tonhowtf/omniget · error · anyhow::Error

Downloaded pdfium archive is too small

Error message

Downloaded pdfium archive is too small ({} bytes) — likely an error page

What it means

After a successful HTTP status, ensure_pdfium_with_variant() sanity-checks the downloaded byte count: archives smaller than 100,000 bytes are rejected because real pdfium archives are always far larger — a small payload is almost certainly an HTML error page or a stub response that slipped through with a 200 status.

Solutions

Inspect what the endpoint actually returns (curl the URL) — usually an HTML error/challenge page
Bypass or correctly configure the proxy/captive portal and retry
Use the global proxy configuration (apply_global_proxy) to point at a working proxy
Install pdfium from a local archive via set_pdfium_from_path instead of downloading

Defensive patterns

Strategy: try-catch

Validate before calling

let resp = client.get(&probe_url).send().await?;
let ct = resp.headers().get(reqwest::header::CONTENT_TYPE);
if let Some(ct) = ct {
    if ct.to_str().unwrap_or("").starts_with("text/html") {
        eprintln!("endpoint returned HTML — captive portal or error page");
    }
}

Try / catch

match ensure_pdfium().await {
    Err(e) if e.to_string().contains("too small") => {
        // network intercepted: switch network/proxy, then retry
    }
    other => other.map(|_| ()),
}

Prevention

Check Content-Type before trusting a 200 response from download endpoints
Detect and handle captive portals before first-run downloads
Fall back to set_pdfium_from_path with a bundled archive when downloads look intercepted

When it happens

Trigger: The server returns HTTP 200 but the body is tiny: a captive-portal/proxy HTML page, an API rate-limit JSON message, or a truncated download.

Common situations: Wi-Fi captive portals intercepting the first request; anti-bot/CDN challenge pages served with 200; misconfigured corporate proxies returning an error page; extremely truncated transfer.

Related errors

AI-assisted analysis of tonhowtf/omniget@8600b91f42 (2026-09-12). Data as JSON: /api/errors/4ca6bda5d1b6ca0b. Report an issue: GitHub.

Appendix: source

Thrown at src-tauri/omniget-core/src/core/pdfium.rs:188

    let target_path = target_dir.join(lib_filename);

    tracing::info!("Downloading pdfium from {}", url);

    let client = crate::core::http_client::apply_global_proxy(reqwest::Client::builder())
        .timeout(std::time::Duration::from_secs(600))
        .build()?;

    let response = client.get(&url).send().await?;
    if !response.status().is_success() {
        return Err(anyhow!(
            "Failed to download pdfium from {}: HTTP {}",
            url,
            response.status()
        ));
    }
    let bytes = response.bytes().await?.to_vec();
    if bytes.len() < 100_000 {
        return Err(anyhow!(
            "Downloaded pdfium archive is too small ({} bytes) — likely an error page",
            bytes.len()
        ));
    }

    let target_path_clone = target_path.clone();
    let target_dir_clone = target_dir.clone();
    let lib_name = lib_filename.to_string();
    let extracted_version =
        tokio::task::spawn_blocking(move || -> anyhow::Result<Option<String>> {
            extract_pdfium_archive(&bytes, &target_path_clone, &target_dir_clone, &lib_name)
        })
        .await
        .map_err(|e| anyhow!("spawn_blocking failed: {}", e))??;

    if let Some(version_marker) = pdfium_version_marker_path() {
        let base = extracted_version.unwrap_or_else(|| "latest".to_string());
        let value = format!("{} ({})", base, archive_name);

View on GitHub (pinned to 8600b91f42)