enh(haskell) Restucture Haskell number parsing

Restructured Haskell's number parsing, taking a cue from Swift's
implementation. As a happy coincidence, this patch adds support for
Haskell's `HexFloatLiterals` language extension.

CHANGES.md

@@ -80,6 +80,7 @@ Grammars:
 - enh(javascript/typescript) highlight object properties (#3072) [Josh Goebel][]
 - enh(haskell) add support for BinaryLiterals (#3150) [Martijn Bastiaan][]
 - enh(haskell) add support for NumericUnderscores (#3150) [Martijn Bastiaan][]
+- enh(haskell) add support for HexFloatLiterals (#3150) [Martijn Bastiaan][]
 
 New Languages:
 

src/languages/haskell.js

@@ -62,6 +62,42 @@ export default function(hljs) {
     contains: LIST.contains
   };
 
+  /* See:
+
+     - https://www.haskell.org/onlinereport/lexemes.html
+     - https://downloads.haskell.org/ghc/9.0.1/docs/html/users_guide/exts/binary_literals.html
+     - https://downloads.haskell.org/ghc/9.0.1/docs/html/users_guide/exts/numeric_underscores.html
+     - https://downloads.haskell.org/ghc/9.0.1/docs/html/users_guide/exts/hex_float_literals.html
+
+  */
+  const decimalDigits = '([0-9]_*)+';
+  const hexDigits = '([0-9a-gA-F]_*)+';
+  const binaryDigits = '([01]_*)+';
+  const octalDigits = '([0-7]_*)+';
+
+  const NUMBER = {
+    className: 'number',
+    relevance: 0,
+    variants: [
+      // decimal floating-point-literal (subsumes decimal-literal)
+      {
+        match: `(-?)\\b(${decimalDigits})(\\.(${decimalDigits}))?` + `([eE][+-]?(${decimalDigits}))?\\b`
+      },
+      // hexadecimal floating-point-literal (subsumes hexadecimal-literal)
+      {
+        match: `(-?)\\b0x_*(${hexDigits})(\\.(${hexDigits}))?` + `([pP][+-]?(${decimalDigits}))?\\b`
+      },
+      // octal-literal
+      {
+        match: `(-?)\\b0b(${octalDigits})\\b`
+      },
+      // binary-literal
+      {
+        match: `(-?)\\b0b(${binaryDigits})\\b`
+      }
+    ]
+  };
+
   return {
     name: 'Haskell',
     aliases: ['hs'],
@@ -154,13 +190,7 @@ export default function(hljs) {
       PREPROCESSOR,
 
       // Literals and names.
-      hljs.inherit(hljs.BINARY_NUMBER_MODE, {
-        begin: '\\b(0b[01_]+)'
-      }),
-
-      hljs.inherit(hljs.C_NUMBER_MODE, {
-        begin: '(-?)(\\b0[xX][a-fA-F0-9_]+|(\\b(\\d|_)+(\\.(\\d|_)*)?|\\.(\\d|_)+)([eE][-+]?(\\d|_)+)?)'
-      }),
+      NUMBER,
 
       // TODO: characters.
       hljs.QUOTE_STRING_MODE,

test/markup/haskell/numbers.expect.txt

@@ -1,3 +1,89 @@
-<span class="hljs-number">0b001_001</span>
-<span class="hljs-number">0x000_0FF</span>
-<span class="hljs-number">132_15</span>
+-- decimal
+million    = 1_000_000
+billion    = 1_000_000_000
+lightspeed = 299_792_458
+version    = 8_04_1
+date       = 2017_12_31
+the_answer = 42
+
+-- hexadecimal
+red_mask = 0xff_00_00
+size1G   = 0x3fff_fffg
+
+-- binary
+bit8th   = 0b01_0000_0000
+packbits = 0b1_11_01_0000_0_111
+bigbits  = 0b1100_1011__1110_1111__0101_0011
+
+-- float
+pi       = 3.141_592_653_589_793
+faraday  = 96_485.332_89
+avogadro = 6.022_140_857e+23
+
+x0 = 1_000_000
+x1 = 1__000000
+
+e0 = 0.0001
+e1 = 0.000_1
+
+f0 = 1e+23
+f1 = 1_e+23
+f2 = 1__e+23
+
+g0 = 1e+23
+
+h0 = 0xffff
+h1 = 0xff_ff
+h2 = 0x_ffff
+h3 = 0x__ffff
+
+-- decimal
+million    = 1_000_000
+billion    = 1_000_000_000
+lightspeed = 299_792_458
+version    = 8_04_1
+date       = 2017_12_31
+
+-- hexadecimal
+red_mask = 0xff_00_00
+size1G   = 0x3fff_fffg
+
+-- binary
+bit8th   = 0b01_0000_0000
+packbits = 0b1_11_01_0000_0_111
+bigbits  = 0b1100_1011__1110_1111__0101_0011
+
+-- float
+pi       = 3.141_592_653_589_793
+faraday  = 96_485.332_89
+avogadro = 6.022_140_857e+23
+
+x0 = 1_000_000
+x1 = 1__000000
+
+e0 = 0.0001
+e1 = 0.000_1
+
+f0 = 1e+23
+f1 = 1_e+23
+f2 = 1__e+23
+
+g0 = 1e+23
+
+h0 = 0xffff
+h1 = 0xff_ff
+h2 = 0x_ffff
+h3 = 0x__ffff
+
+-- hex float
+hf0 = 0x0.1
+hf1 = 0x0.01
+hf2 = 0xF.FF
+hf3 = 0x0.1p41
+hf4 = 0x0.1p-4
+hf5 = 0x0.1p12
+
+-- negative numbers
+n0 = -1
+n0 = -1.0
+n0 = -0xFF0

test/markup/haskell/numbers.txt

@@ -1,3 +1,89 @@
-0b001_001
-0x000_0FF
-132_15
+-- decimal
+million    = 1_000_000
+billion    = 1_000_000_000
+lightspeed = 299_792_458
+version    = 8_04_1
+date       = 2017_12_31
+the_answer = 42
+
+-- hexadecimal
+red_mask = 0xff_00_00
+size1G   = 0x3fff_fffg
+
+-- binary
+bit8th   = 0b01_0000_0000
+packbits = 0b1_11_01_0000_0_111
+bigbits  = 0b1100_1011__1110_1111__0101_0011
+
+-- float
+pi       = 3.141_592_653_589_793
+faraday  = 96_485.332_89
+avogadro = 6.022_140_857e+23
+
+x0 = 1_000_000
+x1 = 1__000000
+
+e0 = 0.0001
+e1 = 0.000_1
+
+f0 = 1e+23
+f1 = 1_e+23
+f2 = 1__e+23
+
+g0 = 1e+23
+
+h0 = 0xffff
+h1 = 0xff_ff
+h2 = 0x_ffff
+h3 = 0x__ffff
+
+-- decimal
+million    = 1_000_000
+billion    = 1_000_000_000
+lightspeed = 299_792_458
+version    = 8_04_1
+date       = 2017_12_31
+
+-- hexadecimal
+red_mask = 0xff_00_00
+size1G   = 0x3fff_fffg
+
+-- binary
+bit8th   = 0b01_0000_0000
+packbits = 0b1_11_01_0000_0_111
+bigbits  = 0b1100_1011__1110_1111__0101_0011
+
+-- float
+pi       = 3.141_592_653_589_793
+faraday  = 96_485.332_89
+avogadro = 6.022_140_857e+23
+
+x0 = 1_000_000
+x1 = 1__000000
+
+e0 = 0.0001
+e1 = 0.000_1
+
+f0 = 1e+23
+f1 = 1_e+23
+f2 = 1__e+23
+
+g0 = 1e+23
+
+h0 = 0xffff
+h1 = 0xff_ff
+h2 = 0x_ffff
+h3 = 0x__ffff
+
+-- hex float
+hf0 = 0x0.1
+hf1 = 0x0.01
+hf2 = 0xF.FF
+hf3 = 0x0.1p41
+hf4 = 0x0.1p-4
+hf5 = 0x0.1p12
+
+-- negative numbers
+n0 = -1
+n0 = -1.0
+n0 = -0xFF0