Where is difference between #[bw] #[br] and #[brw] articulated in the docs?

[vi]

I see #[binread], #[binrw] and #[binwrite] attribute macros all link to binrw::docs::attribute table, which contains r/w column.
Also I see that derive(BinRead) things use #[br] and derive(BinWrite) things use #[bw] in examples.

Why they are distinct attributes? Shall this connection from distinct attribute or derive macros to derive helpers and to "r/w" column in the table be documented explicitly?

[csnover]

Attributes are distinct because reading and writing are distinct. Taking a u32 to/from usize, you need to usize::try_from for reading and u32::try_from for writing. Taking a Vec<T>, you need to take the vector length from the stream when reading and you need to take the length from v.len() when writing. Also, read values are owned and write values are borrowed.

Was there a specific thing you were trying to do that caused this confusion?