1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
// SPDX-FileCopyrightText: 2026 libarchive_oxide contributors
//
// SPDX-License-Identifier: MIT OR Apache-2.0
//! `libarchive_oxide-core` — a frozen trait algebra and sans-IO core.
//!
//! What this crate delivers is not a "working extraction tool" but **the trait algebra itself**.
//! The top-level design criterion is the symmetry, orthogonality, and purity of the abstractions;
//! implementation coverage is subordinate to that.
//!
//! # Invariants of beauty (this crate's acceptance criteria)
//!
//! - **A single sans-IO substrate**: every transformation rides on [`Transform`]. It holds no I/O,
//! and the caller drives the bytes (caller-owned buffers, never forcing an allocation).
//! - **Orthogonal (format ⊥ filter)**: adding compression does not change a single line of the
//! [`format`] layer's code. The converse holds too. format knows nothing of filter, filter nothing of format.
//! - **Dual (read ⇄ write, decode ⇄ encode)**: [`EntryReader`]/[`EntryWriter`] and
//! [`Decoder`]/[`Encoder`] are symmetric at the type level.
//! The design of one side forces the other.
//! - **Purity**: all trait definitions are `no_std`. Only specific impls pull in `std`/`alloc`.
//! - **Origin-opaque**: whether an impl is hand-written or an adapter over a reused crate never leaks into the caller's types.
//!
//! # Implementation status
//!
//! The abstractions (traits and types) are **frozen in their finished form**. The implemented read
//! and write paths for [`format::tar`], [`format::cpio`], [`format::ar`], and [`format::iso9660`]
//! all use the same traits, demonstrating that implementation breadth grows beneath the frozen
//! abstractions rather than changing them.
extern crate alloc;
pub use ;
pub use ;
pub use ;
pub use Sink;
pub use ;
pub use ;