Compression API

Content-Encoding negotiation, compression, and decompression for gzip, deflate, Brotli, and Zstd.

Located in src/compression/ (httpx.compression).

Encoding

Enum representing supported Content-Encoding values.

Variant	Description
.gzip	gzip (RFC 1952)
.deflate	DEFLATE (RFC 1951)
.br	Brotli (RFC 7932)
.zstd	Zstandard
.identity	No encoding (pass-through)

Encoding Constants and Methods

Member	Description
token()	Convert to wire-format string (e.g., .gzip → "gzip")
fromToken(str)	Parse from a string; returns ?Encoding (case-insensitive)
parseAcceptEncoding(allocator, headerValue)	Parse an Accept-Encoding header into quality-sorted entries (caller owns)
negotiate(headerValue)	Pick the best supported encoding for an Accept-Encoding header value

httpx.compression.decompress()

pub fn decompress(allocator: Allocator, encoding: Encoding, data: []const u8) ![]u8

Decompresses body content based on the provided Content-Encoding. The caller owns the returned slice. Use decompressLimited(allocator, encoding, data, maxSize) to cap output (default cap MAX_DECOMPRESSED_SIZE = 64 MiB).

httpx.compression.compress()

pub fn compress(allocator: Allocator, encoding: Encoding, data: []const u8) ![]u8

Compresses data using the specified encoding. The caller owns the returned slice.

Streaming

For chunked data, compress or decompress incrementally with the zstd / brotli packages (httpx.zstd, httpx.brotli). The HTTP client and server negotiate Accept-Encoding automatically: negotiate() picks the best match and responses carry Content-Encoding plus Vary: Accept-Encoding.