thegeneralist01/archivr
1
Fork 0
mirror of https://github.com/thegeneralist01/archivr synced 2026-07-21 18:55:36 +02:00
Code Issues Activity

fix(core): extract HTML title after font stripping, not before

Browse source 
SingleFile embeds fonts as base64 data URIs in <style> blocks in the
<head>, pushing the <title> tag to ~1.2 MB in the raw temp file.
The 256 KiB read window in extract_html_title missed it.

Font extraction rewrites the HTML in-place (2.4 MB → 1.26 MB) before
hashing, so the title is accessible at byte ~106 KB in that content.

Fix: add extract_html_title_str() that operates on a &str; call it on
the in-memory rewritten string after font extraction (server path).
CLI path (no font extraction) falls back to result.title as before.
This commit is contained in:
TheGeneralist 2026-07-19 13:15:35 +02:00
parent f9759c08a3
commit 331fe7fd61
Signed by: thegeneralist01
SSH key fingerprint: SHA256:pp9qddbCNmVNoSjevdvQvM5z0DHN7LTa8qBMbcMq/R4
2 changed files with 22 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

13
crates/archivr-core/src/capture.rs
View file

@ -1065,23 +1065,26 @@ pub fn perform_capture(
// Font extraction: rewrite the HTML in-place before hashing.
// Only runs when archive_id is known (server context). CLI passes
// None and keeps fonts embedded — no behaviour change for CLI.
let (html_hash, byte_size, extracted_fonts) =
let (html_hash, byte_size, extracted_fonts, html_title) =
if let Some(aid) = archive_id {
let content = fs::read_to_string(&temp_html)
.with_context(|| format!("failed to read {}", temp_html.display()))?;
let (rewritten, fonts) =
downloader::font_extractor::extract_and_rewrite(&content, store_path, aid)
.unwrap_or_else(|_| (content.clone(), vec![])); // non-fatal
// Extract title after font-stripping so the title tag is not buried
// behind multi-MB embedded font data that would exceed the 256 KiB window.
let title = downloader::singlefile::extract_html_title_str(&rewritten);
fs::write(&temp_html, rewritten.as_bytes())
.with_context(|| "failed to write rewritten HTML")?;
let size = rewritten.len() as i64;
let new_hash = crate::hash::hash_bytes(rewritten.as_bytes());
(new_hash, size, fonts)
(new_hash, size, fonts, title)
} else {
let size = fs::metadata(&temp_html)
.with_context(|| format!("failed to stat {}", temp_html.display()))?
.len() as i64;
(result.html_hash.clone(), size, vec![])
(result.html_hash.clone(), size, vec![], result.title.clone())
};
// 1. Move HTML to raw store (if this hash hasn't been seen before).
@ -1120,14 +1123,14 @@ pub fn perform_capture(
// 4. Create the entry + primary_media artifact.
// Strip mirror branding from title when fetched via Freedium.
let entry_title = if is_freedium_fetch {
result.title.as_deref().map(|t| {
html_title.as_deref().map(|t| {
t.trim_end_matches(" - Freedium")
.trim_end_matches(" \u{2014} Freedium")
.trim()
.to_string()
})
} else {
result.title
html_title
};
let entry = record_media_entry(
&conn,

15
crates/archivr-core/src/downloader/singlefile.rs
View file

@ -691,7 +691,20 @@ fn extract_html_title(path: &Path) -> Option<String> {
let mut f = std::fs::File::open(path).ok()?;
let mut buf = Vec::new();
f.take(256 * 1024).read_to_end(&mut buf).ok()?;
let lower = String::from_utf8_lossy(&buf).to_ascii_lowercase();
extract_html_title_from_buf(&buf)
}
/// Extracts the `<title>` content from an HTML string.
///
/// Used in the server capture path where the font-extracted HTML is already
/// in memory — avoids a re-read and operates on the smaller post-extraction
/// content where the title is guaranteed to be within range.
pub fn extract_html_title_str(html: &str) -> Option<String> {
extract_html_title_from_buf(html.as_bytes())
}
fn extract_html_title_from_buf(buf: &[u8]) -> Option<String> {
let lower = String::from_utf8_lossy(buf).to_ascii_lowercase();
let start = lower.find("<title>")? + "<title>".len();
let end = lower[start..].find("</title>")? + start;
let title = String::from_utf8_lossy(&buf[start..end])