Add anchor-based TOC navigation: parse #fragment anchors, find exact page within section

src/book.rs

@@ -43,10 +43,33 @@ fn tag_name_from(tag_content: &str) -> &str {
         .trim_end_matches('/')
 }
 
+fn extract_id_from_tag(tag_content: &str) -> Option<String> {
+    if let Some(id_pos) = tag_content.find("id=\"") {
+        let after_quote = &tag_content[id_pos + 4..];
+        if let Some(end_quote) = after_quote.find('\"') {
+            let id_val = &after_quote[..end_quote];
+            if !id_val.is_empty() {
+                return Some(id_val.to_string());
+            }
+        }
+    }
+    if let Some(id_pos) = tag_content.find("id='") {
+        let after_quote = &tag_content[id_pos + 4..];
+        if let Some(end_quote) = after_quote.find('\'') {
+            let id_val = &after_quote[..end_quote];
+            if !id_val.is_empty() {
+                return Some(id_val.to_string());
+            }
+        }
+    }
+    None
+}
+
 pub fn strip_html(input: &str) -> String {
     let mut out = String::with_capacity(input.len());
     let mut pos = 0;
     let mut heading_level: Option<u32> = None;
+    let mut pending_anchor: Option<String> = None;
 
     while pos < input.len() {
         // Find next '<'
@@ -68,8 +91,14 @@ pub fn strip_html(input: &str) -> String {
             if let Some(level) = heading_level {
                 out.push('\x01');
                 out.push(char::from_digit(level, 10).unwrap_or('1'));
+                if let Some(ref anchor) = pending_anchor {
+                    out.push('\x03');
+                    out.push_str(anchor);
+                    out.push('\x04');
+                }
                 out.push_str(&text);
                 out.push('\x02');
+                pending_anchor = None;
             } else {
                 out.push_str(&text);
             }
@@ -101,6 +130,13 @@ pub fn strip_html(input: &str) -> String {
                 }
                 pos = tag_end + 1;
             }
+            "a" => {
+                // Capture anchor id
+                if heading_level.is_none() {
+                    pending_anchor = extract_id_from_tag(tag_content);
+                }
+                pos = tag_end + 1;
+            }
             "br" => {
                 if !out.is_empty() {
                     out.push('\n');
@@ -124,13 +160,18 @@ pub fn strip_html(input: &str) -> String {
             "/li" | "/dd" | "/dt" | "/ol" | "/ul" => {
                 pos = tag_end + 1;
             }
-            "h1" => { heading_level = Some(1); pos = tag_end + 1; }
-            "h2" => { heading_level = Some(2); pos = tag_end + 1; }
-            "h3" => { heading_level = Some(3); pos = tag_end + 1; }
-            "h4" => { heading_level = Some(4); pos = tag_end + 1; }
-            "h5" => { heading_level = Some(5); pos = tag_end + 1; }
-            "h6" => { heading_level = Some(6); pos = tag_end + 1; }
+            "h1" | "h2" | "h3" | "h4" | "h5" | "h6" => {
+                let level = name[1..2].parse::<u32>().unwrap_or(1);
+                heading_level = Some(level);
+                if pending_anchor.is_none() {
+                    pending_anchor = extract_id_from_tag(tag_content);
+                }
+                pos = tag_end + 1;
+            }
             "p" | "div" | "blockquote" => {
+                if pending_anchor.is_none() {
+                    pending_anchor = extract_id_from_tag(tag_content);
+                }
                 pos = tag_end + 1;
             }
             "/p" | "/div" | "/blockquote" => {
@@ -138,10 +179,12 @@ pub fn strip_html(input: &str) -> String {
                     out.push('\n');
                 }
                 out.push('\n');
+                pending_anchor = None;
                 pos = tag_end + 1;
             }
             "/h1" | "/h2" | "/h3" | "/h4" | "/h5" | "/h6" => {
                 heading_level = None;
+                pending_anchor = None;
                 if !out.is_empty() && !out.ends_with('\n') {
                     out.push('\n');
                 }
@@ -149,6 +192,9 @@ pub fn strip_html(input: &str) -> String {
                 pos = tag_end + 1;
             }
             _ => {
+                if pending_anchor.is_none() {
+                    pending_anchor = extract_id_from_tag(tag_content);
+                }
                 pos = tag_end + 1;
             }
         }
@@ -177,6 +223,7 @@ pub fn strip_html(input: &str) -> String {
 pub struct TocEntry {
     pub label: String,
     pub section: usize,
+    pub anchor: Option<String>,
     pub children: Vec<TocEntry>,
 }
 
@@ -184,6 +231,7 @@ pub struct TocEntry {
 pub struct ContentBlock {
     pub text: String,
     pub heading_level: u8, // 0 = body, 1-6 = h1-h6
+    pub anchor: Option<String>,
 }
 
 #[derive(Debug, Clone)]
@@ -307,6 +355,19 @@ fn extract_filename(path: &str) -> &str {
     path.rsplit('/').next().unwrap_or(path)
 }
 
+fn extract_fragment(path: &str) -> Option<String> {
+    if let Some(hash_pos) = path.find('#') {
+        let fragment = &path[hash_pos + 1..];
+        if !fragment.is_empty() {
+            Some(fragment.to_string())
+        } else {
+            None
+        }
+    } else {
+        None
+    }
+}
+
 fn build_toc(
     entries: &[epub::doc::NavPoint],
     spine: &[String],
@@ -315,6 +376,7 @@ fn build_toc(
         .iter()
         .map(|e| {
             let content_str = e.content.to_string_lossy();
+            let anchor = extract_fragment(&content_str);
             let content_file = extract_filename(&content_str);
             let section = spine
                 .iter()
@@ -328,6 +390,7 @@ fn build_toc(
             TocEntry {
                 label: e.label.clone(),
                 section,
+                anchor,
                 children: build_toc(&e.children, spine),
             }
         })