Use alloc for no_std builds

[GeneFerneau]

Replace std structs with alloc equivalents to support no_std builds

[devrandom]

Concept ACK

[TheBlueMatt]

Shouldn't this be dropping every use std... statement? Or are we allowed to use std... on no_std builds?

[devrandom]

If I understand correctly, this is a smaller step towards no_std.

[codecov]

Codecov Report

Merging #924 (735f96c) into main (f074343) will not change coverage.
The diff coverage is 100.00%.

❗ Current head 735f96c differs from pull request most recent head 12461fc. Consider uploading reports for the commit 12461fc to get more accurate results

@@           Coverage Diff           @@
##             main     #924   +/-   ##
=======================================
  Coverage   90.47%   90.47%           
=======================================
  Files          60       60           
  Lines       30572    30572           
=======================================
  Hits        27660    27660           
  Misses       2912     2912           

Impacted Files	Coverage Δ
lightning/src/chain/channelmonitor.rs	91.03% <ø> (ø)
lightning/src/chain/keysinterface.rs	94.96% <ø> (ø)
lightning/src/chain/mod.rs	100.00% <ø> (ø)
lightning/src/chain/onchaintx.rs	93.68% <ø> (ø)
lightning/src/chain/package.rs	92.32% <ø> (ø)
lightning/src/ln/chan_utils.rs	97.34% <ø> (ø)
lightning/src/ln/chanmon_update_fail_tests.rs	97.77% <ø> (ø)
lightning/src/ln/channel.rs	88.24% <ø> (ø)
lightning/src/ln/features.rs	98.83% <ø> (ø)
lightning/src/ln/functional_test_utils.rs	94.90% <ø> (ø)
... and 24 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f074343...12461fc. Read the comment docs.

[TheBlueMatt]

Right, we could take this separately, I suppose. I just think the big std -> core "rename" is half of the work on landing a no_std version, so good to do that in a separate PR as well.

[GeneFerneau]

Right, we could take this separately, I suppose. I just think the big std -> core "rename" is half of the work on landing a no_std version, so good to do that in a separate PR as well.

So, I split these changes per your request in #842. My thought in not enabling #![no_std], and not doing a full rename was to incrementally introduce the changes needed to enable a no_std build.

The next PR I will make will include the std -> core "rename", then another separate PR replacing HashMap and friends with hashbrown equivalents. Another for the literacy::* replacements for std::io, and finally one for the remaining std::sync stuff that relies on OS abstractions.

Basically, I am breaking up #842 into small, digestible chunks to make review easier. This should hopefully allow for an inching toward a no_std build, instead of an all-or-nothing drawn out thing in one PR.

[devrandom]

The following are missing imports of Vec:

• messaging_signing.rs
• ln/mod.rs
• zbase32.rs

[devrandom]

What about String?

[GeneFerneau]

Rebased on latest main, changes now include some std -> core renames.

[devrandom]

ACK

[TheBlueMatt]

LGTM. Can you squash the fixup commit down?

[GeneFerneau]

LGTM. Can you squash the fixup commit down?

Yeah, I just wanted to give credit to @devrandom for the fix