Recursing through the ast to find the lists.

Cargo.lock

@@ -30,18 +30,17 @@ dependencies = [
 
 [[package]]
 name = "axum"
-version = "0.6.20"
+version = "0.6.16"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/tokio-rs/axum.git?rev=52a90390195e884bcc12ff5bd9fd805cac806447#52a90390195e884bcc12ff5bd9fd805cac806447"
-checksum = "3b829e4e32b91e643de6eafe82b1d90675f5874230191a4ffbc1b336dec4d6bf"
 dependencies = [
  "async-trait",
  "axum-core",
- "bitflags",
  "bytes",
  "futures-util",
  "http",
- "http-body",
+ "http-body 0.4.5",
- "hyper",
+ "hyper 0.14.27",
+ "hyper 1.0.0-rc.4",
  "itoa",
  "matchit",
  "memchr",
@@ -56,6 +55,7 @@ dependencies = [
  "sync_wrapper",
  "tokio",
  "tower",
+ "tower-hyper-http-body-compat",
  "tower-layer",
  "tower-service",
 ]
@@ -63,16 +63,17 @@ dependencies = [
 [[package]]
 name = "axum-core"
 version = "0.3.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
+source = "git+https://github.com/tokio-rs/axum.git?rev=52a90390195e884bcc12ff5bd9fd805cac806447#52a90390195e884bcc12ff5bd9fd805cac806447"
-checksum = "759fa577a247914fd3f7f76d62972792636412fbfd634cd452f6a385a74d2d2c"
 dependencies = [
  "async-trait",
  "bytes",
  "futures-util",
  "http",
- "http-body",
+ "http-body 0.4.5",
  "mime",
+ "pin-project-lite",
  "rustversion",
+ "sync_wrapper",
  "tower-layer",
  "tower-service",
 ]
@@ -94,9 +95,9 @@ dependencies = [
 
 [[package]]
 name = "bitflags"
-version = "1.3.2"
+version = "2.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bef38d45163c2f1dde094a7dfd33ccf595c92905c8f8f4fdc18d06fb1037718a"
+checksum = "b4682ae6287fcf752ecaabbfcc7b6f9b72aa33933dc23a554d853aea8eea8635"
 
 [[package]]
 name = "bytes"
@@ -149,6 +150,12 @@ version = "0.3.28"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4bca583b7e26f571124fe5b7561d49cb2868d79116cfa0eefce955557c6fee8c"
 
+[[package]]
+name = "futures-sink"
+version = "0.3.28"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f43be4fe21a13b9781a69afa4985b0f6ee0e1afab2c6f454a8cf30e2b2237b6e"
+
 [[package]]
 name = "futures-task"
 version = "0.3.28"
@@ -201,6 +208,22 @@ dependencies = [
  "pin-project-lite",
 ]
 
+[[package]]
+name = "http-body"
+version = "1.0.0-rc.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "951dfc2e32ac02d67c90c0d65bd27009a635dc9b381a2cc7d284ab01e3a0150d"
+dependencies = [
+ "bytes",
+ "http",
+]
+
+[[package]]
+name = "http-range-header"
+version = "0.3.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "add0ab9360ddbd88cfeb3bd9574a1d85cfdfa14db10b3e21d3700dbc4328758f"
+
 [[package]]
 name = "httparse"
 version = "1.8.0"
@@ -224,7 +247,7 @@ dependencies = [
  "futures-core",
  "futures-util",
  "http",
- "http-body",
+ "http-body 0.4.5",
  "httparse",
  "httpdate",
  "itoa",
@@ -236,6 +259,26 @@ dependencies = [
  "want",
 ]
 
+[[package]]
+name = "hyper"
+version = "1.0.0-rc.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d280a71f348bcc670fc55b02b63c53a04ac0bf2daff2980795aeaf53edae10e6"
+dependencies = [
+ "bytes",
+ "futures-channel",
+ "futures-util",
+ "http",
+ "http-body 1.0.0-rc.2",
+ "httparse",
+ "httpdate",
+ "itoa",
+ "pin-project-lite",
+ "tokio",
+ "tracing",
+ "want",
+]
+
 [[package]]
 name = "itoa"
 version = "1.0.9"
@@ -272,6 +315,16 @@ version = "0.3.17"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "6877bb514081ee2a7ff5ef9de3281f14a4dd4bceac4c09388074a6b5df8a139a"
 
+[[package]]
+name = "mime_guess"
+version = "2.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4192263c238a5f0d0c6bfd21f336a313a4ce1c450542449ca191bb657b4642ef"
+dependencies = [
+ "mime",
+ "unicase",
+]
+
 [[package]]
 name = "minimal-lexical"
 version = "0.2.1"
@@ -339,7 +392,9 @@ version = "0.1.0"
 dependencies = [
  "axum",
  "nom",
+ "serde",
  "tokio",
+ "tower-http",
 ]
 
 [[package]]
@@ -421,6 +476,20 @@ name = "serde"
 version = "1.0.183"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "32ac8da02677876d532745a130fc9d8e6edfa81a269b107c5b00829b91d8eb3c"
+dependencies = [
+ "serde_derive",
+]
+
+[[package]]
+name = "serde_derive"
+version = "1.0.183"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "aafe972d60b0b9bee71a91b92fee2d4fb3c9d7e8f6b179aa99f27203d99a4816"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
 
 [[package]]
 name = "serde_json"
@@ -455,6 +524,15 @@ dependencies = [
  "serde",
 ]
 
+[[package]]
+name = "signal-hook-registry"
+version = "1.4.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d8229b473baa5980ac72ef434c4415e70c4b5e71b423043adb4ba059f89c99a1"
+dependencies = [
+ "libc",
+]
+
 [[package]]
 name = "socket2"
 version = "0.4.9"
@@ -499,14 +577,41 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "17ed6077ed6cd6c74735e21f37eb16dc3935f96878b1fe961074089cc80893f9"
 dependencies = [
  "backtrace",
+ "bytes",
  "libc",
  "mio",
  "num_cpus",
  "pin-project-lite",
+ "signal-hook-registry",
  "socket2 0.5.3",
+ "tokio-macros",
  "windows-sys",
 ]
 
+[[package]]
+name = "tokio-macros"
+version = "2.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "630bdcf245f78637c13ec01ffae6187cca34625e8c63150d424b59e55af2675e"
+dependencies = [
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "tokio-util"
+version = "0.7.8"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "806fe8c2c87eccc8b3267cbae29ed3ab2d0bd37fca70ab622e46aaa9375ddb7d"
+dependencies = [
+ "bytes",
+ "futures-core",
+ "futures-sink",
+ "pin-project-lite",
+ "tokio",
+]
+
 [[package]]
 name = "tower"
 version = "0.4.13"
@@ -523,6 +628,46 @@ dependencies = [
  "tracing",
 ]
 
+[[package]]
+name = "tower-http"
+version = "0.4.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "55ae70283aba8d2a8b411c695c437fe25b8b5e44e23e780662002fc72fb47a82"
+dependencies = [
+ "bitflags",
+ "bytes",
+ "futures-core",
+ "futures-util",
+ "http",
+ "http-body 0.4.5",
+ "http-range-header",
+ "httpdate",
+ "mime",
+ "mime_guess",
+ "percent-encoding",
+ "pin-project-lite",
+ "tokio",
+ "tokio-util",
+ "tower-layer",
+ "tower-service",
+ "tracing",
+]
+
+[[package]]
+name = "tower-hyper-http-body-compat"
+version = "0.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a7ea3e622710ee44a8255baa6fcb2a34d67450e2ffb48e5e58d5a7bd6ff55a21"
+dependencies = [
+ "http",
+ "http-body 0.4.5",
+ "http-body 1.0.0-rc.2",
+ "hyper 1.0.0-rc.4",
+ "pin-project-lite",
+ "tower",
+ "tower-service",
+]
+
 [[package]]
 name = "tower-layer"
 version = "0.3.2"
@@ -562,12 +707,27 @@ version = "0.2.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3528ecfd12c466c6f163363caf2d02a71161dd5e1cc6ae7b34207ea2d42d81ed"
 
+[[package]]
+name = "unicase"
+version = "2.6.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "50f37be617794602aabbeee0be4f259dc1778fabe05e2d67ee8f79326d5cb4f6"
+dependencies = [
+ "version_check",
+]
+
 [[package]]
 name = "unicode-ident"
 version = "1.0.11"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "301abaae475aa91687eb82514b328ab47a211a533026cb25fc3e519b86adfc3c"
 
+[[package]]
+name = "version_check"
+version = "0.9.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "49874b5167b65d7193b8aba1567f5c7d93d001cafc34600cee003eda787e483f"
+
 [[package]]
 name = "want"
 version = "0.3.1"

Cargo.toml

@@ -4,6 +4,8 @@ version = "0.1.0"
 edition = "2021"
 
 [dependencies]
-axum = "0.6.20"
+axum = { git = "https://github.com/tokio-rs/axum.git", rev = "52a90390195e884bcc12ff5bd9fd805cac806447" }
 nom = "7.1.1"
-tokio = { version = "1.30.0", default-features = false, features = ["rt", "rt-multi-thread"] }
+serde = { version = "1.0.183", features = ["derive"] }
+tokio = { version = "1.30.0", default-features = false, features = ["macros", "process", "rt", "rt-multi-thread"] }
+tower-http = { version = "0.4.3", features = ["fs"] }

src/error.rs

@@ -0,0 +1,25 @@
+use nom::error::ErrorKind;
+use nom::error::ParseError;
+use nom::IResult;
+
+pub type Res<T, U> = IResult<T, U, CustomError<T>>;
+
+#[derive(Debug, PartialEq)]
+pub enum CustomError<I> {
+    MyError(MyError<I>),
+    Nom(I, ErrorKind),
+}
+
+#[derive(Debug, PartialEq)]
+pub struct MyError<I>(pub I);
+
+impl<I> ParseError<I> for CustomError<I> {
+    fn from_error_kind(input: I, kind: ErrorKind) -> Self {
+        CustomError::Nom(input, kind)
+    }
+
+    fn append(_input: I, _kind: ErrorKind, /*mut*/ other: Self) -> Self {
+        // Doesn't do append like VerboseError
+        other
+    }
+}

src/main.rs

@@ -1,3 +1,37 @@
-fn main() {
+#![feature(exit_status_error)]
-    println!("Hello, world!");
+use axum::{http::StatusCode, routing::post, Json, Router};
+use owner_tree::{build_owner_tree, OwnerTree};
+use parse::emacs_parse_org_document;
+use tower_http::services::{ServeDir, ServeFile};
+
+mod error;
+mod owner_tree;
+mod parse;
+mod sexp;
+
+#[tokio::main]
+async fn main() {
+    let serve_dir = ServeDir::new("static").not_found_service(ServeFile::new("static/index.html"));
+    let app = Router::new()
+        .route("/parse", post(parse_org_mode))
+        .fallback_service(serve_dir);
+
+    let listener = tokio::net::TcpListener::bind("0.0.0.0:3000").await.unwrap();
+    axum::serve(listener, app).await.unwrap();
+}
+
+async fn parse_org_mode(
+    body: String,
+) -> Result<(StatusCode, Json<OwnerTree>), (StatusCode, String)> {
+    _parse_org_mode(body)
+        .await
+        .map_err(|e| (StatusCode::BAD_REQUEST, e.to_string()))
+}
+
+async fn _parse_org_mode(
+    body: String,
+) -> Result<(StatusCode, Json<OwnerTree>), Box<dyn std::error::Error>> {
+    let ast = emacs_parse_org_document(&body).await?;
+    let owner_tree = build_owner_tree(body.as_str(), ast.as_str()).map_err(|e| e.to_string())?;
+    Ok((StatusCode::OK, Json(owner_tree)))
 }

src/owner_tree.rs

@@ -0,0 +1,217 @@
+use serde::Serialize;
+
+use crate::sexp::{sexp_with_padding, Token};
+
+pub fn build_owner_tree<'a>(
+    body: &'a str,
+    ast_raw: &'a str,
+) -> Result<OwnerTree, Box<dyn std::error::Error + 'a>> {
+    let (_remaining, parsed_sexp) = sexp_with_padding(ast_raw)?;
+    let lists = find_lists_in_document(body, &parsed_sexp)?;
+
+    Ok(OwnerTree {
+        input: body.to_owned(),
+        ast: ast_raw.to_owned(),
+        lists,
+    })
+}
+
+#[derive(Serialize)]
+pub struct OwnerTree {
+    input: String,
+    ast: String,
+    lists: Vec<PlainList>,
+}
+
+#[derive(Serialize)]
+pub struct PlainList {
+    position: SourceRange,
+    items: Vec<PlainListItem>,
+}
+
+#[derive(Serialize)]
+pub struct PlainListItem {
+    position: SourceRange,
+    lists: Vec<PlainList>,
+}
+
+#[derive(Serialize)]
+pub struct SourceRange {
+    start_line: u32,
+    end_line: u32, // Exclusive
+    start_character: u32,
+    end_character: u32, // Exclusive
+}
+
+fn find_lists_in_document<'a>(
+    original_source: &str,
+    current_token: &Token<'a>,
+) -> Result<Vec<PlainList>, Box<dyn std::error::Error>> {
+    // DFS looking for top-level lists
+
+    let mut found_lists = Vec::new();
+    let children = current_token.as_list()?;
+    let token_name = "org-data";
+    assert_name(current_token, token_name)?;
+
+    // skip 2 to skip token name and standard properties
+    for child_token in children.iter().skip(2) {
+        found_lists.extend(recurse_token(original_source, child_token)?);
+    }
+
+    Ok(found_lists)
+}
+
+fn recurse_token<'a>(
+    original_source: &str,
+    current_token: &Token<'a>,
+) -> Result<Vec<PlainList>, Box<dyn std::error::Error>> {
+    match current_token {
+        Token::Atom(_) | Token::TextWithProperties(_) => Ok(Vec::new()),
+        Token::List(_) => {
+            let new_lists = find_lists_in_list(original_source, current_token)?;
+            Ok(new_lists)
+        }
+        Token::Vector(_) => {
+            let new_lists = find_lists_in_vector(original_source, current_token)?;
+            Ok(new_lists)
+        }
+    }
+}
+
+fn find_lists_in_list<'a>(
+    original_source: &str,
+    current_token: &Token<'a>,
+) -> Result<Vec<PlainList>, Box<dyn std::error::Error>> {
+    let mut found_lists = Vec::new();
+    let children = current_token.as_list()?;
+    if assert_name(current_token, "plain-list").is_ok() {
+        // Found a list!
+        let mut found_items = Vec::new();
+        // skip 2 to skip token name and standard properties
+        for child_token in children.iter().skip(2) {
+            found_items.push(get_item_in_list(original_source, child_token)?);
+        }
+
+        found_lists.push(PlainList {
+            position: get_bounds(original_source, current_token)?,
+            items: found_items,
+        });
+    } else {
+        // skip 2 to skip token name and standard properties
+        for child_token in children.iter().skip(2) {
+            found_lists.extend(recurse_token(original_source, child_token)?);
+        }
+    }
+
+    Ok(found_lists)
+}
+
+fn find_lists_in_vector<'a>(
+    original_source: &str,
+    current_token: &Token<'a>,
+) -> Result<Vec<PlainList>, Box<dyn std::error::Error>> {
+    let mut found_lists = Vec::new();
+    let children = current_token.as_vector()?;
+
+    for child_token in children.iter() {
+        found_lists.extend(recurse_token(original_source, child_token)?);
+    }
+
+    Ok(found_lists)
+}
+
+fn get_item_in_list<'a>(
+    original_source: &str,
+    current_token: &Token<'a>,
+) -> Result<PlainListItem, Box<dyn std::error::Error>> {
+    let mut found_lists = Vec::new();
+    let children = current_token.as_list()?;
+    let token_name = "item";
+    assert_name(current_token, token_name)?;
+
+    // skip 2 to skip token name and standard properties
+    for child_token in children.iter().skip(2) {
+        found_lists.extend(recurse_token(original_source, child_token)?);
+    }
+
+    Ok(PlainListItem {
+        position: get_bounds(original_source, current_token)?,
+        lists: found_lists,
+    })
+}
+
+fn assert_name<'s>(emacs: &'s Token<'s>, name: &str) -> Result<(), Box<dyn std::error::Error>> {
+    let children = emacs.as_list()?;
+    let first_child = children
+        .first()
+        .ok_or("Should have at least one child.")?
+        .as_atom()?;
+    if first_child != name {
+        Err(format!(
+            "Expected a {expected} cell, but found a {found} cell.",
+            expected = name,
+            found = first_child
+        ))?;
+    }
+    Ok(())
+}
+
+fn get_bounds<'s>(
+    original_source: &'s str,
+    emacs: &'s Token<'s>,
+) -> Result<SourceRange, Box<dyn std::error::Error>> {
+    let children = emacs.as_list()?;
+    let attributes_child = children
+        .iter()
+        .nth(1)
+        .ok_or("Should have an attributes child.")?;
+    let attributes_map = attributes_child.as_map()?;
+    let standard_properties = attributes_map.get(":standard-properties");
+    let (begin, end) = if standard_properties.is_some() {
+        let std_props = standard_properties
+            .expect("if statement proves its Some")
+            .as_vector()?;
+        let begin = std_props
+            .get(0)
+            .ok_or("Missing first element in standard properties")?
+            .as_atom()?;
+        let end = std_props
+            .get(1)
+            .ok_or("Missing first element in standard properties")?
+            .as_atom()?;
+        (begin, end)
+    } else {
+        let begin = attributes_map
+            .get(":begin")
+            .ok_or("Missing :begin attribute.")?
+            .as_atom()?;
+        let end = attributes_map
+            .get(":end")
+            .ok_or("Missing :end attribute.")?
+            .as_atom()?;
+        (begin, end)
+    };
+    let begin = begin.parse::<u32>()?;
+    let end = end.parse::<u32>()?;
+    let start_line = original_source
+        .chars()
+        .into_iter()
+        .take(usize::try_from(begin)? - 1)
+        .filter(|x| *x == '\n')
+        .count()
+        + 1;
+    let end_line = original_source
+        .chars()
+        .into_iter()
+        .take(usize::try_from(end)? - 1)
+        .filter(|x| *x == '\n')
+        .count()
+        + 1;
+    Ok(SourceRange {
+        start_line: u32::try_from(start_line)?,
+        end_line: u32::try_from(end_line)?,
+        start_character: begin,
+        end_character: end,
+    })
+}

src/parse.rs

@@ -0,0 +1,53 @@
+use tokio::process::Command;
+
+pub async fn emacs_parse_org_document<C>(
+    file_contents: C,
+) -> Result<String, Box<dyn std::error::Error>>
+where
+    C: AsRef<str>,
+{
+    let escaped_file_contents = escape_elisp_string(file_contents);
+    let elisp_script = format!(
+        r#"(progn
+     (erase-buffer)
+     (insert "{escaped_file_contents}")
+     (org-mode)
+     (message "%s" (pp-to-string (org-element-parse-buffer)))
+)"#,
+        escaped_file_contents = escaped_file_contents
+    );
+    let mut cmd = Command::new("emacs");
+    let proc = cmd
+        .arg("-q")
+        .arg("--no-site-file")
+        .arg("--no-splash")
+        .arg("--batch")
+        .arg("--eval")
+        .arg(elisp_script);
+
+    let out = proc.output().await?;
+    out.status.exit_ok()?;
+    Ok(String::from_utf8(out.stderr)?)
+}
+
+fn escape_elisp_string<C>(file_contents: C) -> String
+where
+    C: AsRef<str>,
+{
+    let source = file_contents.as_ref();
+    let source_len = source.len();
+    // We allocate a string 10% larger than the source to account for escape characters. Without this, we would have more allocations during processing.
+    let mut output = String::with_capacity(source_len + (source_len / 10));
+    for c in source.chars() {
+        match c {
+            '"' | '\\' => {
+                output.push('\\');
+                output.push(c);
+            }
+            _ => {
+                output.push(c);
+            }
+        }
+    }
+    output
+}

src/sexp.rs

@@ -0,0 +1,368 @@
+use std::collections::HashMap;
+
+use nom::branch::alt;
+use nom::bytes::complete::escaped;
+use nom::bytes::complete::tag;
+use nom::bytes::complete::take_till1;
+use nom::character::complete::multispace0;
+use nom::character::complete::multispace1;
+use nom::character::complete::one_of;
+use nom::combinator::map;
+use nom::combinator::not;
+use nom::combinator::opt;
+use nom::combinator::peek;
+use nom::multi::separated_list1;
+use nom::sequence::delimited;
+use nom::sequence::preceded;
+use nom::sequence::tuple;
+
+use crate::error::Res;
+
+#[derive(Debug)]
+pub enum Token<'s> {
+    Atom(&'s str),
+    List(Vec<Token<'s>>),
+    TextWithProperties(TextWithProperties<'s>),
+    Vector(Vec<Token<'s>>),
+}
+
+#[derive(Debug)]
+pub struct TextWithProperties<'s> {
+    #[allow(dead_code)]
+    pub text: &'s str,
+    #[allow(dead_code)]
+    pub properties: Vec<Token<'s>>,
+}
+
+impl<'s> TextWithProperties<'s> {
+    pub fn unquote(&self) -> Result<String, Box<dyn std::error::Error>> {
+        let mut out = String::with_capacity(self.text.len());
+        if !self.text.starts_with(r#"""#) {
+            return Err("Quoted text does not start with quote.".into());
+        }
+        if !self.text.ends_with(r#"""#) {
+            return Err("Quoted text does not end with quote.".into());
+        }
+        let interior_text = &self.text[1..(self.text.len() - 1)];
+        let mut state = ParseState::Normal;
+        for current_char in interior_text.chars().into_iter() {
+            state = match (state, current_char) {
+                (ParseState::Normal, '\\') => ParseState::Escape,
+                (ParseState::Normal, _) => {
+                    out.push(current_char);
+                    ParseState::Normal
+                }
+                (ParseState::Escape, 'n') => {
+                    out.push('\n');
+                    ParseState::Normal
+                }
+                (ParseState::Escape, '\\') => {
+                    out.push('\\');
+                    ParseState::Normal
+                }
+                (ParseState::Escape, '"') => {
+                    out.push('"');
+                    ParseState::Normal
+                }
+                _ => todo!(),
+            };
+        }
+
+        Ok(out)
+    }
+}
+
+enum ParseState {
+    Normal,
+    Escape,
+}
+
+impl<'s> Token<'s> {
+    pub fn as_vector<'p>(&'p self) -> Result<&'p Vec<Token<'s>>, Box<dyn std::error::Error>> {
+        Ok(match self {
+            Token::Vector(children) => Ok(children),
+            _ => Err(format!("wrong token type {:?}", self)),
+        }?)
+    }
+
+    pub fn as_list<'p>(&'p self) -> Result<&'p Vec<Token<'s>>, Box<dyn std::error::Error>> {
+        Ok(match self {
+            Token::List(children) => Ok(children),
+            _ => Err(format!("wrong token type {:?}", self)),
+        }?)
+    }
+
+    pub fn as_atom<'p>(&'p self) -> Result<&'s str, Box<dyn std::error::Error>> {
+        Ok(match self {
+            Token::Atom(body) => Ok(*body),
+            _ => Err(format!("wrong token type {:?}", self)),
+        }?)
+    }
+
+    pub fn as_text<'p>(&'p self) -> Result<&'p TextWithProperties<'s>, Box<dyn std::error::Error>> {
+        Ok(match self {
+            Token::TextWithProperties(body) => Ok(body),
+            _ => Err(format!("wrong token type {:?}", self)),
+        }?)
+    }
+
+    pub fn as_map<'p>(
+        &'p self,
+    ) -> Result<HashMap<&'s str, &'p Token<'s>>, Box<dyn std::error::Error>> {
+        let mut hashmap = HashMap::new();
+
+        let children = self.as_list()?;
+        if children.len() % 2 != 0 {
+            return Err("Expecting an even number of children".into());
+        }
+        let mut key: Option<&str> = None;
+        for child in children.iter() {
+            match key {
+                None => {
+                    key = Some(child.as_atom()?);
+                }
+                Some(key_val) => {
+                    key = None;
+                    hashmap.insert(key_val, child);
+                }
+            };
+        }
+
+        Ok(hashmap)
+    }
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+pub fn sexp_with_padding<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, _) = multispace0(input)?;
+    let (remaining, tkn) = token(remaining)?;
+    let (remaining, _) = multispace0(remaining)?;
+    Ok((remaining, tkn))
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+pub fn sexp<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, tkn) = token(input)?;
+    Ok((remaining, tkn))
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn token<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    alt((list, vector, atom))(input)
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn list<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, _) = tag("(")(input)?;
+    let (remaining, children) = delimited(
+        multispace0,
+        separated_list1(multispace1, token),
+        multispace0,
+    )(remaining)?;
+    let (remaining, _) = tag(")")(remaining)?;
+    Ok((remaining, Token::List(children)))
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn vector<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, _) = tag("[")(input)?;
+    let (remaining, children) = delimited(
+        multispace0,
+        separated_list1(multispace1, token),
+        multispace0,
+    )(remaining)?;
+    let (remaining, _) = tag("]")(remaining)?;
+    Ok((remaining, Token::Vector(children)))
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn atom<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    not(peek(one_of(")]")))(input)?;
+    alt((
+        text_with_properties,
+        hash_notation,
+        quoted_atom,
+        unquoted_atom,
+    ))(input)
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn unquoted_atom<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, body) = take_till1(|c| match c {
+        ' ' | '\t' | '\r' | '\n' | ')' | ']' => true,
+        _ => false,
+    })(input)?;
+    Ok((remaining, Token::Atom(body)))
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn quoted_atom<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, _) = tag(r#"""#)(input)?;
+    let (remaining, _) = escaped(
+        take_till1(|c| match c {
+            '\\' | '"' => true,
+            _ => false,
+        }),
+        '\\',
+        one_of(r#""n\\"#),
+    )(remaining)?;
+    let (remaining, _) = tag(r#"""#)(remaining)?;
+    let source = get_consumed(input, remaining);
+    Ok((remaining, Token::Atom(source)))
+}
+
+#[cfg_attr(feature = "tracing", tracing::instrument(ret, level = "debug"))]
+fn hash_notation<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, _) = tag("#<")(input)?;
+    let (remaining, _body) = take_till1(|c| match c {
+        '>' => true,
+        _ => false,
+    })(remaining)?;
+    let (remaining, _) = tag(">")(remaining)?;
+    let source = get_consumed(input, remaining);
+    Ok((remaining, Token::Atom(source)))
+}
+
+fn text_with_properties<'s>(input: &'s str) -> Res<&'s str, Token<'s>> {
+    let (remaining, _) = tag("#(")(input)?;
+    let (remaining, (text, props)) = delimited(
+        multispace0,
+        tuple((
+            map(quoted_atom, |atom| match atom {
+                Token::Atom(body) => body,
+                _ => unreachable!(),
+            }),
+            preceded(multispace1, opt(separated_list1(multispace1, token))),
+        )),
+        multispace0,
+    )(remaining)?;
+    let (remaining, _) = tag(")")(remaining)?;
+    Ok((
+        remaining,
+        Token::TextWithProperties(TextWithProperties {
+            text,
+            properties: props.unwrap_or(Vec::new()),
+        }),
+    ))
+}
+
+/// Get a slice of the string that was consumed in a parser using the original input to the parser and the remaining input after the parser.
+fn get_consumed<'s>(input: &'s str, remaining: &'s str) -> &'s str {
+    assert!(is_slice_of(input, remaining));
+    let source = {
+        let offset = remaining.as_ptr() as usize - input.as_ptr() as usize;
+        &input[..offset]
+    };
+    source
+}
+
+/// Check if the child string slice is a slice of the parent string slice.
+fn is_slice_of(parent: &str, child: &str) -> bool {
+    let parent_start = parent.as_ptr() as usize;
+    let parent_end = parent_start + parent.len();
+    let child_start = child.as_ptr() as usize;
+    let child_end = child_start + child.len();
+    child_start >= parent_start && child_end <= parent_end
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn simple() {
+        let input = "  (foo bar baz )  ";
+        let (remaining, parsed) = sexp_with_padding(input).expect("Parse the input");
+        assert_eq!(remaining, "");
+        assert!(match parsed {
+            Token::Atom(_) => false,
+            Token::List(_) => true,
+            Token::TextWithProperties(_) => false,
+            Token::Vector(_) => false,
+        });
+    }
+
+    #[test]
+    fn quoted() {
+        let input = r#"  ("foo" bar baz )  "#;
+        let (remaining, parsed) = sexp_with_padding(input).expect("Parse the input");
+        assert_eq!(remaining, "");
+        assert!(match parsed {
+            Token::Atom(_) => false,
+            Token::List(_) => true,
+            Token::TextWithProperties(_) => false,
+            Token::Vector(_) => false,
+        });
+        let children = match parsed {
+            Token::List(children) => children,
+            _ => panic!("Should be a list."),
+        };
+        assert_eq!(
+            match children.first() {
+                Some(Token::Atom(body)) => *body,
+                _ => panic!("First child should be an atom."),
+            },
+            r#""foo""#
+        )
+    }
+
+    #[test]
+    fn quoted_containing_paren() {
+        let input = r#"  (foo "b(a)r" baz )  "#;
+        let (remaining, parsed) = sexp_with_padding(input).expect("Parse the input");
+        assert_eq!(remaining, "");
+        assert!(match parsed {
+            Token::List(_) => true,
+            _ => false,
+        });
+        let children = match parsed {
+            Token::List(children) => children,
+            _ => panic!("Should be a list."),
+        };
+        assert_eq!(
+            match children.first() {
+                Some(Token::Atom(body)) => *body,
+                _ => panic!("First child should be an atom."),
+            },
+            r#"foo"#
+        );
+        assert_eq!(
+            match children.iter().nth(1) {
+                Some(Token::Atom(body)) => *body,
+                _ => panic!("Second child should be an atom."),
+            },
+            r#""b(a)r""#
+        );
+        assert_eq!(
+            match children.iter().nth(2) {
+                Some(Token::Atom(body)) => *body,
+                _ => panic!("Third child should be an atom."),
+            },
+            r#"baz"#
+        );
+    }
+
+    #[test]
+    fn string_containing_escaped_characters() {
+        let input = r#"  (foo "\\( x=2 \\)" bar)  "#;
+        let (remaining, parsed) = sexp_with_padding(input).expect("Parse the input");
+        assert_eq!(remaining, "");
+        assert!(match parsed {
+            Token::Atom(_) => false,
+            Token::List(_) => true,
+            Token::TextWithProperties(_) => false,
+            Token::Vector(_) => false,
+        });
+        let children = match parsed {
+            Token::List(children) => children,
+            _ => panic!("Should be a list."),
+        };
+        assert_eq!(
+            match children.get(1) {
+                Some(Token::Atom(body)) => *body,
+                _ => panic!("First child should be an atom."),
+            },
+            r#""\\( x=2 \\)""#
+        )
+    }
+}

static/index.html

@@ -0,0 +1,5 @@
+<html>
+  <body>
+    Test html file.
+  </body>
+</html>