Spaces:
Sleeping
Sleeping
File size: 5,538 Bytes
2f386cc 85d744c 2f386cc 85d744c 2f386cc 85d744c 2f386cc 85d744c 2f386cc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 |
import logging
from textwrap import dedent
from typing import Iterable
import os
from dotenv import load_dotenv
from openai import OpenAI
import streamlit as st
import tiktoken
load_dotenv()
client = OpenAI(api_key=os.environ.get("OPENAI_API_KEY"))
def analyze_code_files(code_files: list[str]) -> Iterable[dict[str, str]]:
"""Analyze the selected code files and return recommendations."""
return (analyze_code_file(code_file) for code_file in code_files)
def analyze_code_file(code_file: str) -> dict[str, str]:
"""Analyze a code file and return a dictionary with file information and recommendations."""
with open(code_file, "r") as f:
code_content = f.read()
if not code_content:
return {
"code_file": code_file,
"code_snippet": code_content,
"recommendation": "No code found in file",
}
try:
logging.info("Analyzing code file: %s", code_file)
analysis = get_code_analysis(code_content)
except Exception as e:
logging.info("Error analyzing code file: %s", code_file)
analysis = f"Error analyzing code file: {e}"
return {
"code_file": code_file,
"code_snippet": code_content,
"recommendation": analysis,
}
def get_num_tokens_from_messages(messages, model="gpt-3.5-turbo-0301"):
"""Returns the number of tokens used by a list of messages."""
# Source: https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb
try:
encoding = tiktoken.encoding_for_model(model)
except KeyError:
logging.debug("Model not found. Using cl100k_base encoding.")
encoding = tiktoken.get_encoding("cl100k_base")
if model == "gpt-3.5-turbo":
logging.debug(
"gpt-3.5-turbo may change over time. Returning num tokens assuming gpt-3.5-turbo-0301."
)
return get_num_tokens_from_messages(
messages, model="gpt-3.5-turbo-0301"
)
elif model == "gpt-4":
logging.debug(
"gpt-4 may change over time. Returning num tokens assuming gpt-4-0314."
)
return get_num_tokens_from_messages(messages, model="gpt-4-0314")
elif model == "gpt-3.5-turbo-0301":
tokens_per_message = (
4 # every message follows <|start|>{role/name}\n{content}<|end|>\n
)
tokens_per_name = -1 # if there's a name, the role is omitted
elif model == "gpt-4-0314":
tokens_per_message = 3
tokens_per_name = 1
else:
raise NotImplementedError(
f"""num_tokens_from_messages() is not implemented for model {model}. See https://github.com/openai/openai-python/blob/main/chatml.md for information on how messages are converted to tokens."""
)
num_tokens = 0
for message in messages:
num_tokens += tokens_per_message
for key, value in message.items():
num_tokens += len(encoding.encode(value))
if key == "name":
num_tokens += tokens_per_name
num_tokens += 3 # every reply is primed with <|start|>assistant<|message|>
return num_tokens
@st.cache_data(show_spinner=False)
def get_code_analysis(code: str) -> str:
"""Get code analysis from the OpenAI API."""
prompt = dedent(
f"""\
Please review the code below and identify any syntax or logical errors, suggest
ways to refactor and improve code quality, enhance performance, address security
concerns, and align with best practices. Provide specific examples for each area
and limit your recommendations to three per category.
Use the following response format, keeping the section headings as-is, and provide
your feedback. Use bullet points for each response. The provided examples are for
illustration purposes only and should not be repeated.
**Syntax and logical errors (example)**:
- Incorrect indentation on line 12
- Missing closing parenthesis on line 23
**Code refactoring and quality (example)**:
- Replace multiple if-else statements with a switch case for readability
- Extract repetitive code into separate functions
**Performance optimization (example)**:
- Use a more efficient sorting algorithm to reduce time complexity
- Cache results of expensive operations for reuse
**Security vulnerabilities (example)**:
- Sanitize user input to prevent SQL injection attacks
- Use prepared statements for database queries
**Best practices (example)**:
- Add meaningful comments and documentation to explain the code
- Follow consistent naming conventions for variables and functions
Code:
```
{code}
```
Your review:"""
)
messages = [{"role": "system", "content": prompt}]
tokens_in_messages = get_num_tokens_from_messages(
messages=messages, model="gpt-3.5-turbo"
)
max_tokens = 4096
tokens_for_response = max_tokens - tokens_in_messages
if tokens_for_response < 200:
return "The code file is too long to analyze. Please select a shorter file."
logging.info("Sending request to OpenAI API for code analysis")
logging.info("Max response tokens: %d", tokens_for_response)
response = client.chat.completions.create(
model="gpt-3.5-turbo",
messages=messages,
max_tokens=tokens_for_response,
n=1,
temperature=0,
)
logging.info("Received response from OpenAI API")
# Get the assistant's response from the API response
assistant_response = response.choices[0].message.content
return assistant_response.strip()
|