From dd59d5caa51b216d2fd2ca27bd2374cb834ec73e Mon Sep 17 00:00:00 2001
From: George Barnett <gbarnett@apple.com>
Date: Mon, 13 Sep 2021 13:29:12 +0100
Subject: [PATCH] Improve performance of
 `HTTPHeaders.subscript(canonicalForm:)`

Motivation:

When getting the canonical form of header values any ascii whitespace
is stripped from the values. The current implementation does this by
doing equality checks on `Character` which is quite expensive.

Modifications:

- Update the `Substring.trimWhitespace()` function to trim on a UTF8
  view

Result:

Retrieving the canonical form of header values is cheaper, significantly
so when values contain whitespace.
---
 Sources/NIOHTTP1/HTTPTypes.swift | 33 ++++++++++++++++++++------------
 1 file changed, 21 insertions(+), 12 deletions(-)

diff --git a/Sources/NIOHTTP1/HTTPTypes.swift b/Sources/NIOHTTP1/HTTPTypes.swift
index ddc7c1430e..50009dff03 100644
--- a/Sources/NIOHTTP1/HTTPTypes.swift
+++ b/Sources/NIOHTTP1/HTTPTypes.swift
@@ -540,28 +540,37 @@ extension HTTPHeaders: RandomAccessCollection {
     }
 }
 
-extension Character {
+extension UTF8.CodeUnit {
     var isASCIIWhitespace: Bool {
-        return self == " " || self == "\t" || self == "\r" || self == "\n" || self == "\r\n"
+        switch self {
+        case UInt8(ascii: " "),
+             UInt8(ascii: "\t"),
+             UInt8(ascii: "\r"),
+             UInt8(ascii: "\n"):
+          return true
+
+        default:
+          return false
+        }
     }
 }
 
 extension String {
     func trimASCIIWhitespace() -> Substring {
-        return self.dropFirst(0).trimWhitespace()
+        return Substring(self).trimWhitespace()
     }
 }
 
-private extension Substring {
-    func trimWhitespace() -> Substring {
-        var me = self
-        while me.first?.isASCIIWhitespace == .some(true) {
-            me = me.dropFirst()
-        }
-        while me.last?.isASCIIWhitespace == .some(true) {
-            me = me.dropLast()
+extension Substring {
+    fileprivate func trimWhitespace() -> Substring {
+        guard let firstNonWhitespace = self.utf8.firstIndex(where: { !$0.isASCIIWhitespace }) else {
+          // The whole substring is ASCII whitespace.
+          return Substring()
         }
-        return me
+
+        // There must be at least one non-ascii whitespace character, so banging here is safe.
+        let lastNonWhitespace = self.utf8.lastIndex(where: { !$0.isASCIIWhitespace })!
+        return Substring(self.utf8[firstNonWhitespace...lastNonWhitespace])
     }
 }