Fix IPP request op handling; add Destination attr fetch convenience API

[Abd002]

This PR includes two related changes:

1. Fix IPP request copying in src/ipp.rs
IppRequest::send() previously copied requests using ippNew() which does not preserve the operation code. This caused cupsDoRequest() to receive an invalid/empty request for custom operations in some cases.
   Now we grab the op via ippGetOperation() and copy using ippNewRequest(op), ensuring the request sent to CUPS has a valid operation code.

2. Convenience API for fetching printer attributes
   Adds AttrKind/AttrSpec and Destination::get_attrs() to request missing printer attributes and automatically populate dest.options. This avoids forcing callers to hand-roll IPP requests just to fetch a couple of fields.

Why

• Makes custom IPP requests behave reliably by preserving the operation code.
• Reduces boilerplate for common “fetch a few attrs” use cases.

[Gmin2]

Hey, little out of sync with this pr, can you explain to me what you are trying to achieve !, I might be completely off track with this !

[Abd002]

The goal is basically to make two things work reliably:
1. ipp.rs fix (must-have):
Right now IppRequest::send() copies the request using ippNew(), which creates an IPP message without the operation code. So when you call cupsDoRequest(), the server can treat it like an empty/invalid request and you don’t get the expected attributes (I hit this when trying GetJobs / custom requests).
The fix is to copy with ippNewRequest(ippGetOperation(original)) so the duplicated request is still a valid IPP request.

2. Destination::get_attrs convenience:
   This is just to reduce boilerplate for common cases where you only need a couple printer attrs (state message, member-names, etc). Instead of every caller building a full Get-Printer-Attributes request manually, get_attrs() fetches whatever is missing and fills dest.options.
   I’m using it in the lpstat example, it makes the implementation cleaner, but it’s optional and doesn’t change existing behavior.

If you want, I can split it into two PRs (one for the ipp.rs fix, one for the get_attrs helper). The ipp.rs one is definitely required.

[Gmin2]

Okay can yu add an example of how previously it was failing and how the new archi fixed that, like working one and give some sc

[Abd002]

test scenario I used (local CUPS):

• Setup: CUPS running with a PDF queue at ipp://localhost/printers/PDF.
• Command: cargo run --example fix_explane (builds a Get-Jobs IPP request via IppRequest::new(IppOperation::GetJobs) and sends it with cupsDoRequest()).

Before the fix: IppRequest::send() duplicated the request with ippNew(), which drops the operation code → CUPS receives an “empty/invalid” request and replies ErrorBadRequest (seen in the first run screenshot).

After the fix: duplicate via ippNewRequest(ippGetOperation(original)) so the op is preserved → the same example returns real job attributes (e.g. job-id / job-state printed in the second run screenshot).

[Abd002]

For the second change: it’s just a small convenience wrapper. Destination::get_attrs() fetches only the missing printer attributes via a Get-Printer-Attributes request and then populates dest.options, so callers don’t have to hand-build IPP requests for common fields.
Example usage:

 let attrs = [
   AttrSpec { name: "printer-state-reasons", kind: AttrKind::StringLike },
   AttrSpec { name: "printer-state-change-time", kind: AttrKind::IntegerLike },
   AttrSpec { name: "member-names", kind: AttrKind::StringLike }, <- likely for classes
 ];
 dest.get_attrs(&conn, &attrs)?;

This keeps examples (lpstat -> Currently working on) much cleaner without changing existing behavior.

[bakayu]

IppRequest::send() previously copied requests using ippNew() which does not preserve the operation code. This caused cupsDoRequest() to receive an invalid/empty request for custom operations in some cases.
Now we grab the op via ippGetOperation() and copy using ippNewRequest(op), ensuring the request sent to CUPS has a valid operation code.

I faced this exact same issue today, @Abd002 your solution seemed to have fixed it, I tested it on my local. Waiting for this to get merged.

cc: @Gmin2

[Gmin2]

#10 (comment)

[Abd002]

@Gmin2 I’ve updated this PR and added tests to make the change easier to review and to show the intended behavior more clearly.