From 9766d7febb5e25d8e1bce622da22ce23e6fd06df Mon Sep 17 00:00:00 2001
From: Daimona Eaytoy <daimona.wiki@gmail.com>
Date: Thu, 15 Jul 2021 14:47:59 +0200
Subject: [PATCH] Improve cache key generation based on image URLs

The protocol is ininfluent; for the size, we can tolerate small
differences, and this way we can prefetch the next image for on-wiki
requests.

Bug: T286356
---
 src/Controller/OcrController.php | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

diff --git a/src/Controller/OcrController.php b/src/Controller/OcrController.php
index 33e7703..7d4f546 100644
--- a/src/Controller/OcrController.php
+++ b/src/Controller/OcrController.php
@@ -225,7 +225,7 @@ private function getText(): string
         $cacheKey = md5(implode(
             '|',
             [
-                $this->imageUrl,
+                self::transformImageURLForCacheKey($this->imageUrl),
                 static::$params['engine'],
                 implode('|', static::$params['langs']),
                 static::$params['psm'],
@@ -237,4 +237,21 @@ private function getText(): string
             return $this->engine->getText($this->imageUrl, static::$params['langs']);
         });
     }
+
+    /**
+     * Make an image URL suitable to be used as a cache key (e.g. strip protocol)
+     * @param string $url
+     * @return string
+     */
+    private static function transformImageURLForCacheKey(string $url): string
+    {
+        return preg_replace_callback(
+            '/(page\d+-)(\d+)px/',
+            static function (array $matches) {
+                // Tolerate ±50px, see T286356.
+                return $matches[1].( round($matches[2] / 100) * 100 ).'px';
+            },
+            preg_replace('/^https?:/i', '', $url)
+        );
+    }
 }