/implibs from:
http://libunicows.sourceforge.net/ by author Vaclav Slavik.
With these libs, it's possible to use UNICODE builds
with both MS unicows.dll and opencows .dll on legacy
Windows-system (Win9x).
These libs will be installed to Harbour core lib directory,
and also used from lib/3rd directory, if they are available.
I've uploaed mingw and bcc libs because these are used
nearly exclusively by users, but they are also available
for older msvc, watcom, dmc and lcc compilers. I don't
think it's worth the extra size to support Win9x using
these legacy compilers, anyhow they can be added easily
if needed (except watcom, which requires more work).
; These changes mean three things:
1. Harbour now has only one build mode on Windows: UNICODE
This will greatly simplify coding and testing and will
allow to focus much better on what's important. It also
allows to drop lots of dual code dealing with separate
UNICODE and non-UNICODE branches.
2. Harbour default builds will now run on Win9x (currently
when using mingw or bcc compilers), though UNICOWS.DLL
is now required. See INSTALL how to get this free .dll
from MS. This also means that Harbour can now theoretically
by built on Win9x systems (I didn't try).
3. Harbour users can create UNICODE builds compatible with
Win9x, by simply adding -lunicows to their .hbp projects.
No extra installation is required.
