119 lines
3.1 KiB
TypeScript
119 lines
3.1 KiB
TypeScript
import endent from 'endent';
|
|
import {
|
|
createParser,
|
|
ParsedEvent,
|
|
ReconnectInterval,
|
|
} from 'eventsource-parser';
|
|
|
|
const createPrompt = (
|
|
inputLanguage: string,
|
|
outputLanguage: string,
|
|
inputCode: string,
|
|
) => {
|
|
const data = (inputCode: any, type: string) => {
|
|
return endent`
|
|
You are an expert programmer in all programming languages.
|
|
You know very well algorithms. You will explain how the code works.
|
|
The explanation must me in markdown format but not rendered, it must include all markdown characteristics.
|
|
Do not include informations about console logs or print messages. Explain the following code ${type !== 'persona' && 'as I am a' + type}:
|
|
|
|
${inputCode}
|
|
|
|
`;
|
|
};
|
|
|
|
switch (outputLanguage) {
|
|
case 'persona':
|
|
return data(inputCode, 'persona');
|
|
case 'teacher':
|
|
return data(inputCode, 'teacher');
|
|
case '5':
|
|
return data(inputCode, '5 years boy');
|
|
case 'beginner':
|
|
return data(inputCode, 'beginner programmer');
|
|
case 'nasa':
|
|
return data(inputCode, 'very intelligent PHD professor at MIT');
|
|
case 'pizza-delivery-guy':
|
|
return data(inputCode, 'pizza delivery guy');
|
|
case 'bus-driver':
|
|
return data(inputCode, 'bus driver');
|
|
case 'magician':
|
|
return data(inputCode, 'magician');
|
|
case 'barista':
|
|
return data(inputCode, 'barista');
|
|
case 'doctor':
|
|
return data(inputCode, 'doctor');
|
|
}
|
|
};
|
|
|
|
export const OpenAIStream = async (
|
|
inputLanguage: string,
|
|
outputLanguage: string,
|
|
inputCode: string,
|
|
model: string,
|
|
key: string,
|
|
) => {
|
|
const prompt = createPrompt(inputLanguage, outputLanguage, inputCode);
|
|
|
|
const system = { role: 'system', content: prompt };
|
|
|
|
const res = await fetch(`https://api.openai.com/v1/chat/completions`, {
|
|
headers: {
|
|
'Content-Type': 'application/json',
|
|
Authorization: `Bearer ${key || process.env.OPENAI_API_KEY}`,
|
|
},
|
|
method: 'POST',
|
|
body: JSON.stringify({
|
|
model,
|
|
messages: [system],
|
|
temperature: 0,
|
|
stream: true,
|
|
}),
|
|
});
|
|
|
|
const encoder = new TextEncoder();
|
|
const decoder = new TextDecoder();
|
|
|
|
if (res.status !== 200) {
|
|
const statusText = res.statusText;
|
|
const result = await res.body?.getReader().read();
|
|
throw new Error(
|
|
`OpenAI API returned an error: ${
|
|
decoder.decode(result?.value) || statusText
|
|
}`,
|
|
);
|
|
}
|
|
|
|
const stream = new ReadableStream({
|
|
async start(controller) {
|
|
const onParse = (event: ParsedEvent | ReconnectInterval) => {
|
|
if (event.type === 'event') {
|
|
const data = event.data;
|
|
|
|
if (data === '[DONE]') {
|
|
controller.close();
|
|
return;
|
|
}
|
|
|
|
try {
|
|
const json = JSON.parse(data);
|
|
const text = json.choices[0].delta.content;
|
|
const queue = encoder.encode(text);
|
|
controller.enqueue(queue);
|
|
} catch (e) {
|
|
controller.error(e);
|
|
}
|
|
}
|
|
};
|
|
|
|
const parser = createParser(onParse);
|
|
|
|
for await (const chunk of res.body as any) {
|
|
parser.feed(decoder.decode(chunk));
|
|
}
|
|
},
|
|
});
|
|
|
|
return stream;
|
|
};
|