Skip to content

decompress cleanups #84

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
folkertdev merged 9 commits into from
Jan 21, 2025
Merged

decompress cleanups #84

Show file tree
Hide file tree
Changes from 8 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
3a5a46a
add 'release with debug info' profile
folkertdev Jan 21, 2025
f74323e
refactor a check
folkertdev Jan 21, 2025
4a2591f
centralize logic in a macro
folkertdev Jan 21, 2025
6d749e7
clean up casting
folkertdev Jan 21, 2025
cbbd6be
inline `make_maps_d` (no need to reborrow any more)
folkertdev Jan 21, 2025
4ab2d44
cleanup cast
folkertdev Jan 21, 2025
a1cc3e7
some more casts
folkertdev Jan 21, 2025
942cdea
make `cftab` and friends store `u32`
folkertdev Jan 21, 2025
a79595c
changes after code review
folkertdev Jan 21, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions Cargo.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,10 @@ rust-version = "1.82" # MSRV
libc = "0.2"
libbz2-rs-sys = { path = "libbz2-rs-sys/", default-features = false }

[profile.relwithdebinfo]
inherits = "release"
debug = true

[package]
name = "libbzip2-rs"
readme.workspace = true
Expand Down
10 changes: 5 additions & 5 deletions libbz2-rs-sys/src/bzlib.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -541,9 +541,9 @@ pub(crate) struct DState {
pub origPtr: i32,
pub tPos: u32,
pub nblock_used: i32,
pub unzftab: [i32; 256],
pub cftab: [i32; 257],
pub cftabCopy: [i32; 257],
pub unzftab: [u32; 256],
pub cftab: [u32; 257],
pub cftabCopy: [u32; 257],
pub tt: DSlice<u32>,
pub ll16: DSlice<u16>,
pub ll4: DSlice<u8>,
Expand Down Expand Up @@ -1484,7 +1484,7 @@ fn un_rle_obuf_to_output_fast(strm: &mut BzStream<DState>, s: &mut DState) -> bo
}

#[inline]
pub(crate) fn index_into_f(index: i32, cftab: &[i32; 257]) -> u8 {
pub(crate) fn index_into_f(index: u32, cftab: &[u32; 257]) -> u8 {
let mut nb = 0u16;
let mut na = 256;
loop {
Expand Down Expand Up @@ -1516,7 +1516,7 @@ macro_rules! BZ_GET_SMALL {
None => return true,
Some(&low_bits) => {
let high_bits = GET_LL4!($s, $s.tPos);
let tmp = index_into_f($s.tPos as i32, &$s.cftab);
let tmp = index_into_f($s.tPos, &$s.cftab);
$s.tPos = u32::from(low_bits) | high_bits << 16;
tmp
}
Expand Down
141 changes: 59 additions & 82 deletions libbz2-rs-sys/src/decompress.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -126,16 +126,6 @@
}
use Block::*;

fn make_maps_d(s: &mut DState) {
s.nInUse = 0;
for (i, in_use) in s.inUse.iter().enumerate() {
if *in_use {
s.seqToUnseq[s.nInUse as usize] = i as u8;
s.nInUse += 1;
}
}
}

pub(crate) fn decompress(
strm: &mut BzStream<DState>,
s: &mut DState,
Expand Down Expand Up @@ -600,9 +590,9 @@
}

// mutable because they need to be reborrowed
let mut tt = s.tt.as_mut_slice();
let mut ll16 = s.ll16.as_mut_slice();
let mut ll4 = s.ll4.as_mut_slice();
let tt = s.tt.as_mut_slice();
let ll16 = s.ll16.as_mut_slice();
let ll4 = s.ll4.as_mut_slice();

'c_10064: loop {
match current_block {
Expand Down Expand Up @@ -752,75 +742,59 @@
current_block = Block24;
}
}
match current_block {
Block24 => {
if zn > 20 {

macro_rules! get_next_sym {
($next_block:ident) => {
if (zn > 20/* the longest code */) {
error!(BZ_DATA_ERROR);
} else if zvec <= s.limit[usize::from(gSel)][zn as usize] {
let index = zvec - s.base[usize::from(gSel)][zn as usize];
nextSym = match s.perm[usize::from(gSel)].get(index as usize) {
match s.perm[usize::from(gSel)].get(index as usize) {
Some(&nextSym) => nextSym,
None => error!(BZ_DATA_ERROR),
};
}
} else {
zn += 1;
current_block = BZ_X_MTF_6;
continue;
current_block = $next_block;
continue 'c_10064;
}
};
}

match current_block {
Block24 => {
nextSym = get_next_sym!(BZ_X_MTF_6);
current_block = Block40;
}
Block52 => {
if zn > 20 {
error!(BZ_DATA_ERROR);
} else if zvec <= s.limit[usize::from(gSel)][zn as usize] {
let index = zvec - s.base[usize::from(gSel)][zn as usize];
nextSym = match s.perm[usize::from(gSel)].get(index as usize) {
Some(&nextSym) => nextSym,
None => error!(BZ_DATA_ERROR),
};
if nextSym == BZ_RUNA || nextSym == BZ_RUNB {
current_block = Block46;
} else {
let uc = s.seqToUnseq[s.mtfa[s.mtfbase[0_usize] as usize] as usize];
s.unzftab[usize::from(uc)] += es as i32;
match s.smallDecompress {
DecompressMode::Small => {
match ll16.get_mut(nblock as usize..(nblock + es) as usize) {
Some(slice) => slice.fill(u16::from(uc)),
None => error!(BZ_DATA_ERROR),
};
nblock += es;
}
DecompressMode::Fast => {
match tt.get_mut(nblock as usize..(nblock + es) as usize) {
Some(slice) => slice.fill(u32::from(uc)),
None => error!(BZ_DATA_ERROR),
};
nblock += es;
}
nextSym = get_next_sym!(BZ_X_MTF_4);

if nextSym == BZ_RUNA || nextSym == BZ_RUNB {
current_block = Block46;
} else {
let uc = s.seqToUnseq[usize::from(s.mtfa[usize::from(s.mtfbase[0])])];
s.unzftab[usize::from(uc)] += es;
match s.smallDecompress {
DecompressMode::Small => {
match ll16.get_mut(nblock as usize..(nblock + es) as usize) {
Some(slice) => slice.fill(u16::from(uc)),
None => error!(BZ_DATA_ERROR),

Check warning on line 781 in libbz2-rs-sys/src/decompress.rs

View check run for this annotation

Codecov / codecov/patch

libbz2-rs-sys/src/decompress.rs#L781 

Added line #L781 was not covered by tests
};
nblock += es;
}
DecompressMode::Fast => {
match tt.get_mut(nblock as usize..(nblock + es) as usize) {
Some(slice) => slice.fill(u32::from(uc)),
None => error!(BZ_DATA_ERROR),

Check warning on line 788 in libbz2-rs-sys/src/decompress.rs

View check run for this annotation

Codecov / codecov/patch

libbz2-rs-sys/src/decompress.rs#L788 

Added line #L788 was not covered by tests
};
nblock += es;
}
current_block = Block40;
}
} else {
zn += 1;
current_block = BZ_X_MTF_4;
continue;
current_block = Block40;
}
}
Block56 => {
if zn > 20 {
error!(BZ_DATA_ERROR);
} else if zvec <= s.limit[usize::from(gSel)][zn as usize] {
let index = zvec - s.base[usize::from(gSel)][zn as usize];
nextSym = match s.perm[usize::from(gSel)].get(index as usize) {
Some(&nextSym) => nextSym,
None => error!(BZ_DATA_ERROR),
};
} else {
zn += 1;
current_block = BZ_X_MTF_2;
continue;
}
nextSym = get_next_sym!(BZ_X_MTF_2);
current_block = Block40;
}
_ => {}
Expand All @@ -836,10 +810,11 @@
error!(BZ_DATA_ERROR);
} else {
let uc = usize::from(initialize_mtfa(&mut s.mtfa, &mut s.mtfbase, nextSym));
s.unzftab[usize::from(s.seqToUnseq[uc])] += 1;
let index = s.seqToUnseq[uc];
s.unzftab[usize::from(index)] += 1;
match s.smallDecompress {
DecompressMode::Small => ll16[nblock as usize] = s.seqToUnseq[uc] as u16,
DecompressMode::Fast => tt[nblock as usize] = s.seqToUnseq[uc] as u32,
DecompressMode::Small => ll16[nblock as usize] = u16::from(index),
DecompressMode::Fast => tt[nblock as usize] = u32::from(index),
}
nblock += 1;
update_group_pos!(s);
Expand All @@ -853,15 +828,15 @@
if s.origPtr < 0 || s.origPtr >= nblock as i32 {
error!(BZ_DATA_ERROR);
} else {
if s.unzftab.iter().any(|e| !(0..=nblock as i32).contains(e)) {
if s.unzftab.iter().any(|e| !(0..=nblock).contains(e)) {
error!(BZ_DATA_ERROR);
}
s.cftab[0] = 0;
s.cftab[1..].copy_from_slice(&s.unzftab);
for i in 1..s.cftab.len() {
s.cftab[i] += s.cftab[i - 1];
}
if s.cftab.iter().any(|e| !(0..=nblock as i32).contains(e)) {
if s.cftab.iter().any(|e| !(0..=nblock).contains(e)) {
error!(BZ_DATA_ERROR);
}
// FIXME: use https://doc.rust-lang.org/std/primitive.slice.html#method.is_sorted
Expand All @@ -870,8 +845,8 @@
error!(BZ_DATA_ERROR);
}
s.state_out_len = 0;
s.state_out_ch = 0_u8;
s.calculatedBlockCRC = 0xffffffffu32;
s.state_out_ch = 0;
s.calculatedBlockCRC = u32::MAX;
s.state = State::BZ_X_OUTPUT;
if s.verbosity >= 2 {
debug_log!("rt+rld");
Expand Down Expand Up @@ -929,7 +904,7 @@
s.tPos = s.origPtr as u32;
s.nblock_used = 0;

s.k0 = index_into_f(s.tPos as i32, &s.cftab);
s.k0 = index_into_f(s.tPos, &s.cftab);
s.tPos = match ll16.get(s.tPos as usize) {
None => error!(BZ_DATA_ERROR),
Some(&low_bits) => {
Expand Down Expand Up @@ -1033,19 +1008,21 @@
}
}
Block18 => {
if i < 16 {
if s.inUse16[i as usize] {
if let Some(&in_use) = s.inUse16.get(i as usize) {
if in_use {
j = 0;
current_block = Block28;
continue;
}
} else {
make_maps_d(s);

// reborrow
tt = s.tt.as_mut_slice();
ll16 = s.ll16.as_mut_slice();
ll4 = s.ll4.as_mut_slice();
// inlined `make_maps_d`
s.nInUse = 0;
for (i, in_use) in s.inUse.iter().enumerate() {
if *in_use {
s.seqToUnseq[usize::from(s.nInUse)] = i as u8;
s.nInUse += 1;
}
}

if s.nInUse == 0 {
current_block = Block11;
Expand Down
Loading