Associativity of + in type parameters.

Commutativity of + in type parameters.

Chooses between alternatives based on a condition, adjusting a type-level parameter appropriately.

Note that while this type class makes sense for bounded builder primitives, it should not be instantiated for fixed-width primitives of differing widths (at least, not without padding to equalize the widths) because the choice between alternatives introduces a run-time variation in width. Instead please use ordinary bool or if _ then _ else _.

The larger of two Nats.

Associativity of Max in type parameters.

Commutativity of Max in type parameters.

Are we restricted to aligned writes only?

StoreUnaligned if the Cabal file defines 1, which it does on architectures where that approach is known to be safe and faster then writing bytes one by one. Otherwise StoreAligned.

Specifies order in which the bytes of an integer are encoded.

The ByteOrder native to the current architecture.

For example, the order of the bytes when you poke a Word32.

A BuildR together with a type-level bound on the number of bytes written and a requirement that the current buffer already contain at least that many bytes.

As in the "bytestring" package, the purpose of a bounded primitive is to improve speed by consolidating the space checks of several small builders.

Executes the given bounded primitive after obtaining the space it requires.

Needed for rewrite rules; normally you would use pmappend or (&<>).

Executes the bounded primitive WITHOUT first ensuring it has enough space.

Similar to a BoundedPrim but also consolidates address updates in order to take advantage of machine instructions that write at an offset.

The additional input is an offset from the current address that specifies the beginning of the region being encoded.

(If GHC learns to consolidate address offsets automatically then we might be able to just use BoundedPrim instead.)

Executes the given fixed primitive and adjusts the current address.

Fixed-width primitive that writes a single byte as-is.

Fixed-width primitive that writes a 16-bit word in the specified byte order.

Fixed-width primitive that writes a 16-bit word in native byte order.

Fixed-width primitive that writes a 16-bit word in big-endian byte order.

Fixed-width primitive that writes a 16-bit word in little-endian byte order.

Fixed-width primitive that writes a 32-bit word in the specified byte order.

Fixed-width primitive that writes a 32-bit word in native byte order.

Fixed-width primitive that writes a 32-bit word in big-endian byte order.

Fixed-width primitive that writes a 32-bit word in little-endian byte order.

Fixed-width primitive that writes a 64-bit word in the specified byte order.

Fixed-width primitive that writes a 64-bit word in native byte order.

Fixed-width primitive that writes a 64-bit word in big-endian byte order.

Fixed-width primitive that writes a 64-bit word in little-endian byte order.

word8 . fromIntegral

bo -> word16 bo . fromIntegral

word16Native . fromIntegral

word16BE . fromIntegral

word16LE . fromIntegral

bo -> word32 bo . fromIntegral

word32Native . fromIntegral

word32BE . fromIntegral

word32LE . fromIntegral

bo -> word64 bo . fromIntegral

word64Native . fromIntegral

word64BE . fromIntegral

word64LE . fromIntegral

Fixed-width primitive that writes a Float in the specified byte order.

Fixed-width primitive that writes a Float in native byte order.

Fixed-width primitive that writes a Float in big-endian byte order.

Fixed-width primitive that writes a Float in little-endian byte order.

Fixed-width primitive that writes a Double in the specified byte order.

Fixed-width primitive that writes a Double in native byte order.

Fixed-width primitive that writes a Double in big-endian byte order.

Fixed-width primitive that writes a Double in little-endian byte order.

Bounded-width primitive that writes a Char according to the UTF-8 encoding.

The bounded primitive implementing wordBase128LEVar.

Like wordBase128LEVar but inlined, possibly bloating your code. On the other hand, inlining an application to a constant may shrink your code.

The bounded primitive implementing word32Base128LEVar.

Like word32Base128LEVar but inlined, which currently means that it is just the same as word32Base128LEVar, which we inline.

The bounded primitive implementing word64Base128LEVar.

Like word64Base128LEVar but inlined, possibly bloating your code. On the other hand, inlining an application to a constant may shrink your code.

The analog of vectorBuildR for when fixed-width primitives encode the elements of the vector. In this special case we can predict the overall length.