node.d.ts

import type { Transform } from 'node:stream';

/**
 * Create a Node.js stream.Transform for zstd compression.
 *
 * Uses Node.js `stream.Transform` to provide chunked compression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param level Compression level (1-22, or negative for fast mode). Default is 3.
 */
export declare function createZstdCompressTransform(level?: number): Transform;

/**
 * Create a Node.js stream.Transform for zstd decompression.
 *
 * Uses Node.js `stream.Transform` to provide chunked decompression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 *
 * The transform emits an error if the input ends before the compressed stream
 * does, including empty input.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createZstdDecompressTransform(maxOutputSize?: number): Transform;

/**
 * Create a Node.js stream.Transform for gzip compression.
 *
 * Uses Node.js `stream.Transform` to provide chunked gzip compression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 * Produces spec-compliant gzip output with proper header and CRC32 footer.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param level Compression level (0-9). Default is 6.
 */
export declare function createGzipCompressTransform(level?: number): Transform;

/**
 * Create a Node.js stream.Transform for gzip decompression.
 *
 * Uses Node.js `stream.Transform` to provide chunked gzip decompression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 * Verifies CRC32 integrity on finalization.
 *
 * The transform emits an error if the input ends before the compressed stream
 * does, including empty input.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createGzipDecompressTransform(maxOutputSize?: number): Transform;

/**
 * Create a Node.js stream.Transform for raw deflate compression.
 *
 * Uses Node.js `stream.Transform` to provide chunked raw deflate compression
 * (no gzip header/footer) compatible with `stream.pipeline()` and pipe-based workflows.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param level Compression level (0-9). Default is 6.
 */
export declare function createDeflateCompressTransform(level?: number): Transform;

/**
 * Create a Node.js stream.Transform for raw deflate decompression.
 *
 * Uses Node.js `stream.Transform` to provide chunked raw deflate decompression
 * compatible with `stream.pipeline()` and pipe-based workflows.
 *
 * The transform emits an error if the input ends before the compressed stream
 * does, including empty input.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createDeflateDecompressTransform(maxOutputSize?: number): Transform;

/**
 * Create a Node.js stream.Transform for brotli compression.
 *
 * Uses Node.js `stream.Transform` to provide chunked brotli compression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param quality Compression quality (0-11). Default is 6.
 */
export declare function createBrotliCompressTransform(quality?: number): Transform;

/**
 * Create a Node.js stream.Transform for brotli decompression.
 *
 * Uses Node.js `stream.Transform` to provide chunked brotli decompression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 *
 * The transform emits an error if the input ends before the compressed stream
 * does, including empty input.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createBrotliDecompressTransform(maxOutputSize?: number): Transform;

/**
 * Create a Node.js stream.Transform for LZ4 frame compression.
 *
 * Uses Node.js `stream.Transform` to provide chunked LZ4 compression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 */
export declare function createLz4CompressTransform(): Transform;

/**
 * Create a Node.js stream.Transform for LZ4 frame decompression.
 *
 * Uses Node.js `stream.Transform` to provide chunked LZ4 decompression compatible
 * with `stream.pipeline()` and pipe-based workflows.
 *
 * The input may hold several concatenated frames, including skippable and
 * legacy frames. The transform emits an error if the input ends inside a
 * frame, including empty input, or if data that is not a frame follows a
 * frame.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createLz4DecompressTransform(maxOutputSize?: number): Transform;

/**
 * Create a Node.js stream.Transform for zstd compression with a pre-trained dictionary.
 *
 * Uses Node.js `stream.Transform` to provide chunked compression with a pre-trained
 * dictionary, compatible with `stream.pipeline()` and pipe-based workflows.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param dict Pre-trained dictionary (from `zstdTrainDictionary`).
 * @param level Compression level (1-22, or negative for fast mode). Default is 3.
 */
export declare function createZstdCompressDictTransform(
  dict: Buffer | Uint8Array,
  level?: number,
): Transform;

/**
 * Create a Node.js stream.Transform for zstd decompression with a pre-trained dictionary.
 *
 * Uses Node.js `stream.Transform` to provide chunked decompression with a pre-trained
 * dictionary, compatible with `stream.pipeline()` and pipe-based workflows.
 *
 * The transform emits an error if the input ends before the compressed stream
 * does, including empty input.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param dict Pre-trained dictionary (must match the one used for compression).
 */
export declare function createZstdDecompressDictTransform(
  dict: Buffer | Uint8Array,
  maxOutputSize?: number,
): Transform;

/**
 * Create a Node.js stream.Transform for brotli compression with a custom dictionary.
 *
 * Uses Node.js `stream.Transform` to provide chunked compression with a custom
 * dictionary, compatible with `stream.pipeline()` and pipe-based workflows.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param dict Custom dictionary bytes.
 * @param quality Compression quality (0-11). Default is 6.
 */
export declare function createBrotliCompressDictTransform(
  dict: Buffer | Uint8Array,
  quality?: number,
): Transform;

/**
 * Create a Node.js stream.Transform for brotli decompression with a custom dictionary.
 *
 * Uses Node.js `stream.Transform` to provide chunked decompression with a custom
 * dictionary, compatible with `stream.pipeline()` and pipe-based workflows.
 *
 * The transform emits an error if the input ends before the compressed stream
 * does, including empty input.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param dict Custom dictionary (must match the one used for compression).
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createBrotliDecompressDictTransform(
  dict: Buffer | Uint8Array,
  maxOutputSize?: number,
): Transform;

/**
 * Create a Node.js stream.Transform for auto-detect decompression.
 *
 * Detects the compression format (zstd, gzip, brotli, or lz4) like
 * `detectFormat` and delegates to the appropriate decompression context.
 * Raw deflate is not supported (no magic bytes to distinguish it).
 *
 * The input is buffered until the format is detected: up to 64 KiB, or the
 * whole input if it is shorter, since brotli has no magic bytes and its
 * detection may need that much. The transform emits an error if the format
 * is still unknown then.
 *
 * The transform emits an error on empty input, which has no format to detect,
 * and on zstd, gzip or brotli input that ends before the compressed stream does.
 *
 * Output chunks hold at most `readableHighWaterMark` bytes (64 KiB by
 * default).
 *
 * @param maxOutputSize Maximum decompressed output size in bytes. Default is 256 MB.
 */
export declare function createDecompressTransform(maxOutputSize?: number): Transform;