Skip to content

PdfCraft: An open-source PDF reader and editor written by scratch in pure Rust

refactor(export): move pptx.rs's tests to pptx/tests.rs

aionggf700dc3f2e31
diff --git a/crates/export/src/pptx.rs b/crates/export/src/pptx.rs
index 2edc0e25c..49765000d 100644
--- a/crates/export/src/pptx.rs
+++ b/crates/export/src/pptx.rs
@@ -358,398 +358,4 @@ xmlns:a=\"http://schemas.openxmlformats.org/drawingml/2006/main\" type=\"blank\"
}
#[cfg(test)]
-mod tests {
- use super::*;
- use crate::Block;
- use crate::tests::part;
-
- fn sample_page() -> Page {
- Page {
- width: 720.0,
- height: 540.0,
- blocks: vec![
- Block {
- text: "Q3 Business Update".into(),
- rect: [72.0, 460.0, 500.0, 500.0],
- size: 28.0,
- bold: true,
- italic: false,
- color: [0.1, 0.2, 0.3],
- rotation: 0.0,
- space_around: [false; 2],
- },
- Block {
- text: "Revenue grew 25% year over year across all regions.".into(),
- rect: [72.0, 380.0, 500.0, 410.0],
- size: 14.0,
- bold: false,
- italic: false,
- color: [0.0; 3],
- rotation: 0.0,
- space_around: [false; 2],
- },
- Block {
- text: "Operational expenses remained flat.".into(),
- rect: [72.0, 340.0, 500.0, 370.0],
- size: 14.0,
- bold: false,
- italic: true,
- color: [0.0; 3],
- rotation: 0.0,
- space_around: [false; 2],
- },
- ],
- images: vec![Image {
- ext: "png",
- bytes: b"\x89PNG\r\n\x1a\n\0\0\0\rIHDR\0\0\0\x02\0\0\0\x03".as_slice().into(),
- rect: [72.0, 150.0, 300.0, 300.0],
- size: [228.0, 150.0],
- rotation: 0.0,
- }],
- background: None,
- }
- }
-
- #[test]
- pub fn pptx_is_a_presentationml_package() {
- let p = sample_page();
- let bytes = pptx(&[p], "Executive Deck");
- assert!(bytes.starts_with(b"PK"));
- for part_name in [
- "[Content_Types].xml",
- "_rels/.rels",
- "docProps/core.xml",
- "ppt/presentation.xml",
- "ppt/_rels/presentation.xml.rels",
- "ppt/slideMasters/slideMaster1.xml",
- "ppt/slideMasters/_rels/slideMaster1.xml.rels",
- "ppt/slideLayouts/slideLayout1.xml",
- "ppt/slideLayouts/_rels/slideLayout1.xml.rels",
- "ppt/theme/theme1.xml",
- "ppt/slides/slide1.xml",
- "ppt/slides/_rels/slide1.xml.rels",
- "ppt/media/image1.png",
- ] {
- assert!(bytes.windows(part_name.len()).any(|w| w == part_name.as_bytes()), "missing {part_name}");
- }
- }
-
- #[test]
- pub fn pptx_slide_has_title_body_and_images_in_order() {
- let p = sample_page();
- let bytes = pptx(&[p], "Deck");
- let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
- assert!(slide_xml.contains("name=\"Title 2\""));
- assert!(slide_xml.contains("txBox=\"1\""));
- assert!(slide_xml.contains("Q3 Business Update"));
- assert!(slide_xml.contains("name=\"TextBox 3\""));
- assert!(slide_xml.contains("Revenue grew 25%"));
- assert!(slide_xml.contains("<p:pic>"));
- assert!(slide_xml.contains("<a:blip r:embed=\"rIdImg1\"/>"));
- assert!(!slide_xml.contains("<p:ph"), "pure text boxes avoid placeholder inheritance repair warnings: {slide_xml}");
-
- let rels_xml = part(&bytes, "ppt/slides/_rels/slide1.xml.rels");
- assert!(rels_xml.contains("Target=\"../media/image1.png\""));
- assert!(rels_xml.contains("Target=\"../slideLayouts/slideLayout1.xml\""));
- }
-
- #[test]
- pub fn pptx_text_colour_and_formatting() {
- let p = sample_page();
- let bytes = pptx(&[p], "Deck");
- let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
- assert!(slide_xml.contains("sz=\"2800\""));
- assert!(slide_xml.contains("b=\"1\""));
- assert!(slide_xml.contains("<a:srgbClr val=\"1A334D\"/>"));
- assert!(slide_xml.contains("i=\"1\""));
- }
-
- /// Every entry of the package in order: (name, uncompressed bytes), read from the local
- /// headers the way an unzipper walks them, so a broken structure fails here.
- fn entries(zip: &[u8]) -> Vec<(String, Vec<u8>)> {
- let at = |i: usize, n: usize| zip[i..i + n].iter().rev().fold(0usize, |v, b| v << 8 | *b as usize);
- let mut out = Vec::new();
- let mut i = 0;
- while zip[i..].starts_with(b"PK\x03\x04") {
- let (method, size, raw, name_len, extra) = (at(i + 8, 2), at(i + 18, 4), at(i + 22, 4), at(i + 26, 2), at(i + 28, 2));
- let data = i + 30 + name_len + extra;
- let name = String::from_utf8(zip[i + 30..i + 30 + name_len].to_vec()).expect("UTF-8 name");
- let body = &zip[data..data + size];
- let bytes = match method {
- 0 => body.to_vec(),
- 8 => {
- let mut v = Vec::new();
- std::io::Read::read_to_end(&mut flate2::read::DeflateDecoder::new(body), &mut v).expect("inflates");
- v
- }
- m => panic!("{name}: unexpected method {m}"),
- };
- assert_eq!(bytes.len(), raw, "{name}: uncompressed size");
- out.push((name, bytes));
- i = data + size;
- }
- assert!(zip[i..].starts_with(b"PK\x01\x02"), "the central directory follows the entries");
- out
- }
-
- /// Checks that every start tag is closed by the matching end tag (text is escaped, so a
- /// `<` always starts markup). Catches a typo such as `<a:lnR>…</a:lnL>`.
- fn assert_tags_balance(name: &str, xml: &str) {
- let mut stack: Vec<&str> = Vec::new();
- for chunk in xml.split('<').skip(1) {
- let tag = chunk.split('>').next().unwrap_or_default();
- if tag.starts_with('?') {
- continue;
- }
- if let Some(close) = tag.strip_prefix('/') {
- assert_eq!(stack.pop(), Some(close.trim()), "{name}: </{close}> closes the wrong element");
- } else if !tag.ends_with('/') {
- stack.push(tag.split_whitespace().next().unwrap_or_default());
- }
- }
- assert!(stack.is_empty(), "{name}: unclosed {stack:?}");
- }
-
- #[test]
- fn pptx_package_parts_are_complete_and_consistent() {
- let cell = |t: &str, x: f64, y: f64| Block {
- text: t.into(),
- rect: [x, y, x + 100.0, y + 14.0],
- size: 12.0,
- bold: false,
- italic: false,
- color: [0.0; 3],
- rotation: 0.0,
- space_around: [false; 2],
- };
- let table_page = Page {
- width: 612.0,
- height: 792.0,
- blocks: vec![
- cell("Metric", 72.0, 400.0),
- cell("Value", 200.0, 400.0),
- cell("Growth", 72.0, 360.0),
- cell("+9%", 200.0, 360.0),
- cell("Target", 72.0, 320.0),
- cell("100%", 200.0, 320.0),
- ],
- images: Vec::new(),
- background: None,
- };
- let bytes = pptx(&[sample_page(), table_page], "Two");
- let parts = entries(&bytes);
- assert_eq!(parts.first().map(|p| p.0.as_str()), Some("[Content_Types].xml"), "content types come first");
- let names: std::collections::HashSet<&str> = parts.iter().map(|p| p.0.as_str()).collect();
- assert_eq!(names.len(), parts.len(), "no duplicate parts");
- for required in [
- "[Content_Types].xml",
- "_rels/.rels",
- "ppt/presentation.xml",
- "ppt/_rels/presentation.xml.rels",
- "ppt/slideMasters/slideMaster1.xml",
- "ppt/slideMasters/_rels/slideMaster1.xml.rels",
- "ppt/slideLayouts/slideLayout1.xml",
- "ppt/slideLayouts/_rels/slideLayout1.xml.rels",
- "ppt/theme/theme1.xml",
- "ppt/slides/slide1.xml",
- "ppt/slides/_rels/slide1.xml.rels",
- "ppt/slides/slide2.xml",
- "ppt/slides/_rels/slide2.xml.rels",
- "ppt/media/image1.png",
- ] {
- assert!(names.contains(required), "missing {required}");
- }
- let text = |n: &str| String::from_utf8(parts.iter().find(|p| p.0 == n).map(|p| p.1.clone()).unwrap_or_default()).expect("UTF-8 XML");
- for (name, body) in &parts {
- if name.ends_with(".xml") || name.ends_with(".rels") {
- assert!(body.starts_with(b"<?xml"), "{name} is XML");
- assert_tags_balance(name, &text(name));
- }
- }
- assert!(text("ppt/slides/slide2.xml").contains("<a:tbl>"), "the second slide carries the table");
- // Text boxes, not placeholders: the blank layout defines none to inherit from.
- for slide in ["ppt/slides/slide1.xml", "ppt/slides/slide2.xml"] {
- assert!(!text(slide).contains("<p:ph"), "{slide} uses no placeholders");
- }
- // Every override names a part that exists, and every XML part has a content type.
- let types = text("[Content_Types].xml");
- for chunk in types.split("PartName=\"/").skip(1) {
- let part = chunk.split('"').next().unwrap_or_default();
- assert!(names.contains(part), "override for missing part {part}");
- }
- for slide in ["ppt/slides/slide1.xml", "ppt/slides/slide2.xml", "ppt/presentation.xml", "ppt/theme/theme1.xml"] {
- assert!(types.contains(&format!("PartName=\"/{slide}\"")), "{slide} has a content type");
- }
- // Relationship targets resolve to parts in the package.
- for (rels, base) in [
- ("_rels/.rels", ""),
- ("ppt/_rels/presentation.xml.rels", "ppt/"),
- ("ppt/slides/_rels/slide1.xml.rels", "ppt/slides/"),
- ("ppt/slides/_rels/slide2.xml.rels", "ppt/slides/"),
- ("ppt/slideMasters/_rels/slideMaster1.xml.rels", "ppt/slideMasters/"),
- ("ppt/slideLayouts/_rels/slideLayout1.xml.rels", "ppt/slideLayouts/"),
- ] {
- for chunk in text(rels).split("Target=\"").skip(1) {
- let target = chunk.split('"').next().unwrap_or_default();
- let mut path: Vec<&str> = base.split('/').filter(|s| !s.is_empty()).collect();
- for seg in target.split('/') {
- if seg == ".." {
- path.pop();
- } else {
- path.push(seg);
- }
- }
- let resolved = path.join("/");
- assert!(names.contains(resolved.as_str()), "{rels}: target {target} is missing");
- }
- }
- // The theme's font scheme carries the schema-required latin, ea and cs faces.
- let theme = text("ppt/theme/theme1.xml");
- for font in ["majorFont", "minorFont"] {
- let inner = theme.split(&format!("<a:{font}>")).nth(1).and_then(|r| r.split(&format!("</a:{font}>")).next()).unwrap_or_default();
- assert!(inner.contains("<a:latin ") && inner.contains("<a:ea ") && inner.contains("<a:cs "), "{font}: {inner}");
- }
- assert!(text("ppt/presentation.xml").contains("r:id=\"rIdSld2\""));
- }
-
- #[test]
- fn hostile_sizes_do_not_overflow_the_slide_position() {
- let huge = |text: &str, size: f64, y: f64| Block {
- text: text.into(),
- rect: [72.0, y, 500.0, y + 10.0],
- size,
- bold: false,
- italic: false,
- color: [0.0; 3],
- rotation: 0.0,
- space_around: [false; 2],
- };
- let mut blocks = vec![huge("Title", f64::MAX, 500.0)];
- for k in 0..40 {
- blocks.push(huge(&"word ".repeat(10_000), 1e300, 400.0 - k as f64));
- blocks.push(huge("nan", f64::NAN, 300.0 - k as f64));
- blocks.push(huge("inf", f64::INFINITY, 200.0 - k as f64));
- }
- let images = (0..40)
- .map(|k| Image {
- ext: "png",
- bytes: [0; 4].as_slice().into(),
- rect: [0.0, -1e300, 1e-300, 1e300 - k as f64],
- size: [1.0, 1.0],
- rotation: 0.0,
- })
- .collect();
- let p = Page { width: f64::MAX, height: f64::NAN, blocks, images, background: None };
- let bytes = pptx(&[p], "Hostile");
- let slide = part(&bytes, "ppt/slides/slide1.xml");
- assert!(slide.contains("Title") && slide.contains("<p:pic>"), "the slide is still written");
- assert!(!slide.contains("y=\"-"), "positions never wrap negative");
- assert!(!slide.contains("<p:ph"), "text boxes, not placeholders");
- }
-
- #[test]
- pub fn pptx_empty_pages_produce_valid_slide() {
- let bytes = pptx(&[], "Blank");
- assert!(bytes.starts_with(b"PK"));
- let pres_xml = part(&bytes, "ppt/presentation.xml");
- assert!(pres_xml.contains("<p:sldId id=\"256\" r:id=\"rIdSld1\"/>"));
- let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
- assert!(slide_xml.contains("<p:spTree>"));
- }
-
- #[test]
- fn pptx_slide_dimensions_and_clamping() {
- // Known dimensions: 720x540 pt (10x7.5 in) = 9144000x6858000 EMU
- let p = Page { width: 720.0, height: 540.0, ..Page::default() };
- let bytes = pptx(&[p], "Known Dimensions");
- let pres_xml = part(&bytes, "ppt/presentation.xml");
- assert!(pres_xml.contains("<p:sldSz cx=\"9144000\" cy=\"6858000\"/>"), "{pres_xml}");
-
- // Tiny page (10x10 pt) clamps to PowerPoint minimum of 1 inch (914400 EMU)
- let tiny = Page { width: 10.0, height: 10.0, ..Page::default() };
- let tiny_bytes = pptx(&[tiny], "Tiny");
- let tiny_pres = part(&tiny_bytes, "ppt/presentation.xml");
- assert!(tiny_pres.contains("<p:sldSz cx=\"914400\" cy=\"914400\"/>"), "{tiny_pres}");
-
- // Poster page (10000x10000 pt) clamps to PowerPoint maximum of 56 inches (51206400 EMU)
- let poster = Page { width: 10_000.0, height: 10_000.0, ..Page::default() };
- let poster_bytes = pptx(&[poster], "Poster");
- let poster_pres = part(&poster_bytes, "ppt/presentation.xml");
- assert!(poster_pres.contains("<p:sldSz cx=\"51206400\" cy=\"51206400\"/>"), "{poster_pres}");
-
- // Non-finite dimensions fall back to default
- let nan_page = Page { width: f64::NAN, height: -1.0, ..Page::default() };
- let nan_bytes = pptx(&[nan_page], "NaN");
- let nan_pres = part(&nan_bytes, "ppt/presentation.xml");
- assert!(nan_pres.contains("<p:sldSz cx=\"9144000\" cy=\"6858000\"/>"), "{nan_pres}");
- }
-
- #[test]
- fn pptx_table_cell_with_largest_font_stays_in_table_and_not_in_title() {
- let cell = |t: &str, x: f64, y: f64, w: f64, sz: f64| Block {
- text: t.into(),
- rect: [x, y, x + w, y + 14.0],
- size: sz,
- bold: false,
- italic: false,
- color: [0.0; 3],
- rotation: 0.0,
- space_around: [false; 2],
- };
- let p = Page {
- width: 720.0,
- height: 540.0,
- blocks: vec![
- // Heading outside the table (short width, does not span table columns)
- Block {
- text: "Monthly Summary".into(),
- rect: [50.0, 480.0, 150.0, 500.0],
- size: 20.0,
- bold: true,
- italic: false,
- color: [0.0; 3],
- rotation: 0.0,
- space_around: [false; 2],
- },
- // Table where one cell has a very large number (e.g. 36 pt callout in the table)
- cell("Metric", 72.0, 400.0, 100.0, 12.0),
- cell("Value", 200.0, 400.0, 100.0, 12.0),
- cell("Growth", 72.0, 360.0, 100.0, 12.0),
- cell("+99%", 200.0, 360.0, 100.0, 36.0),
- cell("Target", 72.0, 320.0, 100.0, 12.0),
- cell("100%", 200.0, 320.0, 100.0, 12.0),
- ],
- images: Vec::new(),
- background: None,
- };
- let bytes = pptx(&[p], "Table Check");
- let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
- // Title should be the block outside the table ("Monthly Summary"), not the "+99%" cell
- assert!(slide_xml.contains("Monthly Summary"));
- assert!(slide_xml.contains("name=\"Title 2\""));
- // The table must contain both the cell and explicit borders
- assert!(slide_xml.contains("<a:tbl>"));
- assert!(slide_xml.contains("+99%"));
- assert!(slide_xml.contains("<a:tcPr><a:lnL w=\"12700\" cmpd=\"s\">"));
- }
-
- #[test]
- fn pptx_opens_in_libreoffice() {
- let p = sample_page();
- let bytes = pptx(&[p], "LibreOffice Check");
- let dir = std::env::temp_dir().join(format!("pdfcraft-pptx-lo-{}", std::process::id()));
- let _ = std::fs::remove_dir_all(&dir);
- let _ = std::fs::create_dir_all(&dir);
- let pptx_path = dir.join("test.pptx");
- std::fs::write(&pptx_path, &bytes).unwrap();
-
- for cmd in ["libreoffice", "soffice"] {
- let status = std::process::Command::new(cmd).args(["--headless", "--convert-to", "pdf", "--outdir"]).arg(&dir).arg(&pptx_path).status();
- if status.is_ok_and(|st| st.success()) {
- assert!(dir.join("test.pdf").is_file(), "LibreOffice produced test.pdf");
- break;
- }
- }
- let _ = std::fs::remove_dir_all(&dir);
- }
-}
+mod tests;
diff --git a/crates/export/src/pptx/tests.rs b/crates/export/src/pptx/tests.rs
new file mode 100644
index 000000000..2e1c1bd04
--- /dev/null
+++ b/crates/export/src/pptx/tests.rs
@@ -0,0 +1,393 @@
+use super::*;
+use crate::Block;
+use crate::tests::part;
+
+fn sample_page() -> Page {
+ Page {
+ width: 720.0,
+ height: 540.0,
+ blocks: vec![
+ Block {
+ text: "Q3 Business Update".into(),
+ rect: [72.0, 460.0, 500.0, 500.0],
+ size: 28.0,
+ bold: true,
+ italic: false,
+ color: [0.1, 0.2, 0.3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ },
+ Block {
+ text: "Revenue grew 25% year over year across all regions.".into(),
+ rect: [72.0, 380.0, 500.0, 410.0],
+ size: 14.0,
+ bold: false,
+ italic: false,
+ color: [0.0; 3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ },
+ Block {
+ text: "Operational expenses remained flat.".into(),
+ rect: [72.0, 340.0, 500.0, 370.0],
+ size: 14.0,
+ bold: false,
+ italic: true,
+ color: [0.0; 3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ },
+ ],
+ images: vec![Image {
+ ext: "png",
+ bytes: b"\x89PNG\r\n\x1a\n\0\0\0\rIHDR\0\0\0\x02\0\0\0\x03".as_slice().into(),
+ rect: [72.0, 150.0, 300.0, 300.0],
+ size: [228.0, 150.0],
+ rotation: 0.0,
+ }],
+ background: None,
+ }
+}
+
+#[test]
+pub fn pptx_is_a_presentationml_package() {
+ let p = sample_page();
+ let bytes = pptx(&[p], "Executive Deck");
+ assert!(bytes.starts_with(b"PK"));
+ for part_name in [
+ "[Content_Types].xml",
+ "_rels/.rels",
+ "docProps/core.xml",
+ "ppt/presentation.xml",
+ "ppt/_rels/presentation.xml.rels",
+ "ppt/slideMasters/slideMaster1.xml",
+ "ppt/slideMasters/_rels/slideMaster1.xml.rels",
+ "ppt/slideLayouts/slideLayout1.xml",
+ "ppt/slideLayouts/_rels/slideLayout1.xml.rels",
+ "ppt/theme/theme1.xml",
+ "ppt/slides/slide1.xml",
+ "ppt/slides/_rels/slide1.xml.rels",
+ "ppt/media/image1.png",
+ ] {
+ assert!(bytes.windows(part_name.len()).any(|w| w == part_name.as_bytes()), "missing {part_name}");
+ }
+}
+
+#[test]
+pub fn pptx_slide_has_title_body_and_images_in_order() {
+ let p = sample_page();
+ let bytes = pptx(&[p], "Deck");
+ let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
+ assert!(slide_xml.contains("name=\"Title 2\""));
+ assert!(slide_xml.contains("txBox=\"1\""));
+ assert!(slide_xml.contains("Q3 Business Update"));
+ assert!(slide_xml.contains("name=\"TextBox 3\""));
+ assert!(slide_xml.contains("Revenue grew 25%"));
+ assert!(slide_xml.contains("<p:pic>"));
+ assert!(slide_xml.contains("<a:blip r:embed=\"rIdImg1\"/>"));
+ assert!(!slide_xml.contains("<p:ph"), "pure text boxes avoid placeholder inheritance repair warnings: {slide_xml}");
+
+ let rels_xml = part(&bytes, "ppt/slides/_rels/slide1.xml.rels");
+ assert!(rels_xml.contains("Target=\"../media/image1.png\""));
+ assert!(rels_xml.contains("Target=\"../slideLayouts/slideLayout1.xml\""));
+}
+
+#[test]
+pub fn pptx_text_colour_and_formatting() {
+ let p = sample_page();
+ let bytes = pptx(&[p], "Deck");
+ let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
+ assert!(slide_xml.contains("sz=\"2800\""));
+ assert!(slide_xml.contains("b=\"1\""));
+ assert!(slide_xml.contains("<a:srgbClr val=\"1A334D\"/>"));
+ assert!(slide_xml.contains("i=\"1\""));
+}
+
+/// Every entry of the package in order: (name, uncompressed bytes), read from the local
+/// headers the way an unzipper walks them, so a broken structure fails here.
+fn entries(zip: &[u8]) -> Vec<(String, Vec<u8>)> {
+ let at = |i: usize, n: usize| zip[i..i + n].iter().rev().fold(0usize, |v, b| v << 8 | *b as usize);
+ let mut out = Vec::new();
+ let mut i = 0;
+ while zip[i..].starts_with(b"PK\x03\x04") {
+ let (method, size, raw, name_len, extra) = (at(i + 8, 2), at(i + 18, 4), at(i + 22, 4), at(i + 26, 2), at(i + 28, 2));
+ let data = i + 30 + name_len + extra;
+ let name = String::from_utf8(zip[i + 30..i + 30 + name_len].to_vec()).expect("UTF-8 name");
+ let body = &zip[data..data + size];
+ let bytes = match method {
+ 0 => body.to_vec(),
+ 8 => {
+ let mut v = Vec::new();
+ std::io::Read::read_to_end(&mut flate2::read::DeflateDecoder::new(body), &mut v).expect("inflates");
+ v
+ }
+ m => panic!("{name}: unexpected method {m}"),
+ };
+ assert_eq!(bytes.len(), raw, "{name}: uncompressed size");
+ out.push((name, bytes));
+ i = data + size;
+ }
+ assert!(zip[i..].starts_with(b"PK\x01\x02"), "the central directory follows the entries");
+ out
+}
+
+/// Checks that every start tag is closed by the matching end tag (text is escaped, so a
+/// `<` always starts markup). Catches a typo such as `<a:lnR>…</a:lnL>`.
+fn assert_tags_balance(name: &str, xml: &str) {
+ let mut stack: Vec<&str> = Vec::new();
+ for chunk in xml.split('<').skip(1) {
+ let tag = chunk.split('>').next().unwrap_or_default();
+ if tag.starts_with('?') {
+ continue;
+ }
+ if let Some(close) = tag.strip_prefix('/') {
+ assert_eq!(stack.pop(), Some(close.trim()), "{name}: </{close}> closes the wrong element");
+ } else if !tag.ends_with('/') {
+ stack.push(tag.split_whitespace().next().unwrap_or_default());
+ }
+ }
+ assert!(stack.is_empty(), "{name}: unclosed {stack:?}");
+}
+
+#[test]
+fn pptx_package_parts_are_complete_and_consistent() {
+ let cell = |t: &str, x: f64, y: f64| Block {
+ text: t.into(),
+ rect: [x, y, x + 100.0, y + 14.0],
+ size: 12.0,
+ bold: false,
+ italic: false,
+ color: [0.0; 3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ };
+ let table_page = Page {
+ width: 612.0,
+ height: 792.0,
+ blocks: vec![
+ cell("Metric", 72.0, 400.0),
+ cell("Value", 200.0, 400.0),
+ cell("Growth", 72.0, 360.0),
+ cell("+9%", 200.0, 360.0),
+ cell("Target", 72.0, 320.0),
+ cell("100%", 200.0, 320.0),
+ ],
+ images: Vec::new(),
+ background: None,
+ };
+ let bytes = pptx(&[sample_page(), table_page], "Two");
+ let parts = entries(&bytes);
+ assert_eq!(parts.first().map(|p| p.0.as_str()), Some("[Content_Types].xml"), "content types come first");
+ let names: std::collections::HashSet<&str> = parts.iter().map(|p| p.0.as_str()).collect();
+ assert_eq!(names.len(), parts.len(), "no duplicate parts");
+ for required in [
+ "[Content_Types].xml",
+ "_rels/.rels",
+ "ppt/presentation.xml",
+ "ppt/_rels/presentation.xml.rels",
+ "ppt/slideMasters/slideMaster1.xml",
+ "ppt/slideMasters/_rels/slideMaster1.xml.rels",
+ "ppt/slideLayouts/slideLayout1.xml",
+ "ppt/slideLayouts/_rels/slideLayout1.xml.rels",
+ "ppt/theme/theme1.xml",
+ "ppt/slides/slide1.xml",
+ "ppt/slides/_rels/slide1.xml.rels",
+ "ppt/slides/slide2.xml",
+ "ppt/slides/_rels/slide2.xml.rels",
+ "ppt/media/image1.png",
+ ] {
+ assert!(names.contains(required), "missing {required}");
+ }
+ let text = |n: &str| String::from_utf8(parts.iter().find(|p| p.0 == n).map(|p| p.1.clone()).unwrap_or_default()).expect("UTF-8 XML");
+ for (name, body) in &parts {
+ if name.ends_with(".xml") || name.ends_with(".rels") {
+ assert!(body.starts_with(b"<?xml"), "{name} is XML");
+ assert_tags_balance(name, &text(name));
+ }
+ }
+ assert!(text("ppt/slides/slide2.xml").contains("<a:tbl>"), "the second slide carries the table");
+ // Text boxes, not placeholders: the blank layout defines none to inherit from.
+ for slide in ["ppt/slides/slide1.xml", "ppt/slides/slide2.xml"] {
+ assert!(!text(slide).contains("<p:ph"), "{slide} uses no placeholders");
+ }
+ // Every override names a part that exists, and every XML part has a content type.
+ let types = text("[Content_Types].xml");
+ for chunk in types.split("PartName=\"/").skip(1) {
+ let part = chunk.split('"').next().unwrap_or_default();
+ assert!(names.contains(part), "override for missing part {part}");
+ }
+ for slide in ["ppt/slides/slide1.xml", "ppt/slides/slide2.xml", "ppt/presentation.xml", "ppt/theme/theme1.xml"] {
+ assert!(types.contains(&format!("PartName=\"/{slide}\"")), "{slide} has a content type");
+ }
+ // Relationship targets resolve to parts in the package.
+ for (rels, base) in [
+ ("_rels/.rels", ""),
+ ("ppt/_rels/presentation.xml.rels", "ppt/"),
+ ("ppt/slides/_rels/slide1.xml.rels", "ppt/slides/"),
+ ("ppt/slides/_rels/slide2.xml.rels", "ppt/slides/"),
+ ("ppt/slideMasters/_rels/slideMaster1.xml.rels", "ppt/slideMasters/"),
+ ("ppt/slideLayouts/_rels/slideLayout1.xml.rels", "ppt/slideLayouts/"),
+ ] {
+ for chunk in text(rels).split("Target=\"").skip(1) {
+ let target = chunk.split('"').next().unwrap_or_default();
+ let mut path: Vec<&str> = base.split('/').filter(|s| !s.is_empty()).collect();
+ for seg in target.split('/') {
+ if seg == ".." {
+ path.pop();
+ } else {
+ path.push(seg);
+ }
+ }
+ let resolved = path.join("/");
+ assert!(names.contains(resolved.as_str()), "{rels}: target {target} is missing");
+ }
+ }
+ // The theme's font scheme carries the schema-required latin, ea and cs faces.
+ let theme = text("ppt/theme/theme1.xml");
+ for font in ["majorFont", "minorFont"] {
+ let inner = theme.split(&format!("<a:{font}>")).nth(1).and_then(|r| r.split(&format!("</a:{font}>")).next()).unwrap_or_default();
+ assert!(inner.contains("<a:latin ") && inner.contains("<a:ea ") && inner.contains("<a:cs "), "{font}: {inner}");
+ }
+ assert!(text("ppt/presentation.xml").contains("r:id=\"rIdSld2\""));
+}
+
+#[test]
+fn hostile_sizes_do_not_overflow_the_slide_position() {
+ let huge = |text: &str, size: f64, y: f64| Block {
+ text: text.into(),
+ rect: [72.0, y, 500.0, y + 10.0],
+ size,
+ bold: false,
+ italic: false,
+ color: [0.0; 3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ };
+ let mut blocks = vec![huge("Title", f64::MAX, 500.0)];
+ for k in 0..40 {
+ blocks.push(huge(&"word ".repeat(10_000), 1e300, 400.0 - k as f64));
+ blocks.push(huge("nan", f64::NAN, 300.0 - k as f64));
+ blocks.push(huge("inf", f64::INFINITY, 200.0 - k as f64));
+ }
+ let images = (0..40)
+ .map(|k| Image {
+ ext: "png",
+ bytes: [0; 4].as_slice().into(),
+ rect: [0.0, -1e300, 1e-300, 1e300 - k as f64],
+ size: [1.0, 1.0],
+ rotation: 0.0,
+ })
+ .collect();
+ let p = Page { width: f64::MAX, height: f64::NAN, blocks, images, background: None };
+ let bytes = pptx(&[p], "Hostile");
+ let slide = part(&bytes, "ppt/slides/slide1.xml");
+ assert!(slide.contains("Title") && slide.contains("<p:pic>"), "the slide is still written");
+ assert!(!slide.contains("y=\"-"), "positions never wrap negative");
+ assert!(!slide.contains("<p:ph"), "text boxes, not placeholders");
+}
+
+#[test]
+pub fn pptx_empty_pages_produce_valid_slide() {
+ let bytes = pptx(&[], "Blank");
+ assert!(bytes.starts_with(b"PK"));
+ let pres_xml = part(&bytes, "ppt/presentation.xml");
+ assert!(pres_xml.contains("<p:sldId id=\"256\" r:id=\"rIdSld1\"/>"));
+ let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
+ assert!(slide_xml.contains("<p:spTree>"));
+}
+
+#[test]
+fn pptx_slide_dimensions_and_clamping() {
+ // Known dimensions: 720x540 pt (10x7.5 in) = 9144000x6858000 EMU
+ let p = Page { width: 720.0, height: 540.0, ..Page::default() };
+ let bytes = pptx(&[p], "Known Dimensions");
+ let pres_xml = part(&bytes, "ppt/presentation.xml");
+ assert!(pres_xml.contains("<p:sldSz cx=\"9144000\" cy=\"6858000\"/>"), "{pres_xml}");
+
+ // Tiny page (10x10 pt) clamps to PowerPoint minimum of 1 inch (914400 EMU)
+ let tiny = Page { width: 10.0, height: 10.0, ..Page::default() };
+ let tiny_bytes = pptx(&[tiny], "Tiny");
+ let tiny_pres = part(&tiny_bytes, "ppt/presentation.xml");
+ assert!(tiny_pres.contains("<p:sldSz cx=\"914400\" cy=\"914400\"/>"), "{tiny_pres}");
+
+ // Poster page (10000x10000 pt) clamps to PowerPoint maximum of 56 inches (51206400 EMU)
+ let poster = Page { width: 10_000.0, height: 10_000.0, ..Page::default() };
+ let poster_bytes = pptx(&[poster], "Poster");
+ let poster_pres = part(&poster_bytes, "ppt/presentation.xml");
+ assert!(poster_pres.contains("<p:sldSz cx=\"51206400\" cy=\"51206400\"/>"), "{poster_pres}");
+
+ // Non-finite dimensions fall back to default
+ let nan_page = Page { width: f64::NAN, height: -1.0, ..Page::default() };
+ let nan_bytes = pptx(&[nan_page], "NaN");
+ let nan_pres = part(&nan_bytes, "ppt/presentation.xml");
+ assert!(nan_pres.contains("<p:sldSz cx=\"9144000\" cy=\"6858000\"/>"), "{nan_pres}");
+}
+
+#[test]
+fn pptx_table_cell_with_largest_font_stays_in_table_and_not_in_title() {
+ let cell = |t: &str, x: f64, y: f64, w: f64, sz: f64| Block {
+ text: t.into(),
+ rect: [x, y, x + w, y + 14.0],
+ size: sz,
+ bold: false,
+ italic: false,
+ color: [0.0; 3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ };
+ let p = Page {
+ width: 720.0,
+ height: 540.0,
+ blocks: vec![
+ // Heading outside the table (short width, does not span table columns)
+ Block {
+ text: "Monthly Summary".into(),
+ rect: [50.0, 480.0, 150.0, 500.0],
+ size: 20.0,
+ bold: true,
+ italic: false,
+ color: [0.0; 3],
+ rotation: 0.0,
+ space_around: [false; 2],
+ },
+ // Table where one cell has a very large number (e.g. 36 pt callout in the table)
+ cell("Metric", 72.0, 400.0, 100.0, 12.0),
+ cell("Value", 200.0, 400.0, 100.0, 12.0),
+ cell("Growth", 72.0, 360.0, 100.0, 12.0),
+ cell("+99%", 200.0, 360.0, 100.0, 36.0),
+ cell("Target", 72.0, 320.0, 100.0, 12.0),
+ cell("100%", 200.0, 320.0, 100.0, 12.0),
+ ],
+ images: Vec::new(),
+ background: None,
+ };
+ let bytes = pptx(&[p], "Table Check");
+ let slide_xml = part(&bytes, "ppt/slides/slide1.xml");
+ // Title should be the block outside the table ("Monthly Summary"), not the "+99%" cell
+ assert!(slide_xml.contains("Monthly Summary"));
+ assert!(slide_xml.contains("name=\"Title 2\""));
+ // The table must contain both the cell and explicit borders
+ assert!(slide_xml.contains("<a:tbl>"));
+ assert!(slide_xml.contains("+99%"));
+ assert!(slide_xml.contains("<a:tcPr><a:lnL w=\"12700\" cmpd=\"s\">"));
+}
+
+#[test]
+fn pptx_opens_in_libreoffice() {
+ let p = sample_page();
+ let bytes = pptx(&[p], "LibreOffice Check");
+ let dir = std::env::temp_dir().join(format!("pdfcraft-pptx-lo-{}", std::process::id()));
+ let _ = std::fs::remove_dir_all(&dir);
+ let _ = std::fs::create_dir_all(&dir);
+ let pptx_path = dir.join("test.pptx");
+ std::fs::write(&pptx_path, &bytes).unwrap();
+
+ for cmd in ["libreoffice", "soffice"] {
+ let status = std::process::Command::new(cmd).args(["--headless", "--convert-to", "pdf", "--outdir"]).arg(&dir).arg(&pptx_path).status();
+ if status.is_ok_and(|st| st.success()) {
+ assert!(dir.join("test.pdf").is_file(), "LibreOffice produced test.pdf");
+ break;
+ }
+ }
+ let _ = std::fs::remove_dir_all(&dir);
+}