Fix #498: EPS images with several data sources read them in turn

An image can take one data source per colour component
(MultipleDataSources, or `colorimage` with multi true). The PostScript
Language Reference (3rd ed., §4.10 "Images" and `image`/`colorimage` in
chapter 8) has the operator read the sources in turn, the same amount
from each. The reader drained each source on its own instead: when the
sources are procedures reading one file a row at a time, the first plane
took the rows of every component and the image came out in sheared
horizontal bands, as in #498.

- Data sources are read in turn by one reader, `read_sources`: a procedure
  once a round, a file or string a row a round, until each plane is full or
  its source ends. Single sources, procedure filters and masked images use
  it too; a string source resumes where it left off.
- Tests (`tests_images.rs`, programs written from the reference):
  procedures reading one file in the dictionary and `colorimage` operand
  forms (fails without the fix), separate string sources, rows padded to a
  byte at 1, 2, 4, 12 and 16 bits with an odd width (one and several
  sources), Flate PNG predictors with every row filter at an odd width.
  The import fuzz vocabulary has several sources reading one file and
  sources of uneven lengths.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
bflatastic
2026-10-08 23:43:49 +07:00
co-authored by Claude Opus 5.5
parent eb255ef379
commit 1d265951df
6 changed files with 182 additions and 32 deletions
+2 -2
View File
File diff suppressed because one or more lines are too long
+4
View File
@@ -953,6 +953,10 @@ fn arb_ps_token() -> impl Strategy<Value = String> {
"/T3 << /FontType 3 /FontMatrix [0.1 0 0 0.1 0 0] /FontBBox [0 0 9 9] /Encoding [/a /b] /BuildChar { pop pop 9 0 setcharwidth 0 0 5 5 rectfill } >> definefont setfont",
"/T4 << /FontType 3 /FontMatrix [1 0 0 1 0 0] /BuildGlyph { pop pop (x) show } /Encoding [/a] >> definefont setfont",
"glyphshow", "xshow", "xyshow", "awidthshow", "kshow", "/a", "[1 2 3]",
// Several data sources reading one file in turn, and sources of uneven lengths.
"/f currentfile /ASCIIHexDecode filter def", "{f 3 string readstring pop}", "{f 0 string readstring pop}", "true 3 colorimage",
"3 2 8 [3 0 0 2 0 0] {f 5 string readstring pop} {(ab)} {f 1 string readstring pop} true 3 colorimage",
"<< /ImageType 1 /Width 3 /Height 2 /BitsPerComponent 4 /Decode [0 1 0 1 0 1 0 1] /ImageMatrix [3 0 0 2 0 0] /MultipleDataSources true /DataSource [(a) (bc) {f 2 string readstring pop} (d)] >>",
// Executable strings and keys other than names (PLRM 3rd ed., `cvx`, `load`).
"(>>) cvx", "(1 2 add) cvx exec", "(x) cvx dup exec", "0 load", "1 { } def", "(mark) cvx cvlit"
])
+43 -29
View File
@@ -635,33 +635,51 @@ impl Interp<'_> {
/// The bytes a data procedure gives, called until it gives an empty string or `n` bytes.
fn gather(&mut self, p: &Obj, n: usize) -> Res<Vec<u8>> {
let mut out = vec![];
while out.len() < n {
self.call(p.clone())?;
let s = self.pop_str()?;
let s = s.borrow();
if s.is_empty() {
break;
}
self.alloc(s.len())?;
out.extend_from_slice(&s);
}
out.truncate(n);
Ok(out)
Ok(self.read_sources(&[Source::Proc(p.clone())], n, n)?.pop().unwrap_or_default())
}
/// `n` bytes of image data from `src` (zeros past its end).
fn source_bytes(&mut self, src: &Source, n: usize) -> Res<Vec<u8>> {
let mut v = match src {
Source::File(f) => self.read(f, n)?,
Source::Str(s) if !s.is_empty() => s.iter().copied().cycle().take(n).collect(),
Source::Str(_) => vec![],
Source::Proc(p) => self.gather(p, n)?,
};
let mut v = self.read_sources(std::slice::from_ref(src), n, n)?.pop().unwrap_or_default();
v.resize(n, 0);
Ok(v)
}
/// Up to `n` bytes from each of `sources`, read in turn as an image reads several data sources
/// (PLRM 3rd ed., §4.10 "Images" and `image`/`colorimage` in chapter 8: the sources are read
/// in turn, the same amount from each): each procedure called once a round (its string,
/// whatever its length), each file and string giving `step` bytes a round, until each has
/// given `n` bytes or ended. Procedures that read one file get their own rows that way.
fn read_sources(&mut self, sources: &[Source], n: usize, step: usize) -> Res<Vec<Vec<u8>>> {
let mut out: Vec<Vec<u8>> = vec![vec![]; sources.len()];
let mut ended = vec![false; sources.len()];
while out.iter().zip(&ended).any(|(o, e)| o.len() < n && !e) {
for ((src, o), e) in sources.iter().zip(out.iter_mut()).zip(ended.iter_mut()) {
if *e || o.len() >= n {
continue;
}
let want = step.min(n - o.len());
let chunk = match src {
Source::File(f) => self.read(f, want)?,
// A string gives its bytes over and over.
Source::Str(s) => {
let at = o.len() % s.len().max(1);
s.get(at..).unwrap_or_default().iter().chain(s.iter().cycle()).take(want).copied().collect()
}
Source::Proc(p) => {
self.call(p.clone())?;
let s = self.pop_str()?.to_vec();
self.alloc(s.len())?;
s
}
};
*e = chunk.is_empty();
o.extend_from_slice(chunk.get(..chunk.len().min(n - o.len())).unwrap_or_default());
}
}
Ok(out)
}
/// `eexec`: an encrypted font program follows. It is skipped (type is drawn in the app's
/// fonts, by name): the font dictionary it would complete is defined as it is.
fn eexec(&mut self) -> Res {
@@ -849,16 +867,12 @@ impl Interp<'_> {
let multi = s.sources.len() > 1;
let row_bits = if multi { s.w * bpc } else { s.w * n * bpc };
let row = row_bits.div_ceil(8);
let planes: Vec<Vec<u8>> = if multi {
let mut v = vec![];
for src in &s.sources {
v.push(self.source_bytes(src, row * s.h)?);
}
v
} else {
let src = s.sources.first().ok_or(PsError::Ps("undefined", "DataSource".into()))?;
vec![self.source_bytes(src, row * s.h)?]
};
// Several sources are read in turn, a row at a time each (zeros past their ends).
let sources = if multi { &s.sources[..] } else { s.sources.get(..1).ok_or(PsError::Ps("undefined", "DataSource".into()))? };
let mut planes = self.read_sources(sources, row * s.h, if multi { row } else { row * s.h })?;
for p in &mut planes {
p.resize(row * s.h, 0);
}
let max = ((1u32 << bpc.min(16)) - 1) as f64;
let sample = |plane: &[u8], y: usize, i: usize| -> u32 {
let bit = y * row * 8 + i * bpc;
+2
View File
@@ -41,6 +41,8 @@ mod tests_fontnames;
mod tests_generators;
#[cfg(test)]
mod tests_illustrator;
#[cfg(test)]
mod tests_images;
use std::sync::Arc;
+1 -1
View File
@@ -42,7 +42,7 @@ fn a85_flate(data: &[u8]) -> String {
}
/// Pixel `(x, y)` (straight RGBA) of the first image in `d`.
fn pixel(d: &Document, x: u32, y: u32) -> [u8; 4] {
pub(super) fn pixel(d: &Document, x: u32, y: u32) -> [u8; 4] {
let im = all(d)
.into_iter()
.find_map(|n| match &n.kind {
+130
View File
@@ -0,0 +1,130 @@
//! Image data read as the PostScript Language Reference (3rd ed.) defines it: several data sources
//! read in turn (§4.10 "Images", `image` and `colorimage` in chapter 8), rows padded to a whole
//! byte at every bit depth and odd widths (§4.10), and Flate predictors undone row by row (§3.13,
//! `FlateDecode` parameters) (#498).
use super::tests::read;
use super::tests_generators::pixel;
/// Each pixel of the first image, row by row.
fn pixels(r: &crate::Imported, w: u32, h: u32) -> Vec<[u8; 4]> {
assert!(!r.preview && r.warnings.is_empty(), "{:?}", r.warnings);
(0..h).flat_map(|y| (0..w).map(move |x| (x, y))).map(|(x, y)| pixel(&r.document, x, y)).collect()
}
/// A CMYK image whose four data sources are procedures reading one file, a row of one ink each:
/// the image operator calls them in turn, so each gets its own rows. Read one after the other,
/// the cyan source took the first four rows of all four inks and the image came out in
/// horizontal bands (#498).
#[test]
fn procedures_reading_one_file_are_read_in_turn() {
// 3 × 2 pixels; rows: C M Y K of row 0, then of row 1.
let rows = ["ff0000", "00ff00", "000000", "000000", "000000", "000000", "ffff00", "0000ff"];
let r = read(&format!(
"/DeviceCMYK setcolorspace 30 20 scale /f currentfile /ASCIIHexDecode filter def \
<< /ImageType 1 /Width 3 /Height 2 /BitsPerComponent 8 /Decode [0 1 0 1 0 1 0 1] /ImageMatrix [3 0 0 -2 0 2] \
/MultipleDataSources true /DataSource [ {{f 3 string readstring pop}} {{f 3 string readstring pop}} \
{{f 3 string readstring pop}} {{f 3 string readstring pop}} ] >> image\n{}>",
rows.concat()
));
let cyan = [0, 174, 239, 255];
let magenta = [236, 0, 140, 255];
let yellow = [255, 242, 0, 255];
let black = [35, 31, 32, 255];
let white = [255, 255, 255, 255];
let px = pixels(&r, 3, 2);
let near = |a: [u8; 4], b: [u8; 4]| a.iter().zip(b).all(|(x, y)| x.abs_diff(y) <= 40);
let want = [cyan, magenta, white, yellow, yellow, black];
assert!(px.iter().zip(want).all(|(a, b)| near(*a, b)), "{px:?}");
// The operand form: `{r} {g} {b} true 3 colorimage`, procedures reading one file.
let r = read(
"30 20 scale /f currentfile /ASCIIHexDecode filter def \
3 2 8 [3 0 0 -2 0 2] {f 3 string readstring pop} {f 3 string readstring pop} {f 3 string readstring pop} true 3 colorimage\n\
ff0000 00ff00 0000ff 102030 405060 708090>",
);
assert_eq!(pixels(&r, 3, 2), [[255, 0, 0, 255], [0, 255, 0, 255], [0, 0, 255, 255], [16, 64, 112, 255], [32, 80, 128, 255], [48, 96, 144, 255]]);
}
/// Separate strings and files as the sources: each is read for its own plane, whatever the row.
#[test]
fn separate_sources_keep_their_planes() {
let r = read("30 20 scale 3 2 8 [3 0 0 -2 0 2] <ff0000102030> <00ff00405060> <0000ff708090> true 3 colorimage");
assert_eq!(pixels(&r, 3, 2), [[255, 0, 0, 255], [0, 255, 0, 255], [0, 0, 255, 255], [16, 64, 112, 255], [32, 80, 128, 255], [48, 96, 144, 255]]);
}
/// Rows of an odd width at 1, 2, 4, 12 and 16 bits a sample start at a byte (their last byte
/// padded), in one source and in several.
#[test]
fn rows_of_every_depth_start_at_a_byte() {
// 3 pixels a row, two rows: 0, max, mid / max, 0, 0 (grey), each row padded.
let cases = [
(1, "a0 80"), // 101(00000) 100(00000)
(2, "cc c0"), // 11 00 11 (00), 11 00 00 (00)
(4, "f0 80 f0 00"), // f 0 8 (0), f 0 0 (0)
(12, "fff0008000 fff0000000"), // fff 000 800 (0), fff 000 000 (0)
(16, "ffff00008000 ffff00000000"),
];
for (bpc, data) in cases {
let data = data.replace(' ', "");
let r = read(&format!("30 20 scale 3 2 {bpc} [3 0 0 -2 0 2] <{data}> image"));
let px: Vec<u8> = pixels(&r, 3, 2).iter().map(|p| p[0]).collect();
let want: &[u8] = match bpc {
1 => &[255, 0, 255, 255, 0, 0],
2 => &[255, 0, 255, 255, 0, 0],
4 => &[255, 0, 136, 255, 0, 0],
12 => &[255, 0, 128, 255, 0, 0],
_ => &[255, 0, 128, 255, 0, 0],
};
assert_eq!(px, want, "{bpc} bits");
}
// Several 4-bit sources, odd width: each plane's rows padded on their own.
let r = read("30 20 scale 3 2 4 [3 0 0 -2 0 2] <f0f000f0> <0f00f000> <00f00ff0> true 3 colorimage");
let (on, off) = (255, 0);
let want = [[on, off, off], [off, on, off], [on, off, on], [off, on, off], [off, off, on], [on, off, on]];
assert_eq!(pixels(&r, 3, 2), want.map(|[r, g, b]| [r, g, b, 255]));
}
/// A PNG-predicted Flate image of an odd width: every row filter (None, Sub, Up, Average, Paeth)
/// undone on its own row.
#[test]
fn predicted_rows_of_an_odd_width() {
// 3 RGB pixels a row, five rows; each row predicted with its own filter.
let raw: Vec<[u8; 9]> = (0..5u8).map(|y| std::array::from_fn(|i| (i as u8).wrapping_mul(29).wrapping_add(y.wrapping_mul(53)))).collect();
let mut encoded = vec![];
let mut prev = [0u8; 9];
for (kind, row) in raw.iter().enumerate() {
encoded.push(kind as u8);
for i in 0..9 {
let a = if i >= 3 { row[i - 3] } else { 0 };
let b = prev[i];
let c = if i >= 3 { prev[i - 3] } else { 0 };
let pred = match kind {
1 => a,
2 => b,
3 => ((u16::from(a) + u16::from(b)) / 2) as u8,
4 => {
let p = i16::from(a) + i16::from(b) - i16::from(c);
let (pa, pb, pc) = ((p - i16::from(a)).abs(), (p - i16::from(b)).abs(), (p - i16::from(c)).abs());
if pa <= pb && pa <= pc {
a
} else if pb <= pc {
b
} else {
c
}
}
_ => 0,
};
encoded.push(row[i].wrapping_sub(pred));
}
prev = *row;
}
let r = read(&format!(
"/DeviceRGB setcolorspace 30 50 scale << /ImageType 1 /Width 3 /Height 5 /BitsPerComponent 8 /Decode [0 1 0 1 0 1] /ImageMatrix [3 0 0 -5 0 5] \
/DataSource currentfile /ASCII85Decode filter << /Predictor 15 /Colors 3 /Columns 3 >> /FlateDecode filter >> image\n{}",
crate::ps::ascii85(&crate::ps::deflate(&encoded))
));
let want: Vec<[u8; 4]> = raw.iter().flat_map(|row| (0..3).map(move |x| [row[3 * x], row[3 * x + 1], row[3 * x + 2], 255])).collect();
assert_eq!(pixels(&r, 3, 5), want);
}