epub-read/src/book.rs

pub fn strip_html(input: &str) -> String {
    let mut result = String::with_capacity(input.len());
    let mut in_tag = false;
    let mut in_entity = false;
    let mut entity = String::new();

    for c in input.chars() {
        match c {
            '<' => in_tag = true,
            '>' if in_tag => in_tag = false,
            '&' if !in_tag => {
                in_entity = true;
                entity.clear();
            }
            ';' if in_entity => {
                in_entity = false;
                let decoded = match entity.as_str() {
                    "amp" => "&",
                    "lt" => "<",
                    "gt" => ">",
                    "quot" => "\"",
                    "nbsp" => " ",
                    _ => "",
                };
                result.push_str(decoded);
            }
            c if !in_tag && !in_entity => result.push(c),
            c if in_entity => entity.push(c),
            _ => {}
        }
    }
    result
}

#[derive(Debug, Clone)]
pub struct TocEntry {
    pub label: String,
    pub section: usize,
    pub children: Vec<TocEntry>,
}

#[derive(Debug, Clone)]
pub struct Section {
    pub title: String,
    pub content: String,
    pub pages: Vec<usize>,
}

#[derive(Debug, Clone)]
pub struct Book {
    pub title: String,
    pub author: String,
    pub cover: Option<Vec<u8>>,
    pub sections: Vec<Section>,
    pub toc: Vec<TocEntry>,
}

use std::path::Path;

pub fn load_epub(path: impl AsRef<Path>) -> Result<Book, String> {
    let path = path.as_ref();
    let mut doc = epub::doc::EpubDoc::new(path)
        .map_err(|e| format!("无法打开文件: {}", e))?;

    let title = doc.mdata("title").unwrap_or_else(|| "未知标题".to_string());
    let author = doc.mdata("creator").unwrap_or_else(|| "未知作者".to_string());
    let cover = doc.get_cover().ok();
    let spine = doc.spine.clone();
    let raw_toc = std::mem::take(&mut doc.toc);

    let mut sections = Vec::new();
    for (i, href) in spine.iter().enumerate() {
        let raw_html = doc.get_resource_str(href)
            .map_err(|e| format!("读取章节失败: {}", e))?;
        let text = strip_html(&raw_html);
        let title = extract_title(&raw_html)
            .unwrap_or_else(|| format!("第{}章", i + 1));
        sections.push(Section {
            title,
            content: text,
            pages: Vec::new(),
        });
    }

    let toc = build_toc(&raw_toc, &spine);

    Ok(Book { title, author, cover, sections, toc })
}

fn extract_title(html: &str) -> Option<String> {
    if let Some(start) = html.find("<title>") {
        let rest = &html[start + 7..];
        if let Some(end) = rest.find("</title>") {
            return Some(strip_html(&rest[..end]).trim().to_string());
        }
    }
    if let Some(start) = html.find("<h1") {
        let rest = &html[start..];
        if let Some(content_start) = rest.find('>') {
            let inner = &rest[content_start + 1..];
            if let Some(end) = inner.find("</h1>") {
                return Some(strip_html(&inner[..end]).trim().to_string());
            }
        }
    }
    None
}

fn build_toc(
    entries: &[epub::doc::NavPoint],
    spine: &[String],
) -> Vec<TocEntry> {
    entries
        .iter()
        .map(|e| {
            let content_str = e.content.to_string_lossy();
            let section = spine
                .iter()
                .position(|s| content_str.contains(s.trim_end_matches('/')))
                .unwrap_or(0);
            TocEntry {
                label: e.label.clone(),
                section,
                children: build_toc(&e.children, spine),
            }
        })
        .collect()
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_epub_loader_nonexistent_file() {
        let result = load_epub("nonexistent.epub");
        assert!(result.is_err());
    }

    #[test]
    fn test_strip_html_plain_text() {
        assert_eq!(strip_html("Hello World"), "Hello World");
    }

    #[test]
    fn test_strip_html_simple_tags() {
        assert_eq!(strip_html("<p>Hello</p>"), "Hello");
    }

    #[test]
    fn test_strip_html_nested_tags() {
        assert_eq!(
            strip_html("<div><p>Hello <b>World</b></p></div>"),
            "Hello World"
        );
    }

    #[test]
    fn test_strip_html_html_entities() {
        assert_eq!(strip_html("Hello &amp; World"), "Hello & World");
        assert_eq!(strip_html("Hello&nbsp;World"), "Hello World");
    }

    #[test]
    fn test_strip_html_empty() {
        assert_eq!(strip_html(""), "");
    }
}