Automattic / Automattic/harper
`Expr` cannot match some `Optional` parts
- Dominant language
- Rust
- Stars
- 15.4k
- Forks
- 627
- Avg merge
- 1d 15h
- Merged PRs (30d)
- 106
Description
I'm working on a linter that needs a pattern with an optional arbitrary word between two set words:
"implement" + (optional word) + "into"
Surprisingly, even though this doesn't seem very different from what I've done several times in other linters, this one only succeeds when the optional word is present but fails when it's absent. Here's the `Expr`:
```rs
SequenceExpr::word_set(&[
"implement",
"implemented",
"implementing",
"implements",
])
.then_optional(
SequenceExpr::whitespace()
.then_any_word()
)
.t_ws()
.t_aco("into")
```
I also tried with the whitespace on the other side:
```rs
SequenceExpr::word_set(&[
"implement",
"implemented",
"implementing",
"implements",
])
.t_ws()
.then_optional(
SequenceExpr::any_word().t_ws()
)
.t_aco("into")
```
Either of these should match both "implement into" and "implement foobar into", but both only match the latter.
I've spent a few hours on this, thinking I'd just overlooked something in my linter. I've also tried an agent on it that often finds things and it hasn't hit it yet. So I think this is a real latent bug.
Contributor guide
Research direction
Start by locating the Rust implementation of SequenceExpr and reproduce the two expressions against both "implement into" and "implement foobar into". Trace how then_optional handles whitespace and word matching; done means both expressions match both inputs without breaking the existing behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- rust
- Domain
- devtools
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 58/100