Major changes

This commit is contained in:
owenqwenstarsky
2026-01-24 16:52:19 -06:00
parent 011e40119d
commit 7181791734
46 changed files with 8371 additions and 127 deletions
+252
View File
@@ -0,0 +1,252 @@
import { NextRequest } from 'next/server';
import { prisma } from '@/lib/db';
interface WikipediaSummary {
title: string;
extract: string;
thumbnail?: { source: string };
originalimage?: { source: string };
}
interface WikiParseResult {
parse: {
title: string;
text: { '*': string };
images: string[];
};
}
async function getWikipediaSummary(title: string): Promise<WikipediaSummary | null> {
try {
const apiTitle = title.replace(/%20/g, '_');
const response = await fetch(
`https://en.wikipedia.org/api/rest_v1/page/summary/${apiTitle}`,
{
headers: { 'User-Agent': 'WikiPlus/1.0' },
}
);
if (!response.ok) return null;
return await response.json();
} catch {
return null;
}
}
async function getWikipediaContent(title: string): Promise<{ text: string; images: string[] } | null> {
try {
const apiTitle = title.replace(/%20/g, '_');
const response = await fetch(
`https://en.wikipedia.org/w/api.php?action=parse&page=${encodeURIComponent(apiTitle)}&format=json&prop=text|images&disableeditsection=true&redirects=true`,
{
headers: { 'User-Agent': 'WikiPlus/1.0' },
}
);
if (!response.ok) return null;
const data: WikiParseResult = await response.json();
if (!data.parse) return null;
// Extract plain text from HTML
const html = data.parse.text?.['*'] || '';
const plainText = html
.replace(/<style[^>]*>[\s\S]*?<\/style>/gi, '')
.replace(/<script[^>]*>[\s\S]*?<\/script>/gi, '')
.replace(/<[^>]+>/g, ' ')
.replace(/&nbsp;/g, ' ')
.replace(/&amp;/g, '&')
.replace(/&lt;/g, '<')
.replace(/&gt;/g, '>')
.replace(/&quot;/g, '"')
.replace(/&#39;/g, "'")
.replace(/\s+/g, ' ')
.trim();
// Get image URLs
const images = (data.parse.images || [])
.filter((img: string) => !img.includes('icon') && !img.includes('logo') && !img.includes('Commons-logo'))
.slice(0, 5)
.map((img: string) => `https://en.wikipedia.org/wiki/Special:FilePath/${encodeURIComponent(img)}`);
return { text: plainText.slice(0, 15000), images };
} catch {
return null;
}
}
export async function GET(request: NextRequest) {
const searchParams = request.nextUrl.searchParams;
const title = searchParams.get('title');
if (!title) {
return new Response('Missing title parameter', { status: 400 });
}
// Normalize the slug for database lookup
const slug = title.replace(/%20/g, '_').replace(/ /g, '_');
// Check if article exists in database
const existingArticle = await prisma.article.findUnique({
where: { slug },
});
if (existingArticle) {
// Return cached article wrapped in <article> tags
const cachedContent = `<article>${existingArticle.content}</article>`;
return new Response(cachedContent, {
headers: {
'Content-Type': 'text/plain; charset=utf-8',
'X-Cache': 'HIT',
},
});
}
const apiKey = process.env.OPENROUTER_API_KEY;
if (!apiKey) {
return new Response('OpenRouter API key not configured', { status: 500 });
}
// Fetch Wikipedia data
const [summary, content] = await Promise.all([
getWikipediaSummary(title),
getWikipediaContent(title),
]);
if (!summary || !content) {
return new Response('Article not found', { status: 404 });
}
const imageUrl = summary.thumbnail?.source || summary.originalimage?.source;
const imageContext = imageUrl
? `\n\nMain image available: ${imageUrl}\nAdditional images: ${content.images.join(', ')}`
: content.images.length > 0
? `\n\nImages available: ${content.images.join(', ')}`
: '';
const systemPrompt = `You are an expert writer creating comprehensive, in-depth articles for an AI-powered Wikipedia alternative. Your goal is to transform encyclopedia content into rich, detailed, and engaging prose that thoroughly covers the topic.
Guidelines:
- Write a COMPREHENSIVE and DETAILED article - aim for depth and thoroughness
- Cover ALL major aspects of the topic: history, significance, key details, related concepts, and impact
- Use markdown formatting extensively:
- Use ## for main sections and ### for subsections
- Use **bold** for key terms and *italic* for emphasis
- Use bullet lists and numbered lists where appropriate
- Use > blockquotes for notable quotes or key facts
- Include the main image at the top using markdown: ![Description](url)
- Structure the article with multiple well-developed sections (5-8 sections minimum)
- Each section should have multiple paragraphs with detailed explanations
- Add context that helps readers understand why this topic matters
- Include interesting facts, historical context, and connections to broader themes
- Maintain factual accuracy - expand on the source material but don't invent facts
- Write at least 1000-1500 words for a thorough treatment of the topic
- Output ONLY the article content wrapped in <article></article> tags
- Do not include any text outside the <article> tags`;
const userPrompt = `Write a comprehensive, detailed article about "${summary.title}" based on this Wikipedia content:
Summary: ${summary.extract}
Full content: ${content.text}${imageContext}
Requirements:
- Write a THOROUGH article with 5-8 well-developed sections minimum
- Each section should have multiple detailed paragraphs
- Cover history, significance, key facts, and broader context
- Use rich markdown formatting throughout
- Aim for 1000-1500+ words total
- Output in markdown wrapped in <article></article> tags`;
// Call OpenRouter API with streaming
const openRouterResponse = await fetch('https://openrouter.ai/api/v1/chat/completions', {
method: 'POST',
headers: {
'Authorization': `Bearer ${apiKey}`,
'Content-Type': 'application/json',
},
body: JSON.stringify({
model: 'minimax/minimax-m2.1',
messages: [
{ role: 'system', content: systemPrompt },
{ role: 'user', content: userPrompt },
],
stream: true,
}),
});
if (!openRouterResponse.ok) {
const error = await openRouterResponse.text();
console.error('OpenRouter error:', error);
return new Response('Failed to generate article', { status: 500 });
}
// Transform the OpenRouter SSE stream to extract content and save to DB
const encoder = new TextEncoder();
const decoder = new TextDecoder();
let fullContent = '';
const transformStream = new TransformStream({
async transform(chunk, controller) {
const text = decoder.decode(chunk);
const lines = text.split('\n');
for (const line of lines) {
if (line.startsWith('data: ')) {
const data = line.slice(6);
if (data === '[DONE]') {
continue;
}
try {
const parsed = JSON.parse(data);
const contentChunk = parsed.choices?.[0]?.delta?.content;
if (contentChunk) {
fullContent += contentChunk;
controller.enqueue(encoder.encode(contentChunk));
}
} catch {
// Skip invalid JSON
}
}
}
},
async flush() {
// Extract content between <article> tags and save to database
const articleMatch = fullContent.match(/<article>([\s\S]*?)<\/article>/);
const articleContent = articleMatch ? articleMatch[1].trim() : fullContent;
if (articleContent) {
try {
const newArticle = await prisma.article.create({
data: {
slug,
title: summary.title,
content: articleContent,
imageUrl: imageUrl || null,
},
});
// Create initial history entry
await prisma.articleHistory.create({
data: {
articleId: newArticle.id,
oldContent: '',
newContent: articleContent,
reason: 'Initial article',
},
});
} catch (error) {
console.error('Failed to save article to database:', error);
}
}
},
});
const stream = openRouterResponse.body?.pipeThrough(transformStream);
return new Response(stream, {
headers: {
'Content-Type': 'text/plain; charset=utf-8',
'Transfer-Encoding': 'chunked',
'Cache-Control': 'no-cache',
'X-Cache': 'MISS',
},
});
}