activitypub-federation-rust/src/fetch/webfinger.rs
Felix Ableitner 071218396b clippy
2023-10-24 10:54:16 +02:00

265 lines
9.4 KiB
Rust

use crate::{
config::Data,
error::{Error, Error::WebfingerResolveFailed},
fetch::{fetch_object_http_with_accept, object_id::ObjectId},
traits::{Actor, Object},
FEDERATION_CONTENT_TYPE,
};
use anyhow::anyhow;
use itertools::Itertools;
use regex::Regex;
use serde::{Deserialize, Serialize};
use std::collections::HashMap;
use tracing::debug;
use url::Url;
/// Takes an identifier of the form `name@example.com`, and returns an object of `Kind`.
///
/// For this the identifier is first resolved via webfinger protocol to an Activitypub ID. This ID
/// is then fetched using [ObjectId::dereference], and the result returned.
pub async fn webfinger_resolve_actor<T: Clone, Kind>(
identifier: &str,
data: &Data<T>,
) -> Result<Kind, <Kind as Object>::Error>
where
Kind: Object + Actor + Send + 'static + Object<DataType = T>,
for<'de2> <Kind as Object>::Kind: serde::Deserialize<'de2>,
<Kind as Object>::Error:
From<crate::error::Error> + From<anyhow::Error> + From<url::ParseError> + Send + Sync,
{
let (_, domain) = identifier
.splitn(2, '@')
.collect_tuple()
.ok_or(WebfingerResolveFailed)?;
let protocol = if data.config.debug { "http" } else { "https" };
let fetch_url =
format!("{protocol}://{domain}/.well-known/webfinger?resource=acct:{identifier}");
debug!("Fetching webfinger url: {}", &fetch_url);
let res: Webfinger =
fetch_object_http_with_accept(&Url::parse(&fetch_url)?, data, "application/jrd+json")
.await?;
debug_assert_eq!(res.subject, format!("acct:{identifier}"));
let links: Vec<Url> = res
.links
.iter()
.filter(|link| {
if let Some(type_) = &link.kind {
type_.starts_with("application/")
} else {
false
}
})
.filter_map(|l| l.href.clone())
.collect();
for l in links {
let object = ObjectId::<Kind>::from(l).dereference(data).await;
if object.is_ok() {
return object;
}
}
Err(WebfingerResolveFailed.into())
}
/// Extracts username from a webfinger resource parameter.
///
/// Use this method for your HTTP handler at `.well-known/webfinger` to handle incoming webfinger
/// request. For a parameter of the form `acct:gargron@mastodon.social` it returns `gargron`.
///
/// Returns an error if query doesn't match local domain.
///
///```
/// # use activitypub_federation::config::FederationConfig;
/// # use activitypub_federation::traits::tests::DbConnection;
/// # use activitypub_federation::fetch::webfinger::extract_webfinger_name;
/// # tokio::runtime::Runtime::new().unwrap().block_on(async {
/// # let db_connection = DbConnection;
/// let config = FederationConfig::builder()
/// .domain("example.com")
/// .app_data(db_connection)
/// .build()
/// .await
/// .unwrap();
/// let data = config.to_request_data();
/// let res = extract_webfinger_name("acct:test_user@example.com", &data).unwrap();
/// assert_eq!(res, "test_user");
/// # Ok::<(), anyhow::Error>(())
/// }).unwrap();
///```
pub fn extract_webfinger_name<T>(query: &str, data: &Data<T>) -> Result<String, Error>
where
T: Clone,
{
// Regex to extract usernames from webfinger query. Supports different alphabets using `\p{L}`.
// TODO: would be nice if we could implement this without regex and remove the dependency
let regex =
Regex::new(&format!(r"^acct:([\p{{L}}0-9_]+)@{}$", data.domain())).map_err(Error::other)?;
Ok(regex
.captures(query)
.and_then(|c| c.get(1))
.ok_or_else(|| Error::other(anyhow!("Webfinger regex failed to match")))?
.as_str()
.to_string())
}
/// Builds a basic webfinger response for the actor.
///
/// It assumes that the given URL is valid both to the view the actor in a browser as HTML, and
/// for fetching it over Activitypub with `activity+json`. This setup is commonly used for ease
/// of discovery.
///
/// ```
/// # use url::Url;
/// # use activitypub_federation::fetch::webfinger::build_webfinger_response;
/// let subject = "acct:nutomic@lemmy.ml".to_string();
/// let url = Url::parse("https://lemmy.ml/u/nutomic")?;
/// build_webfinger_response(subject, url);
/// # Ok::<(), anyhow::Error>(())
/// ```
pub fn build_webfinger_response(subject: String, url: Url) -> Webfinger {
build_webfinger_response_with_type(subject, vec![(url, None)])
}
/// Builds a webfinger response similar to `build_webfinger_response`. Use this when you want to
/// return multiple actors who share the same namespace and to specify the type of the actor.
///
/// `urls` takes a vector of tuples. The first item of the tuple is the URL while the second
/// item is the type, such as `"Person"` or `"Group"`. If `None` is passed for the type, the field
/// will be empty.
///
/// ```
/// # use url::Url;
/// # use activitypub_federation::fetch::webfinger::build_webfinger_response_with_type;
/// let subject = "acct:nutomic@lemmy.ml".to_string();
/// let user = Url::parse("https://lemmy.ml/u/nutomic")?;
/// let group = Url::parse("https://lemmy.ml/c/asklemmy")?;
/// build_webfinger_response_with_type(subject, vec![
/// (user, Some("Person")),
/// (group, Some("Group"))]);
/// # Ok::<(), anyhow::Error>(())
/// ```
pub fn build_webfinger_response_with_type(
subject: String,
urls: Vec<(Url, Option<&str>)>,
) -> Webfinger {
Webfinger {
subject,
links: urls.iter().fold(vec![], |mut acc, (url, kind)| {
let properties: HashMap<Url, String> = kind
.map(|kind| {
HashMap::from([(
"https://www.w3.org/ns/activitystreams#type"
.parse()
.expect("parse url"),
kind.to_string(),
)])
})
.unwrap_or_default();
let mut links = vec![
WebfingerLink {
rel: Some("http://webfinger.net/rel/profile-page".to_string()),
kind: Some("text/html".to_string()),
href: Some(url.clone()),
..Default::default()
},
WebfingerLink {
rel: Some("self".to_string()),
kind: Some(FEDERATION_CONTENT_TYPE.to_string()),
href: Some(url.clone()),
properties,
..Default::default()
},
];
acc.append(&mut links);
acc
}),
aliases: vec![],
properties: Default::default(),
}
}
/// A webfinger response with information about a `Person` or other type of actor.
#[derive(Serialize, Deserialize, Debug, Default)]
pub struct Webfinger {
/// The actor which is described here, for example `acct:LemmyDev@mastodon.social`
pub subject: String,
/// Links where further data about `subject` can be retrieved
pub links: Vec<WebfingerLink>,
/// Other Urls which identify the same actor as the `subject`
#[serde(default, skip_serializing_if = "Vec::is_empty")]
pub aliases: Vec<Url>,
/// Additional data about the subject
#[serde(default, skip_serializing_if = "HashMap::is_empty")]
pub properties: HashMap<Url, String>,
}
/// A single link included as part of a [Webfinger] response.
#[derive(Serialize, Deserialize, Debug, Default)]
pub struct WebfingerLink {
/// Relationship of the link, such as `self` or `http://webfinger.net/rel/profile-page`
pub rel: Option<String>,
/// Media type of the target resource
#[serde(rename = "type")]
pub kind: Option<String>,
/// Url pointing to the target resource
pub href: Option<Url>,
/// Used for remote follow external interaction url
pub template: Option<String>,
/// Additional data about the link
#[serde(default, skip_serializing_if = "HashMap::is_empty")]
pub properties: HashMap<Url, String>,
}
#[cfg(test)]
mod tests {
use super::*;
use crate::{
config::FederationConfig,
traits::tests::{DbConnection, DbUser},
};
#[tokio::test]
async fn test_webfinger() -> Result<(), Error> {
let config = FederationConfig::builder()
.domain("example.com")
.app_data(DbConnection)
.build()
.await
.unwrap();
let data = config.to_request_data();
webfinger_resolve_actor::<DbConnection, DbUser>("LemmyDev@mastodon.social", &data).await?;
// poa.st is as of 2023-07-14 the largest Pleroma instance
webfinger_resolve_actor::<DbConnection, DbUser>("graf@poa.st", &data).await?;
Ok(())
}
#[tokio::test]
async fn test_webfinger_extract_name() -> Result<(), Error> {
use crate::traits::tests::DbConnection;
let data = Data {
config: FederationConfig::builder()
.domain("example.com")
.app_data(DbConnection)
.build()
.await
.unwrap(),
request_counter: Default::default(),
};
assert_eq!(
Ok("test123".to_string()),
extract_webfinger_name("acct:test123@example.com", &data)
);
assert_eq!(
Ok("Владимир".to_string()),
extract_webfinger_name("acct:Владимир@example.com", &data)
);
assert_eq!(
Ok("تجريب".to_string()),
extract_webfinger_name("acct:تجريب@example.com", &data)
);
Ok(())
}
}