From f1b2e1dba74d4996118b14a26afc8ee238d3ff7f Mon Sep 17 00:00:00 2001
From: Nico Burns <nico@nicoburns.com>
Date: Thu, 19 Dec 2024 00:34:58 +1300
Subject: [PATCH] Eliminate `regex` and `once_cell` dependencies.

Replace usage with `unicode_categories` crate which comrak already
depends on.
---
 Cargo.lock        |  2 --
 Cargo.toml        |  2 --
 src/html.rs       | 19 ++++++++++++++-----
 src/parser/mod.rs |  3 +--
 4 files changed, 15 insertions(+), 11 deletions(-)

diff --git a/Cargo.lock b/Cargo.lock
index 3141ab2e..0d0c42f8 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -165,8 +165,6 @@ dependencies = [
  "entities",
  "memchr",
  "ntest",
- "once_cell",
- "regex",
  "shell-words",
  "slug",
  "syntect",
diff --git a/Cargo.toml b/Cargo.toml
index 5f5ff5f8..e9101dbe 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -36,8 +36,6 @@ doc = false
 
 [dependencies]
 typed-arena = "2.0.2"
-regex = "1"
-once_cell = "1.19.0"
 entities = "1.0.1"
 unicode_categories = "0.1.1"
 memchr = "2"
diff --git a/src/html.rs b/src/html.rs
index 07ad630d..c46f7c9a 100644
--- a/src/html.rs
+++ b/src/html.rs
@@ -7,13 +7,12 @@ use crate::nodes::{
 };
 use crate::parser::{Options, Plugins};
 use crate::scanners;
-use once_cell::sync::Lazy;
-use regex::Regex;
 use std::borrow::Cow;
 use std::cell::Cell;
 use std::collections::{HashMap, HashSet};
 use std::io::{self, Write};
 use std::str;
+use unicode_categories::UnicodeCategories;
 
 use crate::adapters::HeadingMeta;
 
@@ -102,11 +101,21 @@ impl Anchorizer {
     /// assert_eq!("ticks-arent-in".to_string(), anchorizer.anchorize(source.to_string()));
     /// ```
     pub fn anchorize(&mut self, header: String) -> String {
-        static REJECTED_CHARS: Lazy<Regex> =
-            Lazy::new(|| Regex::new(r"[^\p{L}\p{M}\p{N}\p{Pc} -]").unwrap());
+        fn is_permitted_char(&c: &char) -> bool {
+            c == ' '
+                || c == '-'
+                || c.is_letter()
+                || c.is_mark()
+                || c.is_number()
+                || c.is_punctuation_connector()
+        }
 
         let mut id = header.to_lowercase();
-        id = REJECTED_CHARS.replace_all(&id, "").replace(' ', "-");
+        id = id
+            .chars()
+            .filter(is_permitted_char)
+            .map(|c| if c == ' ' { '-' } else { c })
+            .collect();
 
         let mut uniq = 0;
         id = loop {
diff --git a/src/parser/mod.rs b/src/parser/mod.rs
index 5fba6c93..d3d11c1b 100644
--- a/src/parser/mod.rs
+++ b/src/parser/mod.rs
@@ -787,8 +787,7 @@ pub struct RenderOptions {
     ///
     /// options.render.full_info_string = true;
     /// let html = markdown_to_html("``` rust extra info\nfn hello();\n```\n", &options);
-    /// let re = regex::Regex::new(r#"data-meta="extra info""#).unwrap();
-    /// assert!(re.is_match(&html));
+    /// assert!(html.contains(r#"data-meta="extra info""#));
     /// ```
     #[builder(default)]
     pub full_info_string: bool,