Struct Compress 

pub struct Compress { /* private fields */ }

Compress responses for clients that ask.

App::new()?.middleware(Compress::default())

JSON is the best case compression has: a list of a hundred users is mostly the same twenty key names repeated a hundred times, and typically shrinks by 70–80%. The cost is CPU on the server, which is why small bodies are left alone — below a kilobyte the headers outweigh the saving — and why a body that is already compressed (an image, a zip, anything with a Content-Encoding) is never touched.

gzip is preferred over deflate when the client accepts both, because browsers historically disagreed about whether “deflate” meant the zlib format or a raw stream, and gzip never had that problem. When deflate is what is asked for, the zlib framing is sent, which is what every modern client means by it.

A strong ETag on the response is weakened, because the compressed bytes are a different representation of the same resource and a strong tag promises byte-for-byte identity. The validator still works; it just says so honestly.

Implementations

impl Compress

pub fn new() -> Self

pub fn min_size(self, bytes: usize) -> Self

The size below which a body is not worth compressing.

Trait Implementations

impl Clone for Compress

fn clone(&self) -> Compress

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Copy for Compress

impl Debug for Compress

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for Compress

fn default() -> Self

Returns the “default value” for a type.

impl Middleware for Compress

fn handle(&self, request: Request, next: Next) -> BoxFuture<Response>