udoprog / ptscan

Hello,

I came across

ptscan/lib/src/encoding.rs

Line 91 in 46a3a76

fn encode_utf16<B>(buf: &mut Vec<u8>, s: &str) -> anyhow::Result<()>

while searching for UTF-16LE encoding in encoding_rs crate and I don't quite understand how it works.

I would like to borrow this function to my project and it would really help if I can understand it better.

Thanks in advance :-)

Steal away, This is the function annotated with comments.

    fn encode_utf16<B>(buf: &mut Vec<u8>, s: &str) -> anyhow::Result<()>
    where
        // Uses the byteorder crate, but could as easily just use u16::to_le_bytes or u16::to_be_bytes
        // and copy the bytes. But that would make it harder to be generic. `byteorder::ByteOrder`
        // already provides that.
        B: byteorder::ByteOrder,
    {
        use std::iter;

        // This is the built-in str::encode_utf16 method which produces an iterator of u16's.
        for c in s.encode_utf16() {
            // Just a silly way to extend buf with two 0 bytes, today I
            // would use `[0, 0]` instead since it directly implements `IntoIterator`.
            buf.extend(iter::repeat(0x0).take(2));
            // Write to the two zeroed bytes we just allocated above in
            // the byte-order specified by the parameter `B`.
            let s = buf.len() - 2;
            B::write_u16(&mut buf[s..], c);
        }

        Ok(())
    }

Can you explain how `encode_utf16` works?