Add bigrams filter

Cargo.lock

@@ -43,6 +43,18 @@ version = "1.3.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "bef38d45163c2f1dde094a7dfd33ccf595c92905c8f8f4fdc18d06fb1037718a"
 
+[[package]]
+name = "bitvec"
+version = "1.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1bc2832c24239b0141d5674bb9174f9d68a8b5b3f2753311927c172ca46f7e9c"
+dependencies = [
+ "funty",
+ "radium",
+ "tap",
+ "wyz",
+]
+
 [[package]]
 name = "bytemuck"
 version = "1.12.1"
@@ -194,6 +206,12 @@ dependencies = [
  "miniz_oxide",
 ]
 
+[[package]]
+name = "funty"
+version = "2.0.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "e6d5a32815ae3f33302d95fdcb2ce17862f8c65363dcfd29360480ba1001fc9c"
+
 [[package]]
 name = "glob"
 version = "0.3.0"
@@ -363,6 +381,7 @@ name = "oxipng"
 version = "6.0.1"
 dependencies = [
  "bit-vec",
+ "bitvec",
  "clap",
  "crossbeam-channel",
  "filetime",
@@ -391,6 +410,12 @@ dependencies = [
  "miniz_oxide",
 ]
 
+[[package]]
+name = "radium"
+version = "0.7.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "dc33ff2d4973d518d823d61aa239014831e521c75da58e3df4840d3f47749d09"
+
 [[package]]
 name = "rayon"
 version = "1.5.3"
@@ -472,6 +497,12 @@ version = "0.10.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "73473c0e59e6d5812c5dfe2a064a6444949f089e20eec9a2e5506596494e4623"
 
+[[package]]
+name = "tap"
+version = "1.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "55937e1799185b12863d447f42597ed69d9928686b8d88a1df17376a097d8369"
+
 [[package]]
 name = "termcolor"
 version = "1.1.3"
@@ -585,6 +616,15 @@ version = "0.36.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c811ca4a8c853ef420abd8592ba53ddbbac90410fab6903b3e79972a631f7680"
 
+[[package]]
+name = "wyz"
+version = "0.5.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "30b31594f29d27036c383b53b59ed3476874d518f0efb151b27a4c275141390e"
+dependencies = [
+ "tap",
+]
+
 [[package]]
 name = "zopfli"
 version = "0.7.1"

Cargo.toml

@@ -32,6 +32,7 @@ libdeflater = "0.11.0"
 log = "0.4.17"
 stderrlog = { version = "0.5.3", optional = true, default-features = false }
 crossbeam-channel = "0.5.6"
+bitvec = "1.0.1"
 
 [dependencies.filetime]
 optional = true

src/filters.rs

@@ -14,6 +14,7 @@ pub enum RowFilter {
     // Heuristic strategies
     MinSum,
     Entropy,
+    Bigrams,
 }
 
 impl TryFrom<u8> for RowFilter {
@@ -40,13 +41,14 @@ impl Display for RowFilter {
                 Self::Paeth => "Paeth",
                 Self::MinSum => "MinSum",
                 Self::Entropy => "Entropy",
+                Self::Bigrams => "Bigrams",
             }
         )
     }
 }
 
 impl RowFilter {
-    pub const LAST: u8 = Self::Entropy as u8;
+    pub const LAST: u8 = Self::Bigrams as u8;
     pub const STANDARD: [Self; 5] = [Self::None, Self::Sub, Self::Up, Self::Average, Self::Paeth];
     pub const SINGLE_LINE: [Self; 2] = [Self::None, Self::Sub];
 

src/png/mod.rs

@@ -4,6 +4,7 @@ use crate::error::PngError;
 use crate::filters::*;
 use crate::headers::*;
 use crate::interlace::{deinterlace_image, interlace_image};
+use bitvec::bitarr;
 use indexmap::IndexMap;
 use rgb::ComponentSlice;
 use rgb::RGBA8;
@@ -367,6 +368,24 @@ impl PngImage {
                             }
                         }
                     }
+                    RowFilter::Bigrams => {
+                        // Count distinct bigrams, from pngwolf
+                        // https://bjoern.hoehrmann.de/pngwolf/
+                        let mut best_size = usize::MAX;
+                        for try_filter in try_filters {
+                            try_filter.filter_line(bpp, line.data, last_line, &mut f_buf);
+                            let mut set = bitarr![0; 0x10000];
+                            for i in 1..f_buf.len() {
+                                let bigram = (f_buf[i - 1] as usize) << 8 | f_buf[i] as usize;
+                                set.set(bigram, true);
+                            }
+                            let size = set.count_ones();
+                            if size < best_size {
+                                best_size = size;
+                                std::mem::swap(&mut best_line, &mut f_buf);
+                            }
+                        }
+                    }
                     _ => unreachable!(),
                 }
                 filtered.extend_from_slice(&best_line);