mORMot Open Source

@flydev — noticed your mention of the mormot.ext.os extension here.

We discussed the stdin pipe need in
https://synopse.info/forum/viewtopic.php?id=7304
and your implementation is spot-on for our multi-FFmpeg scenario.

Really impressed by the mormot2-extensions repo — it's not just a quick patch but a proper library with tests and four demo projects covering different real-world use cases (interactive processes, concurrent FFmpeg, even Claude Code orchestration).

Thanks for the effort!

Thank you Arnaud for the explanation.

That makes perfect sense — GroupRights stored as pointer(IDValue) limits IDs to PtrInt size.

We were assigning TSynUniqueIdentifierGenerator to all tables including TAuthUser/TAuthGroup, which produced 63-bit IDs
incompatible with this design.

Our fix: skip auth tables in SetIDGenerator loop:
if vClass.InheritsFrom(TAuthUser) or vClass.InheritsFrom(TAuthGroup) then
  continue;

Auth tables now use default auto-increment, domain tables keep TSynUniqueIdentifier. Problem solved.

Perhaps worth a note in the SetIDGenerator documentation that it should not be used with TAuthUser/TAuthGroup descendants?

Arnaud, let me bother you again this Sunday!

What's the best way to handle this issue?

Hi, Arnaud.

When you have time and opportunity, please commit steps 7 and 8 to fully understand mORMot2's "Clean Architecture" principle.

I would like to say a special thank you for your teaching, mentoring, and evangelism in modern Pascal.
Respect.

Hello, dear community!

I use IHttpClient/TSimpleHttpClient to access a third-party REST server.
Authentication is performed via a POST request. The request and response are as follows:

C:\Users>curl -i -X POST "http://localhost:8080/j_spring_security_check" -H "Content-Type: application/x-www-form-urlencoded" -H "Cookie: JSESSIONID=F6E4DEBE5FFA22C694907C6AB93B4880" --data "j_username=ODA&password=&j_password=a7f59c966023f1234fe41d69be3a7d43e2e50046&cnonce=09f0cd58db639de3092c1c74f3aa72207f38afa2&_csrf=58ad7adb-fc25-452d-8cad-bad841f619be"
HTTP/1.1 302 Found
Server: Apache-Coyote/1.1
X-Frame-Options: SAMEORIGIN
X-Content-Type-Options: nosniff
X-XSS-Protection: 1; mode=block
Location: /
Content-Length: 0
Date: Wed, 01 Oct 2025 17:17:28 GMT

And if I understand correctly, TSimpleHttpClient perform a redirect occurs followed by a new POST request, to which the server responds with status code 500 — most likely because a POST request to Location: / is not implemented.
I resolved this issue using a workaround.

type
  TSimpleHttpClientAcc = class(TSimpleHttpClient);
...
  HttpClient := TSimpleHttpClientAcc.Create;
...
  HttpClient.fHttp.RedirectMax := 0;

Maybe there's a more elegant way to solve this issue?
On the other hand, it might help someone avoid wasting a ton of time figuring out the cause of a 500 error from third-party REST servers with 30* response

Don't worry.
You're welcome!

Hi!
Two questions:

1. In THttpProxyUrl, there's a property:
/// CSV list of GLOB file or directly names to be rejected as not found
// - e.g. '*.secret'
property RejectCsv: RawUtf8;

If I try to get the content of a file like *.secret, I correctly get:

HTTP 404 Not Found
Wrong route

Which is great!
However, in the folder listing, files matching *.secret are still visible — because the FolderHtmlIndex procedure from mormot.core.search doesn't seem to respect RejectCsv.
Is there a way to override FolderHtmlIndex with a custom version, or hook into the existing one to take into account RejectCsv?

2. What's the best way to reject all files except, say, *.pdf?

Thanks in advance!

Arnaud, please understand me correctly, this is not a call to action, but rather a reflections on this FREE Sunday evening on your thoughts number 28 and 26 in this "Think free as in free speech, not free beer" very inspiring post.

The citation of this article most likely will increased and sooner or later it will be useful for someone in conjunction with mORMot )))
And as always, thank you for your sharing of your knowledge and experience and incredible daily productivity.
Thank you from the bottom of my heart.

Thanks ab, as always!
With this commit, all hidden files and files from hidden subfolders are listed. Checked both Delphi and Lazarus on Windows.
Files with the system attribute are not in the list.

Thanks, ab!
1. Partially. Hidden files in the list. But files from the hidden folders is out, unfortunately.

2. As an option, maybe after executing FindFiles try to check the last file in the list by the FileExists function. And if FileExists returns false, we can assume that the connection was broken.
Of course we should take the last file from the unsorted list and understand that there is a possibility that this last file could have been deleted during the time FindFiles ended and FileExists was called.
But at least it's something, than nothing )))

Let me add my two cents.
For me, there are three things that greatly help in understanding the framework, first two of which I try to do daily:

1. Follow the commits. Honestly, I'm just amazed by Arnaud's productivity throughout the day, especially day after day. Arnaud, please share your secret to such work efficiency!
2. Keep an eye on the discussions  forum.
3. When necessary, look at the test units for examples of how to use various functions, approaches, and usage scenarios, as they cover the entire framework.

Great as always, ab!
Thank you!

Generally speaking also plus FTP, FTPS, and SFTP

ftp://[user[:password]@]server[:port]/path/to/remote/resource.mpeg
sftp://[user[:password]@]server[:port]/path/to/remote/resource.mpeg

From
https://ffmpeg.org/ffmpeg-protocols.html#ftp

Humm, you are rigth. Testing with "ping" works fine.
But with small test program like

 
begin
  for i := 10 to 80 do
    begin
      write(IntToStr(i) + ',');
      sleep(200);
    end;
end. 

I get this log

Start at: 07.02.2024 19:54:58
OnRedirect at: 07.02.2024 19:55:06
10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52
OnRedirect at: 07.02.2024 19:55:12
,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,
Exit Code: 0

function TFoo.OnRedirect(const text: RawByteString; pid: cardinal): boolean;
begin
  Result := False;
//  Writeln(pid);
  if text <> '' then
    begin
      Writeln('OnRedirect at: ' + DateTimeToStr(Now));
      Writeln(text);
    end;
end;

As always, thanks for the lightning-fast response, ab! ))

Add.

Is it possible to add RedirectOutput execution after a certain timeout rather than after the buffer is full?
For long processes with percentage progress, a lot of time passes until the buffer is filled out and the data is transferred to TOnRedirect.
Thanks!

Thank you Arnaud for your daily work and for sharing your knowledge and experience with us!
Great job!

Regarding your question in the blog article about the magic number 1.0594631 is the twelfth root of 2, if I'm not mistaken.
And your hint about programming music suggests that there are 12 semitones in an octave and each semitone differs from its neighbors by 1.0594631 (the twelfth root of 2).
I guessed? )))))

Yes, 48 and below working fine ))). I added the numbers for Ibx firebird 25 to my previous post.

Because, as I wrote earlier, embedded version of fbclient.dll from firebird 5 installation was used instead of connecting to 2.5 firebird server

Yes, that's right, https://www.firebirdfaq.org/faq197/

Yes, but without success.

! EXC   EIBInterBaseError {Message:"Engine Code: 335544351\r\nunsuccessful metadata update"} [Main] at 0100375a46 client/include/wloadlibrary.inc tfbclientapi.ibdatabaseerror (209)

Great, thanks for sharing your experience!

Ок, thanks. Yes, it is clear!

Yes, with my local edits MVCServerFirebirdIbx compiling successful for both (Lazarus and Delphi 12) and working as expected, except FTS.
Tomorrow I'll try to play with ..\ex\extdb-bench\PerfTestConsole.dpr for both (Lazarus and Delphi 12) with the latest ab's commit.

I suspect that the performance figures that I gave a little earlier are not entirely correct.
Most likely the embedded version of fbclient.dll server from firebird 5 installation was used instead of connecting to a particular version of firebird server.
But this is just a guess, I'll try to figure it out tomorrow.

Add.
By the way. Thanks madorin, there is a pretty good library with precise transaction controls for Firebird. Separate for reading and writing transactions, at least.
Maybe this could be a starting point for the next Firebird driver for mORMot 2
https://github.com/madorin/fibplus/
https://github.com/madorin/fibplus/commits/master

Some test results

Running tests using Synopse mORMot framework 2.2.6749, compiled with Free Pascal 3.2.2 64 bit, against SQLite 3.44.2, on Windows 8 64bit (6.2.9200), at 2024-01-19 13:28:17.

Insertion speed								|	Read speed
 			Direct	Batch	Trans	Batch Trans		| 				By one	All Virtual	All Direct
Zeos firebird 25	2265	60781	16408	65852			|	Zeos firebird 25	10884	366770		399840
Zeos firebird 4		2187	55836	6949	63136			|	Zeos firebird 4		6944	373538		412864
Zeos firebird 5		2473	65590	8084	52379			|	Zeos firebird 5		8154	403177		490196
Ibx firebird 25		2366	43955	15503	40030			|	Ibx firebird 25		3850	164076		120290
Ibx firebird 4		1296	37603	3675	49165			|	Ibx firebird 4		1798	106503		99975
Ibx firebird 5		1275	30746	4452	40146			|	Ibx firebird 5		1967	221028		214850

For some reason log shows Windows 8 64bit, but in fact Windows 11 64bit.

fbintf1.4-0 already installed on Delphi.

For Lazarus:
MVCServerFirebirdIbx.dpr(54,17) Error: identifier idents no member "CreateDescendingPK"
MVCServerFirebirdIbx.dpr(95,1) Error: Can't open resource file "D:\Projects\Delphi\Components\Synopse\mORMot2-2.0.stable\ex\mvc-blog\mormot.win.default.manifest.res"

mormot.db.sql.ibx.pas(119):
property FirebirdLibraryPathName: RawUtf8
but
mormot.db.sql.ibx.pas(87):
fFirebirdLibraryPathName: string;

For today's commit, "2.2.6617" works fine. Thank you!
But strangely, I followed https://github.com/synopse/mORMot2/commits/master every day and did not see any changes regarding this issue.
Probably I was inattentive and missed it))))
That's why I decided to bother you today.

Yes. Same results.

Sorry. I do not have permission to publish the log of communication with the corporate AD server. Can I send the log to your email address?

Hi Ab.

The call

TLdapClient.GetGroupInfo('', vBaseDN, vLdapGroup, '', True);

Successfully returns the corresponding group (vLdapGroup) with N members.
One of the member from this group has DN returned by vLdapGroup.member[k] kind of 'CN=Super User,OU=1\+2 MEDIA,OU=Domain Users,DC=1plus2,DC=corp';

To determine sAMAccountName for this user, the next call

TLdapClient.GetUserInfo('', vLdapGroup.member[k], '', vLdapUser);

returns an error
Invalid input name: CN=Super User,OU=1\+2 MEDIA,OU=Domain Users,DC=1plus2,DC=corp

If I remove \ before +, then such a user does not exist.
The actual name of this OU is "1+2 Media" without quotes.

How to correctly call a TLdapClient.GetUserInfo with such a parameter?

In continuation of our conversation, this is not a problem of non ASCII characters, rather this is a problem of special characters for AD.

Thanks a lot.