fixing gitignore activation blunder

.vercelignore

@@ -0,0 +1,42 @@
+# Ignore deployment-irrelevant files
+README.md
+scripts
+utils/*
+!utils/inference.py
+data/embeddings/Alibaba-NLP
+data/emoji-info/*
+!data/emoji-info/characters.txt
+
+# Ignore Python build artifacts
+*.egg
+*.egg-info/
+*.pyc
+
+# Node modules (if applicable)
+node_modules
+
+# Logs
+logs
+*.log
+npm-debug.log*
+
+# Build directories
+dist
+build
+
+# Environment variables
+.env
+.env.local
+.env.*.local
+
+# Temporary files
+*.tmp
+*.swp
+
+# IDE specific files
+.vscode
+.idea
+
+# OS generated files
+.DS_Store
+Thumbs.db

README.md

@@ -0,0 +1,2 @@
+# Emoji Search 🔍
+This is the code for the Vercel backend for [my emoji search engine](https://tim0120.github.io/projects/emoji-search). Try it out!

api/search.py

@@ -0,0 +1,113 @@
+from http.server import BaseHTTPRequestHandler
+from json import dumps
+import os
+import time
+from urllib.parse import parse_qs, urlparse
+
+from dotenv import load_dotenv
+from numpy import argsort, array, dot, einsum, load
+import requests
+
+load_dotenv()
+
+model_id = os.getenv('MODEL_ID')
+embeddings = load(
+    f'./data/embeddings/{model_id}/unicodeNames.npz', allow_pickle=True
+)['embeddings']
+emoji_characters = open('./data/emoji-info/characters.txt', 'r', encoding='utf-8').read().splitlines()
+
+def get_embeddings(text, max_retries=3, initial_delay=1):
+    API_URL = "https://api.openai.com/v1/embeddings"
+    headers = {
+        "Authorization": f"Bearer {os.getenv('OPENAI_API_KEY')}",
+        "Content-Type": "application/json"
+    }
+
+    delay = initial_delay
+    for attempt in range(max_retries):
+        try:
+            response = requests.post(
+                API_URL,
+                headers=headers,
+                json={
+                    "input": text,
+                    "model": "text-embedding-3-small"
+                }
+            )
+            response.raise_for_status()
+            return response.json()["data"][0]["embedding"]
+        except requests.exceptions.RequestException as e:
+            if attempt == max_retries - 1:  # Last attempt
+                return {"error": str(e)}
+            time.sleep(delay)
+            delay *= 2  # Exponential backoff
+
+def k_nearest(queries, k):
+    if len(embeddings.shape) == 2:
+        similarities = dot(embeddings, queries.T)
+    elif len(embeddings.shape) == 3:
+        similarities = einsum('ijk,lk->il', embeddings, queries)
+    else:
+        raise NotImplementedError('Only 2D and 3D embeddings are supported')
+    topk_indices = argsort(-similarities, axis=0)[:k]
+    return topk_indices.flatten().tolist()
+
+def handle_request(query_params):
+    query = query_params.get('query', [None])[0]
+    emb_type = query_params.get('emb_type', ['alternates'])[0]
+
+    if not query:
+        return {
+            "statusCode": 400,
+            "body": "Missing query parameter"
+        }
+
+    if emb_type not in ['unicodeName', 'alternates']:
+        return {
+            "statusCode": 400,
+            "body": f"Unsupported embedding type: {emb_type}"
+        }
+
+    try:
+        response = get_embeddings(query)
+        if isinstance(response, dict) and "error" in response:
+            return {
+                "statusCode": 500,
+                "body": response["error"]
+            }
+        query_embedding = array(response)
+        idxs = k_nearest(query_embedding, int(os.getenv('K_NEAREST')))
+        results = [emoji_characters[idx] for idx in idxs]
+        return {
+            "statusCode": 200,
+            "body": dumps(results),
+            "headers": {
+                "Content-Type": "application/json",
+                "Cache-Control": "s-maxage=3600"
+            }
+        }
+    except Exception as e:
+        return {
+            "statusCode": 500,
+            "body": str(e)
+        }
+
+class handler(BaseHTTPRequestHandler):
+    def do_GET(self):
+        # Parse query parameters
+        query_components = parse_qs(urlparse(self.path).query)
+
+        # Get response from handler
+        response = handle_request(query_components)
+
+        # Set response status code
+        self.send_response(response['statusCode'])
+
+        # Set headers
+        if 'headers' in response:
+            for header, value in response['headers'].items():
+                self.send_header(header, value)
+        self.end_headers()
+
+        # Send response body
+        self.wfile.write(response['body'].encode())