Hello Przemek
This is an xHarbour compiled ( as of Jun 2007 CVS ) application, a very
large one.
RDD = > RDDADS
Here is a code snippet:
select verify
set order to 2
seek substr(cCustId,2) // Generates RT Error and stats are
alias() == verify
recno() == 200120 or like
IndexOrd(
2008-08-01 03:06 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
* harbour/include/hbpp.h
* harbour/include/hbexprb.c
* harbour/source/pp/ppcore.c
* harbour/source/vm/evalhb.c
* harbour/source/vm/runner.c
* harbour/source/vm/hvm.c
* harbour/source/vm/fm.c
* harbour/source/vm/ev
Hi Przemek,
By cleaning the many warnings, two new ones got revealed:
Warning W8056 source\rdd\dbf1.c 2379: Integer arithmetic overflow in
function hb_dbfPutValue
Warning W8056 source\rdd\dbf1.c 2398: Integer arithmetic overflow in
function hb_dbfPutValue
Both lines use the HB_DBL_LIM_64()
2008-08-01 02:09 UTC+0200 Viktor Szakats (harbour.01 syenar hu)
* include/hbdefs.h
! Patch added to workaround following phoney warnings
when using BCC 5.8 compiler (tested with 5.8.2):
Warning W8017 C:\devl\BCC58\Include\stdint.h 77: Redefinition of
'INT16_MIN' is not ident
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> >2) BCC58 old one:
> >Warning W8017 C:\devl\BCC58\Include\stdint.h 77: Redefinition of
> >'INT16_MIN' is not identical
> >Warning W8017 C:\devl\BCC58\Include\stdint.h 78: Redefinition of
> >'INT32_MIN' is not identical
> >Warning W8017 C
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> Okay I got it now, so a special CRC32 is (or may be) used
> to check if the result was indeed correct. [ Now I wonder,
> why the normal CRC32 - which is calculated anyway - wasn't
> good for this purpose for zips, and they've opted for a
> s
Hi Przemek,
2) BCC58 old one:
Warning W8017 C:\devl\BCC58\Include\stdint.h 77: Redefinition of
'INT16_MIN' is not identical
Warning W8017 C:\devl\BCC58\Include\stdint.h 78: Redefinition of
'INT32_MIN' is not identical
Warning W8017 C:\devl\BCC58\Include\stdint.h 79: Redefinition of
'INT64_
Hi Przemek,
Now the remaining mystery is how 7-Zip could extract
the wrong file anyway :)
It was not wrong. Just simply 7-Zip like myunzip does not
compare CRC32 stored in compressed header with CRC32 in
file ZIP entry ignoring it.
Okay I got it now, so a special CRC32 is (or may be) used
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> Now the remaining mystery is how 7-Zip could extract
> the wrong file anyway :)
It was not wrong. Just simply 7-Zip like myunzip does not
compare CRC32 stored in compressed header with CRC32 in
file ZIP entry ignoring it.
> BTW, I wonder i
Hi Przemek,
Many thanks for the hunt and fix, it now works perfectly.
Now the remaining mystery is how 7-Zip could extract
the wrong file anyway :)
BTW, I wonder if we couldn't rather use hb_fs*() functions
instead of fopen/fread. (in places where this wouldn't
badly affect performance, like hb
2008-07-31 23:02 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
* harbour/source/rdd/hbsix/sxcompr.c
* pacified warnings
best regards
Przemek
___
Harbour mailing list
Harbour@harbour-project.org
http://lists.harbour-project.org/mailman/listin
On Thu, 31 Jul 2008, Przemyslaw Czerpak wrote:
Hi Viktor,
> You do not have too. I've just located the reason of problem.
> Just simply I was not writting anything for DOS/Windows for
> too long time. On this systems fopen() without "b" flag strip
> '\r' from the stream so CRC was wrongly calcula
2008-07-31 22:49 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
* harbour/contrib/hbmzip/hbmzip.c
! added "b" to fopen() flags to force binary mode in all systems
during CRC32 calculation
best regards
Przemek
___
Harbour mailing list
H
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> For these tests I was using BCC 5.5.1, with these
> switches (notably debug switches):
> C_USR=-DUNICODE -v -y
> I've now retried with a fully default build, with
> no custom switches, but the effect is still there.
> Even more strangely, I'
Hi Przemek,
For these tests I was using BCC 5.5.1, with these
switches (notably debug switches):
C_USR=-DUNICODE -v -y
I've now retried with a fully default build, with
no custom switches, but the effect is still there.
Even more strangely, I'm getting the same (wrong)
results with an MSVS 2005
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> Now I'm also confused, because myunzip cannot extra it either
> (created a zero length file).
> Now I did the .zip again (theoretically with no difference),
> I'm attaching it. myunzip is able to unzip it, but unzip isn't.
My unzip too. It
Hi Randy,
Thanks a lot. Some of these warnings are known ones, most of them
is complaining about integer size differences, which we have a lot.
I've committed a fix.
Brgds,
Viktor
On 2008.07.30., at 17:24, Randy Portnoff wrote:
Hi Viktor,
Attached are two logs (one for HARBOUR and one for R
2008-07-31 18:58 UTC+0200 Viktor Szakats (harbour.01 syenar hu)
* include/hbextern.ch
+ EXTERNAL __OBJGETPROPERTIES
* doc/whatsnew.txt
+ Updated with new functions.
* source/hbpcre/_hbconf.h
! Silenced a #pragma warning for MSVC 12.x.
; NOTE: Some notable MSVC 12.x wa
Some additional info:
- 7-Zip 4.57 GUI is able to extract myzip.prg from hello.zip with
'abc' password.
- unzip.exe is able to extract sample encrypted .zip created with
zip.exe.
[[[ BTW, 7z support would be nice in Harbour since it hasn't some of the
limits of .zip and std encryption is a l
Hi Przemek,
I'm sorry but I'm also a little bit confused and I think definitively
there's something wrong / strange when I try to use a *password* zip file
created by hbmzip with an external application (WinZip).
My test.zip has about 500 files (dbf, cdx, fpt ...) and they're zipped /
unzipp
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> Now, I was suspecting that an older hbmzip.lib is picked
> up, but I did some test to ensure this is not the case,
> besides renaming my production harbour binary dir
> (but it's not in the path anyway).
I tired to extract hello.zip you att
Hi Viktor,
Attached are two logs (one for HARBOUR and one for RDDADS) which
contain some warnings when built using MSVC 12.00.8804.
Regards,
Randy.
At 09:57 AM 7/30/2008, you wrote:
Hi all,
I'd like to ask everyone to try as many kinds of
builds as possible and report any results on the
li
Please, forget this message and excuse for create confusion.
Now, hbmzip library seems to work very well.
Many thanks
Juan
- Original Message -
From: "Juan Gálvez" <[EMAIL PROTECTED]>
To: "Harbour Project Main Developer List."
Sent: Thursday, July 31, 2008 4:42 PM
Subject: Re: [Harbo
2008-07-31 17:41 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
* harbour/source/vm/classes.c
+ adding 2-nd parameter to __clsGetProperties()
When it's TRUE then __clsGetProperties() returns also exported
messages which have corresponding assign messages (with "_" prefix)
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> I did these, after doing a complete checkout to an empty dir
> and a full rebuild:
Maybe a stupid question but did you recompile myzip.prg?
It's not create automatically.
> del hello.zip
> myzip hello.zip myzip.tds --pass abc
> unzip hello
Hi Przemek, Viktor,
Perhaps these considerations about my test.zip created with hbmzip can help
:
1) Winzip shows all files in zip as cripted ("*").
2) If I set the correct password then Winzip only unzip the first file of
the zip.
3) Rest of files are unzipped with no password.
Best regard
Hi Przemek,
1) I did a quick test, but unzip doesn't seem to accept
any passwords for something zipped with myzip.
Please be sure that you are using current SVN code.
I made test with UNZIP in my linux box and PKUNZIP
in DOS emulator and both nicely decompress password
encrypted archives
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
Hi Viktor,
> 1) I did a quick test, but unzip doesn't seem to accept
>any passwords for something zipped with myzip.
Please be sure that you are using current SVN code.
I made test with UNZIP in my linux box and PKUNZIP
in DOS emulator and both nice
2008-07-31 14:50 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
* harbour/contrib/hbmzip/hbmzip.c
* harbour/contrib/hbmzip/readme.txt
* changed parameters of HB_UnzipFileInfo()
Now in 10-th parameter is @lCrypted and 11-th is @cComment
* harbour/contrib/hbmzip/tests/myunzip.p
Hi Przemek,
1) I did a quick test, but unzip doesn't seem to accept
any passwords for something zipped with myzip.
Also, I've zipped something with a password using zip,
and tried to unzip it with myunzip, but it didn't
work either. Maybe I'm doing something wrong.
One correction: my
Hi Przemek,
1) I did a quick test, but unzip doesn't seem to accept
any passwords for something zipped with myzip.
Also, I've zipped something with a password using zip,
and tried to unzip it with myunzip, but it didn't
work either. Maybe I'm doing something wrong.
2) I'm not sure I
2008-07-31 14:07 UTC+0200 Przemyslaw Czerpak (druzus/at/priv.onet.pl)
* harbour/contrib/hbmzip/zip.c
! set valid ZIP signature in header file (2.0)
* harbour/contrib/hbmzip/hbmzip.c
! fixed ZIP encryption header by adding CRC32 checksum
Now other tools should properly recognize p
2008-07-31 10:10 UTC+0200 Viktor Szakats (harbour.01 syenar hu)
* contrib/xhb/hbcompat.ch
+ Added translations for WildMatch(), hb_DeserialNext(),
hb_Checksum().
--
Brgds,
Viktor
___
Harbour mailing list
Harbour@harbour-project.org
http:/
Hi Przemek,
I've only meant Windows platform this time. If we later
find out any conversion needs for other platforms, we may
add them later.
Brgds,
Viktor
On 2008.07.31., at 10:54, Przemyslaw Czerpak wrote:
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
If it has no side-effects, maybe we could
On Thu, 31 Jul 2008, Szakáts Viktor wrote:
> If it has no side-effects, maybe we could add this conversion
> to these functions.
Not for all platforms. In *nixes file names should be stored in
native ISO CPs (if we want to be compatible with other ZIP tools)
which is more similar to ANSI in Window
If it has no side-effects, maybe we could add this conversion
to these functions.
Brgds,
Viktor
On 2008.07.31., at 10:24, Juan Gálvez wrote:
Hi Przemek,
Don't worry about this subject. Indeed it was a translation issue
and it can be resolved easily :
HB_ZipStoreFile( hZip, cFilePath, HB_A
Hi Przemek,
Don't worry about this subject. Indeed it was a translation issue and it can
be resolved easily :
HB_ZipStoreFile( hZip, cFilePath, HB_AnsiToOem( cFilePath ), cPassword )
HB_UnzipExtractCurrentFile( hZip, HB_OemToAnsi( cFilePath ), cPassword )
Best regards
Juan
- Original Me
Hi Juan,
Okay I'll commit 3 new translations as soon as the SVN
server gets working again.
Brgds,
Viktor
On 2008.07.31., at 9:58, Juan Gálvez wrote:
Hi Viktor
WildMatch() is already implemented as a function in xhb.lib.
Yes. I know it.
BTW, there are a lot of other compatibility functio
On Thu, 31 Jul 2008, Alex Strickland wrote:
Hi Alex,
> To support this syntax TopClass (C4W) has:
> #xcommand RETURN { [] } ;
>=> RETURN ():New( )
> Perhaps someone could add it to HBClass.ch. It would fit with some
> similar #xtranslates under the HB_CLS_VO section.
>
> BTW what does t
Hi Viktor
WildMatch() is already implemented as a function in xhb.lib.
Yes. I know it.
BTW, there are a lot of other compatibility functions (h*(),
inet*(), hb_checksum(), hb_deserialnext()) implemented as
such instead of simple #defines, xhb users would have to
decide which is the better fo
Hi Juan,
WildMatch() is already implemented as a function in xhb.lib.
BTW, there are a lot of other compatibility functions (h*(),
inet*(), hb_checksum(), hb_deserialnext()) implemented as
such instead of simple #defines, xhb users would have to
decide which is the better for them.
[ as opposed
Hi all,
I've noticed equivalence for HB_WildMatch() in xHarbour is WildMatch().
Could someone add the translate in hbcompat.ch ?
#xtranslate WildMatch([]) => HB_WildMatch()
Best regards
Juan
___
Harbour mailing list
Harbour@harbour-project.org
http:
42 matches
Mail list logo
