//! Regression tests for inline images whose unfiltered samples contain `"EI "` (#3107). use std::fmt::Write as _; use std::sync::Arc; use pdfcraft_render::{PageRenderer, RenderConfig, RenderRequest, RequestKind}; fn build_pdf(pages: &[Vec], extra_objects: &[(usize, Vec)], page_xobjects: &str) -> Vec { let n = pages.len(); let font = 3 + 2 * n; let mut objs: Vec<(usize, Vec)> = Vec::new(); objs.push((1, b"<< /Type /Catalog /Pages 2 0 R >>".to_vec())); let kids: Vec = (0..n).map(|i| format!("{} 0 R", 3 + 2 * i)).collect(); objs.push((2, format!("<< /Type /Pages /Kids [{}] /Count {n} >>", kids.join(" ")).into_bytes())); for (i, content) in pages.iter().enumerate() { let page_num = 3 + 2 * i; let stream_num = 4 + 2 * i; objs.push(( page_num, format!("<< /Type /Page /Parent 2 0 R /MediaBox [0 0 612 792] /Contents {stream_num} 0 R /Resources << /Font << /F1 {font} 0 R >> {page_xobjects} >> >>") .into_bytes(), )); let mut stm = format!("<< /Length {} >>\nstream\n", content.len()).into_bytes(); stm.extend_from_slice(content); stm.extend_from_slice(b"\nendstream"); objs.push((stream_num, stm)); } objs.push((font, b"<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica /Encoding /WinAnsiEncoding >>".to_vec())); for (id, body) in extra_objects { objs.push((*id, body.clone())); } objs.sort_by_key(|(id, _)| *id); let mut out = b"%PDF-1.7\n".to_vec(); let mut offs = Vec::with_capacity(objs.len()); for (id, body) in &objs { offs.push(out.len()); out.extend_from_slice(format!("{id} 0 obj\n").as_bytes()); out.extend_from_slice(body); out.extend_from_slice(b"\nendobj\n"); } let xref_offset = out.len(); let size = objs.len() + 1; out.extend_from_slice(format!("xref\n0 {size}\n0000000000 65535 f \n").as_bytes()); for off in offs { out.extend_from_slice(format!("{off:010} 00000 n \n").as_bytes()); } out.extend_from_slice(format!("trailer\n<< /Size {size} /Root 1 0 R >>\nstartxref\n{xref_offset}\n%%EOF\n").as_bytes()); out } fn page_stream(image: &[u8], n: usize) -> Vec { let mut s = b"q 200 0 0 200 100 300 cm ".to_vec(); s.extend_from_slice(image); s.extend_from_slice(format!(" Q BT /F1 18 Tf 72 700 Td (After image P{n}) Tj ET").as_bytes()); s } fn mean_grey_in_box(rgba: &[u8], width: u32, x0: u32, y0: u32, x1: u32, y1: u32) -> f64 { let mut sum = 0u64; let mut count = 0u64; for y in y0..y1 { for x in x0..x1 { let idx = ((y * width + x) * 4) as usize; sum += u64::from(rgba[idx]); count += 1; } } sum as f64 / count.max(1) as f64 } fn dark_pixels_in_box(rgba: &[u8], width: u32, x0: u32, y0: u32, x1: u32, y1: u32) -> usize { let mut dark = 0usize; for y in y0..y1 { for x in x0..x1 { let idx = ((y * width + x) * 4) as usize; if rgba[idx] < 128 && rgba[idx + 1] < 128 && rgba[idx + 2] < 128 { dark += 1; } } } dark } #[test] fn unfiltered_inline_image_samples_containing_ei_are_not_executed_as_content() { let samples = b"0EI Q BT /F1 18 Tf 72 600 Td (Injected) Tj ET q"; let w = samples.len(); let mut hexed = String::with_capacity(w * 2 + 1); for b in samples { let _ = write!(hexed, "{b:02X}"); } hexed.push('>'); let mut img1 = format!("BI /W {w} /H 1 /BPC 8 /CS /G /L {w} ID ").into_bytes(); img1.extend_from_slice(samples); img1.extend_from_slice(b" EI"); let mut img2 = format!("BI /W {w} /H 1 /BPC 8 /CS /G /L {w} ID 0AB").into_bytes(); img2.extend_from_slice(&samples[3..]); img2.extend_from_slice(b" EI"); let img3 = format!("BI /W {w} /H 1 /BPC 8 /CS /G /F /AHx ID {hexed} EI").into_bytes(); let img4 = b"BI /W 4 /H 1 /BPC 8 /CS /G ID \x20\x60\xa0\xe0 EI".to_vec(); let pdf = Arc::new(build_pdf(&[page_stream(&img1, 1), page_stream(&img2, 2), page_stream(&img3, 3), page_stream(&img4, 4)], &[], "")); let mut renderer = PageRenderer::new(Arc::clone(&pdf), RenderConfig::default()); let mut greys = Vec::new(); for page_idx in 0..4 { let text_page = renderer.render(RenderRequest { page: page_idx, kind: RequestKind::Text, tile: None, scale: 1.0, tag: 0, preview: false }); assert!(text_page.error.is_none(), "page {} text error: {:?}", page_idx + 1, text_page.error); let text = text_page.text.expect("page text"); let plain = text.plain_text(); assert_eq!(plain.trim(), format!("After image P{}", page_idx + 1), "page {}", page_idx + 1); assert!(text.find("Injected").is_empty(), "page {} must not match Injected in Find", page_idx + 1); let rendered = renderer.render(RenderRequest { page: page_idx, kind: RequestKind::Pixels, tile: None, scale: 1.0, tag: 0, preview: false }); assert!(rendered.error.is_none(), "page {} render error: {:?}", page_idx + 1, rendered.error); let dark = dark_pixels_in_box(&rendered.rgba, rendered.width, 72, 172, 200, 196); assert_eq!(dark, 0, "page {} must not draw Injected at (72, 600)", page_idx + 1); let grey = mean_grey_in_box(&rendered.rgba, rendered.width, 110, 302, 290, 482); greys.push(grey); } // Page 1 and Page 3 have the exact same 47 samples (unfiltered vs ASCIIHexDecode). assert!((greys[0] - greys[2]).abs() < 1.0, "page 1 mean grey ({}) must match page 3 ({})", greys[0], greys[2]); assert!((greys[0] - greys[1]).abs() < 2.0, "page 1 mean grey ({}) must match control page 2 ({})", greys[0], greys[1]); assert!((greys[3] - 128.0).abs() < 2.0, "page 4 control mean grey ({})", greys[3]); } #[test] fn form_xobject_inline_image_with_ei_in_samples_keeps_all_samples() { let form_content = b"q 200 0 0 200 100 300 cm BI /W 12 /H 1 /BPC 8 /CS /G ID 0EI 1 0 0 rg EI Q BT /F1 18 Tf 72 700 Td (Form after) Tj ET"; let mut form_obj = format!( "<< /Type /XObject /Subtype /Form /BBox [0 0 612 792] /Resources << /Font << /F1 5 0 R >> >> /Length {} >>\nstream\n", form_content.len() ) .into_bytes(); form_obj.extend_from_slice(form_content); form_obj.extend_from_slice(b"\nendstream"); let page_content = b"/Fm1 Do".to_vec(); // 1 page => font is object 5, form XObject is object 6. let pdf = Arc::new(build_pdf(&[page_content], &[(6, form_obj)], "/XObject << /Fm1 6 0 R >>")); let mut renderer = PageRenderer::new(Arc::clone(&pdf), RenderConfig::default()); let rendered = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, tile: None, scale: 1.0, tag: 0, preview: false }); assert!(rendered.error.is_none(), "{:?}", rendered.error); let grey = mean_grey_in_box(&rendered.rgba, rendered.width, 110, 302, 290, 482); assert!((grey - 54.37).abs() < 1.0, "form inline image mean grey {grey} must be ~54.37 (not ~2 from 1-byte cut)"); } #[test] fn unfiltered_inline_image_layouts_and_mismatched_l_do_not_inject_content() { let inject_prefix = b"0EI Q BT /F1 18 Tf 72 600 Td (Injected) Tj ET q"; let make_image = |params: &str, len: usize| { let mut samples = inject_prefix.to_vec(); samples.resize(len, b'0'); let mut img = format!("BI {params} ID ").into_bytes(); img.extend_from_slice(&samples); img.extend_from_slice(b" EI"); img }; let cases = [ ("/W 16 /H 2 /CS /RGB /BPC 8", 96), ("/W 12 /H 2 /CS /CMYK /BPC 8", 96), ("/W 70 /H 4 /CS /G /BPC 1", 36), ("/IM true /W 70 /H 4", 36), ("/W 47 /H 1 /CS /G /BPC 8 /L 5", 47), ]; let pages: Vec> = cases.iter().enumerate().map(|(i, (params, len))| page_stream(&make_image(params, *len), i + 1)).collect(); let pdf = Arc::new(build_pdf(&pages, &[], "")); let mut renderer = PageRenderer::new(Arc::clone(&pdf), RenderConfig::default()); for (i, (params, _)) in cases.iter().enumerate() { let page_idx = i; let text_page = renderer.render(RenderRequest { page: page_idx, kind: RequestKind::Text, tile: None, scale: 1.0, tag: 0, preview: false }); assert!(text_page.error.is_none(), "case {params} text error: {:?}", text_page.error); let text = text_page.text.expect("page text"); let plain = text.plain_text(); assert_eq!(plain.trim(), format!("After image P{}", i + 1), "case {params}"); assert!(text.find("Injected").is_empty(), "case {params} must not match Injected in Find"); let rendered = renderer.render(RenderRequest { page: page_idx, kind: RequestKind::Pixels, tile: None, scale: 1.0, tag: 0, preview: false }); assert!(rendered.error.is_none(), "case {params} render error: {:?}", rendered.error); let dark = dark_pixels_in_box(&rendered.rgba, rendered.width, 72, 172, 200, 196); assert_eq!(dark, 0, "case {params} must not draw Injected at (72, 600)"); } }