test: add extract_pdf_text unit tests (DOC-FULL-023)

- pdf_with_text_stream helper: build minimal PDFs with BT/ET text
- single/multiple text blocks, no BT/ET→empty, empty PDF→empty
- MAX_SCAN_BYTES limit respected, empty text filtered
- page_count from pages dict, default to 1

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

crates/zx_document_core/src/pdf.rs

@@ -344,6 +344,104 @@ mod tests {
         }
     }
 
+    fn pdf_with_text_stream(text_content: &str) -> Vec<u8> {
+        // Build a minimal PDF with a stream object containing BT/ET text
+        let stream_text = format!("BT\n({})\nET", text_content);
+        let obj = format!(
+            "1 0 obj\n<< /Length {} >>\nstream\n{}\nendstream\nendobj\n",
+            stream_text.len(),
+            stream_text
+        );
+        let obj_bytes = obj.as_bytes();
+        let header = b"%PDF-1.4\n%\x80\x80\x80\x80\n";
+        let header_len = header.len() as u64;
+        let obj_len = obj_bytes.len() as u64;
+        let trailer = format!(
+            "xref\n0 2\n0000000000 65535 f \n{:010} 00000 n \ntrailer\n<< /Size 2 /Root 1 0 R >>\nstartxref\n{}\n%%EOF\n",
+            header_len,
+            header_len + obj_len
+        );
+
+        let mut pdf = Vec::new();
+        pdf.extend_from_slice(header);
+        pdf.extend_from_slice(obj_bytes);
+        pdf.extend_from_slice(trailer.as_bytes());
+        pdf
+    }
+
+    #[test]
+    fn test_extract_pdf_text_single_bt_et_block() {
+        let pdf = pdf_with_text_stream("Hello World");
+        let path = write_temp(&pdf, "text_single.pdf");
+        let pages = extract_pdf_text(&path).unwrap();
+        assert_eq!(pages.len(), 1);
+        assert_eq!(pages[0].text, "Hello World");
+    }
+
+    #[test]
+    fn test_extract_pdf_text_multiple_blocks() {
+        let stream = "BT\n(First line)\nET\nBT\n(Second line)\nET";
+        let obj = format!("1 0 obj\n<< /Length {} >>\nstream\n{}\nendstream\nendobj\n", stream.len(), stream);
+        let header = b"%PDF-1.4\n";
+        let pdf: Vec<u8> = [header.as_slice(), obj.as_bytes()].concat();
+        // This is a slightly malformed PDF but we're testing the BT/ET scanner
+        let path = write_temp(&pdf, "text_multi.pdf");
+        let pages = extract_pdf_text(&path).unwrap();
+        // The scanner finds bt\n...\net blocks in sequence
+        assert!(pages.len() >= 1, "Should find at least one text block");
+    }
+
+    #[test]
+    fn test_extract_pdf_text_no_bt_et_returns_empty() {
+        let pdf = b"%PDF-1.4\nNo text blocks here\n%%EOF".to_vec();
+        let path = write_temp(&pdf, "no_text.pdf");
+        let pages = extract_pdf_text(&path).unwrap();
+        assert!(pages.is_empty());
+    }
+
+    #[test]
+    fn test_extract_pdf_text_empty_pdf_returns_empty() {
+        let pdf = b"%PDF-1.4\n".to_vec();
+        let path = write_temp(&pdf, "empty.pdf");
+        let pages = extract_pdf_text(&path).unwrap();
+        assert!(pages.is_empty());
+    }
+
+    #[test]
+    fn test_extract_pdf_text_max_scan_limit_respected() {
+        // Create a PDF larger than MAX_SCAN_BYTES (8MB) by padding zeros
+        let mut pdf = pdf_with_text_stream("Hello");
+        pdf.resize(MAX_SCAN_BYTES + 1024, 0);
+        let path = write_temp(&pdf, "large.pdf");
+        let result = extract_pdf_text(&path);
+        assert!(result.is_ok());
+    }
+
+    #[test]
+    fn test_extract_pdf_text_empty_text_filtered() {
+        let pdf = pdf_with_text_stream("");
+        let path = write_temp(&pdf, "empty_text.pdf");
+        let pages = extract_pdf_text(&path).unwrap();
+        // Empty text blocks are filtered out
+        assert!(pages.is_empty());
+    }
+
+    #[test]
+    fn test_extract_pdf_text_page_count_from_pages_dict() {
+        let data = minimal_pdf(3);
+        let path = write_temp(&data, "count_pages.pdf");
+        let meta = read_pdf_metadata(&path).unwrap();
+        assert_eq!(meta.page_count, 3);
+    }
+
+    #[test]
+    fn test_extract_pdf_text_default_to_one_page() {
+        let pdf = b"%PDF-1.4\n".to_vec();
+        let path = write_temp(&pdf, "default_page.pdf");
+        let meta = read_pdf_metadata(&path).unwrap();
+        assert_eq!(meta.page_count, 1);
+    }
+
     #[test]
     fn test_fixture_scanned_pdf() {
         let root = std::path::Path::new(env!("CARGO_MANIFEST_DIR"))