Struct Builder

source

pub struct Builder { /* private fields */ }

Expand description

A builder to configure a new Client.

§Example

use std::time::Duration;
use hyper_util::client::legacy::Client;
use hyper_util::rt::TokioExecutor;

let client = Client::builder(TokioExecutor::new())
    .pool_idle_timeout(Duration::from_secs(30))
    .http2_only(true)
    .build_http();

Implementations§

source §

impl Builder

source

pub fn new<E>(executor: E) -> Self
where E: Executor<Pin<Box<dyn Future<Output = ()> + Send>>> + Send + Sync + Clone + 'static,

Construct a new Builder.

source

pub fn pool_idle_timeout<D>(&mut self, val: D) -> &mut Self
where D: Into<Option<Duration>>,

Set an optional timeout for idle sockets being kept-alive. A Timer is required for this to take effect. See Builder::pool_timer

Pass None to disable timeout.

Default is 90 seconds.

§Example

use std::time::Duration;
use hyper_util::client::legacy::Client;
use hyper_util::rt::{TokioExecutor, TokioTimer};

let client = Client::builder(TokioExecutor::new())
    .pool_idle_timeout(Duration::from_secs(30))
    .pool_timer(TokioTimer::new())
    .build_http();

source

pub fn pool_max_idle_per_host(&mut self, max_idle: usize) -> &mut Self

Sets the maximum idle connection per host allowed in the pool.

Default is usize::MAX (no limit).

source

pub fn http1_read_buf_exact_size(&mut self, sz: usize) -> &mut Self

Sets the exact size of the read buffer to always use.

Note that setting this option unsets the http1_max_buf_size option.

Default is an adaptive read buffer.

source

pub fn http1_max_buf_size(&mut self, max: usize) -> &mut Self

Set the maximum buffer size for the connection.

Default is ~400kb.

Note that setting this option unsets the http1_read_exact_buf_size option.

§Panics

The minimum value allowed is 8192. This method panics if the passed max is less than the minimum.

source

pub fn http1_allow_spaces_after_header_name_in_responses( &mut self, val: bool, ) -> &mut Self

Set whether HTTP/1 connections will accept spaces between header names and the colon that follow them in responses.

Newline codepoints (\r and \n) will be transformed to spaces when parsing.

You probably don’t need this, here is what RFC 7230 Section 3.2.4. has to say about it:

No whitespace is allowed between the header field-name and colon. In the past, differences in the handling of such whitespace have led to security vulnerabilities in request routing and response handling. A server MUST reject any received request message that contains whitespace between a header field-name and colon with a response code of 400 (Bad Request). A proxy MUST remove any such whitespace from a response message before forwarding the message downstream.

Note that this setting does not affect HTTP/2.

Default is false.

source

pub fn http1_allow_obsolete_multiline_headers_in_responses( &mut self, val: bool, ) -> &mut Self

Set whether HTTP/1 connections will accept obsolete line folding for header values.

You probably don’t need this, here is what RFC 7230 Section 3.2.4. has to say about it:

A server that receives an obs-fold in a request message that is not within a message/http container MUST either reject the message by sending a 400 (Bad Request), preferably with a representation explaining that obsolete line folding is unacceptable, or replace each received obs-fold with one or more SP octets prior to interpreting the field value or forwarding the message downstream.

A proxy or gateway that receives an obs-fold in a response message that is not within a message/http container MUST either discard the message and replace it with a 502 (Bad Gateway) response, preferably with a representation explaining that unacceptable line folding was received, or replace each received obs-fold with one or more SP octets prior to interpreting the field value or forwarding the message downstream.

A user agent that receives an obs-fold in a response message that is not within a message/http container MUST replace each received obs-fold with one or more SP octets prior to interpreting the field value.

Note that this setting does not affect HTTP/2.

Default is false.

source

pub fn http1_ignore_invalid_headers_in_responses( &mut self, val: bool, ) -> &mut Builder

Sets whether invalid header lines should be silently ignored in HTTP/1 responses.

This mimics the behaviour of major browsers. You probably don’t want this. You should only want this if you are implementing a proxy whose main purpose is to sit in front of browsers whose users access arbitrary content which may be malformed, and they expect everything that works without the proxy to keep working with the proxy.

This option will prevent Hyper’s client from returning an error encountered when parsing a header, except if the error was caused by the character NUL (ASCII code 0), as Chrome specifically always reject those.

The ignorable errors are:

empty header names;
characters that are not allowed in header names, except for \0 and \r;
when allow_spaces_after_header_name_in_responses is not enabled, spaces and tabs between the header name and the colon;
missing colon between header name and colon;
characters that are not allowed in header values except for \0 and \r.

If an ignorable error is encountered, the parser tries to find the next line in the input to resume parsing the rest of the headers. An error will be emitted nonetheless if it finds \0 or a lone \r while looking for the next line.

source

pub fn http1_writev(&mut self, enabled: bool) -> &mut Builder

Set whether HTTP/1 connections should try to use vectored writes, or always flatten into a single buffer.

Note that setting this to false may mean more copies of body data, but may also improve performance when an IO transport doesn’t support vectored writes well, such as most TLS implementations.

Setting this to true will force hyper to use queued strategy which may eliminate unnecessary cloning on some TLS backends

Default is auto. In this mode hyper will try to guess which mode to use

source

pub fn http1_title_case_headers(&mut self, val: bool) -> &mut Self

Set whether HTTP/1 connections will write header names as title case at the socket level.

Note that this setting does not affect HTTP/2.

Default is false.

source

pub fn http1_preserve_header_case(&mut self, val: bool) -> &mut Self

Set whether to support preserving original header cases.

Currently, this will record the original cases received, and store them in a private extension on the Response. It will also look for and use such an extension in any provided Request.

Since the relevant extension is still private, there is no way to interact with the original cases. The only effect this can have now is to forward the cases in a proxy-like fashion.

Note that this setting does not affect HTTP/2.

Default is false.

source

pub fn http1_max_headers(&mut self, val: usize) -> &mut Self

Set the maximum number of headers.

When a response is received, the parser will reserve a buffer to store headers for optimal performance.

If client receives more headers than the buffer size, the error “message header too large” is returned.

The headers is allocated on the stack by default, which has higher performance. After setting this value, headers will be allocated in heap memory, that is, heap memory allocation will occur for each response, and there will be a performance drop of about 5%.

Note that this setting does not affect HTTP/2.

Default is 100.

source

pub fn http09_responses(&mut self, val: bool) -> &mut Self

Set whether HTTP/0.9 responses should be tolerated.

Default is false.

source

pub fn timer<M>(&mut self, timer: M) -> &mut Self
where M: Timer + Send + Sync + 'static,

Provide a timer to be used for h2

See the documentation of h2::client::Builder::timer for more details.

source

pub fn pool_timer<M>(&mut self, timer: M) -> &mut Self
where M: Timer + Clone + Send + Sync + 'static,

Provide a timer to be used for timeouts and intervals in connection pools.

source

pub fn retry_canceled_requests(&mut self, val: bool) -> &mut Self

Set whether to retry requests that get disrupted before ever starting to write.

This means a request that is queued, and gets given an idle, reused connection, and then encounters an error immediately as the idle connection was found to be unusable.

When this is set to false, the related ResponseFuture would instead resolve to an Error::Cancel.

Default is true.

source