Pre parse callback #251
Pre parse callback #251
Conversation
Codecov Report
@@ Coverage Diff @@
## master #251 +/- ##
======================================
Coverage 100% 100%
======================================
Files 12 12
Lines 2566 2718 +152
======================================
+ Hits 2566 2718 +152
Continue to review full report at Codecov.
|
Codecov Report
@@ Coverage Diff @@
## master #251 +/- ##
==========================================
- Coverage 100% 99.92% -0.08%
==========================================
Files 12 12
Lines 2566 2668 +102
==========================================
+ Hits 2566 2666 +100
- Misses 0 2 +2
Continue to review full report at Codecov.
|
rework return values from _parse_* function to return true if the value was processed false otherwise, this simplified the logic and got rid of the pulling and clearing of the missing fields from option groups. add TriggerOff and TriggerOn helper functions and some tests for them add shapes example of multiple callbacks in order. allow specification of callbacks that get executed immediately on completion of parsing of subcommand add tests for enabled/disabled by default add _get_fallthrough_parent. To get the most appropriate parent to fallthrough to add enabled and disabled by default functions add positional_arity example Add a pre_parse_callback_ for apps. The Pre parse callback takes an argument for the number of remaining arguments left to process, and will execute prior to parsing for subcommands, and after the first option parse for option_groups.
2115ddc to
2d68c4b
Compare
5241a51 to
9d421e0
Compare
|
Names of the flags, etc. seem fine, at least to me. |
… strings, some cleanup in the README.md Apply suggestions from code review Co-Authored-By: phlptp <top1@llnl.gov> allow callbacks for option_groups, and allow some other characters as flags
822a383 to
3777889
Compare
|
As a note: a number of the issues with callback ordering came up as I am trying to integrate CLI11 into one of our bigger applications. A lot of the capabilities I have worked on over the last few weeks have been driven by the requirements and use cases from those applications, so I think I am nearing the end of integrating it, but we will find out over the next few days as I continue working on that. |
|
We'll wait to do a 1.8 release at least until you have finished integration and verified everything works together. :) |
…ype and update some corresponding test cases and documentation Allow option groups to use ignore_case ignore_underscore for inheritance Allow option groups to specify allow_extras even if the parent app doesn't in which case extra options flow down into the option_group.
3d62b70 to
e5a78ee
Compare
|
I updated based on #252 |
|
Is this ready for review/merge? |
|
I have been slowly integrating CLI11 into our bigger applications, and am probably about half way through that effort. I think with what I did yesterday, I need to add a remove_subcommand function to app and add_subcommand(sub *) to the option_group class. But I haven't done that yet or written test cases for them. So I think you can merge/review as is and I will add the rest and what else I might uncover in another PR or you can wait a few days while I complete the integration. |
|
Will close #256 |
add some test of the remove_excludes functions add test for Issue CLIUtils#256 add remove_subcommand fail test add remove_subcommand function and add_subcommand to option_group and some tests associated with them.
06f1914 to
684e61c
Compare
|
I think this is ready to go, for another look. |
|
Okay, will look at it as soon as I can. Traveling back from a conference in Switzerland tomorrow, then to JLab this weekend for another one. |
|
Looks pretty good, will merge. Standard feedback that I might suggest adjustments later before release. |
This PR adds a pre parse callback and clarifies the callback timing of subcommands and option groups.
It will also address issue #166. The basic concept is to allow each app/subcommand/option_group to have two callbacks, one would execute after the first argument of the subcommand/option group is processed. This allows an ability to effect the parsing dynamically, based on the number of arguments passed like Issue #166, or to dynamically enable or disable groups of options based on which options or subcommands were passed, or whatever other program state is appropriate to an application. The other callback executes after parsing is complete. With
immediate_callbackthis is done immediately following the parse. So callbacks from multiple subcommands can be executed in order and executed multiple times. If that flag is not set all callbacks are executed on completion of all parsing in order of creation.To facilitate adaptive enabling or disabling option_groups some flags were created
enabled_by_defaultanddisabled_by_defaultthat reset the disabled_ flag to a defined value on start of parsing.In the process of this a few issues with subcommand termination were fixed, and the readme updated with a new section on callbacks. Also a subcommand_terminator was added (
++) to enable explicitly specifying that a subcommand is finished.Also added two new examples
shapeswith repeated subcommands and callbacksand positional_arity which is Issue #166 coded up.
TODO:
I am not sure I like all the names of the flags, so suggestions are welcome.