[42,73,102,104,110] Modernization and refactoring

[SwooshyCueb]

Addresses #42 Addresses #73 Addresses #102 Addresses #104 Addresses #110 In service of #95 In service of #103 Supersedes #96 Supersedes #97

This PR has a lot:

• Fix for the RPM transaction check error (#104)
• A little cleanup of CMakeLists.txt
• Use of IrodsRunpathDefaults in CMakeLists.txt
• Refactor to use nlohmann-json (#42)
• Refactor to use qpid-proton's C++ API (#73)
• Removal of json wrapper tokens (#102)
• std::chrono timestamp retrieval (in service of #95)
• Unstringified, @timestamp-keyed milliseconds-since-epoch timestamp (in service of #103)
• Unstringified pid in emitted json (in service of #103)
• Migration from rodsLog to irods::experimental::log (in service of #95)
• Removed unused amqp_options configuration setting (#110)
• Content type and creation timestamp in AMQP message metadata
• hostname from boost (in service of #95)
• better handling of default configuration
• Reformat to our new clang-format and clang-tidy configuration
• Migration from boost::regex to std::regex (in service of #95)
• Migration to std::filesystem::path for test_mode log (in service of #95)
• Removes old stomp-based python tests

[trel]

if this confirms the new qpid-proton packages are good-to-go, i'll get them pushed to packages.irods.org

[korydraughn]

Let's avoid adding the clang-format/tidy files to other projects until we figure out the reusable github workflows. I should have something ready regarding that soon.

[korydraughn]

Let's replace the use of "Resolves" word in the first comment of this PR so that the issues aren't automatically closed.

[SwooshyCueb]

Note to self: add a tick to "maximum oversquash" on the whiteboard

[SwooshyCueb]

How do we feel about main.cpp for the plugin filename? I think I'm okay with it, but just wanted to bring it up for discussion as we try to think about conventions in our plugins.

I'm not married to it. I can change it to audit_amqp.cpp if that is preferred.

[korydraughn]

I think main.cpp is a very clear indicator of where someone should look when coming to an unfamiliar project. I prefer main.cpp, but am not against other ideas if they prove more helpful for us/others.

Logical quotas, metadata guard, and hard links REPs also use main.cpp too.

[trel]

Logical quotas, metadata guard, and hard links REPs also use main.cpp too.

seems main.cpp is both convention and expected

[alanking]

main.cpp seems fine to me, then!

[SwooshyCueb]

Addressed all review comments. 96bca7c01065085129357e13643b305024876881 might be of interest

[korydraughn]

96bca7c looks fine overall. The only thing that stood out was the use of the C-style cast. A static_cast would be safer and better for tools. There may be an integer suffix for that too.

Other than that, was there anything in particular you feel we need to look at regarding that commit?

[korydraughn]

Sadly, the integer suffix zu isn't available until C++23 (or with -std=c++2b).

[SwooshyCueb]

Switched it to a static_cast and squashed. If there are no objections, I'll go ahead and #

[korydraughn]

Do all of the tests pass (python, cpp unit tests)?

[SwooshyCueb]

cpp unit test suite passes. python test suites don't. looks like I've still got some work to do

[SwooshyCueb]

Removed stomp-based python tests (no use updating since everything they covered is already covered by the proton-based python tests) and updated the proton-based python tests. All tests pass except one pulled from the unixfilesystem tests that appears unrelated.

[SwooshyCueb]

If there are no objections, I will #

[korydraughn]

Is there one or more issues we can link the commits without an issue number to?

If not, let's create issues for them. Leave off the pounds so we can eyeball them one last time.

I'm assuming all of the tests pass now (I think you mentioned that this morning?).

[trel]

yes, definitely need some squashing or issuing...

[SwooshyCueb]

Missing issue numbers added. I don't want to squash any more than I already have, as I think it's potentially useful to be able to follow the commit history here.

I'm assuming all of the tests pass now (I think you mentioned that this morning?).

My earlier comment still applies:

All tests pass except one pulled from the unixfilesystem tests that appears unrelated.

[SwooshyCueb]

#'d

[SwooshyCueb]

Fixed the missing #, and added a change I forgot to include in the last commit (postinst script). Should be ready for merge now. Running tests to find the failing unixfilesystem test, will open an issue for it once I find it

[korydraughn]

Don't forget to update/close the issues. Ping @alanking if you need a checkbox checked.