Public
HTTP
Forked from pomdtr/valTownSearch
janpaul123-valtownsemanticsearch.web.val.run
v39
June 17, 2024
- semanticSearchTurso@janpaul123Part of Val Town Semantic Search . Uses Turso to search embeddings of all vals, using the sqlite-vss extension. Call OpenAI to generate an embedding for the search query. Query the vss_vals_embeddings table in Turso using vss_search . The vss_vals_embeddings table has been generated by janpaul123/indexValsTurso . It is not run automatically. This table is incomplete due to a bug in Turso .Script
- codeOnValTown@andreterronCode on Val Town Adds a "Code on Val Town" ribbon to your page. This lets your website visitors navigate to the code behind it. This uses github-fork-ribbon-css under the hood. Usage Here are 2 different ways to add the "Code on Val Town" ribbon: 1. Wrap your fetch handler (recommended) import { modifyFetchHandler } from "https://esm.town/v/andreterron/codeOnValTown?v=45"; import { html } from "https://esm.town/v/stevekrouse/html?v=5"; export default modifyFetchHandler(async (req: Request): Promise<Response> => { return html(`<h2>Hello world!</h2>`); }); Example: @andreterron/openable_handler 2. Wrap your HTML string import { modifyHtmlString } from "https://esm.town/v/andreterron/codeOnValTown?v=45"; import { html } from "https://esm.town/v/stevekrouse/html?v=5"; export default async (req: Request): Promise<Response> => { return html(modifyHtmlString(`<h2>Hello world!</h2>`)); }; Example: @andreterron/openable_html Other ways We made sure this was very modular, so you can also add the ribbon using these methods: Get the element string directly: @andreterron/codeOnVT_ribbonElement Modify an HTTP Response: @andreterron/codeOnVT_modifyResponse Use .pipeThrough to append to a stream: @andreterron/InjectCodeOnValTownStream Customization Linking to the val These functions infer the val using the call stack or the request URL. If the inference isn't working, or if you want to ensure it links to a specific val, pass the val argument: modifyFetchHandler(handler, {val: { handle: "andre", name: "foo" }}) modifyHtmlString("<html>...", {val: { handle: "andre", name: "foo" }}) Styling You can set the style parameter to a css string to customize the ribbon. Check out github-fork-ribbon-css to learn more about how to style the element. modifyFetchHandler(handler, {style: ".github-fork-ribbon:before { background-color: #333; }"}) modifyHtmlString("<html>...", {style: ".github-fork-ribbon:before { background-color: #333; }"}) Here's how you can hide the ribbon on small screens: modifyFetchHandler(handler, {style: `@media (max-width: 768px) { .github-fork-ribbon { display: none !important; } }`}) To-dos [ ] Let users customize the ribbon. Some ideas are the text, color or placement.Script
- semanticSearchBlobs@janpaul123Part of Val Town Semantic Search . Uses Val Town's blob storage to search embeddings of all vals, by downloading them all and iterating through all of them to compute distance. Slow and terrible, but it works! Get metadata from blob storage: allValsBlob${dimensions}EmbeddingsMeta (currently allValsBlob1536EmbeddingsMeta ), which has a list of all indexed vals and where their embedding is stored ( batchDataIndex points to the blob, and valIndex represents the offset within the blob). The blobs have been generated by janpaul123/indexValsBlobs . It is not run automatically. Get all blobs with embeddings pointed to by the metadata, e.g. allValsBlob1536EmbeddingsData_0 for batchDataIndex 0. Call OpenAI to generate an embedding for the search query. Go through all embeddings and compute cosine similarity with the embedding for the search query. Return list sorted by similarity.Script
- semanticSearchNeon@janpaul123Part of Val Town Semantic Search . Uses Neon to search embeddings of all vals, using the pg_vector extension. Call OpenAI to generate an embedding for the search query. Query the vals_embeddings table in Neon using the cosine similarity operator. The vals_embeddings table gets refreshed every 10 minutes by janpaul123/indexValsNeon .Script
v39
June 17, 2024