Merge pull request #2 from barun-saha/pylint-fix

Pylint and CVE fix

app.py

@@ -12,7 +12,7 @@
 from global_config import GlobalConfig
 
 
-APP_TEXT = json5.loads(open(GlobalConfig.APP_STRINGS_FILE, 'r').read())
+APP_TEXT = json5.loads(open(GlobalConfig.APP_STRINGS_FILE, 'r', encoding='utf-8').read())
 GB_CONVERTER = 2 ** 30
 
 
@@ -68,18 +68,6 @@ def get_web_search_results_wrapper(text: str) -> List[Tuple[str, str]]:
     return results
 
 
-@st.cache_data
-def get_ai_image_wrapper(text: str) -> str:
-    """
-    Fetch and cache a Base 64-encoded image by calling an external API.
-
-    :param text: The image prompt
-    :return: The Base 64-encoded image
-    """
-
-    return llm_helper.get_ai_image(text)
-
-
 # def get_disk_used_percentage() -> float:
 #     """
 #     Compute the disk usage.
@@ -111,14 +99,19 @@ def build_ui():
 
     st.title(APP_TEXT['app_name'])
     st.subheader(APP_TEXT['caption'])
-    st.markdown('Powered by [Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.1).')
-    st.markdown('*If the JSON is generated or parsed incorrectly, try again later by making minor changes '
-                'to the input text.*')
+    st.markdown(
+        'Powered by'
+        ' [Mistral-7B-Instruct-v0.2](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2).'
+    )
+    st.markdown(
+        '*If the JSON is generated or parsed incorrectly, try again later by making minor changes'
+        ' to the input text.*'
+    )
 
     with st.form('my_form'):
         # Topic input
         try:
-            with open(GlobalConfig.PRELOAD_DATA_FILE, 'r') as in_file:
+            with open(GlobalConfig.PRELOAD_DATA_FILE, 'r', encoding='utf-8') as in_file:
                 preload_data = json5.loads(in_file.read())
         except (FileExistsError, FileNotFoundError):
             preload_data = {'topic': '', 'audience': ''}
@@ -158,7 +151,8 @@ def build_ui():
     st.text(APP_TEXT['tos2'])
 
     st.markdown(
-        '![Visitors](https://api.visitorbadge.io/api/visitors?path=https%3A%2F%2Fhuggingface.co%2Fspaces%2Fbarunsaha%2Fslide-deck-ai&countColor=%23263759)'
+        '![Visitors]'
+        '(https://api.visitorbadge.io/api/visitors?path=https%3A%2F%2Fhuggingface.co%2Fspaces%2Fbarunsaha%2Fslide-deck-ai&countColor=%23263759)'
     )
 
 
@@ -173,20 +167,17 @@ def generate_presentation(topic: str, pptx_template: str, progress_bar):
     """
 
     topic_length = len(topic)
-    logging.debug(f'Input length:: topic: {topic_length}')
+    logging.debug('Input length:: topic: %s', topic_length)
 
     if topic_length >= 10:
-        logging.debug(
-            f'Topic: {topic}\n'
-        )
-
+        logging.debug('Topic: %s', topic)
         target_length = min(topic_length, GlobalConfig.LLM_MODEL_MAX_INPUT_LENGTH)
 
         try:
             # Step 1: Generate the contents in JSON format using an LLM
             json_str = process_slides_contents(topic[:target_length], progress_bar)
-            logging.debug(f'{topic[:target_length]=}')
-            logging.debug(f'{len(json_str)=}')
+            logging.debug('Truncated topic: %s', topic[:target_length])
+            logging.debug('Length of JSON: %d', len(json_str))
 
             # Step 2: Generate the slide deck based on the template specified
             if len(json_str) > 0:
@@ -196,8 +187,10 @@ def generate_presentation(topic: str, pptx_template: str, progress_bar):
                     icon="💡️"
                 )
             else:
-                st.error('Unfortunately, JSON generation failed, so the next steps would lead to nowhere.'
-                         ' Try again or come back later.')
+                st.error(
+                    'Unfortunately, JSON generation failed, so the next steps would lead'
+                    ' to nowhere. Try again or come back later.'
+                )
                 return
 
             all_headers = generate_slide_deck(json_str, pptx_template, progress_bar)
@@ -225,15 +218,14 @@ def process_slides_contents(text: str, progress_bar: st.progress) -> str:
     json_str = ''
 
     try:
-        logging.info(f'Calling LLM for content generation on the topic: {text}')
+        logging.info('Calling LLM for content generation on the topic: %s', text)
         json_str = get_contents_wrapper(text)
     except Exception as ex:
-        st.error(f'An exception occurred while trying to convert to JSON.'
-                 f' It could be because of heavy traffic or something else.'
-                 f' Try doing it again or try again later.\n'
-                 f' Error message: {ex}')
-
-    # logging.debug(f'JSON: {json_str}')
+        st.error(
+            f'An exception occurred while trying to convert to JSON. It could be because of heavy'
+            f' traffic or something else. Try doing it again or try again later.'
+            f'\nError message: {ex}'
+        )
 
     progress_bar.progress(50, text='Contents generated')
 
@@ -316,6 +308,10 @@ def show_bonus_stuff(ppt_headers: List[str]):
 
 
 def main():
+    """
+    Trigger application run.
+    """
+
     build_ui()
 
 

global_config.py

@@ -1,27 +1,14 @@
+import os
+
 from dataclasses import dataclass
 from dotenv import load_dotenv
-import os
 
 
 load_dotenv()
 
 
 @dataclass(frozen=True)
 class GlobalConfig:
-    # CLARIFAI_PAT = os.environ.get('CLARIFAI_PAT', '')
-    # CLARIFAI_USER_ID = 'meta'
-    # CLARIFAI_APP_ID = 'Llama-2'
-    # CLARIFAI_MODEL_ID = 'llama2-13b-chat'
-    #
-    # CLARIFAI_USER_ID_GPT = 'openai'
-    # CLARIFAI_APP_ID_GPT = 'chat-completion'
-    # CLARIFAI_MODEL_ID_GPT = 'GPT-4'  # 'GPT-3_5-turbo'
-    #
-    # CLARIFAI_USER_ID_SD = 'stability-ai'
-    # CLARIFAI_APP_ID_SD = 'stable-diffusion-2'
-    # CLARIFAI_MODEL_ID_SD = 'stable-diffusion-xl'
-    # CLARIFAI_MODEL_VERSION_ID_SD = '0c919cc1edfc455dbc96207753f178d7'
-
     HF_LLM_MODEL_NAME = 'mistralai/Mistral-7B-Instruct-v0.2'
     LLM_MODEL_TEMPERATURE: float = 0.2
     LLM_MODEL_MIN_OUTPUT_LENGTH: int = 50
@@ -51,4 +38,3 @@ class GlobalConfig:
             'caption': 'Marvel in a monochrome dream'
         }
     }
-

llm_helper.py

@@ -1,8 +1,5 @@
-import json
 import logging
-import time
 import requests
-from langchain.llms import Clarifai
 
 from global_config import GlobalConfig
 
@@ -18,36 +15,6 @@
 # llm = None
 
 
-def get_llm(use_gpt: bool) -> Clarifai:
-    """
-    Get a large language model (hosted by Clarifai).
-
-    :param use_gpt: True if GPT-3.5 is required; False is Llama 2 is required
-    """
-
-    if use_gpt:
-        _ = Clarifai(
-            pat=GlobalConfig.CLARIFAI_PAT,
-            user_id=GlobalConfig.CLARIFAI_USER_ID_GPT,
-            app_id=GlobalConfig.CLARIFAI_APP_ID_GPT,
-            model_id=GlobalConfig.CLARIFAI_MODEL_ID_GPT,
-            verbose=True,
-            # temperature=0.1,
-        )
-    else:
-        _ = Clarifai(
-            pat=GlobalConfig.CLARIFAI_PAT,
-            user_id=GlobalConfig.CLARIFAI_USER_ID,
-            app_id=GlobalConfig.CLARIFAI_APP_ID,
-            model_id=GlobalConfig.CLARIFAI_MODEL_ID,
-            verbose=True,
-            # temperature=0.1,
-        )
-    # print(llm)
-
-    return _
-
-
 def hf_api_query(payload: dict):
     """
     Invoke HF inference end-point API.
@@ -56,9 +23,14 @@ def hf_api_query(payload: dict):
     :return: The output from the LLM
     """
 
-    # logging.debug(f'{payload=}')
-    response = requests.post(HF_API_URL, headers=HF_API_HEADERS, json=payload)
-    return response.json()
+    try:
+        response = requests.post(HF_API_URL, headers=HF_API_HEADERS, json=payload, timeout=15)
+        result = response.json()
+    except requests.exceptions.Timeout as te:
+        logging.error('*** Error: hf_api_query timeout! %s', str(te))
+        result = {}
+
+    return result
 
 
 def generate_slides_content(topic: str) -> str:
@@ -69,7 +41,7 @@ def generate_slides_content(topic: str) -> str:
     :return: The content in JSON format
     """
 
-    with open(GlobalConfig.SLIDES_TEMPLATE_FILE, 'r') as in_file:
+    with open(GlobalConfig.SLIDES_TEMPLATE_FILE, 'r', encoding='utf-8') as in_file:
         template_txt = in_file.read().strip()
         template_txt = template_txt.replace('<REPLACE_PLACEHOLDER>', topic)
 
@@ -98,61 +70,11 @@ def generate_slides_content(topic: str) -> str:
         # logging.debug(f'{json_end_idx=}')
         output = output[:json_end_idx]
 
-    logging.debug(f'{output=}')
+    logging.debug('generate_slides_content: output: %s', output)
 
     return output
 
 
-def get_ai_image(text: str) -> str:
-    """
-    Get a Stable Diffusion-generated image based on a given text.
-
-    :param text: The input text
-    :return: The Base 64-encoded image
-    """
-
-    url = f'''https://api.clarifai.com/v2/users/{GlobalConfig.CLARIFAI_USER_ID_SD}/apps/{GlobalConfig.CLARIFAI_APP_ID_SD}/models/{GlobalConfig.CLARIFAI_MODEL_ID_SD}/versions/{GlobalConfig.CLARIFAI_MODEL_VERSION_ID_SD}/outputs'''
-    headers = {
-        "Content-Type": "application/json",
-        "Authorization": f'Key {GlobalConfig.CLARIFAI_PAT}'
-    }
-    data = {
-        "inputs": [
-            {
-                "data": {
-                    "text": {
-                        "raw": text
-                    }
-                }
-            }
-        ]
-    }
-
-    # print('*** AI image generator...')
-    # print(url)
-
-    start = time.time()
-    response = requests.post(
-        url=url,
-        headers=headers,
-        data=json.dumps(data)
-    )
-    stop = time.time()
-
-    # print('Response:', response, response.status_code)
-    logging.debug('Image generation took', stop - start, 'seconds')
-    img_data = ''
-
-    if response.ok:
-        # print('*** Clarifai SDXL request: Response OK')
-        json_data = json.loads(response.text)
-        img_data = json_data['outputs'][0]['data']['image']['base64']
-    else:
-        logging.error('*** Image generation failed:', response.text)
-
-    return img_data
-
-
 if __name__ == '__main__':
     # results = get_related_websites('5G AI WiFi 6')
     #

pptx_helper.py

@@ -1,10 +1,11 @@
+import logging
 import pathlib
+import re
 import tempfile
 from typing import List, Tuple
+
 import json5
-import logging
 import pptx
-import re
 import yaml
 
 from global_config import GlobalConfig
@@ -57,7 +58,8 @@ def generate_powerpoint_presentation(
     """
     Create and save a PowerPoint presentation file containing the contents in JSON or YAML format.
 
-    :param structured_data: The presentation contents as "JSON" (may contain trailing commas) or YAML
+    :param structured_data: The presentation contents as "JSON" (may contain trailing commas) or
+    YAML
     :param as_yaml: True if the input data is in YAML format; False if it is in JSON format
     :param slides_template: The PPTX template to use
     :param output_file_path: The path of the PPTX file to save as
@@ -69,13 +71,16 @@ def generate_powerpoint_presentation(
         try:
             parsed_data = yaml.safe_load(structured_data)
         except yaml.parser.ParserError as ype:
-            logging.error(f'*** YAML parse error: {ype}')
+            logging.error('*** YAML parse error: %s', str(ype))
             parsed_data = {'title': '', 'slides': []}
     else:
         # The structured "JSON" might contain trailing commas, so using json5
         parsed_data = json5.loads(structured_data)
 
-    logging.debug(f"*** Using PPTX template: {GlobalConfig.PPTX_TEMPLATE_FILES[slides_template]['file']}")
+    logging.debug(
+        "*** Using PPTX template: %s",
+        GlobalConfig.PPTX_TEMPLATE_FILES[slides_template]['file']
+    )
     presentation = pptx.Presentation(GlobalConfig.PPTX_TEMPLATE_FILES[slides_template]['file'])
 
     # The title slide
@@ -84,7 +89,7 @@ def generate_powerpoint_presentation(
     title = slide.shapes.title
     subtitle = slide.placeholders[1]
     title.text = parsed_data['title']
-    logging.debug(f'Title is: {title.text}')
+    logging.debug('Presentation title is: %s', title.text)
     subtitle.text = 'by Myself and SlideDeck AI :)'
     all_headers = [title.text, ]
 

requirements.txt

@@ -1,7 +1,7 @@
 python-dotenv[cli]~=1.0.0
-langchain~=0.0.273
+langchain~=0.1.13
 # huggingface_hub
-streamlit~=1.26.0
+streamlit~=1.32.2
 clarifai==9.7.4
 
 python-pptx