Strict-prefix JSON repair utility for streaming AI output.
A streaming data transport format that aims to support built-in features such as Promises, Dates, RegExps, Maps, Sets and more.
Convert character encodings in pure javascript.
micromark utility to decode numeric character references
micromark utility to decode markdown strings
Decode named character references
A promise based streaming tokenizer
Protocol Buffer encoding/decoding for **Agent-User Interaction (AG-UI) Protocol** events.
Package for formatting JSON data in a coloured YAML-style, perfect for CLI output
Streaming JSON.stringify()
A better decodeURIComponent
Decode WebM audio (Opus, Vorbis) to PCM samples
Decode WMA audio via RockBox WASM
Decode AIFF/AIFF-C audio to PCM samples
Decode CAF (Core Audio Format) audio to PCM samples
Translates between file formats and generates static code as well as TypeScript definitions.
Repair broken JSON documents
Decode AMR-NB/WB audio via opencore-amr WASM
Decode JWT tokens, mostly useful for browser applications.
Decode AAC/M4A audio via FAAD2 WASM
Postgres bytea parser
A micro-library of stream components for building custom JSON and JSONC processing pipelines with a minimal memory footprint — parse, filter, and transform JSON far larger than available memory with a SAX-inspired token API, on Node.js or Web Streams.
Oboe.js reads json, giving you the objects as they are found without waiting for the stream to finish
XRP Ledger binary codec
A full-featured TOON encoder/decoder with JSON feature parity: streaming, hooks, pretty generate, strict parsing, schema hints, CLI and ActiveSupport integration.
Scrapetor is a Ruby HTML parsing + scraping toolkit. The parser is a native C arena DOM with structural indexes built at parse time and NEON SIMD scanners in the SAX hot loop. A streaming extraction engine compiles the schema DSL into a single forward pass — no DOM materialised, one Ruby boundary crossing per document. On builds where libcurl is available, Scrapetor::Fetcher adds an HTTP/2-capable fetch layer with per-thread connection cache, shared DNS + TLS session pool, in-process gzip / deflate / brotli / zstd decoding, iconv charset transcoding, retry + exponential backoff, ETag / Last-Modified disk cache with bulk revalidation, per-host throttle, cookie jar, basic + bearer auth, proxy, and three bulk concurrency models (parallel_fetch / multi_fetch / streaming multi_each). Scrapetor::Session ties the cookie / auth / throttle / retry policies together. Also ships robots.txt + sitemap.xml parsers, a bounded-memory streaming HTML parser, and structured-data extractors (JSON-LD, OpenGraph, Schema.org, Microdata, RDFa, Twitter Cards). The Net::HTTP-based Scrapetor.fetch is preserved as the no-libcurl fallback.