From 2b612dea36da8d6c9aa90ce875c693448e51ca48 Mon Sep 17 00:00:00 2001
From: Hiroshi Nishio <hnishio0105@gmail.com>
Date: Tue, 7 Jan 2025 12:40:59 +0900
Subject: [PATCH] Fix an error in describe_image: You uploaded an unsupported
 image. Please make sure your image has of one the following formats: ['png',
 'jpeg', 'gif', 'webp'].

---
 utils/extract_urls.py | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/utils/extract_urls.py b/utils/extract_urls.py
index 8584e05f..9cc010be 100644
--- a/utils/extract_urls.py
+++ b/utils/extract_urls.py
@@ -4,13 +4,17 @@
 
 @handle_exceptions(default_return_value=[], raise_on_error=False)
 def extract_image_urls(text: str) -> list[dict[str, str]]:
-    """Extract alt text and URLs from img tags in the given text.
+    """Extract alt text and URLs from img tags in the given text. Excludes SVG images.
 
     Example: <img width="1352" alt="Screenshot 2024-12-12 at 6 25 41 PM" src="https://github.com/user-attachments/assets/9f1e8ca9-068e-434d-b2f3-f438638268ef" />
     """
     pattern = r'<img[^>]*alt="([^"]*)"[^>]*src="([^"]*)"[^>]*>'
     matches = findall(pattern, text)
-    return [{"alt": alt, "url": url} for alt, url in matches]
+    return [
+        {"alt": alt, "url": url}
+        for alt, url in matches
+        if not url.lower().endswith(".svg")
+    ]
 
 
 def extract_urls(text: str) -> tuple[list[str], list[str]]: