locust VS semantic-source

Nice!

Have been working on something similar, although my use case is more about learning how code has changed across git commits: https://github.com/bugout-dev/locust

For Javascript/Typescript/React support, like you, I hooked into the Babel toolchain. Can't recommend it highly enough.

There's also a newish project called quick-lint-js which seems to have written their own from-scratch AST parser for JS, but I haven't tried it yet: https://github.com/quick-lint/quick-lint-js

Finally, another project that I know in this space is comby (I believe it is owned/maintained by the folks at Sourcegraph): https://comby.dev/

Don't know why I dumped all those links there. Just figured there may be something useful in them for you. Am also just super passionate about building knowledge about code bases by analyzing their ASTs. Nice to meet a fellow enthusiast. :)

I maintain a free/open source project that does exactly what the author asks for: https://github.com/bugout-dev/locust.

Our tool uses git as the foundation of its functionality. It superimposes git diffs on top of ASTs.

It is insanely powerful.

For example, we use it to power semantic code search and current support Python, Javascript, and Java. We generate a JSON object defining the AST differences between initial and terminal commits on GitHub PRs and doing text search on the JSON objects performs surprisingly well when we want to answer questions like, "When did we add dateutils as a dependency?" or "When did we last change the /journals handler on the API?"

The Python integration currently sees the most use but if you are interested in other languages, we would be happy to support it.

Do drop me a DM if you want help getting started with Locust.

My team has a similar project (Locust: https://github.com/bugout-dev/locust) where the goal is to learn the semantic meanings of code changes in git commits, GitHub PRs, etc.

Since we took git diffs as a target for semantic analysis, we have a different approach to our diffs. We start with line-by-line diffs (specifically using "git diff") and then take a semantic diff by superimposing the git diff information on top of the initial and terminal ASTs.

This makes the diff calculation cheaper because we don't have to do full diff between trees.

Haven't updated the code in a few months, but my team is actively using Locust on public GitHub repos to learn the semantics of those code bases. We do plan to do some work on it soon to make it easier to make Locust easier to use (especially as a library).

Really need to sit down and take a proper look at tree-sitter. We currently support Locust diffs for Python, Javascript, and Java, but each one is custom written and implements the same basic algorithm. It looks like tree sitter might just crush this problem for us.

One production example I know: GitHub code navigation is written in Haskell https://github.com/github/semantic

ah, easy. it's because support has not been added into https://github.com/github/semantic which is the tech that powers the GitHub UI. Adding support is pretty easy/mainly glue code [1] that imports the tree sitter API.

[1] https://github.com/github/semantic/blob/793a876ae45d38a6bd17...

GitHub used Haskell for implementing Semantic, a command-line tool for parsing, analyzing, and comparing source code.

As far as engineering effort, you can read this GitHub comment for an overview of where we’d like to take the project in the future. The tl;dr here is that the open sum type view of the world made it very concise to fold over syntax trees (since such a view of data is ultimately unityped, recursion schemes Just Work), but the tradeoff thus associated—namely, that you have to parse a concrete syntax tree into an open-sum view (a complicated and painful-to-read process), that you can never really be sure how a given syntax tree is shaped, and that the types don’t help you nearly as much as they could—proved to be too onerous to deal with. Going forward, we’re generating syntax types from the AST once per target language, and working on an abstraction (probably via this generated code; I made five separate efforts at using Generics for this, and failed every time) that recovers at least some of the convenience of recursion schemes. It turns out that recursion schemes over a mutually recursive syntax tree—as pretty much every language’s syntax trees are, in practice—are pretty much an unsolved problem, especially when extended to languages like TypeScript, which have hundreds of different syntax nodes.

Meanwhile their Tree-Sitter-based semantic parser[1] looks abandoned. There is even rotting for years pull request[2] adding support of the same stack graphs into it.

[1] https://github.com/github/semantic

[2] https://github.com/github/semantic/pull/535

The semantic team at GitHub uses it for statically analyzing the dozens of languages that end up in GitHub repositories: https://github.com/github/semantic/blob/eaf13783838861fe5eb6cd46d59354774a8eb88d/docs/why-haskell.md

Yesterday I was looking for some examples of projects using tree-sitter (which is C) when I found GitHub's semantic, used to analyze and compare source code, and written in Haskell: https://github.com/github/semantic/