deltachat/net/
http.rs

1//! # HTTP module.
2
3use anyhow::{Context as _, Result, anyhow, bail};
4use bytes::Bytes;
5use http_body_util::BodyExt;
6use hyper_util::rt::TokioIo;
7use mime::Mime;
8use serde::Serialize;
9use tokio::fs;
10
11use crate::blob::BlobObject;
12use crate::context::Context;
13use crate::log::warn;
14use crate::net::proxy::ProxyConfig;
15use crate::net::session::SessionStream;
16use crate::net::tls::wrap_rustls;
17use crate::tools::time;
18
19/// User-Agent for HTTP requests if a resource usage policy requires it.
20/// By default we do not set User-Agent.
21const USER_AGENT: &str = "chatmail/2 (+https://github.com/chatmail/core/)";
22
23/// HTTP(S) GET response.
24#[derive(Debug, Clone, PartialEq, Eq)]
25pub struct Response {
26    /// Response body.
27    pub blob: Vec<u8>,
28
29    /// MIME type extracted from the `Content-Type` header, if any.
30    pub mimetype: Option<String>,
31
32    /// Encoding extracted from the `Content-Type` header, if any.
33    pub encoding: Option<String>,
34}
35
36/// Retrieves the text contents of URL using HTTP GET request.
37pub async fn read_url(context: &Context, url: &str) -> Result<String> {
38    let response = read_url_blob(context, url).await?;
39    let text = String::from_utf8_lossy(&response.blob);
40    Ok(text.to_string())
41}
42
43async fn get_http_sender<B>(
44    context: &Context,
45    parsed_url: hyper::Uri,
46) -> Result<hyper::client::conn::http1::SendRequest<B>>
47where
48    B: hyper::body::Body + 'static + Send,
49    B::Data: Send,
50    B::Error: Into<Box<dyn std::error::Error + Send + Sync>>,
51{
52    let scheme = parsed_url.scheme_str().context("URL has no scheme")?;
53    let host = parsed_url.host().context("URL has no host")?;
54    let proxy_config_opt = ProxyConfig::load(context).await?;
55
56    let stream: Box<dyn SessionStream> = match scheme {
57        "http" => {
58            let port = parsed_url.port_u16().unwrap_or(80);
59
60            // It is safe to use cached IP addresses
61            // for HTTPS URLs, but for HTTP URLs
62            // better resolve from scratch each time to prevent
63            // cache poisoning attacks from having lasting effects.
64            let load_cache = false;
65            if let Some(proxy_config) = proxy_config_opt {
66                let proxy_stream = proxy_config
67                    .connect(context, host, port, load_cache)
68                    .await?;
69                Box::new(proxy_stream)
70            } else {
71                let tcp_stream = crate::net::connect_tcp(context, host, port, load_cache).await?;
72                Box::new(tcp_stream)
73            }
74        }
75        "https" => {
76            let port = parsed_url.port_u16().unwrap_or(443);
77            let (use_sni, load_cache) = (true, true);
78
79            if let Some(proxy_config) = proxy_config_opt {
80                let proxy_stream = proxy_config
81                    .connect(context, host, port, load_cache)
82                    .await?;
83                let tls_stream = wrap_rustls(
84                    host,
85                    port,
86                    use_sni,
87                    "",
88                    proxy_stream,
89                    &context.tls_session_store,
90                )
91                .await?;
92                Box::new(tls_stream)
93            } else {
94                let tcp_stream = crate::net::connect_tcp(context, host, port, load_cache).await?;
95                let tls_stream = wrap_rustls(
96                    host,
97                    port,
98                    use_sni,
99                    "",
100                    tcp_stream,
101                    &context.tls_session_store,
102                )
103                .await?;
104                Box::new(tls_stream)
105            }
106        }
107        _ => bail!("Unknown URL scheme"),
108    };
109
110    let io = TokioIo::new(stream);
111    let (sender, conn) = hyper::client::conn::http1::handshake(io).await?;
112    tokio::task::spawn(conn);
113
114    Ok(sender)
115}
116
117/// Converts the URL to expiration and stale timestamps.
118fn http_url_cache_timestamps(url: &str, mimetype: Option<&str>) -> (i64, i64) {
119    let now = time();
120
121    let expires = now + 3600 * 24 * 35;
122    let stale = if url.ends_with(".xdc") {
123        // WebXDCs are never stale, they just expire.
124        expires
125    } else if url.starts_with("https://tile.openstreetmap.org/")
126        || url.starts_with("https://vector.openstreetmap.org/")
127    {
128        // Policy at <https://operations.osmfoundation.org/policies/tiles/>
129        // requires that we cache tiles for at least 7 days.
130        // Do not revalidate earlier than that.
131        now + 3600 * 24 * 7
132    } else if mimetype.is_some_and(|s| s.starts_with("image/")) {
133        // Cache images for 1 day.
134        //
135        // As of 2024-12-12 WebXDC icons at <https://webxdc.org/apps/>
136        // use the same path for all app versions,
137        // so may change, but it is not critical if outdated icon is displayed.
138        now + 3600 * 24
139    } else {
140        // Revalidate everything else after 1 hour.
141        //
142        // This includes HTML, CSS and JS.
143        now + 3600
144    };
145    (expires, stale)
146}
147
148/// Places the binary into HTTP cache.
149async fn http_cache_put(context: &Context, url: &str, response: &Response) -> Result<()> {
150    let blob =
151        BlobObject::create_and_deduplicate_from_bytes(context, response.blob.as_slice(), "")?;
152
153    let (expires, stale) = http_url_cache_timestamps(url, response.mimetype.as_deref());
154    context
155        .sql
156        .insert(
157            "INSERT OR REPLACE INTO http_cache (url, expires, stale, blobname, mimetype, encoding)
158             VALUES (?, ?, ?, ?, ?, ?)",
159            (
160                url,
161                expires,
162                stale,
163                blob.as_name(),
164                response.mimetype.as_deref().unwrap_or_default(),
165                response.encoding.as_deref().unwrap_or_default(),
166            ),
167        )
168        .await?;
169
170    Ok(())
171}
172
173/// Retrieves the binary from HTTP cache.
174///
175/// Also returns if the response is stale and should be revalidated in the background.
176async fn http_cache_get(context: &Context, url: &str) -> Result<Option<(Response, bool)>> {
177    let now = time();
178    let Some((blob_name, mimetype, encoding, stale_timestamp)) = context
179        .sql
180        .query_row_optional(
181            "SELECT blobname, mimetype, encoding, stale
182             FROM http_cache WHERE url=? AND expires > ?",
183            (url, now),
184            |row| {
185                let blob_name: String = row.get(0)?;
186                let mimetype: Option<String> = Some(row.get(1)?).filter(|s: &String| !s.is_empty());
187                let encoding: Option<String> = Some(row.get(2)?).filter(|s: &String| !s.is_empty());
188                let stale_timestamp: i64 = row.get(3)?;
189                Ok((blob_name, mimetype, encoding, stale_timestamp))
190            },
191        )
192        .await?
193    else {
194        return Ok(None);
195    };
196    let is_stale = now > stale_timestamp;
197
198    let blob_object = BlobObject::from_name(context, &blob_name)?;
199    let blob_abs_path = blob_object.to_abs_path();
200    let blob = match fs::read(blob_abs_path)
201        .await
202        .with_context(|| format!("Failed to read blob for {url:?} cache entry."))
203    {
204        Ok(blob) => blob,
205        Err(err) => {
206            // This should not happen, but user may go into the blobdir and remove files,
207            // antivirus may delete the file or there may be a bug in housekeeping.
208            warn!(context, "{err:?}.");
209            return Ok(None);
210        }
211    };
212
213    let (expires, _stale) = http_url_cache_timestamps(url, mimetype.as_deref());
214    let response = Response {
215        blob,
216        mimetype,
217        encoding,
218    };
219
220    // Update expiration timestamp
221    // to prevent deletion of the file still in use.
222    //
223    // If the response is stale, the caller should revalidate it in the background, so update
224    // `stale` timestamp to avoid revalidating too frequently (and have many parallel revalidation
225    // tasks) if revalidation fails or the HTTP request takes some time. The stale period >= 1 hour,
226    // so 1 more minute won't be a problem.
227    let stale_timestamp = if is_stale { now + 60 } else { stale_timestamp };
228    context
229        .sql
230        .execute(
231            "UPDATE http_cache SET expires=?, stale=? WHERE url=?",
232            (expires, stale_timestamp, url),
233        )
234        .await?;
235
236    Ok(Some((response, is_stale)))
237}
238
239/// Removes expired cache entries.
240pub(crate) async fn http_cache_cleanup(context: &Context) -> Result<()> {
241    // Remove cache entries that are already expired
242    // or entries that will not expire in a year
243    // to make sure we don't have invalid timestamps that are way forward in the future.
244    context
245        .sql
246        .execute(
247            "DELETE FROM http_cache
248             WHERE ?1 > expires OR expires > ?1 + 31536000",
249            (time(),),
250        )
251        .await?;
252    Ok(())
253}
254
255/// Fetches URL and updates the cache.
256///
257/// URL is fetched regardless of whether there is an existing result in the cache.
258async fn fetch_url(context: &Context, original_url: &str) -> Result<Response> {
259    let mut url = original_url.to_string();
260
261    // Follow up to 10 http-redirects
262    for _i in 0..10 {
263        let parsed_url = url
264            .parse::<hyper::Uri>()
265            .with_context(|| format!("Failed to parse URL {url:?}"))?;
266
267        let mut sender = get_http_sender(context, parsed_url.clone()).await?;
268        let authority = parsed_url
269            .authority()
270            .context("URL has no authority")?
271            .clone();
272
273        let req = hyper::Request::builder().uri(parsed_url);
274
275        // OSM usage policy requires
276        // that User-Agent is set for HTTP GET requests
277        // to tile servers:
278        // <https://operations.osmfoundation.org/policies/tiles/>
279        // Same for vectory tiles
280        // at <https://operations.osmfoundation.org/policies/vector/>.
281        let req =
282            if authority == "tile.openstreetmap.org" || authority == "vector.openstreetmap.org" {
283                req.header("User-Agent", USER_AGENT)
284            } else {
285                req
286            };
287
288        let req = req
289            .header(hyper::header::HOST, authority.as_str())
290            .body(http_body_util::Empty::<Bytes>::new())?;
291        let response = sender.send_request(req).await?;
292
293        if response.status().is_redirection() {
294            let header = response
295                .headers()
296                .get_all("location")
297                .iter()
298                .next_back()
299                .ok_or_else(|| anyhow!("Redirection doesn't have a target location"))?
300                .to_str()?;
301            info!(context, "Following redirect to {}", header);
302            url = header.to_string();
303            continue;
304        }
305
306        if !response.status().is_success() {
307            return Err(anyhow!(
308                "The server returned a non-successful response code: {}{}",
309                response.status().as_u16(),
310                response
311                    .status()
312                    .canonical_reason()
313                    .map(|s| format!(" {s}"))
314                    .unwrap_or("".to_string())
315            ));
316        }
317
318        let content_type = response
319            .headers()
320            .get("content-type")
321            .and_then(|value| value.to_str().ok())
322            .and_then(|value| value.parse::<Mime>().ok());
323        let mimetype = content_type
324            .as_ref()
325            .map(|mime| mime.essence_str().to_string());
326        let encoding = content_type.as_ref().and_then(|mime| {
327            mime.get_param(mime::CHARSET)
328                .map(|charset| charset.as_str().to_string())
329        });
330        let body = response.collect().await?.to_bytes();
331        let blob: Vec<u8> = body.to_vec();
332        let response = Response {
333            blob,
334            mimetype,
335            encoding,
336        };
337        info!(context, "Inserting {original_url:?} into cache.");
338        http_cache_put(context, &url, &response).await?;
339        return Ok(response);
340    }
341
342    Err(anyhow!("Followed 10 redirections"))
343}
344
345/// Retrieves the binary contents of URL using HTTP GET request.
346pub async fn read_url_blob(context: &Context, url: &str) -> Result<Response> {
347    if let Some((response, is_stale)) = http_cache_get(context, url).await? {
348        info!(context, "Returning {url:?} from cache.");
349        if is_stale {
350            let context = context.clone();
351            let url = url.to_string();
352            tokio::spawn(async move {
353                // Fetch URL in background to update the cache.
354                info!(context, "Fetching stale {url:?} in background.");
355                if let Err(err) = fetch_url(&context, &url).await {
356                    warn!(context, "Failed to revalidate {url:?}: {err:#}.");
357                }
358            });
359        }
360        return Ok(response);
361    }
362
363    info!(context, "Not found {url:?} in cache, fetching.");
364    let response = fetch_url(context, url).await?;
365    Ok(response)
366}
367
368/// Sends an empty POST request to the URL.
369///
370/// Returns response text and whether request was successful or not.
371///
372/// Does not follow redirects.
373pub(crate) async fn post_empty(context: &Context, url: &str) -> Result<(String, bool)> {
374    let parsed_url = url
375        .parse::<hyper::Uri>()
376        .with_context(|| format!("Failed to parse URL {url:?}"))?;
377    let scheme = parsed_url.scheme_str().context("URL has no scheme")?;
378    if scheme != "https" {
379        bail!("POST requests to non-HTTPS URLs are not allowed");
380    }
381
382    let mut sender = get_http_sender(context, parsed_url.clone()).await?;
383    let authority = parsed_url
384        .authority()
385        .context("URL has no authority")?
386        .clone();
387    let req = hyper::Request::post(parsed_url)
388        .header(hyper::header::HOST, authority.as_str())
389        .body(http_body_util::Empty::<Bytes>::new())?;
390
391    let response = sender.send_request(req).await?;
392
393    let response_status = response.status();
394    let body = response.collect().await?.to_bytes();
395    let text = String::from_utf8_lossy(&body);
396    let response_text = text.to_string();
397
398    Ok((response_text, response_status.is_success()))
399}
400
401/// Posts string to the given URL.
402///
403/// Returns true if successful HTTP response code was returned.
404///
405/// Does not follow redirects.
406#[allow(dead_code)]
407pub(crate) async fn post_string(context: &Context, url: &str, body: String) -> Result<bool> {
408    let parsed_url = url
409        .parse::<hyper::Uri>()
410        .with_context(|| format!("Failed to parse URL {url:?}"))?;
411    let scheme = parsed_url.scheme_str().context("URL has no scheme")?;
412    if scheme != "https" {
413        bail!("POST requests to non-HTTPS URLs are not allowed");
414    }
415
416    let mut sender = get_http_sender(context, parsed_url.clone()).await?;
417    let authority = parsed_url
418        .authority()
419        .context("URL has no authority")?
420        .clone();
421
422    let request = hyper::Request::post(parsed_url)
423        .header(hyper::header::HOST, authority.as_str())
424        .body(body)?;
425    let response = sender.send_request(request).await?;
426
427    Ok(response.status().is_success())
428}
429
430/// Sends a POST request with x-www-form-urlencoded data.
431///
432/// Does not follow redirects.
433pub(crate) async fn post_form<T: Serialize + ?Sized>(
434    context: &Context,
435    url: &str,
436    form: &T,
437) -> Result<Bytes> {
438    let parsed_url = url
439        .parse::<hyper::Uri>()
440        .with_context(|| format!("Failed to parse URL {url:?}"))?;
441    let scheme = parsed_url.scheme_str().context("URL has no scheme")?;
442    if scheme != "https" {
443        bail!("POST requests to non-HTTPS URLs are not allowed");
444    }
445
446    let encoded_body = serde_urlencoded::to_string(form).context("Failed to encode data")?;
447    let mut sender = get_http_sender(context, parsed_url.clone()).await?;
448    let authority = parsed_url
449        .authority()
450        .context("URL has no authority")?
451        .clone();
452    let request = hyper::Request::post(parsed_url)
453        .header(hyper::header::HOST, authority.as_str())
454        .header("content-type", "application/x-www-form-urlencoded")
455        .body(encoded_body)?;
456    let response = sender.send_request(request).await?;
457    let bytes = response.collect().await?.to_bytes();
458    Ok(bytes)
459}
460
461#[cfg(test)]
462mod tests {
463    use super::*;
464    use std::time::Duration;
465
466    use crate::sql::housekeeping;
467    use crate::test_utils::TestContext;
468    use crate::tools::SystemTime;
469
470    #[tokio::test(flavor = "multi_thread", worker_threads = 2)]
471    async fn test_http_cache() -> Result<()> {
472        let t = &TestContext::new().await;
473
474        assert_eq!(http_cache_get(t, "https://webxdc.org/").await?, None);
475
476        let html_response = Response {
477            blob: b"<!DOCTYPE html> ...".to_vec(),
478            mimetype: Some("text/html".to_string()),
479            encoding: None,
480        };
481
482        let xdc_response = Response {
483            blob: b"PK...".to_vec(),
484            mimetype: Some("application/octet-stream".to_string()),
485            encoding: None,
486        };
487        let xdc_editor_url = "https://apps.testrun.org/webxdc-editor-v3.2.0.xdc";
488        let xdc_pixel_url = "https://apps.testrun.org/webxdc-pixel-v2.xdc";
489
490        http_cache_put(t, "https://webxdc.org/", &html_response).await?;
491
492        assert_eq!(http_cache_get(t, xdc_editor_url).await?, None);
493        assert_eq!(http_cache_get(t, xdc_pixel_url).await?, None);
494        assert_eq!(
495            http_cache_get(t, "https://webxdc.org/").await?,
496            Some((html_response.clone(), false))
497        );
498
499        http_cache_put(t, xdc_editor_url, &xdc_response).await?;
500        http_cache_put(t, xdc_pixel_url, &xdc_response).await?;
501        assert_eq!(
502            http_cache_get(t, xdc_editor_url).await?,
503            Some((xdc_response.clone(), false))
504        );
505        assert_eq!(
506            http_cache_get(t, xdc_pixel_url).await?,
507            Some((xdc_response.clone(), false))
508        );
509
510        assert_eq!(
511            http_cache_get(t, "https://webxdc.org/").await?,
512            Some((html_response.clone(), false))
513        );
514
515        // HTML is stale after 1 hour, but .xdc is not.
516        SystemTime::shift(Duration::from_secs(3600 + 100));
517        assert_eq!(
518            http_cache_get(t, "https://webxdc.org/").await?,
519            Some((html_response.clone(), true))
520        );
521        assert_eq!(
522            http_cache_get(t, xdc_editor_url).await?,
523            Some((xdc_response.clone(), false))
524        );
525
526        // Stale cache entry can be renewed
527        // even before housekeeping removes old one.
528        http_cache_put(t, "https://webxdc.org/", &html_response).await?;
529        assert_eq!(
530            http_cache_get(t, "https://webxdc.org/").await?,
531            Some((html_response.clone(), false))
532        );
533
534        // 35 days later pixel .xdc expires because we did not request it for 35 days and 1 hour.
535        // But editor is still there because we did not request it for just 35 days.
536        // We have not renewed the editor however, so it becomes stale.
537        SystemTime::shift(Duration::from_secs(3600 * 24 * 35 - 100));
538
539        // Run housekeeping to test that it does not delete the blob too early.
540        housekeeping(t).await?;
541
542        assert_eq!(
543            http_cache_get(t, xdc_editor_url).await?,
544            Some((xdc_response.clone(), true))
545        );
546        assert_eq!(http_cache_get(t, xdc_pixel_url).await?, None);
547
548        // If we get the blob the second time quickly, it shouldn't be stale because it's supposed
549        // that we've already run a revalidation task which will update the blob soon.
550        assert_eq!(
551            http_cache_get(t, xdc_editor_url).await?,
552            Some((xdc_response.clone(), false))
553        );
554        // But if the revalidation task hasn't succeeded after some time, the blob is stale again
555        // even if we continue to get it frequently.
556        for i in (0..100).rev() {
557            SystemTime::shift(Duration::from_secs(6));
558            if let Some((_, true)) = http_cache_get(t, xdc_editor_url).await? {
559                break;
560            }
561            assert!(i > 0);
562        }
563
564        // Test that if the file is accidentally removed from the blobdir,
565        // there is no error when trying to load the cache entry.
566        for entry in std::fs::read_dir(t.get_blobdir())? {
567            let entry = entry.unwrap();
568            let path = entry.path();
569            std::fs::remove_file(path).expect("Failed to remove blob");
570        }
571
572        assert_eq!(
573            http_cache_get(t, xdc_editor_url)
574                .await
575                .context("Failed to get no cache response")?,
576            None
577        );
578
579        Ok(())
580    }
581}
deltachat/net/http.rs

deltachat/net/
http.rs
