scheme/r7rs

Fork 1

align `substring` and `list-copy` with similar functions with `start` and `end` arguments #206

New issue

Open

opened 2024-09-27 17:43:51 +00:00 by ecraven · 6 comments

ecraven commented

2024-09-27 17:43:51 +00:00

All the other subcollection-accessors (string-copy, vector-copy, bytevector-copy, and so on) take three forms:

(copy object)
(copy object start)
(copy object start end)

I propose that substring be amended to also support this usage, so either the last argument or both can be left off.

All the other subcollection-accessors (`string-copy`, `vector-copy`, `bytevector-copy`, and so on) take three forms: ``` (copy object) (copy object start) (copy object start end) ``` I propose that `substring` be amended to also support this usage, so either the last argument or both can be left off.

Zipheir commented

2024-09-27 17:49:41 +00:00

Agreed.

dpk added the

Foundations

label 2024-09-27 17:52:04 +00:00

dpk added this to the Valued Fascicle milestone 2024-09-27 17:52:08 +00:00

dpk commented

2024-09-27 17:52:42 +00:00

Owner

substring can become a binding alias of string-copy without issue.

`substring` can become a binding alias of `string-copy` without issue.

ecraven commented

2024-09-27 17:59:56 +00:00

Author

note that I don't mention list-copy here. abstractly, it might be fine to give it start and end too, but I think – given there's no efficient implementation for that – that's probably not a good idea ;)

note that I *don't* mention `list-copy` here. abstractly, it might be fine to give it `start` and `end` too, but I think – given there's no efficient implementation for that – that's probably not a good idea ;)

mnw commented

2024-09-27 18:05:44 +00:00

Owner

note that I don't mention list-copy here. abstractly, it might be fine to give it start and end too, but I think – given there's no efficient implementation for that – that's probably not a good idea ;)

It is as efficient as list-copy without start and end.

> note that I *don't* mention `list-copy` here. abstractly, it might be fine to give it `start` and `end` too, but I think – given there's no efficient implementation for that – that's probably not a good idea ;) It is as efficient as `list-copy` without `start` and `end`.

aag commented

2024-09-27 18:15:15 +00:00

Owner

It is as efficient as list-copy without start and end.

Yes. There are times when you need it, and for short lists, the performance will be immaterial in many situations. We should include it.

> It is as efficient as `list-copy` without `start` and `end`. Yes. There are times when you need it, and for short lists, the performance will be immaterial in many situations. We should include it.

dpk changed title from ~~[feature request] align substring with all other similar functions by making the start and end argument optional.~~ to align substring and list-copy with similar functions with start and end arguments

2024-09-27 18:18:36 +00:00

Zipheir commented

2024-09-27 21:03:59 +00:00

The one-argument version will make less sense if “newly allocated” is ever dropped from the spec, but the (substring string start) version will always be nice to have.

The one-argument version will make less sense if “newly allocated” is ever dropped from the spec, but the `(substring string start)` version will always be nice to have.

dpk added the

Cleanup

label 2024-09-27 21:36:20 +00:00

No labels

Macrological Fascicle

align substring and list-copy with similar functions with start and end arguments #206

align `substring` and `list-copy` with similar functions with `start` and `end` arguments #206