From 40477493d328399e747b9e9248430a30b1ea0b31 Mon Sep 17 00:00:00 2001
From: unclecode <unclecode@kidocode.com>
Date: Wed, 31 Jul 2024 16:15:55 +0800
Subject: [PATCH] refactor: Remove image format dot in
 get_content_of_website_optimized

The code change removes the dot from the image format in the `get_content_of_website_optimized` function. This change ensures consistency in the image format and improves the functionality.
---
 crawl4ai/utils.py | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/crawl4ai/utils.py b/crawl4ai/utils.py
index 679c6d1d..e0d5d1ea 100644
--- a/crawl4ai/utils.py
+++ b/crawl4ai/utils.py
@@ -498,6 +498,8 @@ def get_content_of_website_optimized(url: str, html: str, word_count_threshold:
                 width_value, width_unit = parse_dimension(image_width)
                 image_size = 0 #int(fetch_image_file_size(img,base_url) or 0)
                 image_format = os.path.splitext(img.get('src',''))[1].lower()
+                # Remove . from format
+                image_format = image_format.strip('.')
                 score = 0
                 if height_value:
                     if height_unit == 'px' and height_value > 150: