* src/rtl/dirdrive.c
+ Added HB_DIRCREATE(), which is same as MAKEDIR() C5.3 function,
but it's always available, not just when HB_COMPAT_C53 is enabled.
+ Added HB_DIRDELETE(), which is same as DIRREMOVE() C5.3 function,
but it's always available, not just when HB_COMPAT_C53 is enabled.
* contrib/hbnetio/tests/netiotst.prg
* contrib/hbnetio/tests/netiot03.prg
* contrib/hbfimage/tests/fitest.prg
* contrib/hbide/idesaveload.prg
* utils/hbmk2/hbmk2.prg
* utils/hbrun/hbrun.prg
* tests/wcecon.prg
* tests/flink.prg
* bin/hb3rdpat.hbs
* examples/hbdoc2/gentpl.prg
* Replaced MakeDir() with universally available hb_DirCreate().
; TODO: same in config/postinst.hbs (delayed to avoid a million
false bug reports)
* src/rtl/Makefile
+ src/rtl/hbfilehi.prg
+ Added generic file/dir/path handling functions.
They are taken from hbmk2 code, added parameter checking,
did two minor change to DirBuild() to be generic,
renamed to have HB_ prefix.
Please note the parameter list and any details are
subject to change yet, IOW these APIs are not yet finalized:
hb_cwd() -> <cPath>
hb_PathNormalize( <cPath> ) -> <cPath>
hb_PathMakeAbsolute( <cPathRel>, <cPathAbs> ) -> <cPath>
hb_PathMakeRelative( <cPathBase>, <cPathTarget>, <lForceRelative> ) -> <cPath>
hb_DirAddPathSep( <cDir> ) -> <cDir>
hb_DirDelPathSep( <cDir> ) -> <cDir>
hb_DirBuild( <cDir> ) -> <lSuccess>
hb_DirUnbuild( <cDir> ) -> <lSuccess>
hb_FNameDirGet( <cFileName> ) -> <cDir>
hb_FNameNameGet( <cFileName> ) -> <cName>
hb_FNameNameExtGet( <cFileName> ) -> <cNameExt>
hb_FNameExtGet( <cFileName> ) -> <cExt>
hb_FNameExtDef( <cFileName>, <cDefExt> ) -> <cFileName>
hb_FNameExtSet( <cFileName>, <cExt> ) -> <cFileName>
hb_FNameDirExtSet( <cFileName>, <cDirNew>, <cExtNew> ) -> <cFileName>
; TODO: Review, clean interface.
; TODO: Use these functions in .hbs files, utils, hbide.
* include/harbour.hbx
* Updated.
* utils/hbmk2/hbmk2.prg
* One internal function renamed.
* src/rtl/Makefile
- src/rtl/base64.prg
+ src/rtl/base64d.c
+ Replaced HB_BASE64DECODE() with an optimized C implementation.
The core code is borrowed from libb64 project
(http://sourceforge.net/projects/libb64) and has been places
to public domain by author Chris Venter.
I've reformatted the code and made some minor modifications.
I had used the Harbour license for the whole file.
This code is 50-100 times faster than old Harbour code doing
the same job. Please review it, especially the buffer size
calculation.
BTW the C code uses a unique trick called 'Duff's device'.
+ harbour/src/rtl/hbjson.c
+ added JavaScript Object Notation (JSON) encoder, decoder functions
; C level functions:
char * hb_jsonEncode( PHB_ITEM pValue, HB_SIZE * pnLen );
pValue - value to encode;
pnLen - if pnLen is not NULL, length of returned buffer is
stored to *pnLen;
returns pointer to encoded JSON buffer. buffer must be fried
by the caller.
HB_SIZE hb_jsonDecode( const char * szSource, PHB_ITEM pValue );
szSource - JSON source;
pValue - item to store decoded value. Item value is
undetermined in case of error;
returns number of bytes decoded from the buffer. This allows
to use the remaining part of the buffer for some other
purposes. Returns 0 on error.
Harbour level functions:
hb_jsonDecode( cJSON, @xValue ) --> nLengthDecoded
hb_jsonEncode( xValue ) --> cJSON
Note:
- Harbour types unsupported by JSON (date, timestamp, etc.) are
encoded as null values;
- strings are encoded in UTF-8;
- JSON encode functions are safe for recursive arrays and hashes.
Recursive part of array or hash will be stored as null. JSON
encoder still allows to use same structure in the leaves, in
this case content will be duplicate.
I.e.:
xI := {1, NIL}
xI[2] := xI
? hb_jsonEncode( xI ) // [1,null]
but:
xI := {1, .T.}
xI := {2, xI, xI}
? hb_jsonEncode( xI ) // [2,[1,true],[1,true]]
* contrib/make.hbs
+ Generalized the way dynamic lib dependencies are found
to make it work on *nix systems, too.
* include/harbour.hbx
+ src/rtl/libnamec.c
* src/rtl/libname.prg
* src/rtl/Makefile
+ Added HB_LIBEXT() and HB_LIBPREFIX() functions.
* utils/hbmk2/hbmk2.prg
+ Added "platform", "compiler", "cpu", "buildname", "dynsuffix"
to --hbinfo returned data.
* INSTALL
+ Documented HB_BUILD_CONTRIB_DYN setting. The default is
'no', and if set to 'yes', the build system will create
dynamic lib versions of the contribs.
* src/Makefile
* utils/hbformat/Makefile
* utils/hbi18n/Makefile
* utils/hbtest/Makefile
* config/global.mk
* config/dos/watcom.mk
* config/dos/djgpp.mk
* config/dyn.mk
* config/rules.mk
* INSTALL
* Renamed HB_BUILD_DLL to HB_BUILD_DYN.
* ChangeLog
* include/harbour.hbx
* src/rtl/Makefile
+ src/rtl/libname.prg
+ tests/libname.prg
+ Added HB_LIBNAME( <cDynLibName>[, <cDynLibDir> ] ) -> <cOSDynLibName>
Serves to convert a raw dynlib name, f.e. "hbct" to OS specific
notation, by adding platform specific extension, lib name prefix and
optional directory
* utils/hbmk2/hbmk2.prg
! Fixed dynlib prefix for symbian
* include/hbextern.ch
* src/rtl/Makefile
+ src/rtl/hbdoc.prg
+ Added __HBDOC_LOADDIR() internal RTL function to read
NFDOC formatted documentation source directories into
memory. The result will be a list of hashes which contain
the raw content of doc files. The function does format
validation (but not content) and add all agreed-on metadata
to the hashes.
It's a start and I hope others can join to create new
layers of function for below functionality:
- doc content validation
- pre-formatting
- output generators
The goal is to avoid local solutions and standardize
the format.
+ tests/hbdoctst.prg
+ Added small test app which will find and parse all
documentation inside the Harbour source tree and
save them in .hbd (serialized doc hash) format.
Such .hdb files can be generated by the build process
in the future, so tools like HBIDE can use the content
to display intellisense or built-in help.
; TODO: HBQT doc/html should be removed as it's
hbdoc tools' job to convert HBDOC source
format to html or whatever else format needed.
* harbour.spec
* debian/rules
- doc/man
+ src/pp/hbpp.1
+ src/main/harbour.1
+ utils/hbmk2/hbmk2.1
+ utils/hbtest/hbtest.1
+ utils/hbrun/hbrun.1
* Moved man doc files to their own component dirs.
* doc/Makefile
* doc/en/Makefile
+ doc/clipper.txt
- doc/en/clipper.txt
* Moved non-NFDOC formatted doc file to doc dir.
* doc/en/Makefile
! Fixed missing 1stread.txt
* doc/en/hb_apigt.txt
* doc/en/hbinet.txt
* doc/en/tbrowse.txt
* doc/en/hb_apifs.txt
! Fixed NFDOC non-standard entries.
* bin/hbxpatch.hbs
! Fixed detecting dirs. (could cause errors only in rare cases)
* contrib/hbdoc2/tmplates.prg
* Formatting.
* src/common/hbver.c
* src/rtl/Makefile [t]
* src/rdd/Makefile [t]
* include/hbdefs.h
* include/hbsetup.h
+ config/bsd/pcc.mk
* config/rules.mk [t]
+ First step to add support for pcc (Portable C Compiler) on bsd.
; Patch by Tamas Tevesz.
; TODO: ? I guess the -gc3 and -l disabler hacks are temporary.
(in files marked with [t])
* include/hbextern.ch
* src/rtl/Makefile
+ src/rtl/memvarhb.prg
+ Added:
HB_MVSAVE( <cFileName>, [<cMask>], [<lIncludeMask>] )
HB_MVRESTORE( <cFileName>, [<lAdditive>], [<cMask>], [<lIncludeMask>] )
They serve the same purpose with same interface
and similar behavior to __MVSAVE()/__MVRESTORE()
(and __MSAVE()/__MRESTORE(), and 'SAVE TO' command),
only that they support longer than 10 char variable names.
It is not possible to add such support in Clipper-style
.mem files, so these functions will create an incompatible
file format with default '.hbv' extension.
; TOFIX: __MVRESTORE() restores variables as PRIVATEs
while HB_MVRESTORE() as PUBLIC.
; NOTE: GETLIST exclusion is solved with simple trick,
as I couldn't find how it's solved in __MVSAVE().
; Consider this experimental solution to the problem.
* contrib/hbmysql/tmysql.prg
% STOD( "" ) -> STOD()
* contrib/hbide/ideprojmanager.prg
! -hbexec -> -hbexe
* src/rtl/typefile.prg
! __TYPEFILE(): Fixed both RTEs to be Clipper compatible.
* harbour/external/minizip/ioapi.h
! fixed FreeBSD and HAIKU builds
* harbour/src/rtl/fserr.c
* added translation for EPERM
* harbour/include/Makefile
+ harbour/include/hbapicom.h
* harbour/src/rtl/Makefile
+ harbour/src/rtl/hbcom.c
+ added Harbour multiplatform serial port C API (hb_com*() functions).
implemented code for *nixes, MS-Windows and OS2. DOS is not supported
yet. Please make test on different platforms.
; few notes about this code:
termios (*nix) based builds:
- reading LSR is not supported
- duration in break signal is fixed (at least 0.25 seconds, and not
more that 0.5 seconds)
- DTR/DSR hardware flow control is not supported
- setting error character is not supported
- discard card is supported only on few platforms
- checking input/output flow state is not supported
- if some platforms do not support some termio extensions then
compile time warnings TODO_* should be generated.
MS-Windows builds:
- discard card is not supported
OS2 builds:
- checking for input flow state is not supported
- discard card is not supported
In all builds not all RS-232 hardware flags/state are available.
OS2 build supports most of them,
This code needs to be extensively tested. I wrote it without any
serious tests. Especially OS2 builds should be verified by OS2
users. I wrote this port using only documentation which OS2 API
available in the Internet.
DOS is not supported yet though it's possible to compile termio code
for *nixes with DJGPP. Anyhow it would be very nice if we can add
our own DOS serial port code. I would like to ask if any of you
has low level serial port code in C for DOS (i.e. used with Clipper)
and can send it to us so we can use it with Harbour. I have my
own serial port library for Clipper but it's written in assembler
and it's much wider then sth what we need so it cannot be easy
adopted for Harbour.
* harbour/contrib/hbct/Makefile
+ harbour/contrib/hbct/ctcom1.c
+ harbour/contrib/hbct/ctcom.ch
+ harbour/contrib/hbct/ctcom2.c
+ added CT3 compatible COM_*() functions.
+ added CT3 compatible XMOBLOCK(), XMOCHECK(), ZEROINSERT() and
ZEROREMOVE() functions.
+ added new CT COM function:
COM_DEVNAME( <nComPort> [, <cNewName> ] ) -> <cPrevName>
which can be used to specify device name for given port in modern
systems which do not give direct hardware access for the process.
; few notes about CT3 COM_*() functions in Harbour.
In COM_OPEN() only 1-st parameter is significant, input/output buffer
sizes and trap mode parameters are ignored. In modern OS-es such
functionality is controlled by OS and usually buffered read/write
is always enabled. The sizes of IO buffers can be used only as
suggested IO buffer size for OS in some systems. It's possible that
these additional parameters will be used in DOS builds in the future.
COM_READ() does not support 2-nd parameter <lNoDelete>. It can be
implemented but because we do not have direct access to OS buffers
then we will have to emulate it ourselves what is rather inefficient.
Anyhow I can implement it if users think it's critical.
In COM_CRC() I fixed few CT3 bugs so it's not bug compatible with CT3,
i.e. it works correctly for 8 bit and smaller polynomials instead of
returning 0 or supports much larger polynomials up to 64bits.
For 16/17 bit polynomials it gives the same results as CT3 so for most
common usage it should be binary compatible with CT3.
In ZEROREMOVE() I fixed decoding some wrong CCITT strings which does
not have trailing 0 so it returns empty string "" for such corrupted
data. If someone needs original CT3 behavior for some broken code then
is can be enabled by HB_CT3_ZEROREMOVE_BUG macro.
I haven't implemented few functions.
COM_GETIO(), COM_SETIO(), COM_GETIRQ(), COM_SETIRQ() are not portable
and can be implemented only for really few platforms, i.e. for DOS
when/if we add support for serial ports to Harbour DOS builds.
Similar functionality for all other platforms gives our new function
COM_DEVNAME().
COM_KEY(), COM_SKEY() and COM_EVENT() are also not implemented.
To make them well we should add support for asynchronous events to HVM
and such extension is still unavailable. Some simulation can be
implemented using idle tasks but this can be done also by final
Harbour users so I decided to not make it.
Please test it - I've never used CT3 COM_*() API with Clipper.
* harbour/contrib/hbct/dummy.c
* removed dummy CT3 serial communication functions
; such files with dummy functions introduces very serious problems for
user code because they do not respect original module granularity
effectively breaking overloading only chosen functions so it breaks
also some real Clipper code.
* harbour/src/rtl/Makefile
* harbour/src/rtl/hbznet.c
+ harbour/src/rtl/hbinetz.c
* moved HB_INETCOMPRESS() function to separate file
* harbour/include/hbexprb.c
! fixed my very bad C&P typo (2010-02-04 19:14 UTC+0100) which caused
that dummy value was left on HVM stack after post decrementation used
as statement. This dummy value could break construciton like FOR EACH,
WITH OBJECT, BEGIN SEQUENCE. Here is self contain example which can
be used to exploit the problem:
proc main()
begin sequence
p()
end sequence
return
proc p()
local nTemp := 0
begin sequence
nTemp--
end sequence
return
Many thanks to Heinz Bergen who reduced his code to similar small
example.
* src/rtl/Makefile
* src/rtl/getsys.prg
+ src/rtl/getsyshb.prg
! Moved HB_GETREADVAR() to separate source file.
! HB_GETREADVAR() fixed to use HB_ASCIIUPPER() like
TGETLIST():GETREADVAR() instead of UPPER().
* src/rtl/tgetlist.prg
% TGETLIST():GETREADVAR() now can use HB_GETREADVAR()
instead of duplicating the code locally.
* INSTALL
+ Added sort of "disclaimer" note to HB_BUILD_IMPLIB option.
* include/hbextern.ch
+ Added missing TBREADER().
* include/hbsetup.ch
* src/common/hbverdsp.c
- Deleted HB_COMPAT_FLAGSHIP, HB_COMPAT_FOXPRO.
* include/hbextern.ch
* src/rtl/seconds.c
+ Added HB_SECONDSCPU()
(native Harbour version of FlagShip specific SECONDSCPU())
* tests/memtst.prg
* tests/speedold.prg
* tests/speedtst.prg
* tests/vidtest.prg
* Changed to use HB_SECONDSCPU().
+ Added translation from SECONDSCPU() to HB_SECONDSCPU() when
built for FlagShip (where applicable).
* include/hbextern.ch
* src/rtl/Makefile
- src/rtl/strpeek.c
- src/rtl/secondfs.c
* contrib/Makefile
+ contrib/hbfship
+ contrib/hbfship/Makefile
+ contrib/hbfship/hbfship.hbc
+ contrib/hbfship/secondfs.c
+ contrib/hbfship/strpeek.c
- Moved FlagShip specific function from core to new hbfship lib.
* Changed SECONDSCPU() to be just a wrapper over core HB_SECONDSCPU().
; INCOMPATIBLE: If you used SECONDSCPU() function, change it to
HB_SECONDSCPU(), or add hbfship to your lib list.
If you used STRPEEK() or STRPOKE() functions,
add hbfship to your lib list.
* utils/hbmk2/examples/contribf.hbc
+ Added hbfship.
* src/rtl/fstemp.c
! Using hb_fsTempDir() in HB_FTEMPCREATEEX() instead of
old solution. This should fix it for some non-*nix
platforms where *nix specific TEMPDIR envvar was used
to determine temp directory.
* contrib/hbwin/Makefile
* include/Makefile
* src/rtl/Makefile
- contrib/hbwin/hbdyn.ch
+ include/hbdyn.ch
- contrib/hbwin/hbdyn.c
+ src/rtl/hbdyn.c
- contrib/hbwin/win_dllc.c
+ src/rtl/hbdynhb.c
- contrib/hbwin/tests/testdll.c
+ tests/testdyn.c
- contrib/hbwin/tests/testdll.prg
+ tests/testdyn.prg
- contrib/hbwin/tests/testdll1.prg
+ tests/testdyn1.prg
- contrib/hbwin/hbdyn.h
* include/hbapi.h
* Moved dynamic library call related functions to
core (RTL).
* Windows specific test code (currently all) guarded
for platform.
* Changed default calling convention to cdecl.
Tests adaptd accordingly.
* Terminology change: 'dll' -> 'dynamic library'
; HB_DYNCALL() is now part of core Harbour, and it's
platform independent. It's platform (ABI) dependent though,
and currently x64 and x86/stdcall/syscall/cdecl
is supported and these were tested (except OS/2 syscall).
* contrib/hbwin/hbwin.h
* contrib/hbwin/win_misc.c
* contrib/hbwin/legacycd.c
* Low level hbwin_getprocaddress() function moved
to legacy file and made static.
* contrib/xpp/Makefile
* contrib/xpp/dllx.c
* contrib/xpp/xpp.hbc
+ Enabled DLL*() functions for all platforms.
- Deleted dependence on hbwin library.
* contrib/xpp/tests/testdll.prg
* Windows specific parts (currently the whole test) guarded
for platform.
* contrib/hbct/Makefile
- contrib/hbct/diskhb.prg
* contrib/hbct/disk.c
* contrib/xhb/Makefile
+ contrib/xhb/diskhb.prg
* GETVOLINFO() moved from hbct to xhb.
* ChangeLog
* Marked some TODO/TOFIX items as DONE.
* src/rtl/Makefile
* src/rtl/cdpapi.c
+ src/rtl/cdpapihb.c
* include/hbapicdp.h
+ Moved Harbour level functions to a separate file.
I had to rename and publish three static functions.
Eventually the low level part should go to hbcommon lib.
; TODO: HB_CDPLIST() is still to be moved, I couldn't find out elegant
low-level API to solve it.
* src/rtl/Makefile
+ src/rtl/strclear.c
* include/hbextern.ch
+ Added HB_STRCLEAR() function to safely clear the content
of a string variable.
Notice that this method by itself can only work if the
string variable has no other references, and generally
you have to code carefully to avoid creating unwanted
copies of the string when passing it around in an app.
So this is no silver bullet, but rather just one component
to solve this problem.
* contrib/hbnetio/utils/netiosrv.prg
+ Added steps to avoid having the password stored in memory,
while the server is running.
* src/rtl/philes.c
% Minor cleanup.
* harbour/src/rtl/gtwin/gtwin.c
+ added support for HB_GTI_ISUNICODE
* harbour/include/hbapifs.h
* added missing 'extern' in some declarations
* harbour/include/hbzlib.ch
+ added HB_ZLIB_STRATEGY_* constants
+ added HB_ZLIB_COMPRESSION_DISABLE
* harbour/include/Makefile
+ harbour/include/hbznet.h
* harbour/include/hbextern.ch
* harbour/src/rtl/Makefile
* harbour/src/rtl/hbinet.c
+ harbour/src/rtl/hbznet.c
+ added support for ZLIB compression in stream sockets.
+ added .prg function:
HB_INETCOMPRESS( <pSocket>, [<nCompressionLevel>], [<nStrategy>] )
which enables ZLIB compression for given HB_INET*() socket.
<pSocket> is a socket created by one of HB_INET*() functions
<nCompressionLevel> is compression factor between 1 (fastest) and
9 (best) (see HB_ZLIB_COMPRESSION_*)
0 (none) disable compression on output data
but decompression is still working.
<nStrategy> is used to tune compression algorithm,
see HB_ZLIB_STRATEGY_*
The compression must be enabled on both connection sides, i.e.
on the server side:
conn := hb_inetAccept( sock )
hb_inetCompress( conn )
and on the client side:
sock := hb_inetConnect( cServer, nPort )
hb_inetCompress( sock )
in the same moment but it's not necessary to enable it at the
beginning of connection. It can be done later, i.e. when both
sides agree to enable connection using some custom protocol.
The compression has effect only on stream connections, i.e.
TCP and it's ignored in datagram connections like UDP.
This function can be executed more then once changing the compression
parameters but it causes that all data in readahead decompression
buffer is discarded. When called with HB_ZLIB_COMPRESSION_DISABLE
as <nCompressionLevel> then support for stream compression is removed
and sockets works again in raw mode.
The compression level and strategy do not have to be the same on both
connection sides. Each side can chose the best settings for data it's
going to send.
This code was written in a way which allows to easy implement
alternative compression methods or other extensions like encryption
in existing HB_INET*() sockets also by non core code. The public C
functions declared in hbznet.h allows to use this extension with raw
harbour sockets two.
* harbour/src/rtl/Makefile
+ harbour/src/rtl/hbbfish.c
* harbour/include/Makefile
+ harbour/include/hbbfish.h
+ added support for BlowFish encryption at C level:
hb_blowfishInit(), hb_blowfishEncrypt(), hb_blowfishDecrypt()
This code implements BlowFish algorithm designed by Bruce Schneier.
The description of BlowFish algorithm can be found at:
http://www.schneier.com/paper-blowfish-fse.html
This code uses for initial s-boxes and p-array values PI hex digits
taken from tables public at:
http://www.schneier.com/blowfish.html
which can be downloaded from:
http://www.schneier.com/code/constants.txt
This should be quite fast implementation (probably much faster
then reference one) but please remember that in BlowFish algorithm
startup initialization of encryption tables is intentionally
designed to be expensive to strongly reduce the efficiency of
brute force attacks so call hb_blowfishInit() once for each
new password and then reuse initialized encryption tables in
hb_blowfishEncrypt() and hb_blowfishDecrypt().
I'll add .prg functions later.
TODO: add some asymmetric key encryption.
- source
+ src
* src/Makefile
* src/dynlib/mt/Makefile
* src/dynlib/Makefile
* INSTALL
* Makefile
* ChangeLog
* harbour.spec
* mpkg_tgz.sh
* Renamed 'source' dir to 'src' to move closer to other FOSS
projects. It's also easier to type, plus it has the side
effect that some cmdlines will be shorter in the
build process.
