Canonical float equality does not conform to RFC8949

Open
#217 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Assessment

Difficulty
4/5
Estimated time
3-5 days
Newbie friendliness
48/100
Issue type
Bug
Clarity
Mostly clear
Activity status
Active
Tech stack
haskell
Domain
backend

Research direction

Start by locating the canonical float equality and map key equivalence implementation in the Haskell code, then inspect the related tests. The issue requires integers to remain unequal to floats, equal numeric values across float widths to compare equal, and signed zero and NaN significands to follow RFC 8949 §5.6.1. Add or update tests for these cases and run the relevant test suite.

Written by the indexing model from the issue text.

Description

According to the spec:

RFC8949 §2.

Note that integer and floating-point values are distinct in this
model, even if they have the same numeric value.

Also note that serialization variants are not visible at the generic
data model level. This deliberate absence of visibility includes the
number of bytes of the encoded floating-point value. It also
includes the choice of encoding for an "argument" (see Section 3)
such as the encoding for an integer, the encoding for the length of a
text or byte string, the encoding for the number of elements in an
array or pairs in a map, or the encoding for a tag number.

This means that integer types are always unequal to floating point numbers, even if numerically they would be equal. However floating point numbers with different bit widths can still be equal if they represent the same numerical value.

Additionally

RFC8949 §5.6.1.

Within each of these groups, numeric values are distinct unless they
are numerically equal (specifically, -0.0 is equal to 0.0); for the
purpose of map key equivalence, NaN values are equivalent if they
have the same significand after zero-extending both significands at
the right to 64 bits.

Dominant language
Haskell
Stars
8
Forks
5
Avg merge
3d 12h
Merged PRs (30d)
4

Getting set up

This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from IntersectMBO/cuddle

All issues in IntersectMBO/cuddle

Similar issues

More Haskell issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.