[#525] Add support for touch API operation

[korydraughn]

The implementation appears to work.

Just need to implement tests.

[trel]

Why all caps error code? Is there precedent for that in PRC?

[korydraughn]

Just not familiar with all the conventions yet. Also, I think I saw exceptions like that used in other places in the PRC.

[d-w-moore]

This is well done. I like the use of an irods.manager.common for commonly /multiply used code. Wonder now if we should scan for instances where irods.common would be similarly useful.

Approve the changes so far. Guess we'll have one more round once tests are implemented

[d-w-moore]

Why all caps error code? Is there precedent for that in PRC?

I don't know the valid uses of INVALID_OBJECT_TYPE , either. Not currently used otherwise, so may be a relic of something in PRC's past that got deleted. It appears also that it subclasses RuleEngineException, which may be misleading. I would suggest another error (possibly more targeted? Invalid_Path_For_Object_Type?)

[korydraughn]

I added a new exception type, InvalidInputArgument, to replace the use of INVALID_OBJECT_TYPE.

All new tests pass. The only way I could run my tests was by skipping the iRODS version check code in helpers.make_session(). Is there another way to handle that?

[korydraughn]

Forgot to mention this PR also adds modify_time and create_time to the replica object. See aad7740.

[korydraughn]

Looks like I need to address a few more codacy issues.

[alanking]

The meat of the change is good. Just a few suggestions/questions

[korydraughn]

All review comments addressed.

All that's left is to run the full test suite.

[alanking]

Just had one more suggestion. Looks good otherwise

[alanking]

I like it. Please squash to taste and omit #s.

[korydraughn]

Squashed.

[alanking]

# it

[korydraughn]

Running full test suite to make sure nothings broken.

Will pound if everything comes out good.

[korydraughn]

I saw 2 failures and 46 errors. Not sure what to expect.

@d-w-moore Any thoughts? Please try running this PR through the test suite and see what you get.

[d-w-moore]

I saw 2 failures and 46 errors. Not sure what to expect.

@d-w-moore Any thoughts? Please try running this PR through the test suite and see what you get.

Sure, will run now.

[d-w-moore]

I saw 2 failures and 46 errors. Not sure what to expect.
@d-w-moore Any thoughts? Please try running this PR through the test suite and see what you get.

Sure, will run now.
Got 13 errors from the entire suite. (270 tests run), running against iRODS 4.3.2.
All 13 accounted for by:

• errors coming from login_auth_test.py. It requires specific setups, so shouldn't be run as a matter of course in the PRC suite. Unfortunate bad form on that one, mea culpa. That will be fixed this next release (see issue Investigate automating BATS tests #502. we'll be renaming it to run in a BATS test and that should defeat the pattern matching that causes it to be picked up in the routine run of the full suite).
• irods.message.Bad_AVU_Value should be descended from ValueError for irods.test.meta_test.TestMeta.test_add_obj_meta_empty to pass.

@korydraughn - For other errors you may have seen, we should discuss live. Do you have a log dump?

[d-w-moore]

Looks good, can go ahead and #. I've tested using https://github.com/d-w-moore/python-irodsclient/tree/ghrunner-nohealthcheck (currently one of my development branches at ) as we discussed.

[korydraughn]

Excellent.

#'d