ludic/runtime/native/inflate.ludic
Orkuncakilkaya d3301684f1 Phase 8b: modern names for the raw-memory and OS/IO primitives
Groups B and C of the leftover-primitive cleanup — renames, not new machinery,
and deliberately NO unsafe_ prefix (a __-prefix is itself a C convention, and an
`unsafe` marker carries no signal in a fully-manual-memory language with no safe
subset to contrast against).

  memory:  mem_free -> free   mem_realloc -> resize   mem_set -> fill
           ptr_add -> offset
  process: os_argc -> arg_count   os_arg -> arg   os_exit -> exit
           os_system -> run   os_getenv -> getenv   read_byte -> read_char
  dead:    mem_copy, os_time, write_byte (0 uses) — deleted

Two reseeds: accept both old and new names in the intrinsic dispatch, then
migrate every call site and drop the old names. file_open/read/write/seek/tell/
close are left as-is — they're the domain-prefixed syscall layer wrapped by
read_file, not the argc/argv-style C-ness the audit targeted; a `File` type is a
separate, larger design if wanted.

test.sh's CLI smoke updated (os_exit -> exit); check-vocabulary's grammar marker
moved off the deleted names. Reseeded (22243 lines); C-free fixpoint holds;
goldens identical; 18/18; vocab + doc-fences clean.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-08-28 02:55:24 +03:00

276 lines
7.3 KiB
Text

# ============================================================================
# inflate.ludic — DEFLATE decompression (RFC 1951), written in Ludic.
#
# PNG stores its pixels zlib-compressed, so decoding one means implementing
# inflate. The C runtime linked zlib for this. We don't: zlib is not present by
# default on every target Ludic compiles for (Windows especially), and shipping
# a dependency to read a sprite is a poor trade when the algorithm is this
# small. So it lives here, in the language.
#
# The decoder is the canonical-Huffman formulation from Mark Adler's `puff`:
# a symbol table plus per-length counts, walked one bit at a time. Slower than
# a lookup-table decoder, and entirely fast enough to load sprites at startup.
# ============================================================================
# ---- bit reader (DEFLATE packs bits least-significant-first) ---------------
var z_src: ptr = null
var z_len: int = 0
var z_pos: int = 0
var z_bitbuf: int = 0
var z_bitcnt: int = 0
var z_err: int = 0
fn z_start(src: ptr, len: int) -> void {
z_src = src
z_len = len
z_pos = 0
z_bitbuf = 0
z_bitcnt = 0
z_err = 0
}
fn z_bits(need: int) -> int {
var val = z_bitbuf
while z_bitcnt < need {
if z_pos >= z_len {
z_err = 1
return 0
}
val = (val | (z_src[z_pos] << z_bitcnt))
z_pos = z_pos + 1
z_bitcnt = z_bitcnt + 8
}
z_bitbuf = (val >> need)
z_bitcnt = z_bitcnt - need
return (val & ((1 << need) - 1))
}
# ---- Huffman tables -------------------------------------------------------
# A table is a single buffer: 16 length-counts followed by the symbols in
# canonical order. One allocation, no structs.
fn z_table_new(nsym: int) -> ptr {
return words((16 + nsym))
}
# lengths[i] = code length of symbol i (0 = symbol unused)
fn z_table_build(table: words, lengths: words, n: int) -> void {
for i in 0 .. 16 {
table[i] = 0
}
for s in 0 .. n {
let l = lengths[s]
table[l] = table[l] + 1
}
table[0] = 0 # length 0 means "not present"
# offset of each length's first symbol
let offs: words = words(16)
offs[1] = 0
for l in 1 .. 15 {
offs[l + 1] = offs[l] + table[l]
}
for s in 0 .. n {
let l = lengths[s]
if l != 0 {
table[16 + offs[l]] = s
offs[l] = offs[l] + 1
}
}
free(offs)
}
fn z_decode(table: words) -> int {
var code = 0
var first = 0
var index = 0
for len in 1 .. 16 {
code = (code | z_bits(1))
let count = table[len]
if code - first < count {
return table[16 + index + (code - first)]
}
index = index + count
first = ((first + count) << 1)
code = (code << 1)
}
z_err = 1
return -1
}
# ---- length / distance code tables (RFC 1951 section 3.2.5) ---------------
fn z_len_base(sym: int) -> int {
if sym < 8 { return 3 + sym }
if sym == 28 { return 258 }
let extra = (sym - 4) / 4
let group = (1 << extra)
return 3 + ((group - 1) << 2) + 4 + (sym - 4 - extra * 4) * group
}
fn z_len_extra(sym: int) -> int {
if sym < 8 { return 0 }
if sym == 28 { return 0 }
return (sym - 4) / 4
}
fn z_dist_base(sym: int) -> int {
if sym < 4 { return 1 + sym }
let extra = (sym - 2) / 2
let group = (1 << extra)
return 1 + (group << 1) + (sym - 2 - extra * 2) * group
}
fn z_dist_extra(sym: int) -> int {
if sym < 4 { return 0 }
return (sym - 2) / 2
}
# ---- block decoders -------------------------------------------------------
# `out` is the destination window; returns the new write position, or -1.
fn z_stored(out: ptr, at: int, cap: int) -> int {
z_bitbuf = 0
z_bitcnt = 0 # stored blocks are byte-aligned
if z_pos + 4 > z_len { return -1 }
let n = z_src[z_pos] + (z_src[z_pos + 1] << 8)
z_pos = z_pos + 4 # LEN then its one's complement
var w = at
for i in 0 .. n {
if z_pos >= z_len { return -1 }
if w >= cap { return -1 }
out[w] = z_src[z_pos]
w = w + 1
z_pos = z_pos + 1
}
return w
}
fn z_codes(out: ptr, at: int, cap: int, lit: ptr, dist: ptr) -> int {
var w = at
var sym = z_decode(lit)
while sym != 256 {
if z_err != 0 { return -1 }
if sym < 0 { return -1 }
if sym < 256 {
if w >= cap { return -1 }
out[w] = sym
w = w + 1
}
if sym > 256 {
let s = sym - 257
if s >= 29 { return -1 }
let length = z_len_base(s) + z_bits(z_len_extra(s))
let d = z_decode(dist)
if d < 0 { return -1 }
let distance = z_dist_base(d) + z_bits(z_dist_extra(d))
if distance > w { return -1 }
for k in 0 .. length {
if w >= cap { return -1 }
out[w] = out[w - distance]
w = w + 1
}
}
sym = z_decode(lit)
}
return w
}
fn z_fixed_tables(lit: ptr, dist: ptr) -> void {
let lengths: words = words(288)
for i in 0 .. 144 { lengths[i] = 8 }
for i in 144 .. 256 { lengths[i] = 9 }
for i in 256 .. 280 { lengths[i] = 7 }
for i in 280 .. 288 { lengths[i] = 8 }
z_table_build(lit, lengths, 288)
for i in 0 .. 30 { lengths[i] = 5 }
z_table_build(dist, lengths, 30)
free(lengths)
}
fn z_dynamic_tables(lit: ptr, dist: ptr) -> int {
let nlen = z_bits(5) + 257
let ndist = z_bits(5) + 1
let ncode = z_bits(4) + 4
if nlen > 286 { return 0 }
if ndist > 30 { return 0 }
let lengths: words = words(320)
for i in 0 .. 19 { lengths[i] = 0 }
# the code-length alphabet is transmitted in this fixed permutation
# 16,17,18,0,8,7,9,6,10,5,11,4,12,3,13,2,14,1,15 — biased by '0' so it is one literal
let order = "@AB08796:5;4<3=2>1?"
for i in 0 .. ncode {
lengths[order[i] - 48] = z_bits(3)
}
let clen = z_table_new(19)
z_table_build(clen, lengths, 19)
var n = 0
while n < nlen + ndist {
let sym = z_decode(clen)
if sym < 0 { return 0 }
if sym < 16 {
lengths[n] = sym
n = n + 1
}
if sym >= 16 {
var prev = 0
var rep = 0
if sym == 16 {
if n == 0 { return 0 }
prev = lengths[n - 1]
rep = 3 + z_bits(2)
}
if sym == 17 { rep = 3 + z_bits(3) }
if sym == 18 { rep = 11 + z_bits(7) }
for k in 0 .. rep {
if n < 320 {
lengths[n] = prev
n = n + 1
}
}
}
}
z_table_build(lit, lengths, nlen)
# the distance lengths follow the literal ones in the same buffer
let dl: words = words(32)
for i in 0 .. ndist { dl[i] = lengths[nlen + i] }
z_table_build(dist, dl, ndist)
free(dl)
free(lengths)
free(clen)
return 1
}
# Inflate a raw DEFLATE stream. Returns bytes written, or -1.
fn z_inflate(src: ptr, len: int, out: ptr, cap: int) -> int {
z_start(src, len)
let lit = z_table_new(288)
let dist = z_table_new(30)
var w = 0
var final = 0
while final == 0 {
final = z_bits(1)
let btype = z_bits(2)
if z_err != 0 { return -1 }
if btype == 0 { w = z_stored(out, w, cap) }
if btype == 1 {
z_fixed_tables(lit, dist)
w = z_codes(out, w, cap, lit, dist)
}
if btype == 2 {
if z_dynamic_tables(lit, dist) == 0 { return -1 }
w = z_codes(out, w, cap, lit, dist)
}
if btype == 3 { return -1 }
if w < 0 { return -1 }
}
free(lit)
free(dist)
return w
}
# zlib wrapper (RFC 1950): two header bytes, then DEFLATE, then Adler-32.
fn z_uncompress(src: ptr, len: int, out: ptr, cap: int) -> int {
if len < 2 { return -1 }
let cmf = src[0]
if (cmf & 15) != 8 { return -1 }
return z_inflate(offset(src, 2), len - 2, out, cap)
}