feat(core): support org-enforce-todo-dependencies #741
Conversation
Many thanks for the contribution. 👍 Old issues don't mean, there is no interest in a working implementation. It just means, that we need more active contributors. 😄 From a very quick overlook your code looks good - I will try it out as soon as I can.
I wasn't even aware of this, but Orgmode is soo big ... you always learn something new ... 😅 |
add warning for blocked tasks
PannenetsF
force-pushed
the
dev-org-enforce-todo-dependencies
branch
from
08c4490
to
7413717
Compare
|
Haha, glad to hear that! I am migrating my org from emacs to nvim, and will try to implement some daily functions into nvim-orgmode. |
|
Hi there, I really dont want to be pushy. Thanks for your great work here. |
|
@PannenetsF I finally found the time to play around with your contribution. First I was not sure if I misunderstood how the feature is supposed to work, but what was immediately noticeable is that changing the TODO state of a parent headline is completely blocked. And it is actually blocked on pressing "t", so if I want to set it to "Progress" and press "tp" I get a warning and it pastes my current register. On the other hand it does not change the state of the parent headline, when I set a child into "Progress", so I am not able to fix inconsistent states. I then actually took the time to see, how the feature behaves in Emacs. There you can change the state of the parent headline as long as it is not a state considered as "closed" (a state defined on the right hand side of the "|" in the configuration list). To get this merged, we should mimic the behavior of Emacs org mode. |
|
|
||
| local M = {} | ||
| -- @param headline OrgApiHeadline | ||
| -- local M.vis_head = function (headline, indent) |
| function OrgMappings:_has_unfinished_children(headline) | ||
| for _, h in ipairs(headline:get_child_headlines()) do | ||
| local was_done = h:is_done() | ||
| if not was_done then |
There was a problem hiding this comment.
Do we need this local variable? It is not used later on, isn't it?
There was a problem hiding this comment.
| if not was_done then | |
| if not h:is_done() then | |
| return true | |
| end |
| if not was_done then | ||
| return true | ||
| end | ||
| if OrgMappings:_has_unfinished_children(h) then |
There was a problem hiding this comment.
I stumbled over this recursion and after some moments I recognized, that not every headline is a todo. I think this is worth a comment.
| return false | ||
| end | ||
|
|
||
| describe('Todo mappings unfer force dependency', function() |
| }, vim.api.nvim_buf_get_lines(0, 2, 4, false)) | ||
| end) | ||
|
|
||
| it('should change todo state of a headline backward (org_todo_prev)', function() |
There was a problem hiding this comment.
Duplicate test name? Can you reflect the difference to the former test in the name?
| @@ -0,0 +1,329 @@ | |||
| local config = require('orgmode.config') | |||
There was a problem hiding this comment.
I appreciate, that you wrote tests. I saw that some of them are actually testing multiple things. I'm my experience this can make a regression harder to track, although it's much better than having no test. Would you mind to split the tests up a little bit?
Hi Seflue, thanks for your review. I would read the original Emacs orgmode for better mimic behavior. I am kind of busy these weeks. Will try to fetch time to do that |
Hi guys, I just impl the option for
org-enforce-todo-dependencies, which will prevent parent TODOs from toggling when it has some unfinished children.I am not sure if you are still interested in it, as the related issue #274 was updated 2 years ago.
Here comes the reference.
https://orgmode.org/manual/TODO-dependencies.html
:)