fix(pdf): return a conversion error when the PDF parser panics

src/formats/pdf.rs

@@ -12,6 +12,13 @@ use crate::error::ConvertError;
 use pdf_inspector::PdfError;
 
 pub fn to_markdown(bytes: &[u8]) -> Result<String, ConvertError> {
+    // A dependency panic must not unwind across a native binding and abort
+    // its host. No partial conversion state escapes this boundary.
+    std::panic::catch_unwind(|| convert_pdf(bytes))
+        .unwrap_or_else(|_| Err(ConvertError::malformed("PDF parser panicked during conversion")))
+}
+
+fn convert_pdf(bytes: &[u8]) -> Result<String, ConvertError> {
     let result = pdf_inspector::process_pdf_mem(bytes).map_err(map_error)?;
     if !result.pages_needing_ocr.is_empty() {
         // Detection samples content streams and over-reports short or

tests/pdf_panic.rs

@@ -0,0 +1,51 @@
+use anydoc::{ConvertError, Format, to_markdown_bytes};
+
+fn inline_image_pdf(properties: &[u8]) -> Vec<u8> {
+    let mut content = b"q 8 0 0 8 0 0 cm BI /W 8 /H 8 /BPC 8 ".to_vec();
+    content.extend_from_slice(properties);
+    content.extend_from_slice(b" ID ");
+    content.extend_from_slice(&[0; 64]);
+    content.extend_from_slice(b" EI Q\n");
+    let objects = [
+        b"<< /Type /Catalog /Pages 2 0 R >>".to_vec(),
+        b"<< /Type /Pages /Kids [3 0 R] /Count 1 >>".to_vec(),
+        b"<< /Type /Page /Parent 2 0 R /MediaBox [0 0 200 200] /Resources << >> /Contents 4 0 R >>"
+            .to_vec(),
+        [
+            format!("<< /Length {} >>\nstream\n", content.len()).into_bytes(),
+            content,
+            b"\nendstream".to_vec(),
+        ]
+        .concat(),
+    ];
+    let mut pdf = b"%PDF-1.4\n".to_vec();
+    let mut offsets = Vec::new();
+    for (index, object) in objects.iter().enumerate() {
+        offsets.push(pdf.len());
+        pdf.extend_from_slice(format!("{} 0 obj\n", index + 1).as_bytes());
+        pdf.extend_from_slice(object);
+        pdf.extend_from_slice(b"\nendobj\n");
+    }
+    let xref = pdf.len();
+    pdf.extend_from_slice(b"xref\n0 5\n0000000000 65535 f \n");
+    for offset in offsets {
+        pdf.extend_from_slice(format!("{offset:010} 00000 n \n").as_bytes());
+    }
+    pdf.extend_from_slice(
+        format!("trailer\n<< /Size 5 /Root 1 0 R >>\nstartxref\n{xref}\n%%EOF\n").as_bytes(),
+    );
+    pdf
+}
+
+#[test]
+fn inline_image_parser_failure_returns_a_conversion_error() {
+    let result = to_markdown_bytes(&inline_image_pdf(b"/G true"), Format::Pdf);
+    assert!(matches!(result, Err(ConvertError::Malformed { .. })));
+    assert!(to_markdown_bytes(b"first,second\n1,2\n", Format::Csv).is_ok());
+}
+
+#[test]
+fn inline_image_with_color_space_still_requests_ocr() {
+    let result = to_markdown_bytes(&inline_image_pdf(b"/CS /DeviceGray"), Format::Pdf);
+    assert!(matches!(result, Err(ConvertError::NeedsOcr { .. })));
+}