* *
* 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.
* contrib/hbexpat/3rd/expat/_hbconf.h
* contrib/hbexpat/3rd/expat/xmlparse.c
! fixed WINCE builds
! fixed OpenWatcom DOS builds
! fixed OpenWatcom OS2 builds
* contrib/hbexpat/3rd/expat/expat.dif
* rediffed
* include/harbour.hbx
* include/hbvm.h
* src/harbour.def
* src/vm/hvm.c
* src/vm/thread.c
+ added new C function
HB_BOOL hb_vmThreadIsMain( void * );
it checks if given or current thread is main HVM thread
+ added new PRG function
hb_threadIsMain( [ <pThID> ] ) -> <lMainHvmThread>
it returns true if given or current thread is main HVM thread
+ added new PRG function
hb_mutexExists( <pMtx> ) -> <lExists>
it returns true if passed <pMtx> parameter is pointer item to
fully functional mutex
* src/rtl/tpersist.prg
! fixed problem with array item deserialization - many thanks
to Peter Rees for exact information about the problem.
* src/rtl/val.c
* minor simplification
* include/hbcompdf.h
* include/hbexprb.c
* src/common/expropt1.c
* redefined all HB_ET_MACRO_* macros as bitfields
* src/compiler/complex.c
* small modification for future extensions in macro usage
* src/compiler/hbcomp.c
! replaced old temporrary hack with internal error.
* src/compiler/hbmain.c
* code reformating
* src/compiler/harbour.y
* separated DECLARE tokens
* src/compiler/harbour.yyc
* src/compiler/harbour.yyh
* regenerated
* src/rdd/dbffpt/dbffpt1.c
! fixed possible bad interaction (modified default memo type) due to
uninitialized item returned by hb_stackAllocItem(). Such items may
contain any simple value.
* src/vm/memvars.c
* minor variable name cleanup
* utils/hbtest/rt_math.prg
* repeat tests for macro messages and macro <op>=, pre/post ++/--
operations with and without pass by reference (-ks) optimization
* include/hbapifs.h
* src/rtl/filesys.c
+ added new C functions for UNIX and DJGPP builds:
int hb_fsPollFD( PHB_POLLFD pPollSet, int iCount,
HB_MAXINT nTimeOut );
int hb_fsCanRead( HB_FHANDLE hFileHandle, HB_MAXINT nTimeOut );
int hb_fsCanWrite( HB_FHANDLE hFileHandle, HB_MAXINT nTimeOut );
These functions should be used instead of select() in C code to hide
low level access to select()/poll() functionality in *nix builds
(they are supported by DJGPP only to simplify existing code common
for DJGPP and *nix builds). Maximum file handle value which can be
used in select() is limited by FD_SETSIZE. Please note that it's
file handle value not number of file handles in the set. It creates
serious problem for applications which operate on great number of
handles (i.e. servers which have to keep open many sockets, pipes,
files, etc. for their clients) so the new file/socket/pipe/...
handle value can easy exceed FD_SETSIZE limit and in such case
cannot be used with select(). The modification on
2016-04-05 21:24 UTC+0200 Przemyslaw Czerpak
resolved the problem only for sockets and pipes in code which uses
corresponding hb_socket*() and hb_fsPipe*() API but not for all
other cases. This one is for POSIX compilant code which needs pure
POSIX select()/poll() functionality.
Please note that HB_POLLFD structure should is compatible with
struct pollfd defined by POSIX.1-2001 anyhow not all platforms
confirm this standard so portable Harbour code should always use
HB_POLLFD and HB_POLL* constant values instead of POLL* ones.
* include/hbdate.h
* src/common/hbdate.c
+ added new C functions to calculate timeouts:
HB_MAXUINT hb_timerGet( void );
HB_MAXUINT hb_timerInit( HB_MAXINT nTimeOut );
HB_MAXINT hb_timerTest( HB_MAXINT nTimeOut, HB_MAXUINT * pnTimer );
They are designed to be used instead of direct access to
hb_dateMilliSeconds(). Now they internally use hb_dateMilliSeconds()
but it can be easy replaced by any other system monotonic clock by
one local modification inside hb_timerGet() function.
* src/rtl/filesys.c
* use hb_timer*() functions instead of hb_dateMilliSeconds()
* use hb_fsCanRead()/hb_fsCanWrite() instead of select()/poll()
It also fixed timeout processing inside hb_fsPipeIsData() and
hb_fsPipeWrite() in builds using poll()
* src/rtl/filesys.c
* src/rtl/gtcrs/gtcrs.h
* src/rtl/gtcrs/gtcrs.c
* src/rtl/gtpca/gtpca.c
* src/rtl/gtsln/gtsln.c
* src/rtl/gtsln/mousesln.c
* src/rtl/gtstd/gtstd.c
* src/rtl/gttrm/gttrm.c
* src/rtl/gtxwc/gtxwc.c
* use hb_timer*() functions instead of hb_dateMilliSeconds()
* use hb_fsCanRead()/hb_fsCanWrite() instead of select()/poll()
* src/vm/thread.c:
* src/rtl/gtwin/gtwin.c
* src/rtl/hbcom.c
* src/rtl/hbgtcore.c
* src/rtl/hblpp.c
* src/rtl/idle.c
* contrib/hbnetio/netiosrv.c:
* contrib/hbssl/ssl_sock.c:
* use hb_timer*() functions instead of hb_dateMilliSeconds()
* contrib/xhb/hboutdbg.c
* use hb_fsCanWrite() instead of select()
* src/rtl/hbsocket.c
! repeat select() interrupted by signal inside hb_socketSelect()
when poll() function is not available
* src/3rd/hbdossrl/serial.c
! fixed -Wshift-negative-value GCC warnings
* include/hbcomp.h
* include/hbcompdf.h
* include/hberrors.h
* include/hbexprb.c
* src/common/expropt1.c
* src/compiler/hbgenerr.c
* src/compiler/hbmain.c
+ added new macros HB_ET_MACRO_NOLIST and HB_ET_MACRO_NOPARE
% use new HB_ET_MACRO_* macros
+ added new compile time error:
"Code block contains both macro and with object messages ':%s'"
NOTE: -kd compiler switch allows to compile codeblocks with macros
and declared symbols / with object messages
* replaced hb_compErrorCodeblock() with hb_compErrorCodeblockDecl() and
hb_compErrorCodeblockWith()
+ added new C function hb_compPushMacroVar()
* code simplification
; added few comments
* utils/hbtest/rt_math.prg
* extended test code for macro messages and macro <op>=, pre/post ++/--
operations
* src/vm/hvm.c
! protection against executing hb_threadStateNew() during GC pass
inside hb_vmRequestReenterExt()
* src/vm/garbage.c
! add volatile attribute to s_bCollecting variable
* small modification in hb_gcAll() parameter
* src/rtl/errsys.prg
; minor comment cleanup
* doc/xhb-diff.txt
* extended a little bit section STRONG TYPED VARIABLES
* include/hbdefs.h
+ added macros HB_CAST_*( double ) for conversions of double values
to integer types - they are designed to hide some differences
between C compilers and platforms in such conversions
* src/common/expropt2.c
* src/vm/extend.c
* src/vm/itemapi.c
* use new HB_CAST_*() macros
* include/hbapiitm.h
* src/vm/itemapi.c
+ added new C function:
PHB_ITEM hb_itemPutNil( PHB_ITEM pItem );
* src/rdd/dbf1.c
! fixed yet another RDDI_SETHEADER clearing in internal DBF RDD code
* clear few passed variables before internal SELF_INFO() and
SELF_RDDINFO() read calls - they were safe in current code but
potentially may badly interact with 3-rd party DBF RDD wrappers.
* doc/cmpopt.txt
* enumeration
! typos
* src/codepage/uckam.c
+ added info about Kamenicky codepage number: CP895
* include/hbapi.h
* src/common/hbstr.c
+ added new C function:
char * hb_dblToStr( char * szBuf, HB_SIZE nSize,
double dNumber, int iMaxDec );
it converts numeric value to string trying to keep all significant
digits in double numbers
* include/harbour.hbx
* src/rtl/hbntos.c
+ added new PRG function:
hb_ntoc( <nValue> ) -> cValue
it converts numeric value to string trying to keep all significant
digits in double numbers
* src/rtl/hbsocket.c
+ added support for error codes returned by
getaddrinfo() / getnameinfo() and gethostbyname() / gethostbyaddr()
* src/rtl/hbinet.c
+ set HB_SOCKET_ERR_WRONGADDR error code if hb_socketResolveAddr()
returns NULL without setting error code
* src/vm/arrays.c
* src/vm/hashfunc.c
! fix hb_HScan(), AScan(), hb_AScan() and hb_RAScan() for very large
integers with more then 53 significant bits. Such bits were lost
after conversion to double value used in scan process
* src/vm/itemapi.c
* formatting
* src/rtl/gtwvt/gtwvt.c
! do not convert characters received with ALTGR or ALT+CTRL flags to
extended keycodes - some national keyboards may use such combination
for national characters and even pure ASCII ones
* contrib/hbnetio/netiosrv.c
* minor optimization
* src/rdd/dbcmd.c
* src/rdd/wacore.c
* src/rdd/wafunc.c
! added protection against exceeding the limit of workarea numbers.
Current maximum is 65534, WA 65535 is reserved for internal RDD
WA destructor.
! fixed possible memory leak in hb_rddInsertAreaNode()
* src/vm/set.c
! fixed typo in device detection code
+ added NUL to list of known device names in DOS, OS2 and MS-Win builds
+ recognize \\<server>\<device>, \\.\PIPE\*, \\.\<device> as
device names in DOS, OS2 and MS-Win builds.
* src/rtl/gttone.c
* applied patch from Andi Jahja for new BCC versions - many thanks
* src/rtl/strrepl.c
% added optimization code for strings longer then 1024 bytes
For very long strings it can improve replace performance 20 or
even more times.
* src/rtl/sha2.c
* src/rtl/val.c
* src/vm/dynsym.c
* src/vm/eval.c
* src/vm/thread.c
* formatting
* contrib/hbnetio/hbnetio.hbx
* contrib/hbnetio/netiocli.c
* contrib/hbnetio/readme.txt
+ added new client side function:
netio_SetPath( <pConnection> [, <cPath>] ) -> [<cPrevPath>]
It set/get path prefix for automatic file redirection to HBNETIO.
If automatic redirection is activated then <cPath> is removed
from file name passed to HBNETIO server, i.e.:
netio_SetPath( netio_GetConnection(), "//myserver/myshare" )
[...]
/* open "/data/table" on HBNETIO server */
use "//myserver/myshare/data/table"
* allow to pass filename to netio_Decode() with "net:" prefix
* allow to pass to use "net:" prefix in function names passed
to netio_{Proc|Func}*() when <pConnection> parameter is not used
! fixed possible GPF trap when connection cannot be set
* src/vm/memvars.c
% do not create reference chain when memvars are passed by reference
and called function uses PARAMETERS statement
! do not return references from __mvDbgInfo() but destination values
* doc/xhb-diff.txt
! typo
* src/vm/itemapi.c
! typo in debug trace message
* include/hbapifs.h
* src/common/hbfsapi.c
+ added automatic file handle table resizing to hb_fsOS2DosOpenL()
+ added new C function: hb_fsOS2DosOpen() which is wrapper to DosOpen()
with automatic file handle table resizing.
Please remember that unlike hb_fsOS2DosOpenL() this function is
pure wrapper to DosOpen() so it does not make file name conversions
and does not set hb_fsError().
* src/rtl/filesys.c
* src/rtl/hbcom.c
* src/rtl/hbproces.c
* use hb_fsOS2DosOpen() instead of DosOpen()
* src/vm/cmdarg.c
* set the initial size of file handle table to 256 in OS2 OpenWatcom
builds if user does not set it explicitly by //F:<n> switch
* config/global.mk
- removed set HARBOUR=F:100 - it's not necessary with automatic FHT
resizing
; Please test it in real OS2 environment
* contrib/gtqtc/gtqtc.h
* contrib/gtqtc/gtqtc1.cpp
* renamed few variables to avoid -Wshadow warnings
* contrib/hbfimage/fi_wrp.c
* removed unnecessary ;
* contrib/hbodbc/odbc.c
! replaced '\0' with NULL in pointer initialization
* src/3rd/jpeg/Makefile
% added missing HAVE_UNSIGNED_CHAR build macro. Without it LIBJPEG
uses SHORT to hold unsigned 8 bit integers what increases memory
usage and reduces the speed.
* src/pp/hbpp.c
* minor modification to keep string constant C90 comnpatible
* src/rtl/filesys.c
* simplified DOS to POSIX file attribute translation
+ respect hidden and system file attribute in DOS builds of hb_fsOpenEx()
* src/vm/fm.c
! added yet another missing HB_STACK_TLS_PRELOAD
* *
% remove brandings and homepage [1] from copyright header. Pass 1 - using script.
[1] nobody has access to it anymore AFAIK - and it's also just
a redirect since long
! update url in copyright header
; this should make the diff between 3.4 and 3.2 easier to manage
* config/global.mk
* set HARBOUR=F:100 during build process in OS2 builds - it increases
maximum number of open handles per process from 20 to 100 so it should
be possible to use -j<n> switch during Harbour compilation - please
test, i.e.:
os2-make -j3
* contrib/xhb/hbcompat.ch
! fixed pp rules for Str(,,, <lStrip> ) translation
* include/hbapifs.h
* src/rtl/filesys.c
+ added new C function:
HB_FHANDLE hb_fsOpenEx( const char * pszFileName, HB_FATTR nAttr,
HB_USHORT uiFlags );
* minor cleanup
* include/hbmemory.ch
* src/vm/fm.c
+ added two new actions for Memory() function:
HB_MEM_STATISTICS - return non 0 value if FM statistic is enabled
HB_MEM_ISLIMIT - return non 0 value if used memory limit is
supported
+ added new PRG function:
__fm_allocLimit( [ <nNewLimit> ] ) -> <nPrevLimit>
It allows to set limit for maximum used memory in some memory
managers, -1 means no limit.
Now it works with DLMALLOC and in ST mode it's limit for memory
allocated from system for whole application. In MT mode when
HB_FM_DLMT_ALLOC (default for DLMALLOC) is enabled it's limit for
single MSPACE (Harbour allocates 16 MSPACES and balance them
between threads to improve scalability) so it's rather per thread
limit. If memory statistic module is enabled during Harbour build
process then this limit works with any memory manager and defines
limit for total memory allocated by Harbour application regardles
of MT or ST mode.
* src/vm/hvm.c
* use HB_MEM_STATISTICS instead of HB_MEM_USEDMAX to check if memory
statistic module is enabled
* src/vm/hvm.c
* allow to use == operator for codeblocks
* utils/hbtest/rt_hvm.prg
* updated for == operator which accepts codeblocks
* utils/hbtest/rt_main.ch
* do not define __CLIPPER__ macro when ITK Clip compiler is used
* contrib/hbbz2/errint.c
* added HB_EXPORT to bz_internal_error() declaration
* src/vm/hvm.c
+ allow to compare symbol items with == operator, i.e.:
? @QOut() == &( "@QOut()" )
* include/hbapiitm.h
* src/vm/itemapi.c
+ added new C function:
hb_itemCompare( PHB_ITEM pItem1, PHB_ITEM pItem2,
HB_BOOL bForceExact, int * piResult );
For compatible types it compares pItem1 with pItem2 setting piResult
to -1, 0 or 1 if pItem1 is <, == or > then pItem2 and returns true
otherwise it returns false and does not touch piResult.
+ allow to compare symbol items in hb_itemEqual()
* include/harbour.hbx
* src/vm/hvm.c
+ added new PRG function:
__RecoverErrorBlock() -> <bError> | NIL
It extracts previous error block pushed on HVM stack
by BEGIN SEQUENCE WITH ...
* src/rdd/dbf1.c
! fixed detecting VFP tables broken in my previous
modification - thanks to Mario H. Sabado for information
about the problem.
* config/dos/watcom.mk
* config/win/watcom.mk
* config/linux/watcom.mk
! enabled -bd OpenWatcom switch for code compiled as part of dynamic
libraries
* use CauseWay as default DOS extender for dynamic DOS builds
* config/dyn.mk
* config/lib.mk
* src/Makefile
! fixed list of libraries used to create Harbour DLL in DOS builds
* config/bin.mk
* src/vm/Makefile
* create hbmainstd library for OpenWatcom DOS shared builds
* src/vm/maindllh.c
+ added DLL entry function for CauseWay DLLs in OpenWatcom builds
* src/vm/main.c
* include maindllh.c in OpenWatcom DOS shared builds
* utils/hbmk2/hbmk2.prg
+ added support for dynamic binaries in DOS OpenWatcom builds.
Support for CauseWay DLLs is still broken in current OpenWatcom
builds so it will have to wait for the fix to be usable.
* src/vm/dynlibhb.c
+ added support for dynamic libraries to OpenWatcom DOS builds.
Now it's enabled only in harbour.dll due to problems with
CW and current OW.
* src/3rd/zlib/zconf.h
* src/3rd/zlib/zlib.dif
* use OS2 patch for exporting ZLIB symbols also in DOS builds
* src/rdd/dbf1.c
* pacified warning
* include/hbstack.h
* src/vm/estack.c
+ added new C function:
HB_ISIZ hb_stackBaseSymbolOffset( PHB_SYMB pSymbol );
it returns offset of pSymbol function frame on HVM stack or -1
if function/procedure pointed by pSymbol is not called.
* contrib/hbnetio/readme.txt
* contrib/hbnetio/netiosrv.c
* contrib/hbnetio/hbnetio.hbx
+ added new server side function:
netio_ServedConnection() -> <pConnectionSocket>
It returns connection served by netio_Server() in current thread and
can be used inside executed RPC functions to access <pConnectionSocket>
* include/harbour.hbx
* src/rtl/dateshb.c
+ added two new PRG functions:
hb_SecToT( <nSeconds> ) -> <tTime>
hb_MSecToT( <nMilliseconds> ) -> <tTime>
* include/hbset.h
* src/vm/set.c
+ added new C function:
int hb_setUpdateEpoch( int iYear );
if necessary it converts year respecting _SET_EPOCH settings
* src/rtl/dates.c
* use hb_setUpdateEpoch() instead of local code
* src/rdd/dbf1.c
! respect _SET_EPOCH decoding DBF update time from DBF HEADER
(LUpdate() function).
It's not Cl*pper compatible anyhow looks that most of tools
stores only last two digits from the year (just like in the
DBF documentation) and -1900 trick is not used so it helps
to decode correct update date.
* store 'year % 100' instead of 'year - 1900' in DBF header
when DBF table with VFP signature is used (DB_DBF_VFP).
* src/rdd/dbf1.c
+ reserve space for DBC name in newly created DBF files with VFP DBF
signature, Thanks to Nikos Christophi, Jeff Stone and other who
reported this problem.
+ added RTE when size of DBF header is exceed by "_NullFlags" field
* src/vm/thread.c
! typo in comment
* include/hbapigt.h
* src/rtl/inkeyapi.c
+ added new C function:
int hb_inkeyKeyExt( int iKey );
It extracts function/edit key code value HB_KX_* from Harbour extended
key code. If passed keycode is not valid extended keycode which points
to function/edit key then 0 is returned.
* include/harbour.hbx
* src/rtl/inkey.c
+ added new PRG function hb_keyExt() - it's wrapper to hb_inkeyKeyExt()
* include/harbour.hbx
* src/rtl/mlcfunc.c
! restored original Cl*pper behavior in memo line functions,
test shows that inside MemoEdit() Cl*pper internally uses a little
bit different rules to format text then in memo line functions,
It's Cl*pper bug but hard to fix because any differences to Cl*pper
in this area cause that someone reports them as bug. So I decided
to leave original Cl*pper behavior for compatibility and add new
function which can be used in MemoEdit() and related code.
+ added new memo line function:
hb_MLEval( <cString>, <bCode>, [ <nLineLength>=79 ],
[ <nTabSize>=4 ], [ <lWrap>=.T. ],
[ <nPos> ], [ @<nRow> ], [ @<nCol> ] ) -> <nLines>
it formats texts using rules like in Cl*pper MemoEdit() not Cl*pper
memo lines functions. For each processed line <bCode> is executed
with 2 parameters: <cLine> and <lSoftCR>. It recognizes and accepts
any EOLs just like hb_ATokens( <cText>, .T. ) and strips or converts
soft CRs depending on word wrap mode.
* src/rtl/teditor.prg
* Use hb_MLEval() to format text
% minor optimization
* src/rtl/memoedit.prg
* enabled Harbour extended key codes
* small optimization
! fixed ME_BOTTOMRIGHT processing
! fixed initial line size calculation
* src/vm/thread.c
* removed dymmy code
* src/vm/hvm.c
! delay locking of main HVM thread stack until HVM is initialized,
it fixes issue #107
! protect access to HB_STACK_TLS_PRELOAD by s_fHVMActive in
hb_vmLockForce()
* include/hbapicls.h
* src/vm/classes.c
* minor modification in variable name
* src/rtl/filebuf.c
* src/rtl/fscopy.c
* src/rtl/fslink.c
* src/rtl/vfile.c
* cleaned FError() setting. It should not be changed by C file IO
functions because they can be activated from RDD code so accessing
work area field could change the FError() result, i.e. this code
does not work when FError() is changed by C level IO operations:
AnyFunc( FSeek( hFile, 0, FS_END ), field->MEMOVAL, FError() )
Even simple field access may cause file IO operations in RDD code
when relations are set because records in child area are repositioned
only when necessary (when user tries to make some operations on this
area). In the past RDD code was full of constructions like:
uiSaveError = hb_fsFError();
// make some IO operations
hb_fsSetFerror( uiSaveError );
to eliminate such overhead I separated PRG level FError() from
C level IOError(). Please keep it.
* ChangeLog.txt
! few typos
* include/hbvmpub.h
* src/vm/hvm.c
+ added support for hb_vmAtInit(), hb_vmAtExit() and hb_vmAtQuit()
functions for dynamically loaded libraries
This modification should also fix issue #33
* src/vm/estack.c
! protect access to HB_STACK_TLS_PRELOAD by hb_stack_ready() macro,
it's necessary for MT builds which do not use direct TLS access
but need TLS key initialization, i.e. OS2
* src/vm/hvm.c
! protect access to HB_STACK_TLS_PRELOAD by s_fHVMActive
* include/hbdefs.h
* changed reference 32-bit reference counters in Win64 builds to 64-bit
ones. It resolves problems with data alignment in memory allocated
by hb_xgrab().
* src/vm/asort.c
* removed hack for MinGW64 build from new array sort code,
after switching to 64bit reference counters it should not
be longer necessary because allocated that has 64bit alignment
* include/hbapifs.h
* src/rtl/filebuf.c
* added new C function hb_fileLoadData() - it can be used to load data
from already opened regular files and streams
* include/harbour.hbx
* src/rtl/vfile.c
+ added new PRG function:
hb_vfLoad( <cFileName>, [ <nMaxSize> ] ) -> <cFileBody> | NIL
* src/vm/runner.c
* use hb_fileLoadData()
* contrib/hbwin/win_bmp.c
* use hb_fileLoad()
* contrib/hbtcpio/tcpio.c
* contrib/hbtcpio/hbtcpio.hbx
+ added new PRG function
hb_vfFromSocket( <pSocket> ) -> <pFile>
It converts socket created by socket open into TCPIP virtual file
which works just like files created by hb_vfOpen( "tcp:...", ... )
* include/hbapifs.h
* src/rtl/filesys.c
+ added new C function
HB_SIZE hb_fsPipeWrite( HB_FHANDLE hPipe, const void * buffer,
HB_SIZE nSize, HB_MAXINT nTimeOut );
+ implemented hb_fsPipeUnblock() in OS2 builds
+ set IO error in hb_fsPipeCreate(), hb_fsIsPipeOrSock() and
hb_fsPipeUnblock()
* return ( HB_SIZE ) -1 from hb_fsPipeRead() if end of stream is
reached or read error appears on non empty buffer. This modification
also change the behavior of PRG hb_PRead() function so now result -1
can be used to detect end of stream.
* src/rtl/philes.c
* include/harbour.hbx
+ added new PRG function:
hb_PWrite( <nPipeHandle>, <cBuffer>, [<nBytes>], [<nTimeOut>] )
-> <nBytesWritten>
% removed unnecessary assignment
* include/hbapiitm.h
+ added HVM internal macro hb_itemRawSwap()
* src/vm/asort.c
% use hb_itemRawSwap()
* in old code use comparison which prefers < and > operators
instead of <= and >=
* src/vm/asort.c
* restored original loop and added different workaround which for
the problem in 64 bit MinGW we exploited. It's dummy function call
inside a loop so probably link time optimization (LTO is enabled
by -O4 in the newest GCC versions) ignore it and the problem will
reappear. Anyhow it only temporary solution which should give time
to create self contain example and report the problem to MinGW
authors - it's highly possible that also other code exploited this
problem so be careful with MinGW64.
% fixed index range checking to eliminate unnecessary index verification
* src/vm/asort.c
! fixed missing nStart checking during array items reordering
* utils/hbtest/rt_array.prg
* updated Harbour specific results - I thought that someone
invested time to exactly replicate CA-Cl*pper ASort() sorting
but in fact Harbour specific resutls where hardcoded inside
HBTEST so previous code was not Cl*pper compatible in the
same way as current one.
