Don't explicitely pattern-match on the various outputs #4112
Conversation
|
src/libexpr/primops.cc
Outdated
| @@ -1895,26 +1913,26 @@ static RegisterPrimOp primop_path({ | |||
| An enrichment of the built-in path type, based on the attributes | |||
| present in *args*. All are optional except `path`: | |||
|
|
|||
| - path | |||
| - path | |||
There was a problem hiding this comment.
Please keep the trailing white space, it's significant to mdbook (causing it to insert a <br/> between the first line and the rest).
There was a problem hiding this comment.
Oh i was wondering what that was. Is there perhaps another way to do this? At least with Github's markdown
- path
The underlying path.works
There was a problem hiding this comment.
IIRC, I think the other difference is that lowdown treats the first as a definition list so it's rendered differently in manpages. Too bad Markdown doesn't have proper definition lists...
There was a problem hiding this comment.
Too bad Markdown doesn't have proper definition lists...
Amen to that!
I think the other difference is that lowdown treats the first as a definition list so it's rendered differently in manpages
I wonder if they could be convinced to use a more lint-friendly method, like <!-- PRAGMA ... --> or something.
C++ manages to make pattern-matching less convenient to use than a dirty succession of `if`s, so let's use the latter
It doesn't make sense to have one without the other, so it's more logical to query both of them at the same time (and inside a single `std::optional` to ensure that we can't have one without the other).
thufschmitt
force-pushed
the
outputs-by-capability
branch
from
46b6b1f
to
ae55b5d
Compare
This is a refactoring on top of #4056 to remove as much as possible the pattern-matchings on the different kind of outputs that a derivation has. The idea behind that is that we generally don't care about the exact type of output that we have, we're only interested in the answer to one of the two questions: "is this a CA output?" and "Do we know the path of this output?". So I've added a few methods to
DerivationOutputto query this directly without having to introspect the underlying output type and used that as much as possible.(I've also mixed in a few unrelated minor changes, I can remove them if needed)