mirror of
https://github.com/exoticorn/upkr.git
synced 2026-01-21 20:06:43 +01:00
Compare commits
7 Commits
v0.2.0
...
4eab36b9d9
| Author | SHA1 | Date | |
|---|---|---|---|
| 4eab36b9d9 | |||
| 7cec54f62b | |||
| 7fa6be6ff4 | |||
| cab51e06ff | |||
| c4fce626da | |||
| 0d7cda06bb | |||
|
140678ae20 |
253
Cargo.lock
generated
253
Cargo.lock
generated
@@ -10,9 +10,15 @@ checksum = "38d9ff5d688f1c13395289f67db01d4826b46dd694e7580accdc3e8430f2d98e"
|
|||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "autocfg"
|
name = "autocfg"
|
||||||
version = "1.0.1"
|
version = "1.1.0"
|
||||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
checksum = "cdb031dd78e28731d87d56cc8ffef4a8f36ca26c38fe2de700543e627f8a464a"
|
checksum = "d468802bab17cbc0cc575e9b053f41e72aa36bfa6b7f55e3529ffa43161b97fa"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "bitflags"
|
||||||
|
version = "1.3.2"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "bef38d45163c2f1dde094a7dfd33ccf595c92905c8f8f4fdc18d06fb1037718a"
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "cc"
|
name = "cc"
|
||||||
@@ -56,6 +62,31 @@ dependencies = [
|
|||||||
"lazy_static",
|
"lazy_static",
|
||||||
]
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "crossterm"
|
||||||
|
version = "0.25.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "e64e6c0fbe2c17357405f7c758c1ef960fce08bdfb2c03d88d2a18d7e09c4b67"
|
||||||
|
dependencies = [
|
||||||
|
"bitflags",
|
||||||
|
"crossterm_winapi",
|
||||||
|
"libc",
|
||||||
|
"mio",
|
||||||
|
"parking_lot",
|
||||||
|
"signal-hook",
|
||||||
|
"signal-hook-mio",
|
||||||
|
"winapi",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "crossterm_winapi"
|
||||||
|
version = "0.9.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "2ae1b35a484aa10e07fe0638d02301c5ad24de82d310ccbd2f3693da5f09bf1c"
|
||||||
|
dependencies = [
|
||||||
|
"winapi",
|
||||||
|
]
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "lazy_static"
|
name = "lazy_static"
|
||||||
version = "1.4.0"
|
version = "1.4.0"
|
||||||
@@ -70,9 +101,40 @@ checksum = "478ee9e62aaeaf5b140bd4138753d1f109765488581444218d3ddda43234f3e8"
|
|||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "libc"
|
name = "libc"
|
||||||
version = "0.2.108"
|
version = "0.2.135"
|
||||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
checksum = "8521a1b57e76b1ec69af7599e75e38e7b7fad6610f037db8c79b127201b5d119"
|
checksum = "68783febc7782c6c5cb401fbda4de5a9898be1762314da0bb2c10ced61f18b0c"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "lock_api"
|
||||||
|
version = "0.4.9"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "435011366fe56583b16cf956f9df0095b405b82d76425bc8981c0e22e60ec4df"
|
||||||
|
dependencies = [
|
||||||
|
"autocfg",
|
||||||
|
"scopeguard",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "log"
|
||||||
|
version = "0.4.17"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "abb12e687cfb44aa40f41fc3978ef76448f9b6038cad6aef4259d3c095a2382e"
|
||||||
|
dependencies = [
|
||||||
|
"cfg-if",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "mio"
|
||||||
|
version = "0.8.4"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "57ee1c23c7c63b0c9250c339ffdc69255f110b298b901b9f6c82547b7b87caaf"
|
||||||
|
dependencies = [
|
||||||
|
"libc",
|
||||||
|
"log",
|
||||||
|
"wasi 0.11.0+wasi-snapshot-preview1",
|
||||||
|
"windows-sys 0.36.1",
|
||||||
|
]
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "num-traits"
|
name = "num-traits"
|
||||||
@@ -83,6 +145,29 @@ dependencies = [
|
|||||||
"autocfg",
|
"autocfg",
|
||||||
]
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "parking_lot"
|
||||||
|
version = "0.12.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "3742b2c103b9f06bc9fff0a37ff4912935851bee6d36f3c02bcc755bcfec228f"
|
||||||
|
dependencies = [
|
||||||
|
"lock_api",
|
||||||
|
"parking_lot_core",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "parking_lot_core"
|
||||||
|
version = "0.9.4"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "4dc9e0dc2adc1c69d09143aff38d3d30c5c3f0df0dad82e6d25547af174ebec0"
|
||||||
|
dependencies = [
|
||||||
|
"cfg-if",
|
||||||
|
"libc",
|
||||||
|
"redox_syscall",
|
||||||
|
"smallvec",
|
||||||
|
"windows-sys 0.42.0",
|
||||||
|
]
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "pbr"
|
name = "pbr"
|
||||||
version = "1.0.4"
|
version = "1.0.4"
|
||||||
@@ -113,6 +198,15 @@ dependencies = [
|
|||||||
"proc-macro2",
|
"proc-macro2",
|
||||||
]
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "redox_syscall"
|
||||||
|
version = "0.2.16"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "fb5a58c1855b4b6819d59012155603f0b22ad30cad752600aadfcb695265519a"
|
||||||
|
dependencies = [
|
||||||
|
"bitflags",
|
||||||
|
]
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "sacabase"
|
name = "sacabase"
|
||||||
version = "2.0.0"
|
version = "2.0.0"
|
||||||
@@ -122,6 +216,48 @@ dependencies = [
|
|||||||
"num-traits",
|
"num-traits",
|
||||||
]
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "scopeguard"
|
||||||
|
version = "1.1.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "d29ab0c6d3fc0ee92fe66e2d99f700eab17a8d57d1c1d3b748380fb20baa78cd"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "signal-hook"
|
||||||
|
version = "0.3.14"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "a253b5e89e2698464fc26b545c9edceb338e18a89effeeecfea192c3025be29d"
|
||||||
|
dependencies = [
|
||||||
|
"libc",
|
||||||
|
"signal-hook-registry",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "signal-hook-mio"
|
||||||
|
version = "0.2.3"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "29ad2e15f37ec9a6cc544097b78a1ec90001e9f71b81338ca39f430adaca99af"
|
||||||
|
dependencies = [
|
||||||
|
"libc",
|
||||||
|
"mio",
|
||||||
|
"signal-hook",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "signal-hook-registry"
|
||||||
|
version = "1.4.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "e51e73328dc4ac0c7ccbda3a494dfa03df1de2f46018127f60c693f2648455b0"
|
||||||
|
dependencies = [
|
||||||
|
"libc",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "smallvec"
|
||||||
|
version = "1.10.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "a507befe795404456341dfab10cef66ead4c041f62b8b11bbb92bffe5d0953e0"
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "syn"
|
name = "syn"
|
||||||
version = "1.0.101"
|
version = "1.0.101"
|
||||||
@@ -160,7 +296,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
|
|||||||
checksum = "6db9e6914ab8b1ae1c260a4ae7a49b6c5611b40328a735b21862567685e73255"
|
checksum = "6db9e6914ab8b1ae1c260a4ae7a49b6c5611b40328a735b21862567685e73255"
|
||||||
dependencies = [
|
dependencies = [
|
||||||
"libc",
|
"libc",
|
||||||
"wasi",
|
"wasi 0.10.0+wasi-snapshot-preview1",
|
||||||
"winapi",
|
"winapi",
|
||||||
]
|
]
|
||||||
|
|
||||||
@@ -176,6 +312,7 @@ version = "0.2.0"
|
|||||||
dependencies = [
|
dependencies = [
|
||||||
"anyhow",
|
"anyhow",
|
||||||
"cdivsufsort",
|
"cdivsufsort",
|
||||||
|
"crossterm",
|
||||||
"lexopt",
|
"lexopt",
|
||||||
"pbr",
|
"pbr",
|
||||||
"thiserror",
|
"thiserror",
|
||||||
@@ -187,6 +324,12 @@ version = "0.10.0+wasi-snapshot-preview1"
|
|||||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
checksum = "1a143597ca7c7793eff794def352d41792a93c481eb1042423ff7ff72ba2c31f"
|
checksum = "1a143597ca7c7793eff794def352d41792a93c481eb1042423ff7ff72ba2c31f"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "wasi"
|
||||||
|
version = "0.11.0+wasi-snapshot-preview1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "9c8d87e72b64a3b4db28d11ce29237c246188f4f51057d65a7eab63b7987e423"
|
||||||
|
|
||||||
[[package]]
|
[[package]]
|
||||||
name = "winapi"
|
name = "winapi"
|
||||||
version = "0.3.9"
|
version = "0.3.9"
|
||||||
@@ -208,3 +351,103 @@ name = "winapi-x86_64-pc-windows-gnu"
|
|||||||
version = "0.4.0"
|
version = "0.4.0"
|
||||||
source = "registry+https://github.com/rust-lang/crates.io-index"
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
checksum = "712e227841d057c1ee1cd2fb22fa7e5a5461ae8e48fa2ca79ec42cfc1931183f"
|
checksum = "712e227841d057c1ee1cd2fb22fa7e5a5461ae8e48fa2ca79ec42cfc1931183f"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows-sys"
|
||||||
|
version = "0.36.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "ea04155a16a59f9eab786fe12a4a450e75cdb175f9e0d80da1e17db09f55b8d2"
|
||||||
|
dependencies = [
|
||||||
|
"windows_aarch64_msvc 0.36.1",
|
||||||
|
"windows_i686_gnu 0.36.1",
|
||||||
|
"windows_i686_msvc 0.36.1",
|
||||||
|
"windows_x86_64_gnu 0.36.1",
|
||||||
|
"windows_x86_64_msvc 0.36.1",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows-sys"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "5a3e1820f08b8513f676f7ab6c1f99ff312fb97b553d30ff4dd86f9f15728aa7"
|
||||||
|
dependencies = [
|
||||||
|
"windows_aarch64_gnullvm",
|
||||||
|
"windows_aarch64_msvc 0.42.0",
|
||||||
|
"windows_i686_gnu 0.42.0",
|
||||||
|
"windows_i686_msvc 0.42.0",
|
||||||
|
"windows_x86_64_gnu 0.42.0",
|
||||||
|
"windows_x86_64_gnullvm",
|
||||||
|
"windows_x86_64_msvc 0.42.0",
|
||||||
|
]
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_aarch64_gnullvm"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "41d2aa71f6f0cbe00ae5167d90ef3cfe66527d6f613ca78ac8024c3ccab9a19e"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_aarch64_msvc"
|
||||||
|
version = "0.36.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "9bb8c3fd39ade2d67e9874ac4f3db21f0d710bee00fe7cab16949ec184eeaa47"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_aarch64_msvc"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "dd0f252f5a35cac83d6311b2e795981f5ee6e67eb1f9a7f64eb4500fbc4dcdb4"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_i686_gnu"
|
||||||
|
version = "0.36.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "180e6ccf01daf4c426b846dfc66db1fc518f074baa793aa7d9b9aaeffad6a3b6"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_i686_gnu"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "fbeae19f6716841636c28d695375df17562ca208b2b7d0dc47635a50ae6c5de7"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_i686_msvc"
|
||||||
|
version = "0.36.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "e2e7917148b2812d1eeafaeb22a97e4813dfa60a3f8f78ebe204bcc88f12f024"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_i686_msvc"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "84c12f65daa39dd2babe6e442988fc329d6243fdce47d7d2d155b8d874862246"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_x86_64_gnu"
|
||||||
|
version = "0.36.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "4dcd171b8776c41b97521e5da127a2d86ad280114807d0b2ab1e462bc764d9e1"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_x86_64_gnu"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "bf7b1b21b5362cbc318f686150e5bcea75ecedc74dd157d874d754a2ca44b0ed"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_x86_64_gnullvm"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "09d525d2ba30eeb3297665bd434a54297e4170c7f1a44cad4ef58095b4cd2028"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_x86_64_msvc"
|
||||||
|
version = "0.36.1"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "c811ca4a8c853ef420abd8592ba53ddbbac90410fab6903b3e79972a631f7680"
|
||||||
|
|
||||||
|
[[package]]
|
||||||
|
name = "windows_x86_64_msvc"
|
||||||
|
version = "0.42.0"
|
||||||
|
source = "registry+https://github.com/rust-lang/crates.io-index"
|
||||||
|
checksum = "f40009d85759725a34da6d89a94e63d7bdc50a862acf0dbc7c8e488f1edcb6f5"
|
||||||
|
|||||||
@@ -11,4 +11,5 @@ cdivsufsort = "2"
|
|||||||
lexopt = "0.2.1"
|
lexopt = "0.2.1"
|
||||||
anyhow = "1"
|
anyhow = "1"
|
||||||
thiserror = "1.0.36"
|
thiserror = "1.0.36"
|
||||||
pbr = "1"
|
pbr = "1"
|
||||||
|
crossterm = { version = "0.25.0", default-features = false, optional = true }
|
||||||
|
|||||||
14
README.md
14
README.md
@@ -20,18 +20,20 @@ is both about twice as fast and smaller than the Shrinkler unpacker.
|
|||||||
The release includes a reference c unpacker, as well as some optimized asm unpackers (arm and riscv). The unpckers in
|
The release includes a reference c unpacker, as well as some optimized asm unpackers (arm and riscv). The unpckers in
|
||||||
c_unpacker and asm_unpackers unpack the default upkr compressed format. The z80_unpacker
|
c_unpacker and asm_unpackers unpack the default upkr compressed format. The z80_unpacker
|
||||||
is based on some variations to the compressed format. (Use `upkr --z80` to select those variations.)
|
is based on some variations to the compressed format. (Use `upkr --z80` to select those variations.)
|
||||||
An optimized x86 (DOS) unpacker is currently being worked on out of tree.
|
The 16 bit dos unpacker also uses some variations. (`upkr --x86`)
|
||||||
|
|
||||||
## Usage
|
## Usage
|
||||||
|
|
||||||
```
|
```
|
||||||
upkr [-l level(0-9)] [config options] <infile> [<outfile>]
|
upkr [-l level(0-9)] [config options] <infile> [<outfile>]
|
||||||
upkr -u [config options] <infile> [<outfile>]
|
upkr -u [config options] <infile> [<outfile>]
|
||||||
|
upkr --heatmap [config options] <infile> [<outfile>]
|
||||||
upkr --margin [config options] <infile>
|
upkr --margin [config options] <infile>
|
||||||
|
|
||||||
-l, --level N compression level 0-9
|
-l, --level N compression level 0-9
|
||||||
-0, ..., -9 short form for setting compression level
|
-0, ..., -9 short form for setting compression level
|
||||||
-u, --unpack unpack infile
|
-u, --unpack unpack infile
|
||||||
|
--heatmap calculate heatmap from compressed file
|
||||||
--margin calculate margin for overlapped unpacking of a packed file
|
--margin calculate margin for overlapped unpacking of a packed file
|
||||||
|
|
||||||
Config presets for specific unpackers:
|
Config presets for specific unpackers:
|
||||||
@@ -56,3 +58,13 @@ Config options to tailor output to specific optimized unpackers:
|
|||||||
--max-offset N
|
--max-offset N
|
||||||
--max-length N
|
--max-length N
|
||||||
```
|
```
|
||||||
|
|
||||||
|
## Heatmap
|
||||||
|
|
||||||
|
By default, the `--heatmap` flag writes out the heatmap data as a binary file. The heatmap file is
|
||||||
|
the same size as the unpacked data. Each byte can be interpreted like this:
|
||||||
|
|
||||||
|
```
|
||||||
|
is_literal = byte & 1; // whether the byte was encoded as a literal (as opposed to a match)
|
||||||
|
size_in_bits = 2.0 ** (((byte >> 1) - 64) / 8.0); // the size this byte takes up in the compressed data
|
||||||
|
```
|
||||||
|
|||||||
@@ -17,7 +17,7 @@
|
|||||||
upkr_unpack:
|
upkr_unpack:
|
||||||
push { r3-r11, lr }
|
push { r3-r11, lr }
|
||||||
|
|
||||||
mov r2, #384
|
mov r2, #384
|
||||||
mov r3, #128
|
mov r3, #128
|
||||||
.Lclear:
|
.Lclear:
|
||||||
subs r2, r2, #1
|
subs r2, r2, #1
|
||||||
@@ -29,7 +29,7 @@ upkr_unpack:
|
|||||||
bl upkr_decode_bit
|
bl upkr_decode_bit
|
||||||
bcc .Ldata
|
bcc .Ldata
|
||||||
.Lmatch:
|
.Lmatch:
|
||||||
mov r5, #256
|
mov r5, #256
|
||||||
rsbs r6, r4, #0
|
rsbs r6, r4, #0
|
||||||
blcc upkr_decode_bit
|
blcc upkr_decode_bit
|
||||||
bcc .Lskip_offset
|
bcc .Lskip_offset
|
||||||
@@ -39,7 +39,7 @@ upkr_unpack:
|
|||||||
popeq { r3-r11, pc }
|
popeq { r3-r11, pc }
|
||||||
.Lskip_offset:
|
.Lskip_offset:
|
||||||
|
|
||||||
mov r5, #256+64
|
mov r5, #256+64
|
||||||
bl upkr_decode_length
|
bl upkr_decode_length
|
||||||
.Lcopy_loop:
|
.Lcopy_loop:
|
||||||
ldrb r5, [r0, r3]
|
ldrb r5, [r0, r3]
|
||||||
@@ -55,46 +55,46 @@ upkr_unpack:
|
|||||||
.Ldata_loop:
|
.Ldata_loop:
|
||||||
bl upkr_decode_bit
|
bl upkr_decode_bit
|
||||||
adc r5, r5, r5
|
adc r5, r5, r5
|
||||||
movs r4, r5, lsr #8
|
movs r4, r5, lsr #8
|
||||||
beq .Ldata_loop
|
beq .Ldata_loop
|
||||||
b .Lstore
|
b .Lstore
|
||||||
|
|
||||||
.type upkr_decode_length, %function
|
.type upkr_decode_length, %function
|
||||||
upkr_decode_length:
|
upkr_decode_length:
|
||||||
mov r12, lr
|
mov r12, lr
|
||||||
|
|
||||||
mov r4, #0
|
mov r4, #0
|
||||||
mvn r6, #0
|
mvn r6, #0
|
||||||
.Lbit_loop:
|
.Lbit_loop:
|
||||||
bl upkr_decode_bit_inc
|
bl upkr_decode_bit_inc
|
||||||
addcc r4, r4, r6
|
addcc r4, r4, r6
|
||||||
movcc pc, r12
|
movcc pc, r12
|
||||||
|
|
||||||
bl upkr_decode_bit_inc
|
bl upkr_decode_bit_inc
|
||||||
addcs r4, r4, r6
|
addcs r4, r4, r6
|
||||||
mov r6, r6, lsl #1
|
mov r6, r6, lsl #1
|
||||||
b .Lbit_loop
|
b .Lbit_loop
|
||||||
|
|
||||||
.type upkr_decode_bit, %function
|
.type upkr_decode_bit, %function
|
||||||
upkr_decode_bit_inc:
|
upkr_decode_bit_inc:
|
||||||
add r5, r5, #1
|
add r5, r5, #1
|
||||||
upkr_decode_bit:
|
upkr_decode_bit:
|
||||||
cmp r2, #4096
|
cmp r2, #4096
|
||||||
ldrltb r8, [r1], #1
|
ldrltb r8, [r1], #1
|
||||||
orrlt r2, r8, r2, lsl#8
|
orrlt r2, r8, r2, lsl#8
|
||||||
blt upkr_decode_bit
|
blt upkr_decode_bit
|
||||||
|
|
||||||
ldrb r8, [sp, -r5]
|
ldrb r8, [sp, -r5]
|
||||||
and r9, r2, #255
|
and r9, r2, #255
|
||||||
add r9, r9, #1
|
add r9, r9, #1
|
||||||
cmp r8, r9
|
cmp r8, r9
|
||||||
rsbcs r8, r8, #256
|
rsbcs r8, r8, #256
|
||||||
mvn r9, r2, lsr#8
|
mvn r9, r2, lsr#8
|
||||||
addcs r9, r9, #1
|
addcs r9, r9, #1
|
||||||
mla r2, r8, r9, r2
|
mla r2, r8, r9, r2
|
||||||
add r9, r8, #8
|
add r9, r8, #8
|
||||||
sub r8, r8, r9, lsr#4
|
sub r8, r8, r9, lsr#4
|
||||||
rsbcs r8, r8, #256
|
rsbcs r8, r8, #256
|
||||||
strb r8, [sp, -r5]
|
strb r8, [sp, -r5]
|
||||||
mov pc, r14
|
mov pc, r14
|
||||||
|
|
||||||
|
|||||||
13
dos_unpacker/readme.txt
Normal file
13
dos_unpacker/readme.txt
Normal file
@@ -0,0 +1,13 @@
|
|||||||
|
16 bit DOS executable stubs
|
||||||
|
---------------------------
|
||||||
|
|
||||||
|
by pestis and TomCat
|
||||||
|
|
||||||
|
unpack_x86_16_DOS.asm:
|
||||||
|
maximum compatibility, relocates unpacked code to normal start address
|
||||||
|
unpack_x86_16_DOS_no_relocation.asm:
|
||||||
|
saves some bytes by not relocating, unpacked code needs to be assembled to
|
||||||
|
start at 0x3FFE
|
||||||
|
unpack_x86_16_DOS_no_repeated_offset.asm:
|
||||||
|
removes support for repeated offsets, potentially at the cost of some compression ratio.
|
||||||
|
most likely only a win in very narrow circumstances around the 1kb mark
|
||||||
161
dos_unpacker/unpack_x86_16_DOS.asm
Normal file
161
dos_unpacker/unpack_x86_16_DOS.asm
Normal file
@@ -0,0 +1,161 @@
|
|||||||
|
; Contributions from pestis, TomCat and exoticorn
|
||||||
|
;
|
||||||
|
; This is the 16-bit DOS x86 decompression stub for upkr, which is designed for
|
||||||
|
; maximum compatibility: it relocates the compressed data so it can be
|
||||||
|
; decompressed starting at the normal .COM starting address. In other words,
|
||||||
|
; many of the already existing .COM files should be compressable using this
|
||||||
|
; stub.
|
||||||
|
;
|
||||||
|
; How to use:
|
||||||
|
; 1) Pack your intro using upkr into data.bin with the --x86 command line
|
||||||
|
; argument:
|
||||||
|
;
|
||||||
|
; $ upkr --x86 intro.com data.bin
|
||||||
|
;
|
||||||
|
; 2) Compile this .asm file using nasm (or any compatible assembler):
|
||||||
|
;
|
||||||
|
; $ nasm unpack_x86_16_DOS.asm -fbin -o intropck.com
|
||||||
|
;
|
||||||
|
; The packed size of the intro+stub is limited by max_len (see below) bytes.
|
||||||
|
;
|
||||||
|
; In specific cases, the unpacker stub can be further optimized to save a byte
|
||||||
|
; or two:
|
||||||
|
; 1) You can remove CLC before RET, if you don't mind carry being set upon
|
||||||
|
; program entry
|
||||||
|
; 2) You can also move PUSHA before PUSH SI and put POPA as the first
|
||||||
|
; operation of the compressed code.
|
||||||
|
max_len equ 16384
|
||||||
|
prog_start equ (0x100+max_len+510+relocation-upkr_unpack)
|
||||||
|
probs equ (((prog_start+max_len+510)+255)/256)*256
|
||||||
|
|
||||||
|
org 0x100
|
||||||
|
|
||||||
|
; This is will be loaded at 0x100, but relocates the code and data to prog_start
|
||||||
|
relocation:
|
||||||
|
push si ; si = 0x100 at DOS start, so save it for later ret
|
||||||
|
pusha ; pusha to recall all registers before starting intro
|
||||||
|
push si ; for pop di to start writing the output
|
||||||
|
mov di, prog_start ; the depacker & data are relocated from 0x100 to prog_start
|
||||||
|
mov ch, max_len/512
|
||||||
|
rep movsw
|
||||||
|
jmp si ; jump to relocated upkr_unpack
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_unpack unpacks the code to 0x100 and runs it when done.
|
||||||
|
upkr_unpack:
|
||||||
|
xchg ax, bp ; position in input bitstream (bp) = 0
|
||||||
|
cwd ; upkr_state (dx) = 0;
|
||||||
|
xchg ax, cx ; cx = 0x9XX
|
||||||
|
mov al, 128 ; for(int i = 0; i < sizeof(upkr_probs); ++i) upkr_probs[i] = 128;
|
||||||
|
rep stosb
|
||||||
|
pop di ; u8* write_ptr = (u8*)destination;
|
||||||
|
.mainloop:
|
||||||
|
mov bx, probs
|
||||||
|
call upkr_decode_bit
|
||||||
|
jc .else ; if(upkr_decode_bit(0)) {
|
||||||
|
mov bh, (probs+256)/256
|
||||||
|
jcxz .skip_call
|
||||||
|
call upkr_decode_bit
|
||||||
|
jc .skipoffset
|
||||||
|
.skip_call:
|
||||||
|
stc
|
||||||
|
call upkr_decode_number ; offset = upkr_decode_length(258) - 1;
|
||||||
|
loop .notdone ; if(offset == 0)
|
||||||
|
popa
|
||||||
|
clc
|
||||||
|
ret
|
||||||
|
.notdone:
|
||||||
|
mov si, di
|
||||||
|
.sub:
|
||||||
|
dec si
|
||||||
|
loop .sub
|
||||||
|
.skipoffset:
|
||||||
|
mov bl, 128 ; int length = upkr_decode_length(384);
|
||||||
|
call upkr_decode_number
|
||||||
|
rep movsb ; *write_ptr = write_ptr[-offset];
|
||||||
|
jmp .mainloop
|
||||||
|
.byteloop:
|
||||||
|
call upkr_decode_bit ; int bit = upkr_decode_bit(byte);
|
||||||
|
.else:
|
||||||
|
adc bl, bl ; byte = (byte << 1) + bit;
|
||||||
|
jnc .byteloop
|
||||||
|
xchg ax, bx
|
||||||
|
stosb
|
||||||
|
inc si
|
||||||
|
mov cl, 1
|
||||||
|
jmp .mainloop ; prev_was_match = 0;
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_decode_bit decodes one bit from the rANS entropy encoded bit stream.
|
||||||
|
; parameters:
|
||||||
|
; bx = memory address of the context probability
|
||||||
|
; dx = decoder state
|
||||||
|
; bp = bit position in input stream
|
||||||
|
; returns:
|
||||||
|
; dx = new decoder state
|
||||||
|
; bp = new bit position in input stream
|
||||||
|
; carry = bit
|
||||||
|
; trashes ax
|
||||||
|
upkr_load_bit:
|
||||||
|
bt [compressed_data-relocation+prog_start], bp
|
||||||
|
inc bp
|
||||||
|
adc dx, dx
|
||||||
|
upkr_decode_bit:
|
||||||
|
inc dx ; inc dx, dec dx is used to test the top (sign) bit of dx
|
||||||
|
dec dx
|
||||||
|
jns upkr_load_bit
|
||||||
|
movzx ax, byte [bx] ; u16 prob = upkr_probs[context_index]
|
||||||
|
push ax ; save prob, tmp = prob
|
||||||
|
cmp dl, al ; int bit = (upkr_state & 255) < prob ? 1 : 0; (carry = bit)
|
||||||
|
pushf ; save bit flags
|
||||||
|
jc .bit ; (skip if bit)
|
||||||
|
neg al ; tmp = 256 - tmp;
|
||||||
|
.bit:
|
||||||
|
mov [bx], al ; upkr_probs[context_index] = tmp + (256 - tmp + 8) >> 4;
|
||||||
|
neg byte [bx]
|
||||||
|
shr byte [bx], 4
|
||||||
|
adc [bx], al
|
||||||
|
mul dh ; upkr_state = tmp * (upkr_state >> 8) + (upkr_state & 255);
|
||||||
|
mov dh, 0
|
||||||
|
add dx, ax
|
||||||
|
popf
|
||||||
|
pop ax
|
||||||
|
jc .bit2 ; (skip if bit)
|
||||||
|
neg byte [bx] ; tmp = 256 - tmp;
|
||||||
|
sub dx, ax ; upkr_state -= prob; note that this will also leave carry always unset, which is what we want
|
||||||
|
.bit2:
|
||||||
|
ret ; return the bit in carry
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_decode_number loads a variable length encoded number (up to 16 bits) from
|
||||||
|
; the compressed stream. Only numbers 1..65535 can be encoded. If the encoded
|
||||||
|
; number has 4 bits and is 1ABC, it is encoded using a kind of an "interleaved
|
||||||
|
; elias code": 0A0B0C1. The 1 in the end implies that no more bits are coming.
|
||||||
|
; parameters:
|
||||||
|
; cx = must be 0
|
||||||
|
; bx = memory address of the context probability
|
||||||
|
; dx = decoder state
|
||||||
|
; bp = bit position in input stream
|
||||||
|
; carry = must be 1
|
||||||
|
; returns:
|
||||||
|
; cx = length
|
||||||
|
; dx = new decoder state
|
||||||
|
; bp = new bit position in input stream
|
||||||
|
; carry = 1
|
||||||
|
; trashes bl, ax
|
||||||
|
upkr_decode_number_loop:
|
||||||
|
inc bx
|
||||||
|
call upkr_decode_bit
|
||||||
|
upkr_decode_number:
|
||||||
|
rcr cx, 1
|
||||||
|
inc bx
|
||||||
|
call upkr_decode_bit
|
||||||
|
jnc upkr_decode_number_loop ; 0 = there's more bits coming, 1 = no more bits
|
||||||
|
.loop2:
|
||||||
|
rcr cx, 1
|
||||||
|
jnc .loop2
|
||||||
|
ret
|
||||||
|
|
||||||
|
|
||||||
|
compressed_data:
|
||||||
|
incbin "data.bin"
|
||||||
152
dos_unpacker/unpack_x86_16_DOS_no_relocation.asm
Normal file
152
dos_unpacker/unpack_x86_16_DOS_no_relocation.asm
Normal file
@@ -0,0 +1,152 @@
|
|||||||
|
; Contributions from pestis, TomCat and exoticorn
|
||||||
|
;
|
||||||
|
; This is the 16-bit DOS x86 decompression stub for upkr, which decompresses the
|
||||||
|
; code starting at address 0x3FFE (or whatever is defined by the entrypoint
|
||||||
|
; below). Thus, the packed code needs to be assembled with org 0x3FFE to work.
|
||||||
|
;
|
||||||
|
; How to use:
|
||||||
|
; 1) Put POPA as the first instruction of your compiled code and use org
|
||||||
|
; 0x3FFE
|
||||||
|
; 2) Pack your intro using upkr into data.bin with the --x86 command line
|
||||||
|
; argument:
|
||||||
|
;
|
||||||
|
; $ upkr --x86 intro.com data.bin
|
||||||
|
;
|
||||||
|
; 2) Compile this .asm file using nasm (or any compatible assembler) e.g.
|
||||||
|
;
|
||||||
|
; $ nasm unpack_x86_16_DOS_no_relocation.asm -fbin -o intropck.com
|
||||||
|
;
|
||||||
|
; In specific cases, the unpacker stub can be further optimized to save a byte
|
||||||
|
; or two:
|
||||||
|
; 1) If your stub+compressed code is 2k or smaller, you can save 1 byte by
|
||||||
|
; putting probs at 0x900 and initializing DI with SALC; XCHG AX, DI instead
|
||||||
|
; of MOV DI, probs
|
||||||
|
; 2) If you remove the PUSHA (and POPA in the compressed code), then you can
|
||||||
|
; assume the registers as follows: AX = 0x00XX, BX = probs + 0x1XX, CX = 0
|
||||||
|
; DX = (trash), SI = DI = right after your program, SP = as it was when the
|
||||||
|
; program started, flags = carry set
|
||||||
|
;
|
||||||
|
; Note that even with the PUSHA / POPA, carry will be set (!) unlike normal dos
|
||||||
|
; program.
|
||||||
|
entry equ 0x3FFE
|
||||||
|
probs equ entry - 0x1FE ; must be aligned to 256
|
||||||
|
|
||||||
|
org 0x100
|
||||||
|
|
||||||
|
|
||||||
|
; This is will be loaded at 0x100, but relocates the code and data to prog_start
|
||||||
|
upkr_unpack:
|
||||||
|
pusha
|
||||||
|
xchg ax, bp ; position in bitstream = 0
|
||||||
|
cwd ; upkr_state = 0;
|
||||||
|
mov di, probs
|
||||||
|
mov ax, 0x8080 ; for(int i = 0; i < sizeof(upkr_probs); ++i) upkr_probs[i] = 128;
|
||||||
|
rep stosw
|
||||||
|
push di
|
||||||
|
.mainloop:
|
||||||
|
mov bx, probs
|
||||||
|
call upkr_decode_bit
|
||||||
|
jc .else ; if(upkr_decode_bit(0)) {
|
||||||
|
mov bh, (probs+256)/256
|
||||||
|
jcxz .skip_call ; if(prev_was_match || upkr_decode_bit(257)) {
|
||||||
|
call upkr_decode_bit
|
||||||
|
jc .skipoffset
|
||||||
|
.skip_call:
|
||||||
|
stc
|
||||||
|
call upkr_decode_number ; offset = upkr_decode_number(258) - 1;
|
||||||
|
mov si, di
|
||||||
|
loop .sub ; if(offset == 0)
|
||||||
|
ret
|
||||||
|
.sub:
|
||||||
|
dec si
|
||||||
|
loop .sub
|
||||||
|
.skipoffset:
|
||||||
|
mov bl, 128 ; int length = upkr_decode_number(384);
|
||||||
|
call upkr_decode_number
|
||||||
|
rep movsb ; *write_ptr = write_ptr[-offset];
|
||||||
|
jmp .mainloop
|
||||||
|
.byteloop:
|
||||||
|
call upkr_decode_bit ; int bit = upkr_decode_bit(byte);
|
||||||
|
.else:
|
||||||
|
adc bl, bl ; byte = (byte << 1) + bit;
|
||||||
|
jnc .byteloop
|
||||||
|
xchg ax, bx
|
||||||
|
stosb
|
||||||
|
inc si
|
||||||
|
mov cl, 1
|
||||||
|
jmp .mainloop ; prev_was_match = 0;
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_decode_bit decodes one bit from the rANS entropy encoded bit stream.
|
||||||
|
; parameters:
|
||||||
|
; bx = memory address of the context probability
|
||||||
|
; dx = decoder state
|
||||||
|
; bp = bit position in input stream
|
||||||
|
; returns:
|
||||||
|
; dx = new decoder state
|
||||||
|
; bp = new bit position in input stream
|
||||||
|
; carry = bit
|
||||||
|
; trashes ax
|
||||||
|
upkr_load_bit:
|
||||||
|
bt [compressed_data], bp
|
||||||
|
inc bp
|
||||||
|
adc dx, dx
|
||||||
|
upkr_decode_bit:
|
||||||
|
inc dx
|
||||||
|
dec dx ; inc dx, dec dx is used to test the top (sign) bit of dx
|
||||||
|
jns upkr_load_bit
|
||||||
|
movzx ax, byte [bx] ; int prob = upkr_probs[context_index]
|
||||||
|
push ax ; save prob
|
||||||
|
cmp dl, al ; int bit = (upkr_state & 255) < prob ? 1 : 0; (carry = bit)
|
||||||
|
pushf ; save bit flags
|
||||||
|
jc .bit ; (skip if bit)
|
||||||
|
neg al ; tmp = 256 - tmp;
|
||||||
|
.bit:
|
||||||
|
mov [bx], al ; tmp_new = tmp + (256 - tmp + 8) >> 4;
|
||||||
|
neg byte [bx]
|
||||||
|
shr byte [bx], 4
|
||||||
|
adc [bx], al
|
||||||
|
mul dh ; upkr_state = tmp * (upkr_state >> 8) + (upkr_state & 255);
|
||||||
|
mov dh, 0
|
||||||
|
add dx, ax
|
||||||
|
popf
|
||||||
|
pop ax
|
||||||
|
jc .bit2 ; (skip if bit)
|
||||||
|
neg byte [bx] ; tmp = 256 - tmp;
|
||||||
|
sub dx, ax ; upkr_state -= prob; note that this will also leave carry always unset, which is what we want
|
||||||
|
.bit2:
|
||||||
|
ret ; flags = bit
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_decode_number loads a variable length encoded number (up to 16 bits) from
|
||||||
|
; the compressed stream. Only numbers 1..65535 can be encoded. If the encoded
|
||||||
|
; number has 4 bits and is 1ABC, it is encoded using a kind of an "interleaved
|
||||||
|
; elias code": 0A0B0C1. The 1 in the end implies that no more bits are coming.
|
||||||
|
; parameters:
|
||||||
|
; cx = must be 0
|
||||||
|
; bx = memory address of the context probability
|
||||||
|
; dx = decoder state
|
||||||
|
; bp = bit position in input stream
|
||||||
|
; carry = must be 1
|
||||||
|
; returns:
|
||||||
|
; cx = length
|
||||||
|
; dx = new decoder state
|
||||||
|
; bp = new bit position in input stream
|
||||||
|
; carry = 1
|
||||||
|
; trashes bl, ax
|
||||||
|
upkr_decode_number_loop:
|
||||||
|
inc bx
|
||||||
|
call upkr_decode_bit
|
||||||
|
upkr_decode_number:
|
||||||
|
rcr cx, 1
|
||||||
|
inc bx
|
||||||
|
call upkr_decode_bit
|
||||||
|
jnc upkr_decode_number_loop ; while(upkr_decode_bit(context_index)) {
|
||||||
|
.loop2:
|
||||||
|
rcr cx, 1
|
||||||
|
jnc .loop2
|
||||||
|
ret
|
||||||
|
|
||||||
|
|
||||||
|
compressed_data:
|
||||||
|
incbin "data.bin"
|
||||||
155
dos_unpacker/unpack_x86_16_DOS_no_repeated_offset.asm
Normal file
155
dos_unpacker/unpack_x86_16_DOS_no_repeated_offset.asm
Normal file
@@ -0,0 +1,155 @@
|
|||||||
|
; Contributions from pestis, TomCat and exoticorn
|
||||||
|
;
|
||||||
|
; This is the 16-bit DOS x86 decompression stub for upkr, which is designed for
|
||||||
|
; the --no-repeated-offsets option of upkr. The decompression stub is slightly
|
||||||
|
; smaller, but the compressed data might be bigger, so you have to test if
|
||||||
|
; --no-repeated-offsets pays off in the end. This stub relocates the compressed
|
||||||
|
; data so it can be decompressed starting at the normal .COM starting address.
|
||||||
|
;
|
||||||
|
; How to use:
|
||||||
|
; 1) Pack your intro using upkr into data.bin with the --x86b command line
|
||||||
|
; argument: (notice the --x86b, not --x86!)
|
||||||
|
;
|
||||||
|
; $ upkr --x86b intro.com data.bin
|
||||||
|
;
|
||||||
|
; 2) Compile this .asm file using nasm (or any compatible assembler):
|
||||||
|
;
|
||||||
|
; $ nasm unpack_x86_16_DOS_no_repeated_offsets.asm -fbin -o intropck.com
|
||||||
|
;
|
||||||
|
; The packed size of the intro+stub is limited by max_len (see below) bytes.
|
||||||
|
;
|
||||||
|
; In specific cases, the unpacker stub can be further optimized to save a byte
|
||||||
|
; or two:
|
||||||
|
; 1) You can remove CLC before RET, if you don't mind carry being set upon
|
||||||
|
; program entry
|
||||||
|
; 2) You can also move PUSHA before PUSH SI and put POPA as the first
|
||||||
|
; operation of the compressed code.
|
||||||
|
max_len equ 16384
|
||||||
|
prog_start equ (0x100+max_len+510+relocation-upkr_unpack)
|
||||||
|
probs equ (((prog_start+max_len+510)+255)/256)*256
|
||||||
|
|
||||||
|
org 0x100
|
||||||
|
|
||||||
|
|
||||||
|
; This is will be loaded at 0x100, but relocates the code and data to prog_start
|
||||||
|
relocation:
|
||||||
|
push si ; si = 0x100 at DOS start, so save it for later ret
|
||||||
|
pusha ; pusha to recall all registers before starting intro
|
||||||
|
push si ; for pop di to start writing the output
|
||||||
|
mov di, prog_start ; the depacker & data are relocated from 0x100 to prog_start
|
||||||
|
mov ch, max_len/512
|
||||||
|
rep movsw
|
||||||
|
jmp si ; jump to relocated upkr_unpack
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_unpack unpacks the code to 0x100 and runs it when done.
|
||||||
|
upkr_unpack:
|
||||||
|
xchg ax, bp ; position in bitstream = 0
|
||||||
|
cwd ; upkr_state = 0;
|
||||||
|
xchg cx, ax ; cx > 0x0200
|
||||||
|
mov al, 128 ; for(int i = 0; i < sizeof(upkr_probs); ++i) upkr_probs[i] = 128;
|
||||||
|
rep stosb
|
||||||
|
pop di ; u8* write_ptr = (u8*)destination;
|
||||||
|
.mainloop:
|
||||||
|
mov bx, probs
|
||||||
|
call upkr_decode_bit
|
||||||
|
jnc .else ; if(upkr_decode_bit(0)) {
|
||||||
|
inc bh
|
||||||
|
call upkr_decode_number ; offset = upkr_decode_number(258) - 1;
|
||||||
|
loop .notdone ; if(offset == 0)
|
||||||
|
popa
|
||||||
|
clc
|
||||||
|
ret
|
||||||
|
.notdone:
|
||||||
|
mov si, di
|
||||||
|
.sub:
|
||||||
|
dec si
|
||||||
|
loop .sub
|
||||||
|
mov bl, 128 ; int length = upkr_decode_number(384);
|
||||||
|
call upkr_decode_number
|
||||||
|
rep movsb ; *write_ptr = write_ptr[-offset];
|
||||||
|
jmp .mainloop
|
||||||
|
.else:
|
||||||
|
inc bx
|
||||||
|
.byteloop:
|
||||||
|
call upkr_decode_bit ; int bit = upkr_decode_bit(byte);
|
||||||
|
adc bl, bl ; byte = (byte << 1) + bit;
|
||||||
|
jnc .byteloop
|
||||||
|
xchg ax, bx
|
||||||
|
stosb
|
||||||
|
jmp .mainloop ; prev_was_match = 0;
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_decode_bit decodes one bit from the rANS entropy encoded bit stream.
|
||||||
|
; parameters:
|
||||||
|
; bx = memory address of the context probability
|
||||||
|
; dx = decoder state
|
||||||
|
; bp = bit position in input stream
|
||||||
|
; returns:
|
||||||
|
; dx = new decoder state
|
||||||
|
; bp = new bit position in input stream
|
||||||
|
; carry = bit
|
||||||
|
; trashes ax
|
||||||
|
upkr_load_bit:
|
||||||
|
bt [compressed_data-relocation+prog_start], bp
|
||||||
|
inc bp
|
||||||
|
adc dx, dx
|
||||||
|
upkr_decode_bit:
|
||||||
|
inc dx
|
||||||
|
dec dx ; or whatever other test for the top bit there is
|
||||||
|
jns upkr_load_bit
|
||||||
|
movzx ax, byte [bx] ; int prob = upkr_probs[context_index]
|
||||||
|
push ax ; save prob
|
||||||
|
cmp dl, al ; int bit = (upkr_state & 255) < prob ? 1 : 0; (carry = bit)
|
||||||
|
pushf ; save bit flags
|
||||||
|
jc .bit ; (skip if bit)
|
||||||
|
neg al ; tmp = 256 - tmp;
|
||||||
|
.bit:
|
||||||
|
mov [bx], al ; tmp += (256 - tmp + 8) >> 4;
|
||||||
|
neg byte [bx]
|
||||||
|
shr byte [bx],4
|
||||||
|
adc [bx], al ; upkr_probs[context_index] = tmp;
|
||||||
|
mul dh ; upkr_state = tmp * (upkr_state >> 8) + (upkr_state & 255);
|
||||||
|
mov dh, 0
|
||||||
|
add dx, ax
|
||||||
|
popf
|
||||||
|
pop ax
|
||||||
|
jc .bit2 ; (skip if bit)
|
||||||
|
neg byte [bx] ; tmp = 256 - tmp;
|
||||||
|
sub dx, ax ; upkr_state -= prob; note that this will also leave carry always unset, which is what we want
|
||||||
|
.bit2:
|
||||||
|
ret ; flags = bit
|
||||||
|
|
||||||
|
|
||||||
|
; upkr_decode_number loads a variable length encoded number (up to 16 bits) from
|
||||||
|
; the compressed stream. Only numbers 1..65535 can be encoded. If the encoded
|
||||||
|
; number has 4 bits and is 1ABC, it is encoded using a kind of an "interleaved
|
||||||
|
; elias code": 0A0B0C1. The 1 in the end implies that no more bits are coming.
|
||||||
|
; parameters:
|
||||||
|
; cx = must be 0
|
||||||
|
; bx = memory address of the context probability
|
||||||
|
; dx = decoder state
|
||||||
|
; bp = bit position in input stream
|
||||||
|
; carry = must be 1
|
||||||
|
; returns:
|
||||||
|
; cx = length
|
||||||
|
; dx = new decoder state
|
||||||
|
; bp = new bit position in input stream
|
||||||
|
; carry = 1
|
||||||
|
; trashes bl, ax
|
||||||
|
upkr_decode_number_loop:
|
||||||
|
inc bx
|
||||||
|
call upkr_decode_bit
|
||||||
|
upkr_decode_number:
|
||||||
|
rcr cx, 1
|
||||||
|
inc bx
|
||||||
|
call upkr_decode_bit
|
||||||
|
jnc upkr_decode_number_loop ; 0 = there's more bits coming, 1 = no more bits
|
||||||
|
.loop2:
|
||||||
|
rcr cx, 1
|
||||||
|
jnc .loop2
|
||||||
|
ret
|
||||||
|
|
||||||
|
|
||||||
|
compressed_data:
|
||||||
|
incbin "data.bin"
|
||||||
@@ -23,12 +23,12 @@ upkr-windows-$(VERSION).zip: upkr-windows/upkr.exe PHONY
|
|||||||
zip -r -9 $@ upkr-windows
|
zip -r -9 $@ upkr-windows
|
||||||
|
|
||||||
upkr-linux/upkr:
|
upkr-linux/upkr:
|
||||||
cargo build --target x86_64-unknown-linux-musl --release
|
cargo build --target x86_64-unknown-linux-musl --release -F crossterm
|
||||||
mkdir -p upkr-linux
|
mkdir -p upkr-linux
|
||||||
cp ../target/x86_64-unknown-linux-musl/release/upkr upkr-linux/
|
cp ../target/x86_64-unknown-linux-musl/release/upkr upkr-linux/
|
||||||
|
|
||||||
upkr-windows/upkr.exe:
|
upkr-windows/upkr.exe:
|
||||||
cargo build --target x86_64-pc-windows-gnu --release
|
cargo build --target x86_64-pc-windows-gnu --release -F crossterm
|
||||||
mkdir -p upkr-windows
|
mkdir -p upkr-windows
|
||||||
cp ../target/x86_64-pc-windows-gnu/release/upkr.exe upkr-windows/
|
cp ../target/x86_64-pc-windows-gnu/release/upkr.exe upkr-windows/
|
||||||
|
|
||||||
|
|||||||
164
src/heatmap.rs
Normal file
164
src/heatmap.rs
Normal file
@@ -0,0 +1,164 @@
|
|||||||
|
pub struct Heatmap {
|
||||||
|
data: Vec<u8>,
|
||||||
|
cost: Vec<f32>,
|
||||||
|
literal_index: Vec<usize>,
|
||||||
|
}
|
||||||
|
|
||||||
|
impl Heatmap {
|
||||||
|
pub fn new() -> Heatmap {
|
||||||
|
Heatmap {
|
||||||
|
data: Vec::new(),
|
||||||
|
cost: Vec::new(),
|
||||||
|
literal_index: Vec::new(),
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn add_literal(&mut self, byte: u8, cost: f32) {
|
||||||
|
self.data.push(byte);
|
||||||
|
self.cost.push(cost);
|
||||||
|
self.literal_index.push(self.literal_index.len());
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn add_match(&mut self, offset: usize, length: usize, mut cost: f32) {
|
||||||
|
cost /= length as f32;
|
||||||
|
for _ in 0..length {
|
||||||
|
self.data.push(self.data[self.data.len() - offset]);
|
||||||
|
self.literal_index
|
||||||
|
.push(self.literal_index[self.literal_index.len() - offset]);
|
||||||
|
self.cost.push(cost);
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn finish(&mut self) {
|
||||||
|
let mut ref_count = vec![0usize; self.literal_index.len()];
|
||||||
|
for &index in &self.literal_index {
|
||||||
|
ref_count[index] += 1;
|
||||||
|
}
|
||||||
|
|
||||||
|
let mut shifted = vec![];
|
||||||
|
for (&index, &cost) in self.literal_index.iter().zip(self.cost.iter()) {
|
||||||
|
let delta = (self.cost[index] - cost) / ref_count[index] as f32;
|
||||||
|
shifted.push(delta);
|
||||||
|
shifted[index] -= delta;
|
||||||
|
}
|
||||||
|
|
||||||
|
for (cost, delta) in self.cost.iter_mut().zip(shifted.into_iter()) {
|
||||||
|
*cost += delta;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn reverse(&mut self) {
|
||||||
|
self.data.reverse();
|
||||||
|
self.cost.reverse();
|
||||||
|
self.literal_index.reverse();
|
||||||
|
for index in self.literal_index.iter_mut() {
|
||||||
|
*index = self.data.len() - *index;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn len(&self) -> usize {
|
||||||
|
self.cost.len()
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn is_literal(&self, index: usize) -> bool {
|
||||||
|
self.literal_index[index] == index
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn cost(&self, index: usize) -> f32 {
|
||||||
|
self.cost[index]
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn byte(&self, index: usize) -> u8 {
|
||||||
|
self.data[index]
|
||||||
|
}
|
||||||
|
|
||||||
|
#[cfg(feature = "crossterm")]
|
||||||
|
pub fn print_as_hex(&self) -> std::io::Result<()> {
|
||||||
|
use crossterm::{
|
||||||
|
style::{Attribute, Color, Print, SetAttribute, SetBackgroundColor},
|
||||||
|
QueueableCommand,
|
||||||
|
};
|
||||||
|
use std::io::{stdout, Write};
|
||||||
|
|
||||||
|
fn set_color(
|
||||||
|
mut out: impl QueueableCommand,
|
||||||
|
heatmap: &Heatmap,
|
||||||
|
index: usize,
|
||||||
|
num_colors: u16,
|
||||||
|
) -> std::io::Result<()> {
|
||||||
|
let cost = heatmap.cost(index);
|
||||||
|
if num_colors < 256 {
|
||||||
|
let colors = [
|
||||||
|
Color::Red,
|
||||||
|
Color::Yellow,
|
||||||
|
Color::Green,
|
||||||
|
Color::Cyan,
|
||||||
|
Color::Blue,
|
||||||
|
Color::DarkBlue,
|
||||||
|
Color::Black,
|
||||||
|
];
|
||||||
|
let color_index = (3. - cost.log2())
|
||||||
|
.round()
|
||||||
|
.max(0.)
|
||||||
|
.min((colors.len() - 1) as f32) as usize;
|
||||||
|
out.queue(SetBackgroundColor(colors[color_index]))?;
|
||||||
|
} else {
|
||||||
|
let colors = [
|
||||||
|
196, 166, 136, 106, 76, 46, 41, 36, 31, 26, 21, 20, 19, 18, 17, 16,
|
||||||
|
];
|
||||||
|
let color_index = ((3. - cost.log2()) * 2.5)
|
||||||
|
.round()
|
||||||
|
.max(0.)
|
||||||
|
.min((colors.len() - 1) as f32) as usize;
|
||||||
|
out.queue(SetBackgroundColor(Color::AnsiValue(colors[color_index])))?;
|
||||||
|
}
|
||||||
|
out.queue(SetAttribute(if heatmap.is_literal(index) {
|
||||||
|
Attribute::Underlined
|
||||||
|
} else {
|
||||||
|
Attribute::NoUnderline
|
||||||
|
}))?;
|
||||||
|
Ok(())
|
||||||
|
}
|
||||||
|
|
||||||
|
let num_colors = crossterm::style::available_color_count();
|
||||||
|
|
||||||
|
let term_width = crossterm::terminal::size()?.0.min(120) as usize;
|
||||||
|
let bytes_per_row = (term_width - 8) / 4;
|
||||||
|
|
||||||
|
for row_start in (0..self.data.len()).step_by(bytes_per_row) {
|
||||||
|
let row_range = row_start..self.data.len().min(row_start + bytes_per_row);
|
||||||
|
let mut stdout = stdout();
|
||||||
|
|
||||||
|
stdout.queue(Print(&format!("{:04x} ", row_start)))?;
|
||||||
|
|
||||||
|
for i in row_range.clone() {
|
||||||
|
set_color(&mut stdout, self, i, num_colors)?;
|
||||||
|
stdout.queue(Print(&format!("{:02x} ", self.data[i])))?;
|
||||||
|
}
|
||||||
|
|
||||||
|
let num_spaces = 1 + (bytes_per_row - (row_range.end - row_range.start)) * 3;
|
||||||
|
let gap: String = std::iter::repeat(' ').take(num_spaces).collect();
|
||||||
|
stdout
|
||||||
|
.queue(SetAttribute(Attribute::Reset))?
|
||||||
|
.queue(Print(&gap))?;
|
||||||
|
|
||||||
|
for i in row_range.clone() {
|
||||||
|
set_color(&mut stdout, self, i, num_colors)?;
|
||||||
|
let byte = self.data[i];
|
||||||
|
if byte >= 32 && byte < 127 {
|
||||||
|
stdout.queue(Print(format!("{}", byte as char)))?;
|
||||||
|
} else {
|
||||||
|
stdout.queue(Print("."))?;
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
stdout
|
||||||
|
.queue(SetAttribute(Attribute::Reset))?
|
||||||
|
.queue(Print("\n"))?;
|
||||||
|
|
||||||
|
stdout.flush()?;
|
||||||
|
}
|
||||||
|
|
||||||
|
Ok(())
|
||||||
|
}
|
||||||
|
}
|
||||||
66
src/lib.rs
66
src/lib.rs
@@ -1,31 +1,75 @@
|
|||||||
|
#![warn(missing_docs)]
|
||||||
|
|
||||||
|
//! Compression and decompression of the upkr format and variants.
|
||||||
|
//!
|
||||||
|
//! Upkr is a compression format initially designed for the MicroW8 fantasy console,
|
||||||
|
//! with design goals being a competitive compression ratio, reasonable fast
|
||||||
|
//! decompression, low memory overhead and very small decompression code
|
||||||
|
//! when handoptimized in assembler. (An optimized DOS execuable decompressor is <140 bytes.)
|
||||||
|
|
||||||
mod context_state;
|
mod context_state;
|
||||||
mod greedy_packer;
|
mod greedy_packer;
|
||||||
|
mod heatmap;
|
||||||
mod lz;
|
mod lz;
|
||||||
mod match_finder;
|
mod match_finder;
|
||||||
mod parsing_packer;
|
mod parsing_packer;
|
||||||
mod rans;
|
mod rans;
|
||||||
|
|
||||||
pub use lz::{calculate_margin, unpack, UnpackError};
|
pub use heatmap::Heatmap;
|
||||||
|
pub use lz::{calculate_margin, create_heatmap, unpack, UnpackError};
|
||||||
|
|
||||||
|
/// The type of a callback function to be given to the `pack` function.
|
||||||
|
///
|
||||||
|
/// It will be periodically called with the number of bytes of the input already processed.
|
||||||
pub type ProgressCallback<'a> = &'a mut dyn FnMut(usize);
|
pub type ProgressCallback<'a> = &'a mut dyn FnMut(usize);
|
||||||
|
|
||||||
|
/// A configuration of which compression format variation to use.
|
||||||
|
///
|
||||||
|
/// Use `Config::default()` for the standard upkr format.
|
||||||
|
///
|
||||||
|
/// Compression format variants exist to help with micro-optimizations in uncompression
|
||||||
|
/// code on specific platforms.
|
||||||
|
|
||||||
#[derive(Debug)]
|
#[derive(Debug)]
|
||||||
pub struct Config {
|
pub struct Config {
|
||||||
|
/// Shift in bits from a bitstream into the rANS state, rather than whole bytes.
|
||||||
|
/// This decreases the size of the rNAS state to 16 bits which is very useful on
|
||||||
|
/// 8 bit platforms.
|
||||||
pub use_bitstream: bool,
|
pub use_bitstream: bool,
|
||||||
|
/// The number of parity contexts (usually 1, 2 or 4). This can improve compression
|
||||||
|
/// on data that consists of regular groups of 2 or 4 bytes. One example is 32bit ARM
|
||||||
|
/// code, where each instruction is 4 bytes, so `parity_contexts = 4` improves compression
|
||||||
|
/// quite a bit. Defaults to `1`.
|
||||||
pub parity_contexts: usize,
|
pub parity_contexts: usize,
|
||||||
|
|
||||||
|
/// Invert the encoding of bits in the rANS coder. `bit = state_lo >= prob` instead of
|
||||||
|
/// `bit = state_lo < prob`.
|
||||||
pub invert_bit_encoding: bool,
|
pub invert_bit_encoding: bool,
|
||||||
|
/// The boolean value which encodes a match. Defaults to `true`.
|
||||||
pub is_match_bit: bool,
|
pub is_match_bit: bool,
|
||||||
|
/// The boolean value which encodes a new offset (rather than re-using the previous offset).
|
||||||
|
/// Defaults to `true`.
|
||||||
pub new_offset_bit: bool,
|
pub new_offset_bit: bool,
|
||||||
|
/// The boolean value which encodes that there are more bits comming for length/offset values.
|
||||||
|
/// Defaults to `true`.
|
||||||
pub continue_value_bit: bool,
|
pub continue_value_bit: bool,
|
||||||
|
|
||||||
|
/// Reverses the bits in the bitstream.
|
||||||
pub bitstream_is_big_endian: bool,
|
pub bitstream_is_big_endian: bool,
|
||||||
|
/// A slightly less accurate, but slightly simpler variation of the prob update in the
|
||||||
|
/// rANS coder, Used for the z80 uncompressor.
|
||||||
pub simplified_prob_update: bool,
|
pub simplified_prob_update: bool,
|
||||||
|
|
||||||
|
/// Disables support for re-using the last offset in the compression format.
|
||||||
|
/// This might save a few bytes when working with very small data.
|
||||||
pub no_repeated_offsets: bool,
|
pub no_repeated_offsets: bool,
|
||||||
|
/// Standard upkr encodes the EOF marker in the offset. This encodes it in the match length
|
||||||
|
/// instead.
|
||||||
pub eof_in_length: bool,
|
pub eof_in_length: bool,
|
||||||
|
|
||||||
|
/// The maximum match offset value to encode when compressing.
|
||||||
pub max_offset: usize,
|
pub max_offset: usize,
|
||||||
|
/// The maximum match length value to encode when compressing.
|
||||||
pub max_length: usize,
|
pub max_length: usize,
|
||||||
}
|
}
|
||||||
|
|
||||||
@@ -53,7 +97,7 @@ impl Default for Config {
|
|||||||
}
|
}
|
||||||
|
|
||||||
impl Config {
|
impl Config {
|
||||||
pub fn min_length(&self) -> usize {
|
fn min_length(&self) -> usize {
|
||||||
if self.eof_in_length {
|
if self.eof_in_length {
|
||||||
2
|
2
|
||||||
} else {
|
} else {
|
||||||
@@ -62,6 +106,21 @@ impl Config {
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/// Compresses the given data.
|
||||||
|
///
|
||||||
|
/// # Arguments
|
||||||
|
/// - `data`: The data to compress
|
||||||
|
/// - `level`: The compression level (0-9). Increasing the level by one roughly halves the
|
||||||
|
/// compression speed.
|
||||||
|
/// - `config`: The compression format variant to use.
|
||||||
|
/// - `progress_callback`: An optional callback which will periodically be called with
|
||||||
|
/// the number of bytes already processed.
|
||||||
|
///
|
||||||
|
/// # Example
|
||||||
|
/// ```rust
|
||||||
|
/// let compressed_data = upkr::pack(b"Hello, World! Yellow world!", 0, &upkr::Config::default(), None);
|
||||||
|
/// assert!(compressed_data.len() < 27);
|
||||||
|
/// ```
|
||||||
pub fn pack(
|
pub fn pack(
|
||||||
data: &[u8],
|
data: &[u8],
|
||||||
level: u8,
|
level: u8,
|
||||||
@@ -75,6 +134,9 @@ pub fn pack(
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/// Estimate the exact (fractional) size of upkr compressed data.
|
||||||
|
///
|
||||||
|
/// Note that this currently does NOT work for the bitstream variant.
|
||||||
pub fn compressed_size(mut data: &[u8]) -> f32 {
|
pub fn compressed_size(mut data: &[u8]) -> f32 {
|
||||||
let mut state = 0;
|
let mut state = 0;
|
||||||
while state < 4096 {
|
while state < 4096 {
|
||||||
|
|||||||
31
src/lz.rs
31
src/lz.rs
@@ -1,4 +1,5 @@
|
|||||||
use crate::context_state::ContextState;
|
use crate::context_state::ContextState;
|
||||||
|
use crate::heatmap::Heatmap;
|
||||||
use crate::rans::{EntropyCoder, RansDecoder};
|
use crate::rans::{EntropyCoder, RansDecoder};
|
||||||
use crate::Config;
|
use crate::Config;
|
||||||
use thiserror::Error;
|
use thiserror::Error;
|
||||||
@@ -153,21 +154,32 @@ pub fn unpack(
|
|||||||
max_size: usize,
|
max_size: usize,
|
||||||
) -> Result<Vec<u8>, UnpackError> {
|
) -> Result<Vec<u8>, UnpackError> {
|
||||||
let mut result = vec![];
|
let mut result = vec![];
|
||||||
let _ = unpack_internal(Some(&mut result), packed_data, config, max_size)?;
|
let _ = unpack_internal(Some(&mut result), None, packed_data, config, max_size)?;
|
||||||
Ok(result)
|
Ok(result)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn calculate_margin(packed_data: &[u8], config: &Config) -> Result<isize, UnpackError> {
|
pub fn calculate_margin(packed_data: &[u8], config: &Config) -> Result<isize, UnpackError> {
|
||||||
unpack_internal(None, packed_data, config, usize::MAX)
|
unpack_internal(None, None, packed_data, config, usize::MAX)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn unpack_internal(
|
pub fn create_heatmap(
|
||||||
|
packed_data: &[u8],
|
||||||
|
config: &Config,
|
||||||
|
max_size: usize,
|
||||||
|
) -> Result<Heatmap, UnpackError> {
|
||||||
|
let mut heatmap = Heatmap::new();
|
||||||
|
let _ = unpack_internal(None, Some(&mut heatmap), packed_data, config, max_size)?;
|
||||||
|
Ok(heatmap)
|
||||||
|
}
|
||||||
|
|
||||||
|
fn unpack_internal(
|
||||||
mut result: Option<&mut Vec<u8>>,
|
mut result: Option<&mut Vec<u8>>,
|
||||||
|
mut heatmap: Option<&mut Heatmap>,
|
||||||
packed_data: &[u8],
|
packed_data: &[u8],
|
||||||
config: &Config,
|
config: &Config,
|
||||||
max_size: usize,
|
max_size: usize,
|
||||||
) -> Result<isize, UnpackError> {
|
) -> Result<isize, UnpackError> {
|
||||||
let mut decoder = RansDecoder::new(packed_data, &config);
|
let mut decoder = RansDecoder::new(packed_data, &config)?;
|
||||||
let mut contexts = ContextState::new((1 + 255) * config.parity_contexts + 1 + 64 + 64, &config);
|
let mut contexts = ContextState::new((1 + 255) * config.parity_contexts + 1 + 64 + 64, &config);
|
||||||
let mut offset = usize::MAX;
|
let mut offset = usize::MAX;
|
||||||
let mut position = 0usize;
|
let mut position = 0usize;
|
||||||
@@ -198,6 +210,7 @@ pub fn unpack_internal(
|
|||||||
}
|
}
|
||||||
|
|
||||||
loop {
|
loop {
|
||||||
|
let prev_decoder = decoder.clone();
|
||||||
margin = margin.max(position as isize - decoder.pos() as isize);
|
margin = margin.max(position as isize - decoder.pos() as isize);
|
||||||
let literal_base = position % config.parity_contexts * 256;
|
let literal_base = position % config.parity_contexts * 256;
|
||||||
if decoder.decode_with_context(&mut contexts.context_mut(literal_base))?
|
if decoder.decode_with_context(&mut contexts.context_mut(literal_base))?
|
||||||
@@ -231,6 +244,9 @@ pub fn unpack_internal(
|
|||||||
if offset > position {
|
if offset > position {
|
||||||
return Err(UnpackError::OffsetOutOfRange { offset, position });
|
return Err(UnpackError::OffsetOutOfRange { offset, position });
|
||||||
}
|
}
|
||||||
|
if let Some(ref mut heatmap) = heatmap {
|
||||||
|
heatmap.add_match(offset, length, decoder.cost(&prev_decoder));
|
||||||
|
}
|
||||||
if let Some(ref mut result) = result {
|
if let Some(ref mut result) = result {
|
||||||
for _ in 0..length {
|
for _ in 0..length {
|
||||||
if result.len() < max_size {
|
if result.len() < max_size {
|
||||||
@@ -251,6 +267,9 @@ pub fn unpack_internal(
|
|||||||
context_index = (context_index << 1) | bit as usize;
|
context_index = (context_index << 1) | bit as usize;
|
||||||
byte |= (bit as u8) << i;
|
byte |= (bit as u8) << i;
|
||||||
}
|
}
|
||||||
|
if let Some(ref mut heatmap) = heatmap {
|
||||||
|
heatmap.add_literal(byte, decoder.cost(&prev_decoder));
|
||||||
|
}
|
||||||
if let Some(ref mut result) = result {
|
if let Some(ref mut result) = result {
|
||||||
if result.len() < max_size {
|
if result.len() < max_size {
|
||||||
result.push(byte);
|
result.push(byte);
|
||||||
@@ -261,6 +280,10 @@ pub fn unpack_internal(
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
if let Some(heatmap) = heatmap {
|
||||||
|
heatmap.finish();
|
||||||
|
}
|
||||||
|
|
||||||
if position > max_size {
|
if position > max_size {
|
||||||
return Err(UnpackError::OverSize {
|
return Err(UnpackError::OverSize {
|
||||||
size: position,
|
size: position,
|
||||||
|
|||||||
86
src/main.rs
86
src/main.rs
@@ -9,6 +9,9 @@ fn main() -> Result<()> {
|
|||||||
let mut reverse = false;
|
let mut reverse = false;
|
||||||
let mut unpack = false;
|
let mut unpack = false;
|
||||||
let mut calculate_margin = false;
|
let mut calculate_margin = false;
|
||||||
|
let mut create_heatmap = false;
|
||||||
|
#[allow(unused_mut)]
|
||||||
|
let mut do_hexdump = false;
|
||||||
let mut level = 2;
|
let mut level = 2;
|
||||||
let mut infile: Option<PathBuf> = None;
|
let mut infile: Option<PathBuf> = None;
|
||||||
let mut outfile: Option<PathBuf> = None;
|
let mut outfile: Option<PathBuf> = None;
|
||||||
@@ -58,6 +61,9 @@ fn main() -> Result<()> {
|
|||||||
|
|
||||||
Short('u') | Long("unpack") => unpack = true,
|
Short('u') | Long("unpack") => unpack = true,
|
||||||
Long("margin") => calculate_margin = true,
|
Long("margin") => calculate_margin = true,
|
||||||
|
Long("heatmap") => create_heatmap = true,
|
||||||
|
#[cfg(feature = "crossterm")]
|
||||||
|
Long("hexdump") => do_hexdump = true,
|
||||||
Short('l') | Long("level") => level = parser.value()?.parse()?,
|
Short('l') | Long("level") => level = parser.value()?.parse()?,
|
||||||
Short(n) if n.is_ascii_digit() => level = n as u8 - b'0',
|
Short(n) if n.is_ascii_digit() => level = n as u8 - b'0',
|
||||||
Short('h') | Long("help") => print_help(0),
|
Short('h') | Long("help") => print_help(0),
|
||||||
@@ -73,33 +79,46 @@ fn main() -> Result<()> {
|
|||||||
}
|
}
|
||||||
|
|
||||||
let infile = infile.unwrap_or_else(|| print_help(1));
|
let infile = infile.unwrap_or_else(|| print_help(1));
|
||||||
let outfile = outfile.unwrap_or_else(|| {
|
enum OutFileType {
|
||||||
let mut name = infile.clone();
|
Packed,
|
||||||
if unpack {
|
Unpacked,
|
||||||
if name.extension().filter(|&e| e == "upk").is_some() {
|
Heatmap,
|
||||||
name.set_extension("");
|
}
|
||||||
} else {
|
let outfile = |tpe: OutFileType| {
|
||||||
name.set_extension("bin");
|
outfile.clone().unwrap_or_else(|| {
|
||||||
|
let mut name = infile.clone();
|
||||||
|
match tpe {
|
||||||
|
OutFileType::Packed => {
|
||||||
|
let mut filename = name
|
||||||
|
.file_name()
|
||||||
|
.unwrap_or_else(|| OsStr::new(""))
|
||||||
|
.to_os_string();
|
||||||
|
filename.push(".upk");
|
||||||
|
name.set_file_name(filename);
|
||||||
|
}
|
||||||
|
OutFileType::Unpacked => {
|
||||||
|
if name.extension().filter(|&e| e == "upk").is_some() {
|
||||||
|
name.set_extension("");
|
||||||
|
} else {
|
||||||
|
name.set_extension("bin");
|
||||||
|
}
|
||||||
|
}
|
||||||
|
OutFileType::Heatmap => {
|
||||||
|
name.set_extension("heatmap");
|
||||||
|
}
|
||||||
}
|
}
|
||||||
} else {
|
name
|
||||||
let mut filename = name
|
})
|
||||||
.file_name()
|
};
|
||||||
.unwrap_or_else(|| OsStr::new(""))
|
|
||||||
.to_os_string();
|
|
||||||
filename.push(".upk");
|
|
||||||
name.set_file_name(filename);
|
|
||||||
}
|
|
||||||
name
|
|
||||||
});
|
|
||||||
|
|
||||||
if config.parity_contexts != 1 && config.parity_contexts != 2 && config.parity_contexts != 4 {
|
if config.parity_contexts != 1 && config.parity_contexts != 2 && config.parity_contexts != 4 {
|
||||||
eprintln!("--parity has to be 1, 2, or 4");
|
eprintln!("--parity has to be 1, 2, or 4");
|
||||||
process::exit(1);
|
process::exit(1);
|
||||||
}
|
}
|
||||||
|
|
||||||
if !unpack && !calculate_margin {
|
if !unpack && !calculate_margin && !create_heatmap {
|
||||||
let mut data = vec![];
|
let mut data = vec![];
|
||||||
File::open(infile)?.read_to_end(&mut data)?;
|
File::open(&infile)?.read_to_end(&mut data)?;
|
||||||
if reverse {
|
if reverse {
|
||||||
data.reverse();
|
data.reverse();
|
||||||
}
|
}
|
||||||
@@ -126,10 +145,10 @@ fn main() -> Result<()> {
|
|||||||
packed_data.len(),
|
packed_data.len(),
|
||||||
packed_data.len() as f32 * 100. / data.len() as f32
|
packed_data.len() as f32 * 100. / data.len() as f32
|
||||||
);
|
);
|
||||||
File::create(outfile)?.write_all(&packed_data)?;
|
File::create(outfile(OutFileType::Packed))?.write_all(&packed_data)?;
|
||||||
} else {
|
} else {
|
||||||
let mut data = vec![];
|
let mut data = vec![];
|
||||||
File::open(infile)?.read_to_end(&mut data)?;
|
File::open(&infile)?.read_to_end(&mut data)?;
|
||||||
if reverse {
|
if reverse {
|
||||||
data.reverse();
|
data.reverse();
|
||||||
}
|
}
|
||||||
@@ -138,7 +157,28 @@ fn main() -> Result<()> {
|
|||||||
if reverse {
|
if reverse {
|
||||||
unpacked_data.reverse();
|
unpacked_data.reverse();
|
||||||
}
|
}
|
||||||
File::create(outfile)?.write_all(&unpacked_data)?;
|
File::create(outfile(OutFileType::Unpacked))?.write_all(&unpacked_data)?;
|
||||||
|
}
|
||||||
|
if create_heatmap {
|
||||||
|
let mut heatmap = upkr::create_heatmap(&data, &config, max_unpacked_size)?;
|
||||||
|
if reverse {
|
||||||
|
heatmap.reverse();
|
||||||
|
}
|
||||||
|
match do_hexdump {
|
||||||
|
#[cfg(feature = "crossterm")]
|
||||||
|
true => heatmap.print_as_hex()?,
|
||||||
|
_ => {
|
||||||
|
let mut heatmap_bin = Vec::with_capacity(heatmap.len());
|
||||||
|
for i in 0..heatmap.len() {
|
||||||
|
let cost = (heatmap.cost(i).log2() * 8. + 64.)
|
||||||
|
.round()
|
||||||
|
.max(0.)
|
||||||
|
.min(127.) as u8;
|
||||||
|
heatmap_bin.push((cost << 1) | heatmap.is_literal(i) as u8);
|
||||||
|
}
|
||||||
|
File::create(outfile(OutFileType::Heatmap))?.write_all(&heatmap_bin)?;
|
||||||
|
}
|
||||||
|
}
|
||||||
}
|
}
|
||||||
if calculate_margin {
|
if calculate_margin {
|
||||||
println!("{}", upkr::calculate_margin(&data, &config)?);
|
println!("{}", upkr::calculate_margin(&data, &config)?);
|
||||||
@@ -152,11 +192,13 @@ fn print_help(exit_code: i32) -> ! {
|
|||||||
eprintln!("Usage:");
|
eprintln!("Usage:");
|
||||||
eprintln!(" upkr [-l level(0-9)] [config options] <infile> [<outfile>]");
|
eprintln!(" upkr [-l level(0-9)] [config options] <infile> [<outfile>]");
|
||||||
eprintln!(" upkr -u [config options] <infile> [<outfile>]");
|
eprintln!(" upkr -u [config options] <infile> [<outfile>]");
|
||||||
|
eprintln!(" upkr --heatmap [config options] <infile> [<outfile>]");
|
||||||
eprintln!(" upkr --margin [config options] <infile>");
|
eprintln!(" upkr --margin [config options] <infile>");
|
||||||
eprintln!();
|
eprintln!();
|
||||||
eprintln!(" -l, --level N compression level 0-9");
|
eprintln!(" -l, --level N compression level 0-9");
|
||||||
eprintln!(" -0, ..., -9 short form for setting compression level");
|
eprintln!(" -0, ..., -9 short form for setting compression level");
|
||||||
eprintln!(" -u, --unpack unpack infile");
|
eprintln!(" -u, --unpack unpack infile");
|
||||||
|
eprintln!(" --heatmap calculate heatmap from compressed file");
|
||||||
eprintln!(" --margin calculate margin for overlapped unpacking of a packed file");
|
eprintln!(" --margin calculate margin for overlapped unpacking of a packed file");
|
||||||
eprintln!();
|
eprintln!();
|
||||||
eprintln!("Version: {}", env!("CARGO_PKG_VERSION"));
|
eprintln!("Version: {}", env!("CARGO_PKG_VERSION"));
|
||||||
|
|||||||
24
src/rans.rs
24
src/rans.rs
@@ -148,6 +148,7 @@ impl EntropyCoder for CostCounter {
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
#[derive(Clone)]
|
||||||
pub struct RansDecoder<'a> {
|
pub struct RansDecoder<'a> {
|
||||||
data: &'a [u8],
|
data: &'a [u8],
|
||||||
pos: usize,
|
pos: usize,
|
||||||
@@ -166,8 +167,8 @@ const PROB_MASK: u32 = ONE_PROB - 1;
|
|||||||
pub struct UnexpectedEOF;
|
pub struct UnexpectedEOF;
|
||||||
|
|
||||||
impl<'a> RansDecoder<'a> {
|
impl<'a> RansDecoder<'a> {
|
||||||
pub fn new(data: &'a [u8], config: &Config) -> RansDecoder<'a> {
|
pub fn new(data: &'a [u8], config: &Config) -> Result<RansDecoder<'a>, UnexpectedEOF> {
|
||||||
RansDecoder {
|
let mut decoder = RansDecoder {
|
||||||
data,
|
data,
|
||||||
pos: 0,
|
pos: 0,
|
||||||
state: 0,
|
state: 0,
|
||||||
@@ -176,7 +177,9 @@ impl<'a> RansDecoder<'a> {
|
|||||||
bits_left: 0,
|
bits_left: 0,
|
||||||
invert_bit_encoding: config.invert_bit_encoding,
|
invert_bit_encoding: config.invert_bit_encoding,
|
||||||
bitstream_is_big_endian: config.bitstream_is_big_endian,
|
bitstream_is_big_endian: config.bitstream_is_big_endian,
|
||||||
}
|
};
|
||||||
|
decoder.refill()?;
|
||||||
|
Ok(decoder)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn pos(&self) -> usize {
|
pub fn pos(&self) -> usize {
|
||||||
@@ -189,8 +192,7 @@ impl<'a> RansDecoder<'a> {
|
|||||||
Ok(bit)
|
Ok(bit)
|
||||||
}
|
}
|
||||||
|
|
||||||
pub fn decode_bit(&mut self, prob: u16) -> Result<bool, UnexpectedEOF> {
|
fn refill(&mut self) -> Result<(), UnexpectedEOF> {
|
||||||
let prob = prob as u32;
|
|
||||||
if self.use_bitstream {
|
if self.use_bitstream {
|
||||||
while self.state < 32768 {
|
while self.state < 32768 {
|
||||||
if self.bits_left == 0 {
|
if self.bits_left == 0 {
|
||||||
@@ -219,6 +221,13 @@ impl<'a> RansDecoder<'a> {
|
|||||||
self.pos += 1;
|
self.pos += 1;
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
Ok(())
|
||||||
|
}
|
||||||
|
|
||||||
|
pub fn decode_bit(&mut self, prob: u16) -> Result<bool, UnexpectedEOF> {
|
||||||
|
self.refill()?;
|
||||||
|
|
||||||
|
let prob = prob as u32;
|
||||||
|
|
||||||
let bit = (self.state & PROB_MASK) < prob;
|
let bit = (self.state & PROB_MASK) < prob;
|
||||||
|
|
||||||
@@ -231,4 +240,9 @@ impl<'a> RansDecoder<'a> {
|
|||||||
|
|
||||||
Ok(bit ^ self.invert_bit_encoding)
|
Ok(bit ^ self.invert_bit_encoding)
|
||||||
}
|
}
|
||||||
|
|
||||||
|
pub fn cost(&self, prev: &RansDecoder) -> f32 {
|
||||||
|
f32::log2(prev.state as f32) - f32::log2(self.state as f32)
|
||||||
|
+ (self.pos - prev.pos) as f32 * 8.
|
||||||
|
}
|
||||||
}
|
}
|
||||||
|
|||||||
Reference in New Issue
Block a user