From 24ff5a2226aa36e8df90571378b792fdf3d97a69 Mon Sep 17 00:00:00 2001
From: Liam Bigelow <40188355+bglw@users.noreply.github.com>
Date: Sat, 21 May 2022 21:37:10 +1200
Subject: [PATCH] preinitialize slow regex

---
 pagefind/Cargo.toml         |  1 +
 pagefind/src/fossick/mod.rs | 17 ++++++++++-------
 2 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/pagefind/Cargo.toml b/pagefind/Cargo.toml
index 68bedca7..fcb3c85d 100644
--- a/pagefind/Cargo.toml
+++ b/pagefind/Cargo.toml
@@ -29,6 +29,7 @@ minifier = "0.0.43"
 sha-1 = "0.10"
 serde_json = "1"
 serde = { version = "1", features = ["derive"] }
+lazy_static = "1.4.0"
 
 [dev-dependencies]
 json_dotpath = "1.1.0"
diff --git a/pagefind/src/fossick/mod.rs b/pagefind/src/fossick/mod.rs
index 6b26533e..c02318b6 100644
--- a/pagefind/src/fossick/mod.rs
+++ b/pagefind/src/fossick/mod.rs
@@ -1,4 +1,5 @@
 use hashbrown::HashMap;
+use lazy_static::lazy_static;
 use lol_html::html_content::ContentType;
 use lol_html::{element, text, HtmlRewriter, Settings};
 use regex::Regex;
@@ -17,6 +18,12 @@ use crate::fragments::{PageFragment, PageFragmentData};
 use crate::utils::full_hash;
 use crate::SearchOptions;
 
+lazy_static! {
+    static ref EXTRANEOUS_NEWLINES: Regex = Regex::new("(^|\\s)*((\n|\r\n)\\s*)+($|\\s)*").unwrap();
+    static ref TRIM_NEWLINES: Regex = Regex::new("^\n|\n$").unwrap();
+    static ref EXTRANEOUS_SPACES: Regex = Regex::new("\\s{2,}").unwrap();
+}
+
 pub struct FossickedData {
     pub file_path: PathBuf,
     pub fragment: PageFragment,
@@ -205,13 +212,9 @@ fn build_url(page_url: &Path, options: &SearchOptions) -> String {
 }
 
 fn normalize_content(content: &str) -> String {
-    let extraneous_newlines = Regex::new("(^|\\s)*((\n|\r\n)\\s*)+($|\\s)*").unwrap();
-    let trim_newlines = Regex::new("^\n|\n$").unwrap();
-    let extraneous_spaces = Regex::new("\\s{2,}").unwrap();
-
-    let content = extraneous_newlines.replace_all(content, "\n");
-    let content = trim_newlines.replace_all(&content, "");
-    let content = extraneous_spaces.replace_all(&content, " ");
+    let content = EXTRANEOUS_NEWLINES.replace_all(content, "\n");
+    let content = TRIM_NEWLINES.replace_all(&content, "");
+    let content = EXTRANEOUS_SPACES.replace_all(&content, " ");
 
     content.to_string()
 }