Skip to content

Shrink huffman decode tables #78

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 17 commits into from
Jan 14, 2025
Merged

Shrink huffman decode tables #78

Show file tree
Hide file tree
Changes from 6 commits
Commits
Show all changes
17 commits
Select commit Hold shift + click to select a range
aeea83b
add additional guarantees and comments (no functional changes)
folkertdev Jan 14, 2025
7e88785
make `nextSym` a `u16`
folkertdev Jan 14, 2025
76daa30
simplify an if
folkertdev Jan 14, 2025
dfac834
simplify another if
folkertdev Jan 14, 2025
a11ed7b
factor out mtfa stuff
folkertdev Jan 14, 2025
77289cb
move variable declarations into the branches
folkertdev Jan 14, 2025
a0c61ae
use `slice::rotate_right`
folkertdev Jan 14, 2025
432aeaa
make constants unsigned
folkertdev Jan 14, 2025
d3bbdaf
simplify loop
folkertdev Jan 14, 2025
9e339be
simplify inner loop
folkertdev Jan 14, 2025
6d80f2c
clippy
folkertdev Jan 14, 2025
9c00c30
rename variables
folkertdev Jan 14, 2025
f6a4946
remove explicit types
folkertdev Jan 14, 2025
7ba6d1b
use `slice::copy_within`
folkertdev Jan 14, 2025
903c938
make `mtfbase` contain `u16`
folkertdev Jan 14, 2025
2a1cfad
clarify casts
folkertdev Jan 14, 2025
5d3cecf
use `usize::from` in the huffman table code
folkertdev Jan 14, 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
6 changes: 3 additions & 3 deletions libbz2-rs-sys/src/bzlib.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -557,7 +557,7 @@ pub(crate) struct DState {
pub len: [[u8; 258]; 6],
pub limit: [[i32; 258]; 6],
pub base: [[i32; 258]; 6],
pub perm: [[i32; 258]; 6],
pub perm: [[u16; 258]; 6],
pub minLens: [u8; 6],
pub save: SaveArea,
}
Expand All @@ -571,12 +571,12 @@ pub(crate) struct SaveArea {
pub alphaSize: u16,
pub EOB: u16,
pub groupNo: i32,
pub groupPos: u8,
pub nextSym: i32,
pub nblock: u32,
pub es: i32,
pub zvec: i32,
pub nextSym: u16,
pub nSelectors: u16,
pub groupPos: u8,
pub zn: u8,
pub nGroups: u8,
pub curr: u8,
Expand Down
167 changes: 86 additions & 81 deletions libbz2-rs-sys/src/decompress.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -785,7 +785,7 @@ pub(crate) fn decompress(
Some(&nextSym) => nextSym,
None => error!(BZ_DATA_ERROR),
};
if nextSym == BZ_RUNA as i32 || nextSym == BZ_RUNB as i32 {
if nextSym == BZ_RUNA || nextSym == BZ_RUNB {
current_block = Block46;
} else {
es += 1;
Expand Down Expand Up @@ -842,90 +842,30 @@ pub(crate) fn decompress(
_ => {}
}
if current_block == Block40 {
if nextSym == EOB as i32 {
if nextSym == EOB {
current_block = Block41;
} else if nextSym == BZ_RUNA || nextSym == BZ_RUNB {
es = -1;
logN = 0;
current_block = Block46;
} else if nblock >= 100000 * nblockMAX100k as u32 {
error!(BZ_DATA_ERROR);
} else {
if nextSym == 0 || nextSym == 1 {
es = -1;
logN = 0;
} else if nblock >= 100000 * nblockMAX100k as u32 {
error!(BZ_DATA_ERROR);
} else {
let mut ii_0: i32;
let mut jj_0: i32;
let mut kk_0: i32;
let mut pp: i32;
let mut lno: i32;
let off: i32;
let mut nn: u32;
nn = (nextSym - 1) as u32;
if nn < 16 {
pp = s.mtfbase[0_usize];
uc = s.mtfa[(pp as c_uint).wrapping_add(nn) as usize];
while nn > 3 {
let z: i32 = (pp as c_uint).wrapping_add(nn) as i32;
s.mtfa[z as usize] = s.mtfa[(z - 1) as usize];
s.mtfa[(z - 1) as usize] = s.mtfa[(z - 2) as usize];
s.mtfa[(z - 2) as usize] = s.mtfa[(z - 3) as usize];
s.mtfa[(z - 3) as usize] = s.mtfa[(z - 4) as usize];
nn = (nn).wrapping_sub(4);
}
while nn > 0 {
s.mtfa[(pp as c_uint).wrapping_add(nn) as usize] = s.mtfa
[(pp as c_uint).wrapping_add(nn).wrapping_sub(1) as usize];
nn = nn.wrapping_sub(1);
}
s.mtfa[pp as usize] = uc;
} else {
lno = nn.wrapping_div(16) as i32;
off = nn.wrapping_rem(16) as i32;
pp = s.mtfbase[lno as usize] + off;
uc = s.mtfa[pp as usize];
while pp > s.mtfbase[lno as usize] {
s.mtfa[pp as usize] = s.mtfa[(pp - 1) as usize];
pp -= 1;
}
s.mtfbase[lno as usize] += 1;
while lno > 0 {
s.mtfbase[lno as usize] -= 1;
s.mtfa[s.mtfbase[lno as usize] as usize] =
s.mtfa[(s.mtfbase[(lno - 1) as usize] + 16 - 1) as usize];
lno -= 1;
}
s.mtfbase[0_usize] -= 1;
s.mtfa[s.mtfbase[0_usize] as usize] = uc;
if s.mtfbase[0_usize] == 0 {
kk_0 = 4096 - 1;
ii_0 = 256 / 16 - 1;
while ii_0 >= 0 {
jj_0 = 16 - 1;
while jj_0 >= 0 {
s.mtfa[kk_0 as usize] =
s.mtfa[(s.mtfbase[ii_0 as usize] + jj_0) as usize];
kk_0 -= 1;
jj_0 -= 1;
}
s.mtfbase[ii_0 as usize] = kk_0 + 1;
ii_0 -= 1;
}
}
uc = initialize_mtfa(&mut s.mtfa, &mut s.mtfbase, nextSym);
s.unzftab[s.seqToUnseq[uc as usize] as usize] += 1;
match s.smallDecompress {
DecompressMode::Small => {
ll16[nblock as usize] = s.seqToUnseq[uc as usize] as u16
}
s.unzftab[s.seqToUnseq[uc as usize] as usize] += 1;
match s.smallDecompress {
DecompressMode::Small => {
ll16[nblock as usize] = s.seqToUnseq[uc as usize] as u16
}
DecompressMode::Fast => {
tt[nblock as usize] = s.seqToUnseq[uc as usize] as u32
}
DecompressMode::Fast => {
tt[nblock as usize] = s.seqToUnseq[uc as usize] as u32
}
nblock += 1;
update_group_pos!(s);
zn = gMinlen;
current_block = BZ_X_MTF_5;
continue;
}
current_block = Block46;
nblock += 1;
update_group_pos!(s);
zn = gMinlen;
current_block = BZ_X_MTF_5;
continue;
}
match current_block {
Block46 => {}
Expand Down Expand Up @@ -1116,7 +1056,7 @@ pub(crate) fn decompress(
if logN >= LOG_2MB {
error!(BZ_DATA_ERROR);
} else {
let mul = match nextSym as u16 {
let mul = match nextSym {
BZ_RUNA => 1,
BZ_RUNB => 2,
_ => 0,
Expand Down Expand Up @@ -1254,6 +1194,7 @@ pub(crate) fn decompress(

/*--- Create the Huffman decoding tables ---*/
for t in 0..nGroups as usize {
// NOTE: s.nInUse <= 256, alphaSize <= 258
let len = &s.len[t][..alphaSize as usize];

let mut minLen = 32u8;
Expand Down Expand Up @@ -1332,3 +1273,67 @@ pub(crate) fn decompress(

ret_val
}

fn initialize_mtfa(mtfa: &mut [u8; 4096], mtfbase: &mut [i32; 16], nextSym: u16) -> u8 {
let mut nn = (nextSym - 1) as u32;
Copy link
Collaborator

@bjorn3 bjorn3 Jan 14, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

u32::from(nextSym - 1)


if nn < MTFL_SIZE as u32 {
let pp = mtfbase[0_usize];
let uc = mtfa[(pp as c_uint).wrapping_add(nn) as usize];
while nn > 3 {
let z: i32 = (pp as c_uint).wrapping_add(nn) as i32;
mtfa[z as usize] = mtfa[(z - 1) as usize];
mtfa[(z - 1) as usize] = mtfa[(z - 2) as usize];
mtfa[(z - 2) as usize] = mtfa[(z - 3) as usize];
mtfa[(z - 3) as usize] = mtfa[(z - 4) as usize];
nn = (nn).wrapping_sub(4);
}
while nn > 0 {
mtfa[(pp as c_uint).wrapping_add(nn) as usize] =
mtfa[(pp as c_uint).wrapping_add(nn).wrapping_sub(1) as usize];
nn = nn.wrapping_sub(1);
}
mtfa[pp as usize] = uc;

uc
} else {
let mut lno = nn.wrapping_div(16);
let off = nn.wrapping_rem(16);
let mut pp = mtfbase[lno as usize] + off as i32;
let uc = mtfa[pp as usize];
while pp > mtfbase[lno as usize] {
mtfa[pp as usize] = mtfa[(pp - 1) as usize];
pp -= 1;
}
mtfbase[lno as usize] += 1;
while lno > 0 {
mtfbase[lno as usize] -= 1;
mtfa[mtfbase[lno as usize] as usize] =
mtfa[(mtfbase[(lno - 1) as usize] + 16 - 1) as usize];
lno -= 1;
}
mtfbase[0_usize] -= 1;
mtfa[mtfbase[0_usize] as usize] = uc;

let mut ii_0: i32;
let mut jj_0: i32;
let mut kk_0: i32;

if mtfbase[0_usize] == 0 {
kk_0 = 4096 - 1;
ii_0 = 256 / 16 - 1;
while ii_0 >= 0 {
jj_0 = 16 - 1;
while jj_0 >= 0 {
mtfa[kk_0 as usize] = mtfa[(mtfbase[ii_0 as usize] + jj_0) as usize];
kk_0 -= 1;
jj_0 -= 1;
}
mtfbase[ii_0 as usize] = kk_0 + 1;
ii_0 -= 1;
}
}

uc
}
}
10 changes: 6 additions & 4 deletions libbz2-rs-sys/src/huffman.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -176,18 +176,20 @@ pub(crate) fn assign_codes(

#[inline(always)]
pub(crate) fn create_decode_tables(
limit: &mut [i32],
base: &mut [i32],
perm: &mut [i32],
limit: &mut [i32; 258],
base: &mut [i32; 258],
perm: &mut [u16; 258],
length: &[u8],
minLen: u8,
maxLen: u8,
) {
assert!(length.len() <= 258);

let mut pp: i32 = 0;
for i in minLen..=maxLen {
for (j, e) in length.iter().enumerate() {
if *e == i {
perm[pp as usize] = j as i32;
perm[pp as usize] = j as u16;
pp += 1;
}
}
Expand Down