From 035befcfb0d39a85856e17e3e08a4016eb7f5c1d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Florian=20Ho=CC=88nicke?= <hoenicke.florian@gmail.com>
Date: Sat, 27 May 2023 16:31:49 +0200
Subject: [PATCH] =?UTF-8?q?=E2=9E=95=20refactor:=20summarize=20error=20mes?=
 =?UTF-8?q?sage=20without=20line=20number?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../microservice/google_custom_search.py      | 19 +++++++++++++++++--
 test/integration/test_generator.py            | 18 +++++-------------
 2 files changed, 22 insertions(+), 15 deletions(-)

diff --git a/dev_gpt/options/generate/static_files/microservice/google_custom_search.py b/dev_gpt/options/generate/static_files/microservice/google_custom_search.py
index f112129..3f47a98 100644
--- a/dev_gpt/options/generate/static_files/microservice/google_custom_search.py
+++ b/dev_gpt/options/generate/static_files/microservice/google_custom_search.py
@@ -20,8 +20,23 @@ def google_search(search_term, search_type, top_n):
     return response.json()
 
 def search_images(search_term, top_n):
-    response = google_search(search_term, search_type="image", top_n=top_n)
-    return [item["link"] for item in response["items"]]
+    """
+    Returns only images that have a 200 response code.
+    """
+    response = google_search(search_term, search_type="image", top_n=10)
+    image_urls = []
+    for item in response["items"]:
+        if len(image_urls) >= top_n:
+            break
+        try:
+            response = requests.head(item["link"], timeout=2)
+            if response.status_code == 200:
+                image_urls.append(
+                    item["link"]
+                )
+        except requests.exceptions.RequestException:
+            pass
+    return image_urls
 
 def search_web(search_term, top_n):
     response = google_search(search_term, search_type="web", top_n=top_n)
diff --git a/test/integration/test_generator.py b/test/integration/test_generator.py
index 2e8deee..11f663f 100644
--- a/test/integration/test_generator.py
+++ b/test/integration/test_generator.py
@@ -129,17 +129,9 @@ Example input: 'AAPL'
             'y',
             'https://www2.cs.uic.edu/~i101/SoundFiles/taunt.wav',
             f'''\
-import requests
-url = "https://transcribe.whisperapi.com"
-headers = {{
-'Authorization': 'Bearer {os.environ['WHISPER_API_KEY']}'
-}}
-data = {{
-  "url": "URL_OF_STORED_AUDIO_FILE"
-}}
-response = requests.post(url, headers=headers, data=data)
-assert response.status_code == 200
-print('This is the text from the audio file:', response.text)'''
+import openai
+audio_file= open("/path/to/file/audio.mp3", "rb")
+transcript = openai.Audio.transcribe("whisper-1", audio_file)'''
         ]
     ],
     indirect=True
@@ -158,12 +150,12 @@ def test_generation_level_4(microservice_dir, mock_input_sequence):
     generator = Generator(
         f'''Given an audio file (1min wav) of speech, 
 1. convert it to text using the Whisper API.
-2. Summarize the text (~50 words) while still maintaining the key facts.
+2. Summarize the text while still maintaining the key facts.
 3. Create an audio file of the summarized text using a tts library.
 4. Return the the audio file as base64 encoded binary.
 ''',
         str(microservice_dir),
-        'gpt-4',
+        'gpt-3.5-turbo',
         # self_healing=False,
     )
     assert generator.generate() == 0