isahc/redirect.rs
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353
use crate::{
auth::Authentication,
body::AsyncBody,
config::{request::RequestConfig, RedirectPolicy},
error::{Error, ErrorKind},
handler::RequestBody,
interceptor::{Context, Interceptor, InterceptorFuture},
request::RequestExt,
};
use http::{header::ToStrError, uri::Scheme, HeaderMap, HeaderValue, Request, Response, Uri};
use std::{borrow::Cow, convert::TryFrom, fmt::Write, str};
use url::Url;
/// How many redirects to follow by default if a limit is not specified. We
/// don't actually allow infinite redirects as that could result in a dangerous
/// infinite loop, so by default we actually limit redirects to a large amount.
const DEFAULT_REDIRECT_LIMIT: u32 = 1024;
/// Extension containing the final "effective" URI that was visited, after
/// following any redirects.
pub(crate) struct EffectiveUri(pub(crate) Uri);
/// Interceptor that implements automatic following of HTTP redirects.
pub(crate) struct RedirectInterceptor;
impl Interceptor for RedirectInterceptor {
type Err = Error;
fn intercept<'a>(
&'a self,
mut request: Request<AsyncBody>,
ctx: Context<'a>,
) -> InterceptorFuture<'a, Self::Err> {
Box::pin(async move {
// Store the effective URI to include in the response.
let mut effective_uri = request.uri().clone();
// Get the redirect policy for this request.
let policy = request
.extensions()
.get::<RequestConfig>()
.and_then(|config| config.redirect_policy.as_ref())
.cloned()
.unwrap_or_default();
// No redirect handling, just proceed normally.
if policy == RedirectPolicy::None {
let mut response = ctx.send(request).await?;
response
.extensions_mut()
.insert(EffectiveUri(effective_uri));
return Ok(response);
}
let auto_referer = request
.extensions()
.get::<RequestConfig>()
.and_then(|config| config.auto_referer)
.unwrap_or(false);
let limit = match policy {
RedirectPolicy::Limit(limit) => limit,
_ => DEFAULT_REDIRECT_LIMIT,
};
// Keep track of how many redirects we've done.
let mut redirect_count: u32 = 0;
loop {
// Preserve a clone of the request before sending it.
let mut request_builder = request.to_builder();
// Send the request to get the ball rolling.
let mut response = ctx.send(request).await?;
// Check for a redirect.
if let Some(redirect_location) = get_redirect_location(&effective_uri, &response) {
// If we've reached the limit, return an error as requested.
if redirect_count >= limit {
return Err(Error::with_response(ErrorKind::TooManyRedirects, &response));
}
// Set referer header.
if auto_referer {
if let Some(referer) = create_referer(&effective_uri, &redirect_location) {
if let Some(headers) = request_builder.headers_mut() {
headers.insert(http::header::REFERER, referer);
}
}
}
// Check if we should change the request method into a GET. HTTP
// specs don't really say one way or another when this should
// happen for most status codes, so we just mimic curl's
// behavior here since it is so common.
if response.status() == 301
|| response.status() == 302
|| response.status() == 303
{
request_builder = request_builder.method(http::Method::GET);
}
// If we are redirecting to a different authority, scrub
// sensitive headers from subsequent requests.
if !is_same_authority(&effective_uri, &redirect_location) {
if let Some(headers) = request_builder.headers_mut() {
scrub_sensitive_headers(headers);
}
// Remove auth configuration.
if let Some(extensions) = request_builder.extensions_mut() {
extensions.remove::<Authentication>();
}
}
// Grab the request body back from the internal handler, as we
// might need to send it again (if possible...)
let mut request_body = response
.extensions_mut()
.remove::<RequestBody>()
.map(|v| v.0)
.unwrap_or_default();
// Redirect handling is tricky when we are uploading something.
// If we can, reset the body stream to the beginning. This might
// work if the body to upload is an in-memory byte buffer, but
// for arbitrary streams we can't do this.
//
// There's not really a good way of handling this gracefully, so
// we just return an error so that the user knows about it.
if !request_body.reset() {
return Err(Error::with_response(
ErrorKind::RequestBodyNotRewindable,
&response,
));
}
// Update the request to point to the new URI.
effective_uri = redirect_location.clone();
request = request_builder
.uri(redirect_location)
.body(request_body)
.map_err(|e| Error::new(ErrorKind::InvalidRequest, e))?;
redirect_count += 1;
}
// No more redirects; set the effective URI we finally settled on and return.
else {
response
.extensions_mut()
.insert(EffectiveUri(effective_uri));
return Ok(response);
}
}
})
}
}
fn get_redirect_location<T>(request_uri: &Uri, response: &Response<T>) -> Option<Uri> {
if response.status().is_redirection() {
let location = response.headers().get(http::header::LOCATION)?;
match parse_location(location) {
Ok(location) => match resolve(request_uri, location.as_ref()) {
Ok(uri) => return Some(uri),
Err(e) => {
tracing::debug!("invalid redirect location: {}", e);
}
},
Err(e) => {
tracing::debug!("invalid redirect location: {}", e);
}
}
}
None
}
/// Parse the given `Location` header value into a string.
fn parse_location(location: &HeaderValue) -> Result<Cow<'_, str>, ToStrError> {
match location.to_str() {
// This will return a `str` if the header value contained only legal
// US-ASCII characters.
Ok(s) => Ok(Cow::Borrowed(s)),
// Try to parse the location as UTF-8 bytes instead of US-ASCII. This is
// technically against the URI spec which requires all literal
// characters in the URI to be US-ASCII (see [RFC 3986, Section
// 4.1](https://tools.ietf.org/html/rfc3986#section-4.1)).
//
// This is also more or less against the HTTP spec, which historically
// allowed for ISO-8859-1 text in header values but since was restricted
// to US-ASCII plus opaque bytes. Never has UTF-8 been encouraged or
// allowed as-such. See [RFC 7230, Section
// 3.2.4](https://tools.ietf.org/html/rfc7230#section-3.2.4) for more
// info.
//
// However, some bad or misconfigured web servers will do this anyway,
// and most web browsers recover from this by allowing and interpreting
// UTF-8 characters as themselves even though they _should_ have been
// percent-encoded. The third-party URI parsers that we use have no such
// leniency, so we percent-encode such bytes (if legal UTF-8) ahead of
// time before handing them off to the URI parser.
Err(e) => {
if str::from_utf8(location.as_bytes()).is_ok() {
let mut s = String::with_capacity(location.len());
for &byte in location.as_bytes() {
if byte.is_ascii() {
s.push(byte as char);
} else {
write!(&mut s, "%{:02x}", byte).unwrap();
}
}
Ok(Cow::Owned(s))
} else {
// Header value isn't legal UTF-8 either, not much we can
// reasonably do.
Err(e)
}
}
}
}
/// Resolve one URI in terms of another.
fn resolve(base: &Uri, target: &str) -> Result<Uri, Box<dyn std::error::Error>> {
// Optimistically check if this is an absolute URI.
match Url::parse(target) {
Ok(url) => Ok(Uri::try_from(url.as_str())?),
// Relative URI, resolve against the base.
Err(url::ParseError::RelativeUrlWithoutBase) => {
let base = Url::parse(base.to_string().as_str())?;
Ok(Uri::try_from(base.join(target)?.as_str())?)
}
Err(e) => Err(Box::new(e)),
}
}
/// Create a `Referer` header value to include in a redirected request, if
/// possible and appropriate.
fn create_referer(uri: &Uri, target_uri: &Uri) -> Option<HeaderValue> {
// Do not set a Referer header if redirecting to an insecure location from a
// secure one.
if uri.scheme() == Some(&Scheme::HTTPS) && target_uri.scheme() != Some(&Scheme::HTTPS) {
return None;
}
let mut referer = String::new();
if let Some(scheme) = uri.scheme() {
referer.push_str(scheme.as_str());
referer.push_str("://");
}
if let Some(authority) = uri.authority() {
referer.push_str(authority.host());
if let Some(port) = authority.port() {
referer.push(':');
referer.push_str(port.as_str());
}
}
referer.push_str(uri.path());
if let Some(query) = uri.query() {
referer.push('?');
referer.push_str(query);
}
HeaderValue::try_from(referer).ok()
}
fn is_same_authority(a: &Uri, b: &Uri) -> bool {
a.scheme() == b.scheme() && a.host() == b.host() && a.port() == b.port()
}
fn scrub_sensitive_headers(headers: &mut HeaderMap) {
headers.remove(http::header::AUTHORIZATION);
headers.remove(http::header::COOKIE);
headers.remove("cookie2");
headers.remove(http::header::PROXY_AUTHORIZATION);
headers.remove(http::header::WWW_AUTHENTICATE);
}
#[cfg(test)]
mod tests {
use http::Response;
use test_case::test_case;
#[test_case("http://foo.com", "http://foo.com", "http://foo.com/")]
#[test_case("http://foo.com", "/two", "http://foo.com/two")]
#[test_case("http://foo.com", "http://foo.com#foo", "http://foo.com/")]
fn resolve_redirect_location(request_uri: &str, location: &str, resolved: &str) {
let response = Response::builder()
.status(301)
.header("Location", location)
.body(())
.unwrap();
assert_eq!(
super::get_redirect_location(&request_uri.parse().unwrap(), &response)
.unwrap()
.to_string(),
resolved
);
}
#[test_case(
"http://example.org/Overview.html",
"http://example.org/Overview.html",
Some("http://example.org/Overview.html")
)]
#[test_case(
"http://example.org/#heading",
"http://example.org/#heading",
Some("http://example.org/")
)]
#[test_case(
"http://user:pass@example.org",
"http://user:pass@example.org",
Some("http://example.org/")
)]
#[test_case("https://example.com", "http://example.org", None)]
fn create_referer_from_uri(uri: &str, target_uri: &str, referer: Option<&str>) {
assert_eq!(
super::create_referer(&uri.parse().unwrap(), &target_uri.parse().unwrap())
.as_ref()
.and_then(|value| value.to_str().ok()),
referer
);
}
#[test_case("http://example.com", "http://example.com", true)]
#[test_case("http://example.com", "http://example.com/foo", true)]
#[test_case("http://example.com", "http://user:pass@example.com", true)]
#[test_case("http://example.com", "http://example.com:9000", false)]
#[test_case("http://example.com:9000", "http://example.com:9000", true)]
#[test_case("http://example.com", "http://example.org", false)]
#[test_case("http://example.com", "https://example.com", false)]
#[test_case("http://example.com", "http://www.example.com", false)]
fn is_same_authority(a: &str, b: &str, expected: bool) {
assert_eq!(
super::is_same_authority(&a.parse().unwrap(), &b.parse().unwrap()),
expected
);
}
}