* 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()
* src/rtl/hbcom.c
* src/rtl/gtcrs/gtcrs.c
* src/rtl/gtpca/gtpca.c
* src/rtl/gtstd/gtstd.c
* src/rtl/gttrm/gttrm.c
* added workaround for bug in some Linux kernels (i.e.
3.13.0-64-generic Ubuntu) in which select() unconditionally
accepts terminal device for reading if c_cc[ VMIN ] = 0
It's very serious problem, i.e. in GTTRM no input is possible
in such kernels without this workaround.
* src/rtl/dirscan.prg
* modified hb_DirScan() and hb_DirRemoveAll() to work with Harbour
File IO API (hb_vf*() functions)
+ added new PRG function:
hb_FileDelete( <cFileMask> [, <cAttr> ] ) -> <lResult>
this function removes files which match given <cFileMask>
(it may contain path) and returns .T. if at least one file
was deleted. Optional <cAttr> parameter can be used to include
system ("S") and hidden ("H") files. If <cAttr> contains "R"
letter then before deleting READONLY attribute is removed from
files (it's necessary to remove files with READONLY attribute
in systems like DOS, MS-Windows or OS2). This function uses
Harbour File IO API (hb_vf*() functions)
* remove READONLY attribute from files in hb_DirRemoveAll()
* include/harbour.hbx
+ added hb_DirRemoveAll() and hb_FileDelete()
* contrib/hbct/files.c
! do not allow to remove empty directories by FileDelete() function
even if user explicitly sets directory attribute - CT3 compatible fix.
! remove READONLY attribute before deleting READONLY files if 2-nd
parameter of FileDelete() contains READONLY bit
* src/rtl/filesys.c
* formatting
* ChangeLog.txt
! typo
* src/common/hbffind.c
+ added support for link attribute (HB_FA_LINK -> "L") passed to
Directory() and other functions using hb_fsFindFirst()/hb_fsFindNext()
When this attribute is passed then softlinks are presented "as is"
without following target.
* src/rtl/dirscan.prg
+ added new PRG function hb_DirRemoveAll( <cDir> ) -> <lResult>
BE CAREFUL! This function removed recursively whole directories
with it's body so hb_DirRemoveAll( hb_ps() ) can remove all
files and directories from your disk.
* src/rtl/hbproces.c
! added protection against writing more then 4096 bytes to unblocked
pipe in MS-Windows builds of hb_fsProcessRun()
! close write side of child stdin handle if all data is sent in OS2
builds of hb_fsProcessRun()
* src/rtl/hbproces.c
* In OS2 builds use fGentle parameter of hb_fsProcessClose() to
control if kill signal should be send only to child process or
also to its all descendant processes.
* src/rtl/fscopy.c
* pacified BCC warning
* src/rtl/hbproces.c
* added HB_PROCESS_USEFILES macro which enable child process
standard IO handles redirection to files instead of pipes
in hb_fsProcessRun() function. It's enabled by default in
DOS and WinCE builds
* enable HB_PROCESS_USEFILES for tests in OS2 builds
* 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/rtl/gttrm/gttrm.c
- removed unused macros and definitions
* formatting and minor cleanup
* src/rtl/gtcrs/gtcrs.h
* src/rtl/gtcrs/gtcrs.c
+ added support for extended key codes, GTCRS does not support
unicode so it's not fully functional just like in GTTRM
* src/rtl/filesys.c
* clear ERROR_BROKEN_PIPE (109) in MS-Windows and OS2 builds reported
when pear closes the write end of the pipe.
Now hb_PRead() returns -1 with FError() = 0 in such case in all builds.
* src/rtl/filesys.c
+ added code for anonymous pipes emulation by named pipes,
now nonblocking access to pipe streams in OS2 builds should work
* src/rtl/hbproces.c
* pacified warnings in OS2 builds
* 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
* src/rtl/filesys.c
* implemented hb_fsSetDevMode() only for DOS builds,
all other platforms always use O_BINARY mode
* utils/hbmk2/hbmk2.prg
! fixed EOL handling in stdout and stderr output
* src/nortl/nortl.c
+ added hb_fsSetError()
* include/hbapifs.h
* src/common/hbfsapi.c
+ added new C function for OS2 builds: hb_fsOS2QueryPathInfo()
Unlike DosQueryPathInfo() it uses dir entries so it does not need to
open the file what can create file sharing violation problem.
+ added new C function for OS2 builds: hb_isWSeB()
It returns true if OS2 version supports long file API
(IBM OS/2 Warp Server for e-Business)
+ added new C functions for OS2 builds:
hb_fsOS2DosOpenL(), hb_fsOS2DosSetFileLocksL(),
hb_fsOS2DosSetFilePtrL(), hb_fsOS2DosSetFileSizeL()
These are dynamically initialized wrappers to OS2 long file API
functions if such functions exists in given OS2 version.
* src/common/hbfsapi.c
* use hb_fsQueryPathInfo() in OS2 builds of hb_fsNameExists(),
hb_fsFileExists() and hb_fsDirExists()
* src/common/hbffind.c
+ added support for long files (longer then 4GB) in OS2 builds.
It also detects OS2 version to check if long file API is supported.
! fixed file error setting in OS2 builds
* src/rtl/filesys.c
* use hb_fsQueryPathInfo() in OS2 builds of hb_fsGetFileTime() and
hb_fsGetAttr()
* use hb_fsOS2*() functions for long files
* src/rtl/fssize.c
* use hb_fsQueryPathInfo() in OS2 builds of hb_fsFSize()
; please make real life tests in OS2
* src/rtl/gtos2/gtos2.c
+ added support for HB_GTI_CODEPAGE
! removed code which sets CP437 at application start
Now the default codepage is the same set by system.
If someone wants to use CP437 regardless of default OS2 settings
then he should add to his code:
hb_gtInfo( HB_GTI_CODEPAGE, 437 )
* src/rtl/hbproces.c
* use native OS2 system functions in hb_fsProcessOpen(),
hb_fsProcessValue(), hb_fsProcessClose() - please test it
; TODO: change hb_fsProcessRun() to use real assynchrous execution in
OS2 builds
* src/rtl/disksphb.c
* src/rtl/filesys.c
* use native system functions for disk and file IO operations in OS2
builds.
+ added support for long files (> 2GB) in OS2 builds
! fixed file IO error setting in OS2 builds
; For above modifications I had to change nearly all file IO functions
in OS2 build. I would like to OS2 users to make tests.
In next step I'll update hb_process*() functions.
* 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()
* include/hbapifs.h
* src/rtl/filebufd.c
* moved HB_FILE_ERR_UNSUPPORTED macro to header file
* include/hbapifs.h
* src/rtl/filebuf.c
* added new C function hb_fileLoad() - it can be used to load data
from regular files and streams
* src/rtl/memofile.c
* modified [hb_]Memo{Read|Writ}() to work with streams, i.e. now
this code works:
REQUEST HB_PIPEIO
cDir := hb_memoRead( "|ls -la" )
? upper( cDir )
or:
cData := hb_memoRead( "|xz -c data.gz" )
? hb_memoWrit( "data2.txt", cData )
? hb_memoWrit( "|sh -c 'xz -9 -e > data2.xz'", cData )
* src/rtl/fscopy.c
* removed not used in current code unix header files
* ChangeLog.txt
! fixed typo in ChangeLog entry. I used hb_socketNew() instead of
hb_socketSetFilter()
* include/hbapifs.h
* src/rtl/filebuf.c
+ added helper C function:
HB_SIZE hb_fileResult( HB_SIZE nSize );
It converts ( HB_SIZE ) FS_ERROR to 0 so it can be used to wrap
hb_fileRead()/hb_fileWrite() to force previous results.
* src/rdd/dbffpt/dbffpt1.c
* reduced variable scope
* src/rtl/copyfile.c
! fixed typo in last commit - thanks to Viktor.
* src/rtl/hbsockhb.c
+ added new PRG function:
hb_socketAutoShutdown( hSocket, [ lNewSetting ] ) --> lPrevSetting
it allows to enable/disable automatic shutdown when connected socket
is closed.
* contrib/hbpipeio/pipeio.c
* removed trailing space
* contrib/hbpipeio/tests/test.prg
* cleaned the code and messages
* contrib/hbtcpio/tcpio.c
* contrib/hbpipeio/pipeio.c
* modified hb_fileWrite() to return 0 in case of timeout or unblocking
write and -1 on other errors.
* contrib/hbcomio/comio.c
* modified hb_fileRead() to return 0 in case of timeout or unblocking
read and -1 on other errors.
* src/rtl/filesys.c
* return -1 instead of 0 from hb_fsPipeWrite() in MS-Windows and OS2
builds if PIPE state cannot be read
* ChangeLog.txt
! c&p typo in previous ChangeLog entry
* include/hbsocket.h
* src/rtl/hbsockhb.c
+ added socket filters to standard socket API.
At C level hb_sockex*() function with PHB_SOCKEX handler can be
used to operate on socket filters. At PRG level standard hb_socket*()
functions can be used.
The following things has been changed in PRG hb_socket*() functions:
hb_socketErorrString() can accept <pSocket> as 1-st or 2-nd parameter
and redirect call to socket filter errorStr() method.
hb_socketClose() executes automatically shutdown() for connected
sockets - it is important in windows only where without explicit
call to shutdown() before close transmitted data can be lost.
hb_socketSend() and hb_socketRecv() can be redirected to filter
streams if filter set such redirection. If filter does not redirect
them then they operate on raw sockets. If hb_socketSend() is
redirected then sent data is flushed automatically.
The following new PRG functions has been added:
Add/replace socket filter:
hb_socketSetFilter( <pSocket>, [<cFilterName>], [<hParams>] )
-> <pSocket> | NIL
<cFilterName> is filter name, It's possible to set many filters
in single hb_socketSetFilter() call separating filter names
with "|" character, i.e.:
pSock := hb_socketSetFilter( pSock, "ZSOCK|BFSOCK", hParams )
<hParams> is hash array with initialization parameters used by
given socket filter. The core implementation recognize the
following settings:
"readahead" - numeric value with size of read ahead buffer
"flush" - numeric value with auto flush parameter (for more
information look at hb_socketAutoFlush() below)
"redir" - logical value which can be use to enable/disable
hb_socketSend() and hb_socketRecv() redirection
to filter stream.
Return filter name used by socket:
hb_socketGetFilter( <pSocket> ) -> <cFilterName>
Read from socket stream:
hb_socketRead( <pSocket>, @<cData>, [<nLen> = Len( cData )],
[<nTimeout> = FOREVER] ) -> <nRead>
this function is similar to hb_socketRecv() but is always
redirected to socket stream filters.
Write to socket stream:
hb_socketWrite( <pSocket>, <cData>, [<nLen> = Len( cData )],
[<nTimeout> = FOREVER] ) -> <nWritten>
this function is similar to hb_socketSend() but it is always
redirected to socket stream filters. Written data is not flushed
by default and it should be flushed explicitly by hb_socketFlush().
Automatic flushing can be enabled by hb_socketAutoFlush() function.
Flush data written to socket:
hb_socketFlush( <pSocket>, [<nTimeout> = FOREVER], [<lSync>] )
-> <nNotFlushed>
<lSync> parameter is logical value which can be used to force
special synchronization method in some filters. Usually users
do not have to use it in normal code.
Enable/disable automatic flushing of written data.
hb_socketAutoFlush( <pSocket>, [ <nTimeout> ] ) -> <nTimeout>
<nTimeout> is timeout for automatic flush operation on written
data in milliseconds. <nTimeout> = -1 means wait forever and
<nTimeout> = 0 disables auto flush.
automatic flushing can help in adopting existing code anyhow it
may strongly reduce the performance in some filters, i.e.
compression filters like ZSOCK have to add special data to the
stream after each flush operation so it's suggested to call
flush explicitly when we want to force delivering written data
to the peer.
* include/hbznet.h
* src/rtl/hbznet.c
+ added ZNET socket filter - compressed and encrypted streams are
compatible with hb_znet*() streams. The old hb_znet*() interface
is obsolete for pure socket communication and if not used as
hb_inet*() filter then should be replaced by hb_sockex*() in
user programs.
ZNET socket filter can be created by new PRG functions:
hb_socketNewZNet( <pSocket>, [<cPass>], [<nCompressionLevel>], ;
[<nStrategy>] ) -> <pSocket> | NIL
or by standard socket API with "ZNET" as filter name.
ZNET filter recognize the following settings in initialization
hash array:
"key" or "pass" - string with encryption password
"zlib" - numeric compression level (HB_ZLIB_COMPRESSION_*)
"zs" - numeric ZLIB compression strategy (HB_ZLIB_STRATEGY_*)
ZNET filter always disables any other filters and operates on raw
socket.
Please remember that it's optional module. If programmer does not
use hb_socketNewZNet() explicitly and prefers using hb_socketNew()
then he should force linking this module by REQUEST hb_socketNewZNet
+ added fSync parameter to hb_znetFlush()
[INCOMPATIBLE]
* src/rtl/hbinet.c
* call flush filter function before socket is closed
* src/rtl/Makefile
+ src/rtl/hbzsock.c
+ added ZSOCK socket filter - ZLIB and GZIP compression for socket
streams.
ZSOCK socket filter can be created by new PRG functions:
hb_socketNewZSock( <pSocket>, [<hParams>] ) -> <pSocket> | NIL
or by standard socket API with "ZSOCK" as filter name.
Programmers using hb_socketNew() can force linking this module by
REQUEST hb_socketNewZSock
ZSOCK filter can be used with other filters.
ZSOCK filter recognize the following settings in initialization
hash array:
"zlib" - numeric compression level (HB_ZLIB_COMPRESSION_*)
"zs" - numeric ZLIB compression strategy (HB_ZLIB_STRATEGY_*)
"zin" - logical value which allow to enable/disable ZLIB
decompression on input stream (default)
"gzin" - logical value which allow to enable/disable GZIP
decompression on input stream - it's possible to
enable both ZLIB and GZIP decompression together
so both streams can be decompress
"zout" - logical value which allow to enable/disable ZLIB
compression on output stream (default)
"gzout" - logical value which allow to enable/disable GZIP
compression on output stream - if both "zout" and
"gzout" are enabled GZIP compression is used.
* src/rtl/Makefile
+ src/rtl/hbbfsock.c
+ added BFSOCK socket filter - BlowFish input and output stream
encryption in CTR mode.
BFSOCK socket filter can be created by new PRG functions:
hb_socketNewBFSock( <pSocket>, [<hParams>] ) -> <pSocket> | NIL
or by standard socket API with "BFSOCK" as filter name.
Programmers using hb_socketNew() can force linking this module by
REQUEST hb_socketNewBFSock
BFSOCK filter can be used with other filters, i.e. with ZSOCK.
Please only remember that good encryption algorithms have to
generate data which cannot be compressed so using "BFSOCK|ZSOCK"
only wastes resources and correct filter order is "ZSOCK|BFSOCK".
BFSOCK filter recognize the following settings in initialization
hash array:
"key" or "pass" - string with encryption password
"iv" - string with initialization vector for CTR mode
* contrib/hbssl/hbssl.ch
* contrib/hbssl/hbssl.h
* contrib/hbssl/hbssl.hbm
* contrib/hbssl/hbssl.hbx
* contrib/hbssl/ssl.c
* contrib/hbssl/ssl_inet.c
+ contrib/hbssl/ssl_sock.c
+ added SSL socket filter
SSL socket filter can be created by new PRG functions:
hb_socketNewSSL_connect( <pSocket>, <pSSL> [, <nTimeout> ] )
-> <pSocketSSL> | NIL
hb_socketNewSSL_accept( <pSocket>, <pSSL> [, <nTimeout> ] )
-> <pSocketSSL> | NIL
or by standard socket API with "SSL" as filter name.
Programmers using hb_socketNew() can force linking this module by
REQUEST hb_socketNewSSL_connect
or
REQUEST hb_socketNewSSL_accept
SSL filter always disables any other filters and operates on raw
socket.
SSL filter recognize the following settings in initialization hash
array:
"ctx" or "key" - pointer SSL item <pSSL>
"timeout" - timeout (numeric)
"client" - logical value indicating client mode (SSL_connect())
"server" - logical value indicating server mode (SSL_accept())
* contrib/hbssl/tests/inetssl.prg
! cleaned typo in local function name
* contrib/hbnetio/netiocli.c
* contrib/hbnetio/netiosrv.c
* use new Harbour extended socket API (hb_sockex*()) instead of
raw sockets and hb_znet*()
* contrib/hbtcpio/tcpio.c
* use new Harbour extended socket API (hb_sockex*()) instead of
raw sockets
+ implemented hb_fileFlush()
* 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.
* src/rtl/gtxwc/gtxwc.c
* src/rtl/mlcfunc.c
* pacified new GCC C++ warnings
* src/rtl/hbsocket.c
* do not execute select() if HVM exception was generated when
socket handlers are extracted but return an error immediately
* src/vm/asort.c
% added new code for ASort()
Warning: new sorting algorithm is stable (does not change the
order of equal items). It means is not strictly Cl*pper
compatible and it can be seen some dummy ASort() tests
in HBTEST results.
2015-07-31 14:04 UTC+0200 Viktor Szakats (vszakats users.noreply.github.com)
* include/hbapifs.h
* src/rtl/filebuf.c
* src/rtl/vfile.c
+ add C-level hb_fileSizeGet() function
* update HB_VFSIZE() to use hb_fileSizeGet() when passed a filename
It means now HB_VFSIZE() works for non-virtual filenames
just like HB_FSIZE() did.
