fixed parser bug

Add import sorting options to .rustfmt.toml.

.rustfmt.toml

@@ -9,3 +9,5 @@ brace_style = "PreferSameLine"
 blank_lines_upper_bound = 5
 combine_control_expr = false
 wrap_comments = false
+group_imports = "StdExternalCrate"
+imports_granularity = "Crate"

Cargo.lock

@@ -225,6 +225,27 @@ dependencies = [
  "syn 2.0.104",
 ]
 
+[[package]]
+name = "dirs"
+version = "6.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c3e8aa94d75141228480295a7d0e7feb620b1a5ad9f12bc40be62411e38cce4e"
+dependencies = [
+ "dirs-sys",
+]
+
+[[package]]
+name = "dirs-sys"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e01a3366d27ee9890022452ee61b2b63a67e6f13f58900b651ff5665f0bb1fab"
+dependencies = [
+ "libc",
+ "option-ext",
+ "redox_users",
+ "windows-sys 0.60.2",
+]
+
 [[package]]
 name = "displaydoc"
 version = "0.2.5"
@@ -558,6 +579,7 @@ name = "htmlua-parser"
 version = "0.1.0"
 dependencies = [
  "anyhow",
+ "dirs",
  "html5ever 0.35.0",
  "httptest",
  "kuchikiki",
@@ -569,6 +591,7 @@ dependencies = [
  "serde_json",
  "syntect",
  "tendril",
+ "toml",
 ]
 
 [[package]]
@@ -919,6 +942,16 @@ version = "0.2.174"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "1171693293099992e19cddea4e8b849964e9846f4acee11b3948bcc337be8776"
 
+[[package]]
+name = "libredox"
+version = "0.1.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1580801010e535496706ba011c15f8532df6b42297d2e471fec38ceadd8c0638"
+dependencies = [
+ "bitflags 2.9.1",
+ "libc",
+]
+
 [[package]]
 name = "linked-hash-map"
 version = "0.5.6"
@@ -1209,6 +1242,12 @@ dependencies = [
  "vcpkg",
 ]
 
+[[package]]
+name = "option-ext"
+version = "0.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "04744f49eae99ab78e0d5c0b603ab218f515ea8cfe5a456d7629ad883a3b6e7d"
+
 [[package]]
 name = "ordered-float"
 version = "2.10.1"
@@ -1397,9 +1436,9 @@ checksum = "7edddbd0b52d732b21ad9a5fab5c704c14cd949e5e9a1ec5929a24fded1b904c"
 
 [[package]]
 name = "plist"
-version = "1.7.3"
+version = "1.7.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "546b279bf0638ee811d9e47de2ca5b66575a543035d79fdf83959dd2f5c3b4c3"
+checksum = "3af6b589e163c5a788fab00ce0c0366f6efbb9959c2f9874b224936af7fce7e1"
 dependencies = [
  "base64",
  "indexmap 2.10.0",
@@ -1474,9 +1513,9 @@ checksum = "007d8adb5ddab6f8e3f491ac63566a7d5002cc7ed73901f72057943fa71ae1ae"
 
 [[package]]
 name = "quick-xml"
-version = "0.37.5"
+version = "0.38.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "331e97a1af0bf59823e6eadffe373d7b27f485be8748f71471c662c1f269b7fb"
+checksum = "8927b0664f5c5a98265138b7e3f90aa19a6b21353182469ace36d4ac527b7b1b"
 dependencies = [
  "memchr",
 ]
@@ -1586,6 +1625,17 @@ dependencies = [
  "bitflags 2.9.1",
 ]
 
+[[package]]
+name = "redox_users"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "dd6f9d3d47bdd2ad6945c5015a226ec6155d0bcdfd8f7cd29f86b71f8de99d2b"
+dependencies = [
+ "getrandom 0.2.16",
+ "libredox",
+ "thiserror 2.0.12",
+]
+
 [[package]]
 name = "regex"
 version = "1.11.1"
@@ -1865,6 +1915,15 @@ dependencies = [
  "serde",
 ]
 
+[[package]]
+name = "serde_spanned"
+version = "1.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "40734c41988f7306bb04f0ecf60ec0f3f1caa34290e4e8ea471dcd3346483b83"
+dependencies = [
+ "serde",
+]
+
 [[package]]
 name = "serde_urlencoded"
 version = "0.7.1"
@@ -2023,7 +2082,7 @@ dependencies = [
  "serde",
  "serde_derive",
  "serde_json",
- "thiserror",
+ "thiserror 1.0.69",
  "walkdir",
  "yaml-rust",
 ]
@@ -2085,7 +2144,16 @@ version = "1.0.69"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b6aaf5339b578ea85b50e080feb250a3e8ae8cfcdff9a461c9ec2904bc923f52"
 dependencies = [
- "thiserror-impl",
+ "thiserror-impl 1.0.69",
+]
+
+[[package]]
+name = "thiserror"
+version = "2.0.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "567b8a2dae586314f7be2a752ec7474332959c6460e02bde30d702a66d488708"
+dependencies = [
+ "thiserror-impl 2.0.12",
 ]
 
 [[package]]
@@ -2099,6 +2167,17 @@ dependencies = [
  "syn 2.0.104",
 ]
 
+[[package]]
+name = "thiserror-impl"
+version = "2.0.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7f7cf42b4507d8ea322120659672cf1b9dbb93f8f2d4ecfd6e51350ff5b17a1d"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn 2.0.104",
+]
+
 [[package]]
 name = "time"
 version = "0.3.41"
@@ -2202,6 +2281,45 @@ dependencies = [
  "tokio",
 ]
 
+[[package]]
+name = "toml"
+version = "0.9.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ed0aee96c12fa71097902e0bb061a5e1ebd766a6636bb605ba401c45c1650eac"
+dependencies = [
+ "indexmap 2.10.0",
+ "serde",
+ "serde_spanned",
+ "toml_datetime",
+ "toml_parser",
+ "toml_writer",
+ "winnow",
+]
+
+[[package]]
+name = "toml_datetime"
+version = "0.7.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bade1c3e902f58d73d3f294cd7f20391c1cb2fbcb643b73566bc773971df91e3"
+dependencies = [
+ "serde",
+]
+
+[[package]]
+name = "toml_parser"
+version = "1.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "97200572db069e74c512a14117b296ba0a80a30123fbbb5aa1f4a348f639ca30"
+dependencies = [
+ "winnow",
+]
+
+[[package]]
+name = "toml_writer"
+version = "1.0.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fcc842091f2def52017664b53082ecbbeb5c7731092bad69d2c63050401dfd64"
+
 [[package]]
 name = "tower"
 version = "0.5.2"
@@ -2675,6 +2793,12 @@ version = "0.53.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "271414315aff87387382ec3d271b52d7ae78726f5d44ac98b4f4030c91880486"
 
+[[package]]
+name = "winnow"
+version = "0.7.12"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f3edebf492c8125044983378ecb5766203ad3b4c2f7a922bd7dd207f6d443e95"
+
 [[package]]
 name = "winsafe"
 version = "0.0.19"

cspell.json

@@ -2,7 +2,9 @@
   "words": [
     "htmlua",
     "headermap",
-    "reqwest"
+    "reqwest",
+    "includeelement",
+    "exportelement"
   ],
   "userWords": [
     "mlua"

htmlua-apache-cgi/src/main.rs

@@ -1,6 +1,7 @@
-use htmlua_parser::serve::serve_content;
 use std::env;
 
+use htmlua_parser::serve::serve_content;
+
 fn main() {
     println!("Content-Type: text/html\n");
 

htmlua-parser/Cargo.toml

@@ -5,6 +5,7 @@ edition = "2024"
 
 [dependencies]
 anyhow = "1.0.98"
+dirs = "6.0.0"
 html5ever = "0.35.0"
 httptest = "0.16.3"
 kuchikiki = "0.8.2"
@@ -16,7 +17,7 @@ serde = { version = "1.0.219", features = ["derive"] }
 serde_json = "1.0.140"
 syntect = "5.2.0"
 tendril = "0.4.3"
-
+toml = "0.9.2"
 
 [lints.clippy]
 pedantic = { level = "warn", priority = -1 }

htmlua-parser/src/config.rs

@@ -0,0 +1,100 @@
+use std::{fs, path::PathBuf};
+
+use anyhow::{Context, Result};
+use serde::{Deserialize, Serialize};
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct Config {
+    pub paths: PathConfig,
+    pub server: ServerConfig,
+    pub syntax_highlighting: SyntaxConfig,
+}
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct PathConfig {
+    pub pages: PathBuf,
+    pub components: PathBuf,
+    pub themes: PathBuf,
+}
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct ServerConfig {
+    pub host: String,
+    pub port: u16,
+}
+
+#[derive(Debug, Deserialize, Serialize, Clone)]
+pub struct SyntaxConfig {
+    pub default_theme: String,
+    pub load_custom_themes: bool,
+}
+
+impl Default for Config {
+    fn default() -> Self {
+        Self {
+            paths: PathConfig {
+                pages: PathBuf::from("/var/www/htmlua/pages"),
+                components: PathBuf::from("/var/www/htmlua/components"),
+                themes: PathBuf::from("/var/www/htmlua/themes"),
+            },
+            server: ServerConfig {
+                host: "127.0.0.1".to_string(),
+                port: 8080,
+            },
+            syntax_highlighting: SyntaxConfig {
+                default_theme: "base16-ocean.dark".to_string(),
+                load_custom_themes: true,
+            },
+        }
+    }
+}
+
+impl Config {
+    pub fn load() -> Result<Self> {
+        let config_path = Self::get_config_path();
+        if config_path.exists() {
+            let config_content = fs::read_to_string(&config_path)
+                .with_context(|| format!("Failed to read config file: {}", config_path.display()))?;
+            let config: Config = toml::from_str(&config_content)
+                .with_context(|| format!("Failed to parse config file: {}", config_path.display()))?;
+            Ok(config)
+        } else {
+            let default_config = Config::default();
+            if let Some(parent) = config_path.parent() {
+                fs::create_dir_all(parent)
+                    .with_context(|| format!("Failed to create config directory: {}", parent.display()))?;
+            }
+            let config_content =
+                toml::to_string_pretty(&default_config).context("Failed to serialize default config")?;
+            fs::write(&config_path, config_content)
+                .with_context(|| format!("Failed to write default config to: {}", config_path.display()))?;
+            Ok(default_config)
+        }
+    }
+
+    fn get_config_path() -> PathBuf {
+        if cfg!(windows) {
+            dirs::config_dir()
+                .unwrap_or_else(|| PathBuf::from("C:\\ProgramData"))
+                .join("htmlua")
+                .join("config.toml")
+        } else {
+            PathBuf::from("/etc/htmlua.toml")
+        }
+    }
+
+    pub fn save(&self) -> Result<()> {
+        let config_path = Self::get_config_path();
+        if let Some(parent) = config_path.parent() {
+            fs::create_dir_all(parent)
+                .with_context(|| format!("Failed to create config directory: {}", parent.display()))?;
+        }
+        let config_content = toml::to_string_pretty(self).context("Failed to serialize config")?;
+        fs::write(&config_path, config_content)
+            .with_context(|| format!("Failed to write config to: {}", config_path.display()))?;
+        Ok(())
+    }
+
+    #[must_use]
+    pub fn config_file_path() -> PathBuf { Self::get_config_path() }
+}

htmlua-parser/src/helpers.rs

@@ -6,12 +6,10 @@ use std::{
 
 use anyhow::Result;
 use kuchikiki::NodeRef;
-use markup5ever::{LocalName, Namespace, QualName};
+use markup5ever::{LocalName, QualName, namespace_url, ns};
 use tendril::TendrilSink;
 
 
-
-
 pub fn read_doc_from_file(path: PathBuf) -> Result<NodeRef> {
     let mut file = File::open(path)?;
     let mut reader = BufReader::new(&mut file);
@@ -29,7 +27,7 @@ pub fn read_doc_from_file(path: PathBuf) -> Result<NodeRef> {
     if is_whole_doc {
         Ok(kuchikiki::parse_html().one(page_string))
     } else {
-        let ctx_name = QualName::new(None, Namespace::from("http://www.w3.org/1999/xhtml"), LocalName::from("div"));
+        let ctx_name = QualName::new(None, ns!(), LocalName::from("div"));
         Ok(kuchikiki::parse_fragment(ctx_name, Vec::new()).one(page_string))
     }
 }

htmlua-parser/src/lib.rs

@@ -1,3 +1,4 @@
+pub mod config;
 pub mod helpers;
 pub mod htmlua_stdlib;
 pub mod render;

htmlua-parser/src/render.rs

@@ -7,6 +7,7 @@ use std::{
 
 use anyhow::{Result, anyhow};
 use kuchikiki::{NodeRef, traits::TendrilSink};
+use markup5ever::{LocalName, Namespace, QualName};
 use mlua::Lua;
 use pulldown_cmark::{Options, Parser, html};
 use syntect::{
@@ -17,7 +18,7 @@ use syntect::{
     util::LinesWithEndings,
 };
 
-use crate::{helpers::read_doc_from_file, htmlua_stdlib::create_htmlua_stdlib};
+use crate::{helpers::read_doc_from_file, htmlua_stdlib::create_htmlua_stdlib, serve::get_config};
 
 
 fn build_lua_with_stdout(stdout: &Rc<RefCell<String>>) -> Result<Lua> {
@@ -134,16 +135,15 @@ pub fn expand_template(document: NodeRef, component_path: &PathBuf, include_from
 }
 
 pub fn process_syntax_highlighting(document: NodeRef) -> Result<NodeRef> {
+    let config = get_config();
     let ps = SyntaxSet::load_defaults_newlines();
     let mut ts = ThemeSet::load_defaults();
     let syntax_elements: Vec<_> = match document.select("syntaxhighlight") {
         Ok(e) => e.collect(),
         Err(()) => return Err(anyhow!("Unable to find syntaxhighlight elements")),
     };
-    if !syntax_elements.is_empty() {
-        // TODO: read from config
-        let themes = PathBuf::from("/var/www/htmlua/themes");
-        let _ = ts.add_from_folder(themes);
+    if !syntax_elements.is_empty() && config.syntax_highlighting.load_custom_themes {
+        let _ = ts.add_from_folder(&config.paths.themes);
     }
     for node in syntax_elements {
         let attrs = match node.as_node().as_element() {
@@ -151,7 +151,7 @@ pub fn process_syntax_highlighting(document: NodeRef) -> Result<NodeRef> {
             None => continue,
         };
         let language = attrs.get("lang").unwrap_or("text");
-        let theme_name = attrs.get("theme").unwrap_or("base16-ocean.dark");
+        let theme_name = attrs.get("theme").unwrap_or(&config.syntax_highlighting.default_theme);
         if let Some(text_node) = node.as_node().first_child() {
             if let Some(code_text) = text_node.as_text() {
                 let syntax = ps
@@ -181,9 +181,41 @@ pub fn process_syntax_highlighting(document: NodeRef) -> Result<NodeRef> {
     Ok(document)
 }
 
+pub fn generate_footnotes(document: NodeRef) -> Result<NodeRef> {
+    let Ok(footnote_container) = document.select_first("footnotecontainer") else {
+        return Ok(document);
+    };
+
+    let ctx_name = QualName::new(None, Namespace::from("http://www.w3.org/1999/xhtml"), LocalName::from("div"));
+    for (i, footnote) in document
+        .select("footnote")
+        .map_err(|()| anyhow!("Failed to get footnote"))?
+        .enumerate()
+    {
+        let i = i + 1;
+        let fn_text = footnote.text_contents();
+        let sup_tag = kuchikiki::parse_fragment(ctx_name.clone(), Vec::new())
+            .one(format!("<a href=#ft-text-{i}><sup id=\"ft-sup-{i}\" title=\"{fn_text}\">{i}</sup></a>"))
+            .select_first("a")
+            .map_err(|()| anyhow!("parse err"))?;
+        footnote.as_node().insert_after(sup_tag.as_node().clone());
+        let text_tag = kuchikiki::parse_fragment(ctx_name.clone(), Vec::new())
+            .one(format!("<p id=\"ft-text-{i}\"><a href=#ft-sup-{i}>{i}:</a> {fn_text}</p>"))
+            .select_first("p")
+            .map_err(|()| anyhow!("parse err"))?;
+        footnote_container.as_node().insert_before(text_tag.as_node().clone());
+    }
+    while let Ok(i) = document.select_first("footnote") {
+        i.as_node().detach();
+    }
+    footnote_container.as_node().detach();
+    Ok(document)
+}
+
 #[cfg(test)]
 mod tests {
     use httptest::{Expectation, ServerPool, matchers::*, responders::*};
+    use markup5ever::{namespace_url, ns};
 
     use super::*;
 
@@ -315,7 +347,6 @@ mod tests {
         let document = kuchikiki::parse_html().one(page);
         let d = execute_lua(expand_template(document, &p, None).unwrap()).unwrap();
         let text = d.select_first("span").unwrap().as_node().text_contents();
-        println!("{}", d.to_string());
         assert_eq!(text, "Test from Lua!\n");
         assert!(d.select_first("lua").is_err());
         let text = d.select_first("#inc1").unwrap().as_node().text_contents();
@@ -373,6 +404,74 @@ mod tests {
         assert!(d.select_first("includeelement").is_err());
     }
 
+    #[test]
+    fn include_base() {
+        let page = r#"
+            <include path="include_base.html">
+                <exportelement class="t"><title>title_test</title></exportelement>
+                <exportelement class="b"><span id="tb">body_el</span></exportelement>
+            </include>
+            "#;
+        let mut p = PathBuf::from(env!("CARGO_MANIFEST_DIR"));
+        p.push("tests/components");
+        let ctx_name = QualName::new(None, ns!(html), LocalName::from("div"));
+        let document = kuchikiki::parse_fragment(ctx_name, Vec::new()).one(page);
+        let d = expand_template(document, &p, None).unwrap();
+        let text = d
+            .select_first("head")
+            .unwrap()
+            .as_node()
+            .select_first("title")
+            .unwrap()
+            .as_node()
+            .text_contents();
+        assert_eq!(text, "title_test");
+        let text = d.select_first("#tb").unwrap().as_node().text_contents();
+        assert_eq!(text, "body_el");
+        assert!(d.select_first("exportelement").is_err());
+        assert!(d.select_first("includeelement").is_err());
+    }
+
+    #[test]
+    fn footnotes() {
+        let page = r"
+            <!DOCTYPE html>
+            <html>
+            <head>
+                <title>Basic HTML Page</title>
+            </head>
+            <body>
+                <div>
+                    <p>asdf</p>
+                    <p>asdf<footnote>um actually</footnote></p>
+                    <p>asdf</p>
+                    <p>asdf<footnote>no</footnote></p>
+                </div>
+                <div>
+                <footnotecontainer></footnotecontainer>
+                </div>
+                </body>
+                </html>";
+        let document = kuchikiki::parse_html().one(page);
+        let d = generate_footnotes(document).unwrap();
+        let sup1 = d.select_first("#ft-sup-1").unwrap();
+        assert_eq!(sup1.attributes.borrow().get("title").unwrap(), "um actually");
+        assert_eq!(sup1.text_contents(), "1");
+
+        let sup2 = d.select_first("#ft-sup-2").unwrap();
+        assert_eq!(sup2.attributes.borrow().get("title").unwrap(), "no");
+        assert_eq!(sup2.text_contents(), "2");
+
+        let ft_text_1 = d.select_first("#ft-text-1").unwrap();
+        assert!(ft_text_1.text_contents().contains("um actually"));
+
+        let ft_text_2 = d.select_first("#ft-text-2").unwrap();
+        assert!(ft_text_2.text_contents().contains("no"));
+
+        assert!(d.select_first("footnote").is_err());
+        assert!(d.select_first("footnotecontainer").is_err());
+    }
+
     #[test]
     fn basic_markdown() {
         let page = r"

htmlua-parser/src/serve.rs

@@ -1,19 +1,33 @@
-use crate::helpers::read_doc_from_file;
+use std::{path::Path, sync::OnceLock};
 
-use super::render::{execute_lua, expand_template, process_markdown, process_syntax_highlighting};
-use anyhow::Result;
-use std::path::{Path, PathBuf};
+use anyhow::{Result, anyhow};
+
+use crate::{
+    config::Config,
+    helpers::read_doc_from_file,
+    render::{execute_lua, expand_template, process_markdown, process_syntax_highlighting},
+};
+
+static CONFIG: OnceLock<Config> = OnceLock::new();
+
+pub fn get_config() -> &'static Config {
+    CONFIG.get_or_init(|| {
+        Config::load().unwrap_or_else(|e| {
+            eprintln!("Warning: Failed to load config: {e}");
+            eprintln!("Using default configuration");
+            Config::default()
+        })
+    })
+}
 
 pub fn serve_content(request_uri: &str) -> Result<String> {
-    // TODO: read from config
-    let pages = PathBuf::from("/var/www/htmlua/pages");
-    let components = PathBuf::from("/var/www/htmlua/components");
+    let config = get_config();
     let safe_path = Path::new(request_uri).strip_prefix("/")?;
-    let page_path = pages.join(safe_path);
-
-    let doc = read_doc_from_file(page_path)?;
-
-    let full_doc = expand_template(doc, &components, None)?;
+    let page_path = config.paths.pages.join(safe_path);
+    let doc = read_doc_from_file(page_path)?
+        .select_first("html")
+        .map_err(|()| anyhow!("failed to read doc"))?;
+    let full_doc = expand_template(doc.as_node().to_owned(), &config.paths.components, None)?;
     let markdown_doc = process_markdown(full_doc)?;
     let highlighted_doc = process_syntax_highlighting(markdown_doc)?;
     let executed_doc = execute_lua(highlighted_doc)?;

htmlua-parser/tests/components/include_base.html

@@ -0,0 +1,9 @@
+<html>
+  <head>
+    <includeelement name="t"></includeelement>
+  </head>
+  <body>
+    <h1>Hello World</h1>
+    <includeelement name="b"></includeelement>
+  </body>
+</html>