can there be tuple for-comprehension too?

[goyalyashpal]

• hy has for-comprehensions outputting in various types like list, dict, generator, & set
• ... but not in tuple.
• tuples are great for their immutability & are particularly suited for work with generator / map-filter pipelines where you don't need to mutate the elements

---

conclusion

• the "marked answer" (& its parent comment) reflects the current state python's at regarding tuple comprehensions
• but motivating use cases still exist, and the request for such a feature still stands open

---

motivating use cases

• Interest for #, (comma tuple) tag macro in core? hyrule#44

• (abridged citations from elsewhere)

  • from stackoverflow:

    ... [comprehension is] not technically syntactic sugar as the processes are actually different, even if the end product is the same.
    - jpp

    ... where you're processing large datasets of text records such as CSV, using the tuple constructor here instead of a comprehension has performance implications
    - Tom

    ... Tuple is lighter-weight than list after it's constructed; so it's a bit annoying that its construction is actually heavier-weight than list's due to lack of direct tuple comprehension syntax.
    - Don Hatch

    Comprehensions     -- ^{faster construction than} -->     generator passed to a constructor [function].
    ... functions are expensive in Python. [thing for thing in things] is much faster than list(thing for thing in things).
    A tuple comprehension would not be useless; tuple(thing for thing in things) has latency issues and tuple([thing for thing in things]) could have memory issues.

    - Justin Turner Arthur

  • the reason I'm opening this pull request is that people are already recommending the usage of tuple([...]) instead of tuple(...). While tuple([…]) is faster (for now), its peak memory consumption is (up to) twice that of tuple(…), as mentioned in the comment in the linked StackOverflow answer above.
    - tuple unpacking is slower than tuple(list comprehension) python/cpython#138325 (comment)

[atisharma]

I'm guessing because there is no tuple generator in python.

Are you after something equivalent to (tuple (lfor x (range 10) (* x x)))? Could be a simple macro. The implementations, from my very quick look, appear to be in core/result_macros.py.

[wrobell]

in python that would be
tuple(x * x for ...
so in hy
(tuple (gfor ...
?

[Kodiologist]

I was curious, once I thought about it, why there are no tuple comprehensions. A core Python developer discusses it here: https://discuss.python.org/t/why-no-tuple-comprehension/2820

[goyalyashpal]

_{as the stackoverflow citations (linked in description) say that comprehensions are faster than using generator, so ...}

is the process of list comprehension in python (at levels lower than source-code, say AST etc.) made up of modular components which can be used to spit out a similar optimum process for tuple comprehension via hylang?

my reasoning is that, maybe the deeper components to construct the similar direct/minimal process for tuple-comprehensions already exist, but just can't be exposed at syntax level due to py's syntax being already saturated for this case.

[Kodiologist]

is the process of list comprehension in python (at levels lower than source-code, say AST etc.) made up of modular components which can be used to spit out a similar optimum process for tuple comprehension via hylang?

No, each comprehension type is its own AST node type, and there is no API to add and implement your own node types.