LemmyNet/lemmy
2
0
Fork 0
mirror of https://github.com/LemmyNet/lemmy.git synced 2025-01-03 08:41:34 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Some cleanup.

Browse source
This commit is contained in:
Dessalines 2024-12-03 23:08:31 -05:00
parent e33e72729c
commit c5e4430c38
1 changed files with 33 additions and 21 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

54
crates/api_common/src/request.rs
View file

@ -73,34 +73,46 @@ pub async fn fetch_link_metadata(url: &Url, context: &LemmyContext) -> LemmyResu
// then try to infer the content_type from the file extension. // then try to infer the content_type from the file extension.
.or(mime_guess::from_path(url.path()).first()); .or(mime_guess::from_path(url.path()).first());
let opengraph_data = 'ograph: { let opengraph_data = {
if let Some(c) = &content_type { let is_html = content_type
// application/xhtml+xml is a subset of HTML .as_ref()
let application_xhtml: Mime = "application/xhtml+xml".parse()?; .map(|c| {
if c.essence_str() == TEXT_HTML.essence_str() // application/xhtml+xml is a subset of HTML
|| c.essence_str() == application_xhtml.essence_str() let application_xhtml: Mime = "application/xhtml+xml".parse::<Mime>().unwrap_or(TEXT_HTML);
{ let allowed_mime_types = [TEXT_HTML.essence_str(), application_xhtml.essence_str()];
// Can't use .text() here, because it only checks the content header, not the actual bytes allowed_mime_types.contains(&c.essence_str())
// https://github.com/LemmyNet/lemmy/issues/1964 })
// So we want to do deep inspection of the actually returned bytes but need to be careful .unwrap_or_default();
// not spend too much time parsing binary data as HTML
// only take first bytes regardless of how many bytes the server returns if is_html {
let html_bytes = collect_bytes_until_limit(response, bytes_to_fetch).await?; // Can't use .text() here, because it only checks the content header, not the actual bytes
break 'ograph extract_opengraph_data(&html_bytes, url) // https://github.com/LemmyNet/lemmy/issues/1964
.map_err(|e| info!("{e}")) // So we want to do deep inspection of the actually returned bytes but need to be careful
.unwrap_or_default(); // not spend too much time parsing binary data as HTML
}
// If a server is serving `application/octet-stream`, it's likely a mistake, // only take first bytes regardless of how many bytes the server returns
// so we try to guess the file type from its magic number. let html_bytes = collect_bytes_until_limit(response, bytes_to_fetch).await?;
else if c.subtype() == "octet-stream" { extract_opengraph_data(&html_bytes, url)
.map_err(|e| info!("{e}"))
.unwrap_or_default()
} else {
let is_octet_type = content_type
.as_ref()
.map(|c| c.subtype() == "octet-stream")
.unwrap_or_default();
// Overwrite the content type if its an octet type
if is_octet_type {
// Don't need to fetch as much data for this as we do with opengraph // Don't need to fetch as much data for this as we do with opengraph
let octet_bytes = collect_bytes_until_limit(response, 512).await?; let octet_bytes = collect_bytes_until_limit(response, 512).await?;
// content_type = infer::get(&octet_bytes).or(&content_type, |t|
// t.mime_type().parse().ok());
content_type = content_type =
infer::get(&octet_bytes).map_or(content_type, |t| t.mime_type().parse().ok()); infer::get(&octet_bytes).map_or(content_type, |t| t.mime_type().parse().ok());
} }
Default::default()
} }
Default::default()
}; };
Ok(LinkMetadata { Ok(LinkMetadata {