FediSearch

[
    {
        "id":"109244873849193955",
        "username":"jikodesu",
        "acct":"jikodesu",
        "display_name":"Jiko",
        ...
        "created_at":"2022-10-28T00:00:00.000Z",
        "note":"\u003cp\u003eTrying out Mastodon. Part of 2022 Twitter migration\u003c/p\u003e\u003cp\u003eDuolingo user: Nihongo, ...
    },
    {
        "id":"......"
    }

@Grab(group='org.apache.commons', module='commons-lang3', version='3.12.0')

import groovy.json.*
import java.net.http.*
import static org.apache.commons.lang3.StringEscapeUtils.escapeHtml4

json = new JsonSlurper().parse("${args[0]}/api/v1/directory".toURL())

// extraemos los campos de interes de cada usuario
users = json.collect { user->
	[username: user.username,
	 note: user.display_name +"|"+escapeHtml4(user.note)+"|"+user.fields.collect{it.name+'|'+it.value}.join('|'),
	 url: user.url]
}

body = [
	input: users.collect{it.note},
	model: "text-embedding-ada-002"
]
// se los mandamos a openai para que los vectorize
request = HttpRequest.newBuilder(new URI("https://api.openai.com/v1/embeddings"))
                .headers(
                        "Content-Type", "application/json",
                        "Authorization", "Bearer ${args[1]}"
                )
                .POST(HttpRequest.BodyPublishers.ofString(JsonOutput.toJson(body).toString()))
                .build()
response = HttpClient.newHttpClient().send(request, HttpResponse.BodyHandlers.ofInputStream())
result = response.body().text

// añadimos a cada usuario su vector
embeddings= new JsonSlurper().parseText(result) as Map
embeddings.data.each{ def entry ->
	users[entry.index].embedding = entry.embedding
}

// se los mandamos a Supabase con un POST
request = HttpRequest.newBuilder(new URI("https://TUPROJECTO.supabase.co/rest/v1/users"))
		.headers(
				"Content-Type", "application/json",
				"apikey", args[2]
		)
		.POST(HttpRequest.BodyPublishers.ofString(JsonOutput.toJson(users).toString()))
		.build()
HttpClient.newHttpClient().send(request, HttpResponse.BodyHandlers.ofInputStream())

import { serve } from 'https://deno.land/std@0.170.0/http/server.ts'
import 'https://deno.land/x/xhr@0.2.1/mod.ts'
import { createClient } from 'https://esm.sh/@supabase/supabase-js@2.5.0'
import { Configuration, OpenAIApi } from 'https://esm.sh/openai@3.1.0'

export const corsHeaders = {
  'Access-Control-Allow-Origin': '*',
  'Access-Control-Allow-Headers': 'authorization, x-client-info, apikey, content-type',
}

serve(async (req) => {
  // Handle CORS
  if (req.method === 'OPTIONS') {
    return new Response('ok', { headers: corsHeaders })
  }

  // Search query is passed in request payload
  const { query } = await req.json()

  // OpenAI recommends replacing newlines with spaces for best results
  const input = query.replace(/\n/g, ' ')
  console.log(`Searching users for ${input}`)
  const configuration = new Configuration({ apiKey: 'sk-----------' })
  const openai = new OpenAIApi(configuration)

  // Generate a one-time embedding for the query itself
  const embeddingResponse = await openai.createEmbedding({
    model: 'text-embedding-ada-002',
    input,
  })

  const [{ embedding }] = embeddingResponse.data.data

  const supabaseClient = createClient('https://pqlvxllouutlnnnwnlxp.supabase.co', 'TU_API_TOKEN');

  // In production we should handle possible errors
  const { data: documents } = await supabaseClient.rpc('match_users', {
    query_embedding: embedding,
    match_threshold: 0.75, // Choose an appropriate threshold for your data
    match_count: 50, // Choose the number of matches
  })

  return new Response(JSON.stringify(documents.sort((a,b)=>a.similarity - b.similarity)), {
    headers: { ...corsHeaders, 'Content-Type': 'application/json' },
  })
})