From dfd6849e773f17deb60dcf2dcd18128248991e46 Mon Sep 17 00:00:00 2001
From: Anthony Shoumikhin <anthony@shoumikh.in>
Date: Mon, 2 Jun 2025 14:50:34 +0000
Subject: [PATCH] Update lint_urls.sh (#154838)

Do not match empty urls pieces like "https://"
Add headers for better handling urls like "https://www.amd.com/content/dam/amd/en/documents/instinct-tech-docs/data-sheets/amd-instinct-mi300x-data-sheet.pdf"

Pull Request resolved: https://github.com/pytorch/pytorch/pull/154838
Approved by: https://github.com/Skylion007
---
 scripts/lint_urls.sh | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/scripts/lint_urls.sh b/scripts/lint_urls.sh
index c8f4e183e177..0a8fddce00b3 100755
--- a/scripts/lint_urls.sh
+++ b/scripts/lint_urls.sh
@@ -19,7 +19,7 @@ while IFS=: read -r filepath url; do
     code=$(curl -k -gsLm30 --retry 3 --retry-delay 3 --retry-connrefused -o /dev/null -w "%{http_code}" -I "$url") || code=000
     if [ "$code" -lt 200 ] || [ "$code" -ge 400 ]; then
       sleep 1
-      code=$(curl -k -gsLm30 --retry 3 --retry-delay 3 --retry-connrefused -o /dev/null -w "%{http_code}" -r 0-0 -A "$user_agent" "$url") || code=000
+      code=$(curl -k -gsLm30 --retry 3 --retry-delay 3 --retry-connrefused -o /dev/null -w "%{http_code}" -r 0-0 -A "$user_agent" -H "Accept-Language: en-US,en" -H "Connection: keep-alive" "$url") || code=000
     fi
     if [ "$code" -lt 200 ] || [ "$code" -ge 400 ]; then
       sleep 1
@@ -62,7 +62,7 @@ while IFS=: read -r filepath url; do
     sleep 1
   done
 done < <(
-  pattern='(?!.*@lint-ignore)(?<!git\+)(?<!\$\{)https?://(?![^/]*@)(?![^\s<>\")]*[<>\{\}\$])[^[:space:]<>")\[\]\\|]+'
+  pattern='(?!.*@lint-ignore)(?<!git\+)(?<!\$\{)https?://(?![^/]*@)(?![^\s<>\")]*[<>\{\}\$])[[:alnum:]][^[:space:]<>")\[\]\\|]*'
   excludes=(
     ':(exclude,glob)**/.*'
     ':(exclude,glob)**/*.lock'