Add summary cache

.gitignore

@@ -4,3 +4,4 @@ __pycache__
 /search-results
 /web-pages
 /page-content
+/summaries

main.py

@@ -8,6 +8,7 @@ from dotenv import load_dotenv
 from googleapiclient.discovery import build
 from slugify import slugify
 from transformers import pipeline
+import uuid
 
 from beautiful_soup.app import get_url_content
 
@@ -60,15 +61,27 @@ def main():
 
         for result in results:
             st.write(result['link'])
+
             try:
                 content = get_url_content( result['link'] )
-                summarizer = pipeline("summarization", model="sshleifer/distilbart-cnn-12-6")
-                summary = summarizer(content, max_length=130, min_length=30, do_sample=False, truncation=True)
-                for sentence in summary:
-                    st.write(sentence['summary_text'])
             except Exception as exception:
                 st.exception(exception)
 
+            file_path = 'summaries/' + uuid.uuid5( uuid.NAMESPACE_URL, result['link'] ).hex + '.json'
+            if exists( file_path ):
+                with open( file_path, 'r' ) as file:
+                    summary = json.load( file )
+            else:
+                try:
+                    summarizer = pipeline("summarization", model="sshleifer/distilbart-cnn-12-6")
+                    summary = summarizer(content, max_length=130, min_length=30, do_sample=False, truncation=True)
+                except Exception as exception:
+                    raise exception
+                with open( file_path, 'w' ) as file:
+                    json.dump( summary, file )
+    
+            for sentence in summary:
+                st.write(sentence['summary_text'])
 
 if __name__ == '__main__':
     main()