* *
* 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/rddads/ads1.c
* src/rdd/workarea.c
* simplified the code and eliminated wrong /* fallthrough */ comments
inside case block reported by Viktor
* 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
* src/compiler/hbgenerr.c
* changed "with object" in last error message to upper cases
* src/rdd/dbf1.c
+ added assigned value to error object ARGS array when data type or
data width error is generated inside PUTVALUE() method
* utils/hbtest/hbtest.prg
* utils/hbtest/rt_array.prg
* utils/hbtest/rt_hvma.prg
* utils/hbtest/rt_main.ch
* utils/hbtest/rt_math.prg
* utils/hbtest/rt_misc.prg
+ added support for alternative results to HBTEST code.
It's enabled by default and can be turned off by -noalt hbtest
parameter.
+ added alternative results for tests which return different then
Cl*pper results but they are expected and not Harbour bugs.
* package/harbour.spec
* contrib/hbbz2/hbbz2.hbp
* package/mpkg_rpm.sh
+ added support for '--with localbz2' RPM build parameter
* include/hbpp.h
* src/pp/ppcore.c
* src/compiler/hbmain.c
* modified hb_pp_inBuffer() parameters. Now it supports optional
szFileName parameter.
Warning: it's incompatible with previous version but I do not think
that anyone used this function outside core code so I decided to not
add new function but simply change the existing one.
[INCOMPATIBLE]
* src/rtl/filebuf.c
! do not resepct SET DEFAULT/PATH in hb_fileExists() when 2nd parameter
is NULL in default file IO drivers. It also changes the behvior of
PRG function hb_vfOpen() which now honors SET DEFAULT/PATH only if
its 2nd parameter is passed by reference - this modification is
compatible with 2016-01-25 18:16 UTC+0100 Viktor Szakats in his fork.
* src/rdd/dbf1.c
* src/rdd/dbfcdx/dbfcdx1.c
* src/rdd/dbfnsx/dbfnsx1.c
* src/rdd/dbfntx/dbfntx1.c
* updated to work correctly with modified hb_fileExists() behavior
; CL5.2 DBFCDX and Six3 CDX/NSX RDDs looking for production indexes
respect SET PATH but Harbour in core DBF* index RDDs is CL5.3/COMIX
compatible and looks for production indexes only in the directory
where DBF file is located and only SIXCDX Harbour RDD is CL5.2/Six3
compatible.
Please also note that it's significant only for situations when
DBF file is located in current directory. If it's open from
SET DEFAULT/PATH directory then corresponding path is added
to DBF file name what effectively disables SET DEFAULT/PATH
processing when production index is open.
* src/rtl/filebuf.c
* src/rtl/filebufd.c
* src/rtl/iousr.c
* contrib/hbmemio/memio.c
; cleaned name of hb_fileCopy() paramter
* include/hbapifs.h
* src/rtl/filebuf.c
* src/harbour.def
+ added new C function:
HB_BOOL hb_fileMove( const char * pszSrcFile,
const char * pszDstFile );
* src/rtl/vfile.c
* include/harbour.hbx
* src/harbour.def
+ added new PRG function:
hb_vfMoveFile( <cFileSrc>, <cFileDst> ) -> <nResult>
; unlike hb_fileRename()/hb_vfRename() new hb_fileMove()/hb_vfMoveFile()
functions allow to move files between different file systems/drives
and also different Harbour file IO drivers. Please note that it means
that file name IO driver prefix is significant in both source and
destination file names unless it's local FS operation. New functions
when both file names point to the same Harbour file IO driver try to
use simple rename operation. If it fails then they try to copy the
file and remove the source one.
* ChangeLog.txt
! typo
* include/hbfloat.h
* contrib/xhb/xhbctbit.c
* src/compiler/hbcmplib.c
* src/compiler/hbcomp.c
* src/compiler/hbusage.c
! pacified new GCC warnings and partially synced with Viktor's fork
* src/compiler/hbusage.c
* added Klas Engwall to credits list
* added Rolf to credits list
* updated April White email address
; please verify it and update if necessary
* include/harbour.hbx
* src/harbour.def
* src/rtl/dateshb.c
+ added new PRG functions which convert time difference to hours,
minutes, seconds and milliseconds. Except milliseconds the results
are floating point numbers with fractional part representing the
rest of conversion:
hb_NToHour( <nTimeDiff> ) => <nHours>
hb_NToMin ( <nTimeDiff> ) => <nMinutes>
hb_NToSec ( <nTimeDiff> ) => <nSeconds>
hb_NToMSec( <nTimeDiff> ) => <nMilliseconds>
+ added new PRG functions which convert timestamp value to hours,
minutes, seconds and milliseconds. Except milliseconds the results
are floating point numbers with fractional part representing the
rest of conversion:
hb_TToHour( <tTimeStamp> ) => <nHours>
hb_TToMin ( <tTimeStamp> ) => <nMinutes>
hb_TToSec ( <tTimeStamp> ) => <nSeconds>
hb_TToMSec( <tTimeStamp> ) => <nMilliseconds>
* src/rtl/gtsln/gtsln.c
! fixed cleanup string sent to terminal to remove test UTF-8 character
* src/rtl/gtsln/kbsln.c
+ added workaround for some broken Linux kernel which wrongly sets
tty device as ready for reading
* src/rtl/hbsocket.c
* added comment
* src/rdd/dbfnsx/dbfnsx1.c
* src/rdd/dbfntx/dbfntx1.c
! added missing tag root page restoring cleared after discarding buffers
due to concurrent access to multitag indexes.
Many thanks for Karsten Stucker for self contain example.
* src/rdd/dbfntx/dbfntx1.c
* discard buffers when next free page is changed even without touching
version number - minor protection for potentially broken 3-rd party
NTX code or very old [x]Harbour NTX RDDs which do not increment
version number on update.
* contrib/hbssl/hbssl.h
* contrib/hbssl/evppkey.c
+ added new C function
void hb_EVP_PKEY_ret( EVP_PKEY * pkey );
* contrib/hbssl/hbssl.hbx
* contrib/hbssl/pem.c
+ added new PRG functions:
PEM_READ_PRIVATEKEY()
PEM_READ_PUBKEY()
* contrib/hbssl/ssl_sock.c
! added missing HB_SOCKET_ERR_TIMEOUT setting in hb_ssl_socketRead()
! fixed cleared to early SSL error code in hb_ssl_socketWrite()
* contrib/hbssl/tests/inetssl.prg
! interrupt test program if server thread stops
* load server certificates before listen socket is created
* src/rdd/dbfnsx/dbfnsx1.c
! code cleanup to pacify warnings in new GCC versions
* src/rtl/filesys.c
! check if child's fork() was executed successfully
* src/rtl/hbproces.c
! use hb_fsCloseRaw() instead of close() in hb_fsProcessExec() to avoid
lost handles when close() operation is interrupted by signal.
! repeat waitpid() calls interrupted by signal
* include/hbapi.h
* include/hbvmpub.h
* removed casting C style casting from commonly used macros
* contrib/gtqtc/gtqtc1.cpp
* use C++ casting instead of C casting
* contrib/rddads/ads1.c
* contrib/rddads/adsx.c
* contrib/rddsql/sqlmix.c
* src/rdd/dbfcdx/dbfcdx1.c
* src/rdd/dbfnsx/dbfnsx1.c
* src/rdd/dbfntx/dbfntx1.c
* src/rdd/dbsql.c
* src/rdd/workarea.c
* cleaned HB_IT_* usage
* contrib/rddads/ads1.c
! fixed typo in low level code of ordDescend() / DBOI_ISDESC,
now ordDescend() should work with ADS* RDDs just like with
native RDDs
* contrib/hbfoxpro/relfunc.c
! fixed typo in Between() function - thanks to Petr Chorney
* src/rtl/gtwvt/gtwvt.c
! added workaround for keyboard drivers which sets CTRL key modifier
flag when only ALTGR is pressed - thanks to Maurizio la Cecilia
* src/rdd/dbfcdx/dbfcdx1.c
* src/rdd/dbffpt/dbffpt1.c
* src/rdd/dbfnsx/dbfnsx1.c
* src/rdd/dbfntx/dbfntx1.c
* src/rdd/delim1.c
* src/rdd/sdf1.c
* src/rdd/hsx/hsx.c
* src/rdd/dbcmd.c
* src/rdd/dbcmd53.c
* next SELF_INFO() and SELF_RDDINFO() cleanups
* src/rtl/gtwvt/gtwvt.c
! fixed reverted LEFT and RIGHT ALT modifiers in my last commit
* 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.
* 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.
* include/hbclass.ch
! use more restrictive PP rules to avoid possible wrong
interactions with code using 'var' and 'data' identifiers
* force class name to be valid identifier
* commented unused rules
* include/std.ch
! fixed CODEPAGE clause in COPY TO and APPEND FROM commands
+ added VIA and CODEPAGE clauses to JOIN command
* src/rdd/dbcmd.c
* updated comments with some function syntax
* include/harbour.hbx
* src/rdd/dbcmd.c
+ added new function:
hb_dbGetFilter() -> <bCode> | NIL
it returns codeblock used as work area filter or NIL
* src/debug/dbgwa.prg
* use hb_dbGetFilter() to check if WA filter was set without
filter text representation
; above modifications created by Rolf - thanks
* src/rdd/dbcmd.c
! clear filter if dbGetFilter( , "" ) is called instead of setting
dummy one - thanks to Rolf
* include/hbclass.ch
* disabled DECLARE to not generate ignored by compiler noise in .ppo
files
* src/rdd/dbcmd.c
! allow to use dbSetFilter() with string only parameter without
codeblock - Cl*pper compatibility fix.
Cl*pper allows to set text filter without codeblock. In local
RDDs it effectively does nothing and only dbFilter() returns it
but RDDs with automatic filter optimization like CL53/DBFCDX /
COMIX/ClipMore or RDDs working with remote data base servers
may use only text version of filter and ignore or use with
lower priority the codeblock so Harbour has to work like
Cl*pper here.
* src/rdd/workarea.c
* typo in comment
* contrib/rddads/ads1.c
* do not validate empty text filters
* *
% 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
* include/hbrdddel.h
* src/rdd/delim1.c
+ added to DELIM RDD support for defined by user default filename
extenssion. It can be changed by:
hb_rddInfo( RDDI_TABLEEXT, ".csv", "DELIM" )
+ added to DELIM RDD support for exporting initial record with field
names. It can be enabled by:
hb_rddInfo( RDDI_SETHEADER, 1, "DELIM" )
and disabled by:
hb_rddInfo( RDDI_SETHEADER, 0, "DELIM" )
Importing from CSV files with such header is not supported yet but
I plan to add it so in the future RDDI_SETHEADER=1 will also effect
import (APPEND FROM).
* src/rtl/hbznet.c
* src/rtl/hbzsock.c
! cleanup
* include/hbrdddel.h
* include/hbrddsdf.h
* src/rdd/delim1.c
* src/rdd/sdf1.c
+ added support for any standard EOL combination to DELIM and SDF RDDs,
Now LF, CRLF, CR and LFCR EOLs are supported and automatically
recognized during import if one of such EOLs is also set in HVM by
_SET_EOL.
% use 8192 bytes read ahead buffer to speed up import from DELIM and
SDF RDDs
+ modified file IO operations to work with stream only (no support for
file seek) Harbour File IO drivers
% optimized some File IO operations
* strip all EOF (^Z) characters from imported data. It's not strictly
Cl*pper compatible because Cl*pper interrupts import on first ^Z
character anyhow it was always the source of problems when .txt
files are concatenated by tools like 'cat' in *nixes so I decided
to change it.
+ added support RECORD clause command in commands like:
APPEND FROM "data.txt" RECORD (nRec) DELIMITED
APPEND FROM "data.txt" RECORD (nRec) SDF
It's Harbour extension - Cl*pper does not support it.
+ generate RTE when some unsupported in transfer RDDs methods are
called
* 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/dbinfo.ch
* include/hbrdddbf.h
* src/rdd/dbf1.c
+ added new RDD switch RDDI_SETHEADER
It allows to define how DBF header is updated in tables which
will be opened later, i.e.:
// maximize performance reducing number of DBF header updating
// header is updated only when table is closed if before new
// record was added
hb_rddInfo( RDDI_SETHEADER, DB_SETHEADER_MINIMAL )
or:
// update header in the same moment as Clipper and xBase++
// and store only last two digit from the year for compatibility
// with old dBase and FoxPro versions
hb_rddInfo( RDDI_SETHEADER, ;
hb_bitOr( DB_SETHEADER_CLIPPER, DB_SETHEADER_YYEAR ) )
// DB_SETHEADER_CLIPPER
By default Harbour uses DB_SETHEADER_APPENDSYNC.
For more information look at comments for DB_SETHEADER_* switches
in dbinfo.ch
+ added new table switch DBI_SETHEADER
It has the same meaning as RDDI_SETHEADER but changes the header
updated mode of already open table by dbInfo() function.
* src/rtl/filebuf.c
! added missing redirection for Configure() method.
* src/rtl/hbproces.c
! fixed sysconf() emulation for OpenWatcom 1.9 and earlier to work with
_SC_PAGESIZE
* src/rdd/dbf1.c
* update DBF header only after modification in COMMIT and CLOSE methods.
It's Cl*pper and xBase++ compatible behavior anyhow it also means that
number of records in DBF header is not modified just after APPEND
so it cannot be used to synchronization in concurrent file access.
* src/rtl/hbproces.c
! safe and restore hb_fsError() when unused pipe handles are closed
in hb_fsProcessOpen()
! add extension (.com or .exe) in hb_fsProcessOpen()/hb_fsProcessRun()
when executed command is given with PATH part and without extension.
It's documented that DosExecPgm() needs full filename in such case.
It fixes problem with contrib/make.hb which calls hbmk2 with directory
part but without extension.
! enabled pipes instead of temporary files for stdio redirection in OS2
builds of hb_fsProcessRun() - now it should work correctly.
* src/rtl/hbprocfn.c
+ update FError() in hb_Proces*() functions
* 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).
* include/hbapirdd.h
* src/rdd/wacore.c
* changed 4-th parameter of hb_rddRequestArea() from HB_BOOL fWait to
HB_ULONG ulMilliSec
[INCOMPATIBLE]
* src/rdd/dbdetach.c
+ accept optionally as 4-th parameter of hb_dbRequest() numeric value
with timeout in seconds. The previous version using xBase++ compatible
parameters (logical value <lWait> in 4-th parameter) is still working.
; above modifications are based on request and patch sent to harbour-devel
list by Rolf (many thanks), ref:
https://groups.google.com/forum/?hl=pl#!topic/harbour-devel/Pu2b1M5VXjU
* 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
* 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.
* include/hbapifs.h
* src/rtl/filebuf.c
* src/rtl/filebufd.c
* src/rtl/filesys.c
* src/rtl/vfile.c
* src/rtl/iousr.c
* src/rdd/dbfcdx/dbfcdx1.c
* src/rdd/dbffpt/dbffpt1.c
* src/rdd/dbfnsx/dbfnsx1.c
* src/rdd/dbfntx/dbfntx1.c
* src/rdd/hsx/hsx.c
* contrib/hbcomio/comio.c
* contrib/hbmemio/memio.c
* contrib/hbtcpio/tcpio.c
* contrib/hbnetio/netio.h
* contrib/hbnetio/netiocli.c
* contrib/hbnetio/netiosrv.c
* changed HB_USHORT uiAttr parameter to HB_FATTR nAttr
in hb_fsExtOpen() and hb_fileExtOpen()
* include/fileio.ch
* src/rtl/vfile.c
+ added support for FO_DEFAULTS in hb_vfOpen().
When this flag is used in <nModeAttr> parameters then new
_SET_DEFAULT is respected when new file is created and
_SET_DEFAULT + _SET_PATH when file is opened.
+ added support for checking file size by it's in hb_vfSize()
so it can work like hb_FSize() but for virtual FS not only
for local files. Now this function has the following syntax:
hb_vfSize( <pHandle> | <cFileName> [, <lUseDirEntry> ] ) -> <nSize>
The 2-nd parameter <lUseDirEntry> is significant only when
the 1-st one <cFileName> is character value.
* src/rtl/gttrm/gttrm.c
* replaced C++ comments with ANSI ones
* src/rtl/vfile.c
* updated casing of function names in comments
* return F_ERROR instead of NIL hb_vfHandle() when wrong parameter
is given (borrowed from Viktor's branch)
* added C function hb_fileParamGet() borrowed from Viktor's branch
* src/rtl/filesys.c
* variable localization (synced with Viktor's branch)
* src/rtl/disksphb.c
! fixed bug with potentially uninitialized buffer
(fix borrowed from Viktor's branch)
* src/rdd/workarea.c
! fixed typo in RDDI_AUTOORDER type - it should be numeric
just like in _SET_AUTORDER. Many thanks to Rolf for locating
the problem
* src/debug/dbgentry.c
! fixed typo in my previous commit for AltD() function detection
Thanks to Franček Prijatelj for reporting the problem and Abe
for self contain example
* src/rdd/dbfcdx/dbfcdx1.c
* reverted byte order in Harbour CDX signature - by mistake I used
little endian. When existsing index is open then reverted signatures
are accepted for backward compatiblity though they will be removed
in the future.
* src/rdd/workarea.c
* move results of EVALBLOCK() RDD method from Harbour stack return item
to workarea valResult item instead of copping it.
; Warning: This modification may exploit problems in code which
wrongly looks for result of EVALBLOCK() RDD method in
HVM stack return item instead of workarea valResult item.
* include/hbapirdd.h
* include/hbusrrdd.ch
+ added DBTF_RECALL flag which disables transferring record DELETED flag
* src/rdd/dbcmd.c
* src/rdd/workarea.c
! disable transferring DELETED flag to destination area in SORT TO /
__dbArrange() operations
* src/rdd/delim1.c
* src/rdd/sdf1.c
+ added dummy RECALL() methods - now RECALL() can be executed by workarea
TRANSREC() method when DBTF_RECALL flag is set
* src/rdd/dbf1.c
! restored the line with initialization of allocated memory buffer
I deleted by mistake removing my private comments before commit.
Without this line sort could randomly crash on small tables (less
then 65537 sorted records)
* include/Makefile
- include/hbdbsort.h
* src/rdd/Makefile
- src/rdd/hbdbsort.c
* src/rdd/dbf1.c
- completely removed old code used in DBF* RDDs as low level backend
for SORT TO ... command and __dbArrange() function.
It was buggy and extremely inefficient in some cases, i.e. after
20 hours I killed process which was sorting table with 2'000'000
records in reverted order so I cannot even say how much time it
needed.
* src/rdd/dbf1.c
+ added new code for table sorting in DBF* RDDs
(SORT TO ... / __dbArrange() backend)
New code fixes many different problems which existed in previous one
like missing support for national collation, wrong descending orders,
wrong sorting of numeric fields with negative values, missing support
for sorting many field types, missing support for transferring MEMO
fields, missing support for transferring records to table with different
field structure or serving by different RDD, etc.
New code is also many times faster then the old one. In practice it
means is now usable for tables with more then few thousands records,
i.e. the test table with 2'000'000 records I used with old code was
copied in sorted order in 13 secs. when pure COPY TO needed 10 secs.
Now it's possible to export sorted tables to different RDDs, i.e.
using DELIM or SDF RDDs in desitnation area.
New code is written in general form without any local to DBF* RDDs
extensions so it can be adopted as base in any other RDD. Probably
I'll move it to default workarea methods so it will be inherited by
all Harbour RDDs and only if necessary authors of some RDDs may
overload it, i.e. to move the operation to the server side in remote
RDDs when source and destination tables are processed by the same
server.
This modification closes the last known for years bug or rather bag
of bugs ;-) in Harbour.
; NOTE: For large tables new sorting algorithm may access source records
more then once. It means that results may be wrongly sorted when
sorted fields in exported records are modified concurrently by
other station during exporting. This can be easy eliminated by
copping source records to temporary file anyhow it will introduce
additional overhead in all cases and user can easy eliminate the
problem by simple FLOCK before sort or making export to temporary
file and then sorting this file or he can simply ignore this
problem as unimportant in the specific situation so I decided to
not implement double copping.
I haven't tested what Cl*pper exactly does in such case so
I cannot say if current behavior is or isn't Cl*pper compatible.
* src/rdd/dbcmd.c
! fixed __dbArrange() to allow record transfer with the same conditions
as __dbTrans() just like Cl*pper does.
Low level RDD code should respect it, i.e. it should be possible to
sort using fields which exist only in source workarea, the order
and number of fields in source and destination workareas can be
completely different, etc.
Low level DBF* SORT() method is completely broken in such cases and
has to be rewritten.
