Add non-spread overload for Result.all #125
Conversation
| // Utility types | ||
| type Head<T extends any[]> = T extends [any, ...infer R] ? | ||
| T extends [...infer F, ...R] ? F : never : never | ||
| type Tail<T extends any[]> = T extends [any, ...infer R] ? R : never |
There was a problem hiding this comment.
These could likely be removed since the arrays I'm parsing above are typed as any[]. I could just as easily type the arguments as arg0: T[0] | T, argN: ...T but using these types are technically truer to what's going on.
jest.config.js
Outdated
| '^.*\.ts$': ['ts-jest', { | ||
| tsconfig: 'test/tsconfig.json' | ||
| }] | ||
| }, |
There was a problem hiding this comment.
This silences the warnings that global ts-jest configuration is deprecated
|
Hey, I think this is a nice idea, thank you very much for this. I'll do a full review later but in the meantime please can you remove unrelated changes from this PR? (reformatting, whitespace, jest.config changes, anything else that I missed) |
ConnorSinnott
force-pushed
the
add-non-spread-result-all
branch
2 times, most recently
from
08f3d0b
to
6141535
Compare
|
Done! Thank you @jstasiak. Had to wrestle with the editor to omit the whitespace changes. All tests pass now as well. |
|
The jest changes have been moved to #126 |
|
Thank you, I think this is a great addition. Can you update the documentation too? It's in |
ConnorSinnott
force-pushed
the
add-non-spread-result-all
branch
6 times, most recently
from
440f751
to
f6035ce
Compare
ConnorSinnott
force-pushed
the
add-non-spread-result-all
branch
from
f6035ce
to
a604ea8
Compare
|
Ok @jstasiak, how do those docs read for you? |
ConnorSinnott
force-pushed
the
add-non-spread-result-all
branch
from
a604ea8
to
0cb8cb2
Compare
ConnorSinnott
force-pushed
the
add-non-spread-result-all
branch
from
0cb8cb2
to
fc0bafd
Compare
|
Very nice, thank you |
|
Released in version 4.2.0 just now. |
Howdy! I'm bringing this PR over from the decidedly dead vultix/ts-results repo.
This PR adds non-spread variants of
Result.allandResult.anywhich should be preferred over the parameter spread variants. For instances whereResult.allorResult.anyare invoked with very large arrays, the spread operation could lead to stack overflows.Additionally, speaking from personal experience in an enterprise environment; the usage of
Result.allvia a spread array (Result.all(...myResults)) grossly exceeds the usage ofResult.allwith multiple parameters. I'd imagine this is the same for others as well.This new behavior has been added as an overload of
Result.allto offer backwards compatibility for those spreading arrays in, or passing each argument individually. Though based on the stack overflow issue mentioned above, I'd recommend users migrate from array spreading to just passing the array.