Small comments

[douweschulte]

Nice to see how this works. It looks very nicely contained and structured in this way. Also the opening of the files looks nice with this conversion to a BufReader.

Small comments:

• Technically speaking the residue sequence number is not unique, if coupled to the insertion code it is unique. This can be retrieved by using the Residue::id method. It is hard to return this type with wasm_bindgen, but be prepared to find a single PDB with duplicate residue sequence numbers. (Idea: just crash when you detect one of these, so it cannot go wrong in any other way.)
• I found a nice crate in the meantime: https://crates.io/crates/serde-wasm-bindgen, this allows any struct that implements Serde::Serialize to be passed to the Java/typescript side. So that would enable you to directly throw the full PDB to the other side, or use more intricate structures Vec<String> in your PDBInfo.

[douweschulte]

I hope you do not mind the uninvited commentary, I am just very interested in how this Wasm stuff works and if changes to pdbtbx are needed. And feel free to reach out (or tag me in an issue/PR) if you have questions about Rust stuff I would be happy to help.

[sverhoeven]

Comments are welcome,

• I just needed a list of valid numbers, for doing form validation, I do not yet care about uniqueness.
• I was a bit disappointed that wasm-bindgen does not support Vec<String>, but indeed serde seems the way to go according to https://rustwasm.github.io/docs/wasm-bindgen/reference/arbitrary-data-with-serde.html .