Cooper Pierce

Looks like this arises because we lower `@expr` to `__builtin__at(expr)`, so the matcher sees it as a function call. We could probably resolve this by translating it to an `OtherExpr`...

This seems like its a segfault inside of a tree sitter parser. Partial backtrace: ``` Process 39066 launched: '/Users/cooper/.opam/4.14.0/bin/semgrep-core' (arm64) Process 39066 stopped * thread #1, queue = 'com.apple.main-thread', stop...

This seems like an upstream problem; I'm able to reproduce the segfault by running the tree sitter parser against the target file directly.

The easiest workaround I can think of is to simply add a newline to the end of the Dockerfile. Could you share information about the machine(s) you're encountering this on?...

Relevant upstream issues: - https://github.com/camdencheek/tree-sitter-dockerfile/issues/22 - https://github.com/tree-sitter/tree-sitter/issues/1842

My initial reaction, without looking into this, is that it is unlikely to be the exact same error. That error message is just communicating that semgrep was expecting more output...

While our C support may currently be a bit lacking still, I believe this should be enabled by a combination of #5621 and using `paths > include` (to include `*.sc`...

I believe most use cases listed should be addressed by #5621 --- some are not, e.g., Jupyter Notebook, but we've got a separate issue for that one.

This is possible with the rule schema; it would be something like: ```yml pattern-not: type: object properties: pattern-not: title: Do not return finding where Semgrep pattern matches exactly oneOf: -...

@ievans I would be reasonably confident in asserting that's probably a different underlying issue. If the semgrep-core binary aborts abnormally with the CLI is expecting output, you'll see this same...