kreuzberg 4.0.0.pre.rc.6

data/vendor/kreuzberg/src/extraction/markdown.rs

@@ -0,0 +1,213 @@
+//! Markdown table formatting utilities
+//!
+//! This module provides utilities for converting tabular data into GitHub-Flavored Markdown (GFM) tables.
+//! It's used by multiple extractors (DOCX, HTML) that need to represent structured table data in markdown format.
+
+/// Converts a 2D vector of cell strings into a GitHub-Flavored Markdown table.
+///
+/// # Behavior
+///
+/// - The first row is treated as the header row
+/// - A separator row is inserted after the header
+/// - Pipe characters (`|`) in cell content are automatically escaped with backslash
+/// - Irregular tables (rows with varying column counts) are padded with empty cells to match the header
+/// - Returns an empty string for empty input
+///
+/// # Arguments
+///
+/// * `cells` - A slice of vectors representing table rows, where each inner vector contains cell values
+///
+/// # Returns
+///
+/// A `String` containing the GFM markdown table representation
+///
+/// # Examples
+///
+/// ```
+/// # use kreuzberg::extraction::cells_to_markdown;
+/// let cells = vec![
+///     vec!["Name".to_string(), "Age".to_string()],
+///     vec!["Alice".to_string(), "30".to_string()],
+///     vec!["Bob".to_string(), "25".to_string()],
+/// ];
+///
+/// let markdown = cells_to_markdown(&cells);
+/// assert!(markdown.contains("| Name | Age |"));
+/// assert!(markdown.contains("|------|------|"));
+/// ```
+pub fn cells_to_markdown(cells: &[Vec<String>]) -> String {
+    if cells.is_empty() {
+        return String::new();
+    }
+
+    let mut markdown = String::new();
+
+    let num_cols = cells.first().map(|r| r.len()).unwrap_or(0);
+    if num_cols == 0 {
+        return String::new();
+    }
+
+    if let Some(header) = cells.first() {
+        markdown.push('|');
+        for cell in header {
+            markdown.push(' ');
+            let escaped = cell.replace('|', "\\|");
+            markdown.push_str(&escaped);
+            markdown.push_str(" |");
+        }
+        markdown.push('\n');
+
+        markdown.push('|');
+        for _ in 0..num_cols {
+            markdown.push_str("------|");
+        }
+        markdown.push('\n');
+    }
+
+    for row in cells.iter().skip(1) {
+        markdown.push('|');
+        for (idx, cell) in row.iter().enumerate() {
+            if idx >= num_cols {
+                break;
+            }
+            markdown.push(' ');
+            let escaped = cell.replace('|', "\\|");
+            markdown.push_str(&escaped);
+            markdown.push_str(" |");
+        }
+        for _ in row.len()..num_cols {
+            markdown.push_str(" |");
+        }
+        markdown.push('\n');
+    }
+
+    markdown
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_markdown_formatting_from_simple_table() {
+        let cells = vec![
+            vec!["Header1".to_string(), "Header2".to_string()],
+            vec!["Row1Col1".to_string(), "Row1Col2".to_string()],
+            vec!["Row2Col1".to_string(), "Row2Col2".to_string()],
+        ];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("| Header1 | Header2 |"));
+        assert!(markdown.contains("|------|------|"));
+        assert!(markdown.contains("| Row1Col1 | Row1Col2 |"));
+        assert!(markdown.contains("| Row2Col1 | Row2Col2 |"));
+
+        let lines: Vec<&str> = markdown.lines().collect();
+        assert_eq!(lines.len(), 4);
+    }
+
+    #[test]
+    fn test_markdown_handles_empty_input() {
+        let cells: Vec<Vec<String>> = vec![];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert_eq!(markdown, "");
+    }
+
+    #[test]
+    fn test_markdown_escapes_pipe_characters() {
+        let cells = vec![vec!["Header".to_string()], vec!["Cell with | pipe".to_string()]];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("Cell with \\| pipe"));
+
+        for line in markdown.lines() {
+            if !line.is_empty() {
+                assert!(line.starts_with('|'));
+                assert!(line.ends_with('|'));
+            }
+        }
+    }
+
+    #[test]
+    fn test_markdown_pads_irregular_tables() {
+        let cells = vec![
+            vec!["H1".to_string(), "H2".to_string(), "H3".to_string()],
+            vec!["R1C1".to_string(), "R1C2".to_string()],
+            vec!["R2C1".to_string(), "R2C2".to_string(), "R2C3".to_string()],
+        ];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("| H1 | H2 | H3 |"));
+
+        assert!(markdown.contains("| R1C1 | R1C2 | |"));
+
+        let lines: Vec<&str> = markdown.lines().filter(|l| !l.is_empty()).collect();
+        let pipe_counts: Vec<usize> = lines
+            .iter()
+            .map(|line| line.chars().filter(|c| *c == '|').count())
+            .collect();
+        assert!(pipe_counts.iter().all(|&count| count == pipe_counts[0]));
+    }
+
+    #[test]
+    fn test_markdown_single_row_table() {
+        let cells = vec![vec!["OnlyHeader".to_string()]];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("| OnlyHeader |"));
+        assert!(markdown.contains("|------|"));
+
+        let lines: Vec<&str> = markdown.lines().collect();
+        assert_eq!(lines.len(), 2);
+    }
+
+    #[test]
+    fn test_markdown_single_column_table() {
+        let cells = vec![
+            vec!["Header".to_string()],
+            vec!["Data1".to_string()],
+            vec!["Data2".to_string()],
+        ];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("| Header |"));
+        assert!(markdown.contains("|------|"));
+        assert!(markdown.contains("| Data1 |"));
+        assert!(markdown.contains("| Data2 |"));
+    }
+
+    #[test]
+    fn test_markdown_special_characters() {
+        let cells = vec![
+            vec!["*Header*".to_string(), "#Title".to_string()],
+            vec!["**Bold**".to_string(), "~~Strike~~".to_string()],
+        ];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("*Header*"));
+        assert!(markdown.contains("#Title"));
+        assert!(markdown.contains("**Bold**"));
+        assert!(markdown.contains("~~Strike~~"));
+    }
+
+    #[test]
+    fn test_markdown_unicode_content() {
+        let cells = vec![
+            vec!["Emoji".to_string(), "Accents".to_string()],
+            vec!["🎉 Party".to_string(), "Café".to_string()],
+        ];
+
+        let markdown = cells_to_markdown(&cells);
+
+        assert!(markdown.contains("🎉 Party"));
+        assert!(markdown.contains("Café"));
+    }
+}

data/vendor/kreuzberg/src/extraction/mod.rs

@@ -0,0 +1,81 @@
+pub mod structured;
+pub mod text;
+
+#[cfg(feature = "ocr")]
+pub mod image;
+
+#[cfg(feature = "archives")]
+pub mod archive;
+
+#[cfg(feature = "email")]
+pub mod email;
+
+#[cfg(feature = "excel")]
+pub mod excel;
+
+#[cfg(feature = "html")]
+pub mod html;
+
+#[cfg(feature = "office")]
+pub mod docx;
+
+#[cfg(feature = "office")]
+pub mod libreoffice;
+
+#[cfg(feature = "office")]
+pub mod office_metadata;
+
+#[cfg(feature = "office")]
+pub mod pptx;
+
+#[cfg(feature = "excel")]
+pub mod table;
+
+#[cfg(feature = "xml")]
+pub mod xml;
+
+#[cfg(any(feature = "office", feature = "html"))]
+pub mod markdown;
+
+pub use structured::{JsonExtractionConfig, StructuredDataResult, parse_json, parse_toml, parse_yaml};
+pub use text::parse_text;
+
+#[cfg(feature = "ocr")]
+pub use image::{ImageMetadata, extract_image_metadata};
+
+#[cfg(feature = "archives")]
+pub use archive::{
+    ArchiveEntry, ArchiveMetadata, extract_7z_metadata, extract_7z_text_content, extract_tar_metadata,
+    extract_tar_text_content, extract_zip_metadata, extract_zip_text_content,
+};
+
+#[cfg(feature = "email")]
+pub use email::{build_email_text_output, extract_email_content, parse_eml_content, parse_msg_content};
+
+#[cfg(feature = "excel")]
+pub use excel::{excel_to_markdown, read_excel_bytes, read_excel_file};
+
+#[cfg(feature = "html")]
+pub use html::{convert_html_to_markdown, process_html};
+
+#[cfg(feature = "office")]
+pub use libreoffice::{check_libreoffice_available, convert_doc_to_docx, convert_ppt_to_pptx};
+
+#[cfg(feature = "office")]
+pub use office_metadata::{
+    CoreProperties, CustomProperties, DocxAppProperties, OdtProperties, PptxAppProperties, XlsxAppProperties,
+    extract_core_properties, extract_custom_properties, extract_docx_app_properties, extract_odt_properties,
+    extract_pptx_app_properties, extract_xlsx_app_properties,
+};
+
+#[cfg(feature = "office")]
+pub use pptx::{extract_pptx_from_bytes, extract_pptx_from_path};
+
+#[cfg(feature = "excel")]
+pub use table::table_from_arrow_to_markdown;
+
+#[cfg(feature = "xml")]
+pub use xml::parse_xml;
+
+#[cfg(any(feature = "office", feature = "html"))]
+pub use markdown::cells_to_markdown;

data/vendor/kreuzberg/src/extraction/office_metadata/app_properties.rs

@@ -0,0 +1,398 @@
+//! Application properties extraction from docProps/app.xml
+//!
+//! Extracts format-specific metadata from Office Open XML documents.
+
+use crate::error::{KreuzbergError, Result};
+use roxmltree::Node;
+use std::io::Read;
+use zip::ZipArchive;
+
+/// Application properties from docProps/app.xml for DOCX
+///
+/// Contains Word-specific document statistics and metadata.
+#[derive(Debug, Clone, Default, PartialEq)]
+pub struct DocxAppProperties {
+    /// Application name (e.g., "Microsoft Office Word")
+    pub application: Option<String>,
+    /// Application version
+    pub app_version: Option<String>,
+    /// Template filename
+    pub template: Option<String>,
+    /// Total editing time in minutes
+    pub total_time: Option<i32>,
+    /// Number of pages
+    pub pages: Option<i32>,
+    /// Number of words
+    pub words: Option<i32>,
+    /// Number of characters (excluding spaces)
+    pub characters: Option<i32>,
+    /// Number of characters (including spaces)
+    pub characters_with_spaces: Option<i32>,
+    /// Number of lines
+    pub lines: Option<i32>,
+    /// Number of paragraphs
+    pub paragraphs: Option<i32>,
+    /// Company name
+    pub company: Option<String>,
+    /// Document security level
+    pub doc_security: Option<i32>,
+    /// Scale crop flag
+    pub scale_crop: Option<bool>,
+    /// Links up to date flag
+    pub links_up_to_date: Option<bool>,
+    /// Shared document flag
+    pub shared_doc: Option<bool>,
+    /// Hyperlinks changed flag
+    pub hyperlinks_changed: Option<bool>,
+}
+
+/// Application properties from docProps/app.xml for XLSX
+///
+/// Contains Excel-specific document metadata.
+#[derive(Debug, Clone, Default, PartialEq)]
+pub struct XlsxAppProperties {
+    /// Application name (e.g., "Microsoft Excel")
+    pub application: Option<String>,
+    /// Application version
+    pub app_version: Option<String>,
+    /// Document security level
+    pub doc_security: Option<i32>,
+    /// Scale crop flag
+    pub scale_crop: Option<bool>,
+    /// Links up to date flag
+    pub links_up_to_date: Option<bool>,
+    /// Shared document flag
+    pub shared_doc: Option<bool>,
+    /// Hyperlinks changed flag
+    pub hyperlinks_changed: Option<bool>,
+    /// Company name
+    pub company: Option<String>,
+    /// Worksheet names
+    pub worksheet_names: Vec<String>,
+}
+
+/// Application properties from docProps/app.xml for PPTX
+///
+/// Contains PowerPoint-specific document metadata.
+#[derive(Debug, Clone, Default, PartialEq)]
+pub struct PptxAppProperties {
+    /// Application name (e.g., "Microsoft Office PowerPoint")
+    pub application: Option<String>,
+    /// Application version
+    pub app_version: Option<String>,
+    /// Total editing time in minutes
+    pub total_time: Option<i32>,
+    /// Company name
+    pub company: Option<String>,
+    /// Document security level
+    pub doc_security: Option<i32>,
+    /// Scale crop flag
+    pub scale_crop: Option<bool>,
+    /// Links up to date flag
+    pub links_up_to_date: Option<bool>,
+    /// Shared document flag
+    pub shared_doc: Option<bool>,
+    /// Hyperlinks changed flag
+    pub hyperlinks_changed: Option<bool>,
+    /// Number of slides
+    pub slides: Option<i32>,
+    /// Number of notes
+    pub notes: Option<i32>,
+    /// Number of hidden slides
+    pub hidden_slides: Option<i32>,
+    /// Number of multimedia clips
+    pub multimedia_clips: Option<i32>,
+    /// Presentation format (e.g., "Widescreen", "Standard")
+    pub presentation_format: Option<String>,
+    /// Slide titles
+    pub slide_titles: Vec<String>,
+}
+
+/// Extract DOCX application properties from an Office Open XML document
+///
+/// Parses `docProps/app.xml` and extracts Word-specific metadata.
+pub fn extract_docx_app_properties<R: Read + std::io::Seek>(archive: &mut ZipArchive<R>) -> Result<DocxAppProperties> {
+    let mut xml_content = String::new();
+
+    match archive.by_name("docProps/app.xml") {
+        Ok(mut file) => {
+            file.read_to_string(&mut xml_content)
+                .map_err(|e| KreuzbergError::parsing(format!("Failed to read app.xml: {}", e)))?;
+        }
+        Err(_) => {
+            return Ok(DocxAppProperties::default());
+        }
+    }
+
+    let doc = roxmltree::Document::parse(&xml_content)
+        .map_err(|e| KreuzbergError::parsing(format!("Failed to parse app.xml: {}", e)))?;
+
+    let root = doc.root_element();
+
+    Ok(DocxAppProperties {
+        application: super::parse_xml_text(root, "Application"),
+        app_version: super::parse_xml_text(root, "AppVersion"),
+        template: super::parse_xml_text(root, "Template"),
+        total_time: super::parse_xml_int(root, "TotalTime"),
+        pages: super::parse_xml_int(root, "Pages"),
+        words: super::parse_xml_int(root, "Words"),
+        characters: super::parse_xml_int(root, "Characters"),
+        characters_with_spaces: super::parse_xml_int(root, "CharactersWithSpaces"),
+        lines: super::parse_xml_int(root, "Lines"),
+        paragraphs: super::parse_xml_int(root, "Paragraphs"),
+        company: super::parse_xml_text(root, "Company"),
+        doc_security: super::parse_xml_int(root, "DocSecurity"),
+        scale_crop: super::parse_xml_bool(root, "ScaleCrop"),
+        links_up_to_date: super::parse_xml_bool(root, "LinksUpToDate"),
+        shared_doc: super::parse_xml_bool(root, "SharedDoc"),
+        hyperlinks_changed: super::parse_xml_bool(root, "HyperlinksChanged"),
+    })
+}
+
+/// Extract XLSX application properties from an Office Open XML document
+///
+/// Parses `docProps/app.xml` and extracts Excel-specific metadata including worksheet names.
+pub fn extract_xlsx_app_properties<R: Read + std::io::Seek>(archive: &mut ZipArchive<R>) -> Result<XlsxAppProperties> {
+    let mut xml_content = String::new();
+
+    match archive.by_name("docProps/app.xml") {
+        Ok(mut file) => {
+            file.read_to_string(&mut xml_content)
+                .map_err(|e| KreuzbergError::parsing(format!("Failed to read app.xml: {}", e)))?;
+        }
+        Err(_) => {
+            return Ok(XlsxAppProperties::default());
+        }
+    }
+
+    let doc = roxmltree::Document::parse(&xml_content)
+        .map_err(|e| KreuzbergError::parsing(format!("Failed to parse app.xml: {}", e)))?;
+
+    let root = doc.root_element();
+
+    let worksheet_names = extract_titles_of_parts(root);
+
+    Ok(XlsxAppProperties {
+        application: super::parse_xml_text(root, "Application"),
+        app_version: super::parse_xml_text(root, "AppVersion"),
+        doc_security: super::parse_xml_int(root, "DocSecurity"),
+        scale_crop: super::parse_xml_bool(root, "ScaleCrop"),
+        links_up_to_date: super::parse_xml_bool(root, "LinksUpToDate"),
+        shared_doc: super::parse_xml_bool(root, "SharedDoc"),
+        hyperlinks_changed: super::parse_xml_bool(root, "HyperlinksChanged"),
+        company: super::parse_xml_text(root, "Company"),
+        worksheet_names,
+    })
+}
+
+/// Extract PPTX application properties from an Office Open XML document
+///
+/// Parses `docProps/app.xml` and extracts PowerPoint-specific metadata including slide information.
+pub fn extract_pptx_app_properties<R: Read + std::io::Seek>(archive: &mut ZipArchive<R>) -> Result<PptxAppProperties> {
+    let mut xml_content = String::new();
+
+    match archive.by_name("docProps/app.xml") {
+        Ok(mut file) => {
+            file.read_to_string(&mut xml_content)
+                .map_err(|e| KreuzbergError::parsing(format!("Failed to read app.xml: {}", e)))?;
+        }
+        Err(_) => {
+            return Ok(PptxAppProperties::default());
+        }
+    }
+
+    let doc = roxmltree::Document::parse(&xml_content)
+        .map_err(|e| KreuzbergError::parsing(format!("Failed to parse app.xml: {}", e)))?;
+
+    let root = doc.root_element();
+
+    let slide_titles = extract_titles_of_parts(root);
+
+    let presentation_format = super::parse_xml_text(root, "PresentationFormat");
+
+    Ok(PptxAppProperties {
+        application: super::parse_xml_text(root, "Application"),
+        app_version: super::parse_xml_text(root, "AppVersion"),
+        total_time: super::parse_xml_int(root, "TotalTime"),
+        company: super::parse_xml_text(root, "Company"),
+        doc_security: super::parse_xml_int(root, "DocSecurity"),
+        scale_crop: super::parse_xml_bool(root, "ScaleCrop"),
+        links_up_to_date: super::parse_xml_bool(root, "LinksUpToDate"),
+        shared_doc: super::parse_xml_bool(root, "SharedDoc"),
+        hyperlinks_changed: super::parse_xml_bool(root, "HyperlinksChanged"),
+        slides: super::parse_xml_int(root, "Slides"),
+        notes: super::parse_xml_int(root, "Notes"),
+        hidden_slides: super::parse_xml_int(root, "HiddenSlides"),
+        multimedia_clips: super::parse_xml_int(root, "MMClips"),
+        presentation_format,
+        slide_titles,
+    })
+}
+
+/// Extract titles from TitlesOfParts vt:vector element
+///
+/// Handles the vt:vector/vt:lpstr structure used for worksheet/slide names.
+fn extract_titles_of_parts(root: Node) -> Vec<String> {
+    let mut titles = Vec::new();
+
+    if let Some(titles_node) = root.descendants().find(|n| n.has_tag_name("TitlesOfParts"))
+        && let Some(vector_node) = titles_node.descendants().find(|n| n.has_tag_name("vector"))
+    {
+        for lpstr_node in vector_node.descendants().filter(|n| n.has_tag_name("lpstr")) {
+            if let Some(text) = lpstr_node.text() {
+                let text = text.trim();
+                if !text.is_empty() {
+                    titles.push(text.to_string());
+                }
+            }
+        }
+    }
+
+    titles
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+    use std::io::{Cursor, Write};
+
+    fn create_test_zip_with_app_xml(app_xml: &str) -> ZipArchive<Cursor<Vec<u8>>> {
+        let buffer = Vec::new();
+        let cursor = Cursor::new(buffer);
+        let mut zip = zip::ZipWriter::new(cursor);
+
+        let options = zip::write::FileOptions::<()>::default().compression_method(zip::CompressionMethod::Stored);
+
+        zip.start_file("docProps/app.xml", options).unwrap();
+        zip.write_all(app_xml.as_bytes()).unwrap();
+
+        let cursor = zip.finish().unwrap();
+        ZipArchive::new(cursor).unwrap()
+    }
+
+    #[test]
+    fn test_extract_docx_app_properties() {
+        let app_xml = r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Properties xmlns="http://schemas.openxmlformats.org/officeDocument/2006/extended-properties">
+    <Application>Microsoft Office Word</Application>
+    <AppVersion>16.0000</AppVersion>
+    <TotalTime>120</TotalTime>
+    <Pages>5</Pages>
+    <Words>1000</Words>
+    <Characters>5500</Characters>
+    <CharactersWithSpaces>6500</CharactersWithSpaces>
+    <Lines>100</Lines>
+    <Paragraphs>50</Paragraphs>
+    <Company>Acme Corp</Company>
+    <DocSecurity>0</DocSecurity>
+    <ScaleCrop>false</ScaleCrop>
+</Properties>"#;
+
+        let mut archive = create_test_zip_with_app_xml(app_xml);
+        let props = extract_docx_app_properties(&mut archive).unwrap();
+
+        assert_eq!(props.application, Some("Microsoft Office Word".to_string()));
+        assert_eq!(props.app_version, Some("16.0000".to_string()));
+        assert_eq!(props.total_time, Some(120));
+        assert_eq!(props.pages, Some(5));
+        assert_eq!(props.words, Some(1000));
+        assert_eq!(props.characters, Some(5500));
+        assert_eq!(props.characters_with_spaces, Some(6500));
+        assert_eq!(props.lines, Some(100));
+        assert_eq!(props.paragraphs, Some(50));
+        assert_eq!(props.company, Some("Acme Corp".to_string()));
+        assert_eq!(props.doc_security, Some(0));
+        assert_eq!(props.scale_crop, Some(false));
+    }
+
+    #[test]
+    fn test_extract_xlsx_app_properties_with_worksheets() {
+        let app_xml = r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Properties xmlns="http://schemas.openxmlformats.org/officeDocument/2006/extended-properties"
+            xmlns:vt="http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes">
+    <Application>Microsoft Excel</Application>
+    <AppVersion>16.0300</AppVersion>
+    <Company>Test Company</Company>
+    <TitlesOfParts>
+        <vt:vector size="3" baseType="lpstr">
+            <vt:lpstr>Sheet1</vt:lpstr>
+            <vt:lpstr>Sheet2</vt:lpstr>
+            <vt:lpstr>Sheet3</vt:lpstr>
+        </vt:vector>
+    </TitlesOfParts>
+</Properties>"#;
+
+        let mut archive = create_test_zip_with_app_xml(app_xml);
+        let props = extract_xlsx_app_properties(&mut archive).unwrap();
+
+        assert_eq!(props.application, Some("Microsoft Excel".to_string()));
+        assert_eq!(props.app_version, Some("16.0300".to_string()));
+        assert_eq!(props.company, Some("Test Company".to_string()));
+        assert_eq!(props.worksheet_names, vec!["Sheet1", "Sheet2", "Sheet3"]);
+    }
+
+    #[test]
+    fn test_extract_pptx_app_properties() {
+        let app_xml = r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<Properties xmlns="http://schemas.openxmlformats.org/officeDocument/2006/extended-properties"
+            xmlns:vt="http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes">
+    <Application>Microsoft Office PowerPoint</Application>
+    <AppVersion>16.0000</AppVersion>
+    <TotalTime>45</TotalTime>
+    <Slides>10</Slides>
+    <Notes>5</Notes>
+    <HiddenSlides>2</HiddenSlides>
+    <MMClips>3</MMClips>
+    <PresentationFormat>Widescreen</PresentationFormat>
+    <TitlesOfParts>
+        <vt:vector size="2" baseType="lpstr">
+            <vt:lpstr>Title Slide</vt:lpstr>
+            <vt:lpstr>Agenda</vt:lpstr>
+        </vt:vector>
+    </TitlesOfParts>
+</Properties>"#;
+
+        let mut archive = create_test_zip_with_app_xml(app_xml);
+        let props = extract_pptx_app_properties(&mut archive).unwrap();
+
+        assert_eq!(props.application, Some("Microsoft Office PowerPoint".to_string()));
+        assert_eq!(props.slides, Some(10));
+        assert_eq!(props.notes, Some(5));
+        assert_eq!(props.hidden_slides, Some(2));
+        assert_eq!(props.multimedia_clips, Some(3));
+        assert_eq!(props.presentation_format, Some("Widescreen".to_string()));
+        assert_eq!(props.slide_titles, vec!["Title Slide", "Agenda"]);
+    }
+
+    #[test]
+    fn test_extract_app_properties_missing_file() {
+        let buffer = Vec::new();
+        let cursor = Cursor::new(buffer);
+        let zip = zip::ZipWriter::new(cursor);
+        let cursor = zip.finish().unwrap();
+        let mut archive = ZipArchive::new(cursor).unwrap();
+
+        let docx = extract_docx_app_properties(&mut archive).unwrap();
+        assert_eq!(docx, DocxAppProperties::default());
+
+        let buffer = Vec::new();
+        let cursor = Cursor::new(buffer);
+        let zip = zip::ZipWriter::new(cursor);
+        let cursor = zip.finish().unwrap();
+        let mut archive = ZipArchive::new(cursor).unwrap();
+
+        let xlsx = extract_xlsx_app_properties(&mut archive).unwrap();
+        assert_eq!(xlsx, XlsxAppProperties::default());
+    }
+
+    #[test]
+    fn test_extract_titles_of_parts_empty() {
+        let xml = r#"<Properties xmlns:vt="http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes">
+            <TitlesOfParts><vt:vector size="0" baseType="lpstr"></vt:vector></TitlesOfParts>
+        </Properties>"#;
+
+        let doc = roxmltree::Document::parse(xml).unwrap();
+        let titles = extract_titles_of_parts(doc.root_element());
+        assert_eq!(titles, Vec::<String>::new());
+    }
+}