* include/hbapifs.h
! added missing declaration for hb_fileParamGet()
* include/hbgtcore.h
* added HB_EXPORT attribute for hb_gt_getClipboard(), hb_gt_setClipboard()
and hb_gt_dos_keyCodeTranslate()
* include/hbrdddbf.h
! removed declaration for hb_dbfTranslateRec() function - this functions
was removed many years ago
* include/harbour.hbx
* src/harbour.def
! fixed order of declared functions - they should be sorted.
* contrib/xhb/hbcompat.ch
! fixed HSetAACompatibility() #xtranslate
* contrib/hbzebra/code39.c
* pacified warning, it's also fix for possible buffer overflow.
* include/hbcomp.h
* src/compiler/genc.c
* src/compiler/hbmain.c
* replaced function hb_compFunctionMarkStatic() with
hb_compFunctionSetScope()
* src/compiler/genc.c
* detect public prg functions written in C inside
#pragma BEGINDUMP / #pragma ENDDUMP and mark them as LOCAL ones
* src/vm/hvm.c
! removed my pseudo fix from 2023-01-30 15:18 UTC+0100 Przemyslaw Czerpak.
It was ugly hack which can work only with some linkers.
This modification fixes FILE() function used with C functions written
without own symbol table
* src/rdd/dbf1.c
* minor code formatting
* .github/workflows/windows-ci.yml
* update windows MSYS2 integration runner set of
requested packages. PGSQL, Firebird and MariaDB libs
are no longer available in the 32-bit environment
+ add clang x86_64 compiler for testing under this CI
* include/harbour.hbx
* include/hbapi.h
* src/common/hbver.c
* src/harbour.def
* src/rtl/version.c
! fixed Windows 11 detection, by vendor likings it has
NT Kernel Version stamped with 10.0, so now Build Number
is more significant as a recognition factor.
+ new C helper functions hb_winisbuild(), hb_iswin11()
+ added PRG function hb_osIsWin11()
* also try to detect new Windows Server versions
* config/win/clang.mk
* utils/hbmk2/hbmk2.prg
* ChangeLog.txt
! recursion happening in the Makefiles rendered recently
introduced compatibility option unreliable. Changed to
use separate variable: HB_USER_FIXES=--mingw-script
You don't need it, the problem shouldn't exist, but
anyway...
* config/global.mk
! detect ARM64 CPU on Windows also under non-native shell (MSYS2 sh)
* on unix also recognize aarch64 from uname
* include/hbsetup.h
* __MINGW64__ define is not x86_64 specific anymore
* config/win/clang.mk
* utils/hbmk2/hbmk2.prg
* reworked Clang on Windows detection (ARM64, x86_64, x86),
now it recognizes those flavours in PATH as distributed
by MSYS2 project and also Visual Studio 2022 Build tools.
Starting Harbour build process should be now possible
both from MSYS2 shell (with the special note to use "make"
command instead of "win-make" from sh) and regular batch
script/cmd shell.
Building on Clang distributed by MS x86_64:
PATH=<InstallPath>\BuildTools\VC\Tools\Llvm\x64\bin;%PATH%
win-make
Clang/MS ARM64:
PATH=<InstallPath>\BuildTools\VC\Tools\Llvm\ARM64\bin;%PATH%
win-make
Building on Clang x86_64 distributed by MSYS2 from cmd:
PATH=C:\msys64\clang64\bin;%PATH%
win-make
Clang/MSYS ARM64 called from cmd:
PATH=C:\msys64\clangarm64\bin;%PATH%
win-make
* src/common/hbver.c
* patched clang version string builder to skip duplicate version
number in some builds
* append processor architecture to clang compiler string
on non-Intel systems
+ config/win/msvcarm.mk
+ config/win/msvcarm64.mk
* config/global.mk
* include/hbsetup.h
* utils/hbmk2/hbmk2.prg
+ introduced support for Windows (10,11) ARM 64-bit platform
native compilers using Visual Studio 2022 Build Tools
before calling win-make, setup environment with:
<InstallPath>\BuildTools\VC\Auxiliary\Build\vcvarsarm64.bat
toolchain should be autodetected as following:
! HB_HOST_PLAT: win (arm64) HB_SHELL: nt
! HB_PLATFORM: win (arm64) (auto-detected)
! HB_COMPILER: msvcarm64 (auto-detected: C:/Program[...]/ARM64/)
to build final application with hbmk2 adding a valid Windows
Platform SDK / CRT paths to INCLUDE= and/or LIB= variable
is needed
; cross compilation from AMD64 hosts is possible:
vcvars64.bat
win-make (build Harbour for your host first)
set HB_HOST_BIN=<HarbourPath>\bin\win\msvc64
vcvarsamd64_arm64.bat
win-make
+ also added support for Windows for ARM 32-bit
before calling win-make, setup environment using:
<InstallPath>\BuildTools\VC\Auxiliary\Build\vcvarsarm.bat
or when cross compiling from ARM64 host:
<InstallPath>\BuildTools\VC\Auxiliary\Build\vcvarsarm64_arm.bat
; please test Windows CE builds if you're using compiler newer
than VC2003 for ARM for possible regressions
* contrib/hbmlzo/3rd/minilzo/lzodefs.h
! recognize ARM64 architecture on Windows to fix build error
; consider updating to upstream minilzo 2.10, issue is fixed there
* src/common/hbver.c
* changed to add processor architecture for MSVC compiler string
on non-Intel systems, as exposed by hb_compiler(), harbour /build,
finalapp.exe //build
* tests/speedtst.prg
* workaround that processor architecture can now appear
in hb_compiler(), so it's not printed by speed test twice
* contrib/xhb/xhberror.c
! fixed very bad bug introduced in this modification:
2023-01-31 23:19 UTC+0100 Przemyslaw Czerpak
It corrupted dynamic symbol table due to stupid mistake.
I overloaded whole function symbol instead of its address only.
Many thanks to Juan Gálvez who found the problem.
* include/hbdefs.h
! fixed INT64_MIN definition
* src/vm/dynsym.c
* minor change in function order
* src/vm/memvars.c
+ allow to use symbol items in the following functions:
__mvScope(), __mvExist(), __mvGet(), __mvGetDef(), __mvPut()
* include/hbapifs.h
* src/rtl/fscopy.c
+ added new C function:
HB_BOOL hb_fileCopyEx( const char * pszSource, const char * pszDest,
HB_SIZE nBufSize, HB_BOOL fTime,
PHB_ITEM pCallBack );
Unlike hb_fileCopy() it is never redirected to remote server and copy
operation is always done locally.
pCallBack is codeblock or function symbol, it's executed at the
beginning and then on each nBufSize bytes written and receives two
parameters: nBytesWritten, nTotalSize.
Warning: nTotalSize could be 0 when non regular files like pipes or
sockets are copied.
* src/rtl/vfile.c
+ added new PRG function:
hb_vfCopyFile( <cFileSrc>, <cFileDst>, [<nBufSize>=65536], ;
[<lTimePreserve>=.t.], [<bCallBack>] ) --> <nResult>
It's wrapper to hb_fileCopyEx() C function.
For very big files setting <nBufSize> to greater value, i.e. 16777216
may increase performance.
* contrib/hbexpat/3rd/expat/expat.diff
* contrib/hbexpat/3rd/expat/loadlibr.c
! use HB_WINAPI_GETPROCADDRESS() instead of GetProcAddress()
in WinCE builds
* contrib/hbhpdf/3rd/libhpdf/hpdfstre.c
* contrib/hbhpdf/3rd/libhpdf/libhpdf.diff
! include <windows.h> in WinCE builds, it's necessary for GetLastError()
* src/3rd/jpeg/Makefile
! do not use getenv() in all WinCE builds
* src/common/hbver.c
* cast revision number to HB_ULONG to pacify warning
* src/vm/extrap.c
! do not compile unused in WinCE builds hb_winExceptionHandler()
* include/hbapi.h
* added HB_EXPORT attribute to hb_macroTextValue() function
* include/hbvm.h
* src/vm/hvm.c
+ added C function:
PHB_ITEM hb_vmCompileMacro( const char * szExpr, PHB_ITEM pDest );
It returns Harbour item or NULL at error (wrong macro text).
The result should be used with hb_vmEvalBlockOrMacro() and then freed
by hb_vmDestroyBlockOrMacro(). It hides low level macro representation
so we can change it in the future. Now it's still simple pointer item
without autodestructor.
* src/rdd/hsx/hsx.c
* src/rdd/workarea.c
* use hb_vmCompileMacro() instead of hb_macroCompile()
* src/vm/runner.c
* changed 6101 error message for "Unknown or unregistered symbol"
to "Unknown or unregistered function symbol".
* report above error when HRB module is loaded and its external function
is not registered or not marked as DYNAMIC by earlier registered modules
in HVM symbol table.
Warning. This modification may exploit problems in code loading HRB files
with cross references which has worked so far. In such case user should
pass to hb_hrbLoad()/hb_hrbRun() as first parameters flags containing
HB_HRB_BIND_LAZY or declare the problematic function in his main code
as dynamic, i.e.:
DYNAMIC <funcName>
* src/rdd/dbf1.c
* declare internal function hb_dbfTransCheckCounters() as static one
* src/harbour.def
! added missing exported Harbour functions
* contrib/hbcurl/core.c
* contrib/hbcurl/hbcurl.ch
+ added HB_CURLOPT_XFERINFODATA and HB_CURLOPT_XFERINFOFUNCTION
! use CURLOPT_XFERINFO* instead of depreciated CURLOPT_PROGRESS*
to implement HB_CURLOPT_PROGRESSBLOCK
! do not use depreciated CURLOPT_RANDOM_FILE and CURLOPT_EGDSOCKET,
they serve no purpose anymore
+ added macros for new protocols
+ added HB_CURLOPT_PROTOCOLS_STR and HB_CURLOPT_REDIR_PROTOCOLS_STR
! use CURLOPT_PROTOCOLS_STR and CURLOPT_REDIR_PROTOCOLS_STR instead
of depreciated CURLOPT_PROTOCOLS and CURLOPT_REDIR_PROTOCOLS
* emulate CURLOPT_PROTOCOLS and CURLOPT_REDIR_PROTOCOLS using
CURLOPT_PROTOCOLS_STR and CURLOPT_REDIR_PROTOCOLS_STR in new curl
versions
+ added support for HB_CURLOPT_MIMEPOST
! use CURLOPT_MIMEPOST to emulate depreciated in new curl versions
CURLOPT_HTTPPOST
+ added support for HB_CURLOPT_PROXY_SERVICE_NAME
! use CURLOPT_PROXY_SERVICE_NAME to emulate depreciated in new curl
versions CURLOPT_SOCKS5_GSSAPI_SERVICE
+ added HB_CURLINFO_ACTIVESOCKET
! use CURLINFO_ACTIVESOCKET instead of depreciated CURLINFO_LASTSOCKET
+ added HB_CURLINFO_SIZE_UPLOAD_T, HB_CURLINFO_SIZE_DOWNLOAD_T,
HB_CURLINFO_SPEED_DOWNLOAD_T, HB_CURLINFO_SPEED_UPLOAD_T,
HB_CURLINFO_CONTENT_LENGTH_DOWNLOAD_T and
HB_CURLINFO_CONTENT_LENGTH_UPLOAD_T
! use CURLINFO_*_T actions instead of depreciated ones which operate
double as file offset and emulate old actions using new ones in
new curl versions
* contrib/hbssl/evp.c
! Do no use EVP_cleanup() in OpenSSL 1.1.0 and newer.
It no longer has any effect.
* contrib/hbssl/hbssl.h
* set OPENSSL_API_COMPAT to 1.2.0 to pacify OpenSSL 3.0 API.
It hides OpenSSL 3.0 warnings but we should update the code to use
new suggested API.
* contrib/hbwin/olecore.c
! invoke assign methods with DISPATCH_PROPERTYPUTREF instead of
DISPATCH_PROPERTYPUT if assigned value is OLE object. If such
functionality is not implemented by the object (some OLE
implementations do not support it and returns DISP_E_MEMBERNOTFOUND)
then call it again but in previous form with DISPATCH_PROPERTYPUT
* include/hbapirdd.h
* src/rdd/dbcmd.c
! fixed error codes set by DbSetRelations() to be Cl*pper compatible
* src/rdd/dbcmd.c
+ accept symbol items in Select() and DbSelectArea() just like in
( <alias> ) -> <exp>
* include/hbcompdf.h
* include/hbexprb.c
* src/common/funcid.c
+ added compile time optimization for Select() function without parameters
+ added compile time optimization DbSelectArea( <nNum> | <sSym> )
+ added support for hb_PIsByRef( @localVarName ) -> <lPassedByRef>
When the parameter is passed by reference Harbour verifies if it's
existing local variable and change it to its index in parameter list
so effectively it works like hb_IsByRef( @localVarName ) in xHarbour
* contrib/xhb/hbcompat.ch
+ added translations for
hb_PIsByRef( @<localVar> ) <=> hb_IsByRef( @<localVar> )
* contrib/hbct/ctwin.c
! typo in comment
* include/hbexprb.c
* include/hbexprop.h
* src/common/expropt2.c
* src/harbour.def
* moved code for reducing NOT expression to new separate function
hb_compExprReduceNot()
; added comments to mark places for possible compiletime warnings when
incompatible types are used inside in logical expressions
(.AND. / .OR. / .NOT.)
2023-11-11 18:38 UTC+0100 Phil Krylov (phil a t krylov.eu)
* include/hbwinuni.h
* src/vm/maindllp/dllpcode.c
* In `HB_WINAPI_GETPROCADDRESS[T]`, cast the return value to `void *`
to pacify further cast warnings, as the exact function signature is almost
never a `FARPROC` anyway.
* src/compiler/harbour.yyc
* src/macro/macro.yyc
* Added `HB_SYMBOL_UNUSED( yynerrs )` to pacify compiler warnings.
* src/rtl/replic.c
* Added a cast to a comparison to pacify compiler warning.
* src/vm/classes.c
! Fixed outdated format specifier in a trace message.
* include/hbclass.ch
! fixed declaration stripping when HB_CLS_NO_DECLARATIONS is defined
* src/compiler/hbmain.c
! do not generate warnings for undeclared methods when -w[12] is used.
They should be emitted only for -w3 or higher (if any in the future).
Thanks to Toninho for the information about the problem.
* include/harbour.hbx
* src/harbour.def
* src/rtl/vfile.c
+ added new PRG function:
hb_vfIsLocal( <cFileName> ) --> <lLocalFileSystem>
It returns TRUE if <cFileName> is not redirected to any Harbour File IO
driver but access local file system API.
* contrib/xhb/traceprg.prg
* merged with Victor's branch so now Harbour File IO API is used to
access log file.
* convert relative log file name to absolute file path name
% write whole log entry in single IO operation.
* include/hbapierr.h
* src/harbour.def
* src/rtl/errapi.c
+ added new C function:
void hb_errReinit( PHB_ITEM pError );
it allows to replace default error object with user custom one which
support :Init() method
* contrib/xhb/xhb.hbp
+ contrib/xhb/xhb.h
+ contrib/xhb/xhberror.prg
+ added code which extends Harbour error objects adding functionality
known from xHarbour:
oError:ProcName
oError:ProcLine
oError:ProcModule
Above instance variables are initialized automatically when error
object is created.
To enable this functionality it's enough to add to PRG code:
REQUEST xhb_ErrorNew
* contrib/xhb/xhberr.prg
* use error object :Proc*() methods if they are available
- removed PRG version of xhb_ErrorNew()
* contrib/xhb/xhbmemo.prg
* contrib/xhb/xhbtedit.prg
! use xhb_ErrorNew() instead of ErrorNew() which does not support
any parameters
* include/hbapicls.h
; typo in comment
* include/harbour.hbx
! added missing hb_socketSetNoSigPipe()
* src/rdd/dbfcdx/dbfcdx1.c
* redirect all debug output to stderr
* reenabled light debug code for early detection of index corruption
* contrib/gtqtc/gtqtc.h
* contrib/gtqtc/gtqtc1.cpp
* replaced depreciated in 5.15 methods for mouse wheel events processing
by modern ones
* replaced depreciated in 5.15 methods for accessing screen geometry by
modern ones
* contrib/gtqtc/gtqtc1.cpp
+ added support for TAB key processing when other widgets are shown and
TAB is use to switch focus
* src/rtl/alert.prg
* src/rtl/hbgtcore.c
+ added support for hb_Alert() parameters passed in hash array, i.e.
hb_Alert( { "TXT" => <cMessage>, ;
"BTN" => <aButtons>, ;
[ "TIM" => <nTimeOut> ] } ) => <nButton>
* contrib/gtqtc/gtqtc.h
* contrib/gtqtc/gtqtc1.cpp
* include/hbgtinfo.ch
+ added new hb_gtInfo() action: HB_GTI_MSGBOX
It opens QMessageBox with given by user text and buttons, i.e.
hb_gtInfo( HB_GTI_MSGBOX, <cMessage>, <aButtons> ) => <nButton>
or:
hb_gtInfo( HB_GTI_MSGBOX, ;
{ "TXT" => <cMessage>, ;
"BTN" => <aButtons>, ;
[ "TIM" => <nTimeOut>, ] ;
[ "TIT" => <cTitle>, ] ;
[ "INF" => <cInfoText>, ] ;
[ "DTL" => <cDetailedText> ] } ) => <nButton>
+ added support for redirecting Alert() and hb_Alert() calls to QT GUI
QMessageBox(). This functionality can be enabled by:
hb_gtInfo( HB_GTI_MSGBOX, .t. )
and then
Alert( <cMessage>, <aButtons> )
is redirected to QT GUI QMessageBox().
* include/hbgtinfo.ch
* contrib/gtqtc/gtqtc.h
* contrib/gtqtc/gtqtc.hbc
* contrib/gtqtc/gtqtc.hbp
* contrib/gtqtc/gtqtc1.cpp
+ added new hb_gtInfo() action: HB_QT_SOUND
It allows to play sound using the low level GUI library, i.e.
hb_gtInfo( HB_QT_SOUND, <cSoundFile> )
This modification needs additional QT5 library Qt5Multimedia so I decided
to enable it conditionally by HB_QT_SOUND=yes environment variable.
* include/hbapi.h
* src/vm/cmdarg.c
! changed returned type in hb_verSvnID() form int to HB_MAXINT
* src/nortl/nortl.c
! fixed void* pointer incrementation
! pacified warnings
* src/rdd/dbcmd.c
* accept NIL value in hb_FieldPut() according to Aleksander recommendation
* include/hbapi.h
* src/common/hbver.c
* src/main/harbour.c
* src/pp/hbpp.c
* src/rtl/version.c
* src/vm/cmdarg.c
! fixed integer overflow in revision number calculated from
the last ChangeLog entry date
* include/harbour.hbx
* src/harbour.def
* src/rdd/dbcmd.c
+ added two new functions:
hb_FieldGet( <cFieldName> | <nFieldPos> )
-> <xFieldValue> | NIL
hb_FieldPut( <cFieldName> | <nFieldPos>, <xFieldValue> )
-> <xFieldValue> | NIL
They works like FieldGet() and FieldPut() but allows to use field
name instead of field index.
* include/hbvm.h
* include/harbour.hbx
* src/harbour.def
* src/vm/hvm.c
+ added new C function:
extern HB_EXPORT HB_BOOL hb_vmSetKeyPool( HB_BOOL fEnable );
It allows to disable keyboard pooling by GT driver in main HVM loop.
It's important in programs which mix GT terminal with GUI library
which use common event loop. Many GUI objects are not reentrant
safe and activating event loop during big changes may cause crash.
Such things can happen in applications which try to mix HBQT
and GTQTC. To avoid such problems before PRG code is called it's
enough to disable keyboard pooling in main HVM loop, eg.
if( hb_vmRequestReenter() )
{
HB_BOOL fKeyPool = hb_vmSetKeyPool( HB_FALSE );
hb_vmPushEvalSym();
hb_vmPush( pBlockItm );
hb_vmSend( 0 );
hb_vmSetKeyPool( fKeyPool );
hb_vmRequestRestore();
}
+ added new PRG function:
__vmKeyPool( [<fEnable>] ) -> <fPrevState>
It's PRG interface to above C function. If application uses GT driver
and GUI library using the same event loop and such GUI library does not
disable keyboard pooling in main HVM loop before PRG code is activated
then it's enough to call this function at the beginning of application,
eg.
PROCEDURE INIT Clip()
__vmKeyPool( .f. )
RETURN
With above peace of code you can mix HBQT or other QT wrapper with GTQTC.
+ src/rtl/vfilehi.prg
* src/rtl/Makefile
* src/harbour.def
* include/harbour.hbx
+ added FILE API counterparts to hb_DirBuild() and hb_FNameExists()
hb_vfDirBuild( <cDir> ) -> <lSuccess>
hb_vfDirUnbuild( <cDir> ) -> <lSuccess>
hb_vfNameExists( <cName> ) -> <lExists>
; their implementations are imported from 3.4
https://github.com/vszakats/hb/blob/master/src/rtl/hbfilehi.prg
but were renamed to fit in the hb_vf* namespace
* src/rtl/hbfilehi.prg
* use hb_LeftEq() instead of Left() ==
; partial merge of 3.4
2015-07-27 14:26 UTC+0200 Viktor Szakats
! fix several sloppy checks where Empty() were
used on strings while in fact a zero length
check would have been correct.
; Empty() is a "weasel-word" when used on strings,
should be used carefully and sparingly.
* ChangeLog.txt
% avoid spaces before EOLs
* contrib/hbct/dattime2.c
% minor optimization
* src/rdd/dbsql.c
; added somment with syntax info
* include/inkey.ch
* src/rtl/hbgtcore.c
+ added support for mouse wheel left and right events to base GT code
* config/linux/clang.mk
! fixed rule for dynamic library
* src/3rd/png/Makefile
+ added -DPNG_ARM_NEON_OPT=0 to build flags
* contrib/3rd/sqlite3/sqlite3.c
* contrib/3rd/sqlite3/sqlite3.diff
! pacified warning
* contrib/gtwvg/gtwvgd.c
* contrib/gtwvg/wvgwing.c
! fixed missing break/return in case statements - please verify it.
* contrib/hbct/dattime3.c
* added #define _DEFAULT_SOURCE necessay in new Linux distors
* contrib/hblzf/3rd/liblzf/liblzf.diff
* contrib/hblzf/3rd/liblzf/lzfP.h
* do not use nested #define in #if statements - some C compilers do not
support it
* contrib/hbssl/bio.c
! tuned #if condition
* contrib/hbmisc/hbeditc.c
* simpliefied for condition and pacified warning
* contrib/hbodbc/hbodbc.hbp
* contrib/sddodbc/sddodbc.hbp
+ added check for iodbc library
* utils/hbmk2/hbmk2.prg
+ added support for clang in android builds
* include/hbdefs.h
+ added check for __BYTE_ORDER__ macro used in some new lib C
implementations
* include/hbapi.h
* include/hbdefs.h
* include/hbstack.h
* include/hbvmpub.h
* src/vm/classes.c
* src/vm/dynsym.c
* src/vm/estack.c
* src/vm/memvars.c
+ extended the size of dynamic symbol table from 65535 to 4294967295.
Adopting class code I decided to keep current algorithm of method indexes
hashing with only some minor modifications. It's very fast anyhow it may
cause noticeable (though static) quite big memory allocation for class
definitions in applications using millions of symbols and which increase
dynamic symbol table at runtime loading new classes dynamically form .hrb,
.dll, .so or other dynamic libraries supported by Harbour. It's random
and rather impossible to exploit situation in real life anyhow I cannot
exclude it so I'd like to report it in ChangeLog. The solution is very
simple, i.e. it's enough to use classic divide et impera algorithm using
symbol numbers to find method definition anyhow it will be slower then
current one and address only very seldom hypothetical situations so I
decided to not implement it. Such static memory cost begins to be
completely unimportant in the world of 64-bit architectures and extremely
big memory address space.
The modification was sponsored by TRES company.
* src/vm/estack.c
! fixed __mvClear() in MT builds - due to stupid typo GetList variable
was removed in MT programs by CLEAR MEMORY command (__mvClear())
So far noone reported it and I've found it analyzing the code before
increasing symbol table size.
* contrib/hbwin/hbolesrv.c
* updated for new size of dynamic symbol table
+ src/codepage/cpde858.c
* src/codepage/Makefile
* include/hbcpage.hbx
* src/harbour.def
+ added codepage DE858, it is essentially DE850 but with EURO SIGN
(U+20AC) at position 0xD5 instead of Turkish dotless-i (U+0131)
Issue #201
* src/vm/fm.c
* OpenWatcom remaining pragmas related to 201 warning (unreachable code)
assigned to C mode (wcc386) build only - issue #202
* include/dbinfo.ch
* src/rdd/dbf1.c
+ added DB_SETHEADER_EOL flag, it's used to force setting EOL marker
when header is written. In Harbour's DBF* RDDs is set in CLOSE()
method so just like in Clipper when DBF is closed and header has to
be updated the EOL() marker is set.
As side effect reducing header updates to minimal level (in such
case DBF header is not updated after APPEND what is safe for Harbour,
Clipper and compatible RDDs because they use file size to calculate
number of records but some other DBF drivers may be confused)
increase the APPEND speed and also forces EOL setting in all cases
when CLOSE() method is called. Header updates can be reduce to minimal
level by:
hb_rddInfo( RDDI_SETHEADER, DB_SETHEADER_MINIMAL )
* src/rdd/usrrdd/usrrdd.c
! fixed GPF in UsrRDD redirector for DROP(), EXISTS() and RENAME() methods
* src/vm/cmdarg.c
* use HB_MEM_STATISTICS instead of HB_MEM_USEDMAX to check if memory
statistic module is enabled in //info message
* bin/commit.hb
! fixed UTC offset formatting
* contrib/hbwin/win_os.prg
* updated win_osNetRegOk() for modern MS-Windows versions:
- on Win7 and upper set
System\CurrentControlSet\Services\LanmanServer\Parameters\DisableLeasing
to disable opportunistic locks.
- do not force SMB1 to disable oplocks on Win7 and upper - new MS-Win10
does not support SMB1 at all so this setting on the server with
such system completely disables SMB network and forcing SMB1 on the
client side blocks access to new Win10 servers.
Warning! this setting is still activated on Vista so it cannot
connect work with new Win10 but I do not know any other
working method to disable oplocks in Windows Vista.
* synced with Viktor's branch
* contrib/xhb/hbserv.c
! added missing return
* include/harbour.hbx
* src/harbour.def
* src/rtl/version.c
+ added new PRG functions:
hb_osIsWin7(), hb_osIsWin8(), hb_osIsWin81(), hb_osIsWin10()
* src/rtl/gttrm/gttrm.c
+ added autodetection for few other XTerm compatible terminals
+ respect color extension in TERM name of all XTerm compatible
terminals
* utils/hbmk2/hbmk2.prg
+ added support for -cpp=isoXX borowed from Viktor's branch
* contrib/gtqtc/gtqtc.hbc
* contrib/gtqtc/gtqtc.hbp
* use -cpp=iso11 required for QT 5.7.0 or upper
* extended QT detection and partial syncing with Viktor's branch
* include/harbour.hbx
* include/hbchksum.h
* src/harbour.def
* src/rtl/hbmd5.c
+ added new C function
void hb_hmac_md5( const void * key, HB_SIZE nKeyLen,
const void * message, HB_SIZE nMsgLen,
char * digest );
+ added new PRG function
HB_HMAC_MD5( <cMessage>, <cKey> ) -> <cMAC>
; HMAC MD5 is used in some authentication methods i.e. CRAM-MD5
authentication in SMTP protocol
* src/rtl/gttrm/gttrm.c
+ enable extended XTERM colors automatically when xterm-16color,
xterm-256color, xterm+256color, xterm-88color, rxvt-16color,
rxvt-256color, rxvt-98color or putty terminal is detected
+ added support for HB_GTI_SCREENDEPTH
* contrib/gtqtc/gtqtc1.cpp
+ set clipboard data also into selection buffer on platforms which
support it
+ when main clipboard buffer is empty retrieve clipboard data from
selection buffer and if it's empty too from find buffer
* include/harbour.hbx
* src/harbour.def
* src/vm/memvars.c
+ added new PRG function
__mvGetDef( <cMemvar> [, <xDefault> ] ) -> <xValue>
it works in similar way to __mvGet() but if <cMemvar> does
not exist then it returns <xDefault> or NIL instead of RTE
* contrib/hbwin/hbwin.hbx
+ added missing declarations for win_osIs10() and win_osIs81()
* contrib/hbssl/ssl.c
! fixed compilation with OpenSSL < 0.9.8k
* contrib/sddodbc/core.c
* pacified warning
* contrib/xhb/hbxml.c
* contrib/xhb/hbxml.h
! fixed CDATA terminator, it should be "]]>" instead of " ]]>"
+ added support for HBXML_STYLE_NONEWLINE flag
* src/3rd/pcre/Makefile
! typo wince -> wce
* src/main/Makefile
* use harbour.rc only in MS-Windows builds - it fixes OS2 OpenWatcom
builds
* config/global.mk
! in DOS builds check if the 'uname' exists before running it
* include/hbapigt.h
* include/hbgtcore.h
* src/rtl/hbgtcore.c
* src/rtl/inkey.c
* src/rtl/inkeyapi.c
* contrib/xhb/xhbfunc.c
+ added 3-rd parameter to hb_inkeySetText() function. When this parameter
is true then ';' is not translated to Chr( 13 ) but DOS and UNIX EOLs
are translated to Chr( 13 )
+ added logical parameter to HB_GTI_CLIPBOARDPASTE which allows to set
text without ';' translation but with platform independent EOLs
* src/rtl/memoedit.prg
* utils/hbmk2/hbmk2.prg
* contrib/hbnetio/utils/hbnetio/_console.prg
* extras/dbu/dbu52.patch
* extras/dbu/dbu53.patch
* use hb_gtInfo( HB_GTI_CLIPBOARDPASTE, .T. ) instead of
hb_gtInfo( HB_GTI_CLIPBOARDPASTE ) for pasting data from clipboard
with platform independent EOLs and ';' characters
* include/hbapi.h
* src/common/hbver.c
* src/harbour.def
* contrib/hbwin/win_osc.c
* synced Windows version detection code with Viktor's work in 3.4
* src/main/Makefile
+ src/main/harbour.rc
* include Windows manifest in standalone compiler - "harbour.exe /build"
should now correctly report Windows 10. "yourapp.exe //build" plus
in-app results still depend on manual manifest inclusion within .rc
* include/hbapiitm.h
* src/rtl/hbdef.c
* src/vm/itemapi.c
* src/harbour.def
+ added new public C function
HB_BOOL hb_itemTypeCmp( PHB_ITEM pItem1, PHB_ITEM pItem2 );
which uses low level item type comparison code taken from hb_default()
and hb_defaultValue() PRG functions
* include/harbour.hbx
* src/harbour.def
* src/vm/hashfunc.c
+ added new PRG function
hb_HSetDef( <hVal>, <xKey> [, <xDefVal> ] )
it checks if <xKey> exists and if not then adds it to hash array
and optionally sets key value to <xDefVal> otherwise (the key
exists in hash array) and <xDefVal> is given then it checks if
the type of key's value is compatible with <xDefVal> and if not
then replaces key's value with <xDefVal>.
In other words it's combination of hb_HGetDef() and hb_default()
and works like this PRG code:
FUNCTION hb_HSetDef( hVal, xKey, xDefVal )
IF xKey $ hVal
hb_default( @hVal[ xKey ], xDefVal )
ELSE
hVal[ xKey ] := xDefVal
ENDIF
RETURN hVal
but much faster.
* src/rtl/dirscan.prg
* remove READONLY attribute from deleted directories.
Be careful. Now this function allows to recursively remove
all directories even if they have READONLY attribute.
* config/darwin/clang.mk
* config/darwin/gcc.mk
* config/darwin/global.mk
* config/darwin/icc.mk
* config/darwin/libs.mk
* utils/hbmk2/hbmk2.prg
* synced with Viktor's modifications in 3.4 branch:
; 2017-11-11 13:14 UTC Viktor Szakats:
* switch to call the C compiler to link dynamic libs on darwin,
which is the recommended way to do this, as suggested by Apple.
This also aligns the platform better with other *nix platforms.
'libtool' was used before, but that started having intermittent
issues around Sierra (mitigated by disabling parallel build),
which returned and got worse in High Sierra (with no remedy).
The symptom was 'ld: file not found: ' errors with the filename
not shown or appearing as garbage, then a 'libtool: internal link
edit command failed'. This was reported and will be fixed in a
future Xcode release.
Ref: Apple Radar 34944562
* config/global.mk
* config/rules.mk
+ added new user build envvar HB_USER_DCFLAGS
It allows to set C compiler parameters to compile .c code for
dynamic libraries.
* src/rtl/hbsocket.c
! fixed fcntl(F_SETFL) 3-rd parameter in hb_socketSetBlockingIO(),
By mistake I used long instead of int. it created problem on big
endian 64 bit machines.
* include/hbgtcore.h
* src/rtl/hbgtcore.c
+ added new C functions for GT programmers:
void hb_gt_BaseUnlock( PHB_GT pGT );
void hb_gt_BaseLock( PHB_GT pGT );
void hb_gtSleep( PHB_GT pGT, double dSeconds );
* src/rtl/hbgtcore.c
* src/rtl/gtcrs/gtcrs.c
* src/rtl/gtdos/gtdos.c
* src/rtl/gtgui/gtgui.c
* src/rtl/gtos2/gtos2.c
* src/rtl/gtpca/gtpca.c
* src/rtl/gtsln/gtsln.c
* src/rtl/gtstd/gtstd.c
* src/rtl/gttrm/gttrm.c
* src/rtl/gtwin/gtwin.c
* src/rtl/gtwvt/gtwvt.c
* src/rtl/gtxwc/gtxwc.c
* use new functions to unblock GT when low level TONE() code is
executed. It allows other threads to access shared GT driver
when one of them executes TONE()
* contrib/hbexpat/hbexpat.hbx
* regenerated automatically
* contrib/hbamf/amfenc.c
* update amf3_Encode() serialize function with great idea of Przemek
implemented recenly in hb_Serialize(). Checking garbage collector
reference count can save time here too, unique arrays and hashes won't
be indexed as references.
* contrib/hbexpat/3rd/expat/*
! updated to 2.2.5 (from 2.2.1) using 3rdpatch.hb, expat.diff from
Viktor's 3.4 fork was used - but (again) adapted for DOS 8.3 naming
scheme. Also i've kept local patches for WinCE, OpenWatcom DOS/OS2,
please test. Compilation in CPP mode is explicitly disabled for libexpat
now, as upstream decided to ignore such use cases completly.
* contrib/hbexpat/*
* synced with Viktor's 3.4 fork
+ include/hbarc4.h
* src/harbour.def
+ export ARC4 core routines like in 3.4 fork, updated expat lib reuses it.
* contrib/hbtip/hbtip.hbp
! adapt *.hbx file specifier to hbmk2 3.2 syntax
* contrib/hbnf/fttext.c
* contrib/hbnf/hbnf.hbx
* synced with Viktor's modifications:
- use FILE API instead of FS API
- protection against GPF when FT area has no valid file handle
- alternative function names which exceeds 10 character Cl*pper limit
! fixed handle and memory leak when new file is open in FT area which
already has open file
* include/harbour.hbx
* src/harbour.def
* src/vm/hvm.c
+ added new PRG function: __vmItemRefs( <xVal> ) -> <nRefs>
Please remember that passed <xVal> parameter increases the counter so
if you want to eliminate it then pass it by reference, i,e.:
aVal := {}
? __vmItemRefs( aVal ), __vmItemRefs( @aVal )
* include/hbapi.h
* src/harbour.def
* src/vm/arrays.c
* src/vm/codebloc.c
* src/vm/hashes.c
+ added new C functions:
HB_COUNTER hb_arrayRefs( PHB_ITEM pArray );
HB_COUNTER hb_hashRefs( PHB_ITEM pHash );
HB_COUNTER hb_codeblockRefs( PHB_ITEM pItem );
* src/rtl/itemseri.c
* added new internal function to replace 3 times repeated code for
resizing array of references.
% use new functions to check number of references to hashes, arrays and
objects to ignore in system detecting multiple references the items
which have reference counter smaller then 2. Such items cannot have
multiple references.
In practice this modification resolves the time problem when very big
items are serialized to the level comparable to HB_SERIALIZE_IGNOREREF
so now this flag is useless though I'll keep it because only in such
mode serialization code can be used as filter in streams.
I also created binary tree to store references but after this
modifications it is not necessary so I decided to not commit it yet.
Please only remember that practice limit of such serialization code
is created by physical memory attached by OS to the process. When this
limit is exceed then swap is actively used what completely kill the
performance. The CPU usage is reduced to less then 1% because it waits
for restoring swapped memory pages by extremely heavy used disks. In
such case I cannot help.
* include/hbjson.h
* src/rtl/hbjson.c
* changed 3-rd parameter 'HB_BOOL fHuman' in hb_jsonEncode() and
hb_jsonEncodeCP() C functions to 'int iIndent'
char * hb_jsonEncode( PHB_ITEM pValue, HB_SIZE * pnLen,
int iIndent );
char * hb_jsonEncodeCP( PHB_ITEM pValue, HB_SIZE * pnLen,
int iIndent, PHB_CODEPAGE cdp );
Positive iIndent value defines number of spaces used for indenting,
0 disables indenting and -1 means TAB ( ASCII:9 ) indenting.
This modification is binary compatible though in C code using HB_TRUE
as 3-rd parameter it changes indenting from 2 spaces to 1 so please
update the code if it's significant.
+ added optional support for numeric value in 2-nd parameter of
hb_jsonEncode() PRG function. Current syntax is:
hb_jsonEncode( <xValue>, [ <lHuman> | <nIndent> ], [ <cDestCP> ] )
-> <cJSONString>
* contrib/rddads/ads1.c
* minor simplification
* include/hbapigt.h
* formatting
* src/pp/ppcore.c
! fixed line numbering in unclosed extended block
* interrupt extended block processing if it contains internal
FUNCTION or PROCEDURE declaration
* include/hbcompdf.h
* src/compiler/complex.c
* src/compiler/harbour.y
* src/compiler/hbmain.c
! fixed possible GPF when FUNCTION/PROCEDURE starts in unclosed
control structure - thanks to Abe Buchinger for information
about the problem
* report all unclosed control structures when FUNCTION/PROCEDURE starts
* src/compiler/harbour.yyc
* src/compiler/harbour.yyh
* regenerated
* contrib/hbmemio/memio.c
! fixed FO_TRUNC attribute in MEM: file IO - thanks to Miroslav Georgiev
for information about the problem
* include/hbgtinfo.ch
* src/rtl/gtwvt/gtwvt.h
* src/rtl/gtwvt/gtwvt.c
+ added new hb_gtInfo() action HB_GTI_SYSMENUADD
It allows to add new items to window system menu which
selected add defined by user key code into keyboard queue, i.e.
hb_gtInfo( HB_GTI_SYSMENUADD, HB_K_MENU, "Settings" )
* In WinCE builds use DeleteMenu() + AppendMenu() instead of
missing ModifyMenu()
! pacified warning in non UNICODE builds
* *
* partial sync with the 3.4 fork codebase. These are the things
synces for the most part:
- copyright headers
- grammar/typos in comments and some readmes
- comment/whitespace/decorations
- variable scoping in C files
- DO CASE/SWITCH and some other alternate syntax usage
- minimal amount of human readable text in strings
- minor code updates
- HB_TRACE() void * casts for pointers and few other changes to
avoid C compiler warnings
- various other, minor code cleanups
- only Harbour/C code/headers were touched in src, utils, contrib,
include. No 3rd party code, no make files, and with just a few
exceptions, no 'tests' code was touched.
- certain components were not touched were 3.4 diverged too much
already, like f.e. hbmk2, hbssl, hbcurl, hbexpat
- the goal was that no actual program logic should be altered by
these changes. Except some possible minor exceptions, any such
change is probably a bug in this patch.
It's a massive patch, if you find anything broken after it, please
open an Issue with the details. Build test was done on macOS.
The goal is make it easier to see what actual code/logic was changed
in 3.4 compared to 3.2 and to make patches easier to apply in both
ways.
