v0.22.0 of Recife ( https://github.com/pfeodrippe/recife ), a Clojure model checker (using the TLA+/TLC engine) Now, instead of creating a new shell process per invocation (we were doing it because TLC has lots of globals... it was just easier to have a new JVM), it runs locally in the same JVM! This makes startup much much faster and it's REPL-friendly. • Updated with latest TLC • Improve model checking perf (~2.8x)
I updated my zero dependencies Clojure csv reader with ClojureScript (works in both Node and Browser envs) support. Hope someone finds it useful: https://github.com/oneness/csvx. And blogged about it here as well: https://www.birkey.co/2025-12-28-once-csv-parser-to-rule-them-all.html.
https://github.com/oneness/csvx/blob/master/src/csvx/core.cljc#L17-L34 why did you use explicit conditional readers if the implementation is identical? https://github.com/oneness/csvx/blob/master/src/csvx/core.cljc#L12-L13 why do you need to have a safe guard for maximum number of lines?
@delaguardo Thanks for checking it out. For your first question, Yes, I can refactor to use one impl by just using explicit conditional readers where I do the type hints for JVM. As for second questions, it is to control how many lines to read so I can experiment in the REPL more and I like putting bounds to resource hungry code. 🙂 . Again, thanks for your comments and PR is welcome and pretty easy: 1. git checkout 2. change the code 3. run tests and push Have fund coding!