From 4a16993bec1318b611027fd5f2432e6605da1f85 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Maciej=20Jastrze=CC=A8bski?= <mdjastrzebski@gmail.com>
Date: Sun, 22 Sep 2024 21:16:17 +0200
Subject: [PATCH] refactor: numeric precedence

---
 src/constructs/anchors.ts             |  8 ++++----
 src/constructs/capture.ts             |  6 +++---
 src/constructs/char-class.ts          |  2 +-
 src/constructs/char-escape.ts         | 14 +++++++-------
 src/constructs/choice-of.ts           |  2 +-
 src/constructs/lookahead.ts           |  2 +-
 src/constructs/lookbehind.ts          |  2 +-
 src/constructs/negative-lookahead.ts  |  2 +-
 src/constructs/negative-lookbehind.ts |  2 +-
 src/constructs/quantifiers.ts         |  6 +++---
 src/constructs/repeat.ts              |  4 ++--
 src/constructs/unicode.ts             |  4 ++--
 src/encoder.ts                        | 12 +++++-------
 src/types.ts                          |  2 +-
 14 files changed, 33 insertions(+), 35 deletions(-)

diff --git a/src/constructs/anchors.ts b/src/constructs/anchors.ts
index df9f276..e32ca84 100644
--- a/src/constructs/anchors.ts
+++ b/src/constructs/anchors.ts
@@ -4,7 +4,7 @@ import type { EncodedRegex } from '../types';
  * Start of string anchor. Matches the start of of string. In `multiline` mode, also matches immediately following a newline.
  */
 export const startOfString: EncodedRegex = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '^',
 };
 
@@ -12,7 +12,7 @@ export const startOfString: EncodedRegex = {
  * End of string anchor. Matches the end of a string. In `multiline` mode, also matches immediately preceding a newline.
  */
 export const endOfString: EncodedRegex = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '$',
 };
 
@@ -20,7 +20,7 @@ export const endOfString: EncodedRegex = {
  * Word boundary anchor. Matches the position where one side is a word character (alphanumeric or underscore) and the other side is a non-word character (anything else).
  */
 export const wordBoundary: EncodedRegex = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\b',
 };
 
@@ -28,7 +28,7 @@ export const wordBoundary: EncodedRegex = {
  * Non-word boundary anchor. Matches the position where both sides are word characters.
  */
 export const nonWordBoundary: EncodedRegex = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\B',
 };
 
diff --git a/src/constructs/capture.ts b/src/constructs/capture.ts
index 3814866..0db37fd 100644
--- a/src/constructs/capture.ts
+++ b/src/constructs/capture.ts
@@ -21,13 +21,13 @@ export function capture(sequence: RegexSequence, options?: CaptureOptions): Enco
   const name = options?.name;
   if (name) {
     return {
-      precedence: 'atom',
+      precedence: 0,
       pattern: `(?<${name}>${encode(sequence).pattern})`,
     };
   }
 
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `(${encode(sequence).pattern})`,
   };
 }
@@ -43,7 +43,7 @@ export function capture(sequence: RegexSequence, options?: CaptureOptions): Enco
  */
 export function ref(name: string): Reference {
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `\\k<${name}>`,
     name,
   };
diff --git a/src/constructs/char-class.ts b/src/constructs/char-class.ts
index a2b475d..c81bcb7 100644
--- a/src/constructs/char-class.ts
+++ b/src/constructs/char-class.ts
@@ -81,7 +81,7 @@ function encodeCharClass(
   isNegated?: boolean,
 ): EncodedRegex {
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `[${isNegated ? '^' : ''}${this.elements.join('')}]`,
   };
 }
diff --git a/src/constructs/char-escape.ts b/src/constructs/char-escape.ts
index ee2854b..4533e7e 100644
--- a/src/constructs/char-escape.ts
+++ b/src/constructs/char-escape.ts
@@ -5,7 +5,7 @@ import type { CharacterEscape, EncodedRegex } from '../types';
  * Specifically this one is NOT a character escape.
  */
 export const any: EncodedRegex = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '.',
 };
 
@@ -13,7 +13,7 @@ export const any: EncodedRegex = {
  * Matches any digit (0-9).
  */
 export const digit: CharacterEscape = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\d',
   elements: ['\\d'],
 };
@@ -22,7 +22,7 @@ export const digit: CharacterEscape = {
  * Matches any non-digit (0-9) character.
  */
 export const nonDigit: CharacterEscape = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\D',
   elements: ['\\D'],
 };
@@ -31,7 +31,7 @@ export const nonDigit: CharacterEscape = {
  * Matches any word character (alphanumeric or underscore).
  */
 export const word: CharacterEscape = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\w',
   elements: ['\\w'],
 };
@@ -40,7 +40,7 @@ export const word: CharacterEscape = {
  * Matches any non-word (alphanumeric or underscore) character.
  */
 export const nonWord: CharacterEscape = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\W',
   elements: ['\\W'],
 };
@@ -49,7 +49,7 @@ export const nonWord: CharacterEscape = {
  * Matches any whitespace character (space, tab, newline, etc.).
  */
 export const whitespace: CharacterEscape = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\s',
   elements: ['\\s'],
 };
@@ -58,7 +58,7 @@ export const whitespace: CharacterEscape = {
  * Matches any non-whitespace (space, tab, newline, etc.) character.
  */
 export const nonWhitespace: CharacterEscape = {
-  precedence: 'atom',
+  precedence: 0,
   pattern: '\\S',
   elements: ['\\S'],
 };
diff --git a/src/constructs/choice-of.ts b/src/constructs/choice-of.ts
index 0434174..170ae0d 100644
--- a/src/constructs/choice-of.ts
+++ b/src/constructs/choice-of.ts
@@ -18,7 +18,7 @@ export function choiceOf(...alternatives: RegexSequence[]): EncodedRegex {
   }
 
   return {
-    precedence: 'disjunction',
+    precedence: 2,
     pattern: encodedAlternatives.map((n) => n.pattern).join('|'),
   };
 }
diff --git a/src/constructs/lookahead.ts b/src/constructs/lookahead.ts
index 6180033..d9fb582 100644
--- a/src/constructs/lookahead.ts
+++ b/src/constructs/lookahead.ts
@@ -17,7 +17,7 @@ import type { EncodedRegex, RegexSequence } from '../types';
  */
 export function lookahead(sequence: RegexSequence): EncodedRegex {
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `(?=${encode(sequence).pattern})`,
   };
 }
diff --git a/src/constructs/lookbehind.ts b/src/constructs/lookbehind.ts
index 9187bed..85f1bc2 100644
--- a/src/constructs/lookbehind.ts
+++ b/src/constructs/lookbehind.ts
@@ -17,7 +17,7 @@ import type { EncodedRegex, RegexSequence } from '../types';
  */
 export function lookbehind(sequence: RegexSequence): EncodedRegex {
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `(?<=${encode(sequence).pattern})`,
   };
 }
diff --git a/src/constructs/negative-lookahead.ts b/src/constructs/negative-lookahead.ts
index 5694ca6..54e054a 100644
--- a/src/constructs/negative-lookahead.ts
+++ b/src/constructs/negative-lookahead.ts
@@ -17,7 +17,7 @@ import type { EncodedRegex, RegexSequence } from '../types';
  */
 export function negativeLookahead(sequence: RegexSequence): EncodedRegex {
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `(?!${encode(sequence).pattern})`,
   };
 }
diff --git a/src/constructs/negative-lookbehind.ts b/src/constructs/negative-lookbehind.ts
index b0264f3..debc5f3 100644
--- a/src/constructs/negative-lookbehind.ts
+++ b/src/constructs/negative-lookbehind.ts
@@ -17,7 +17,7 @@ import type { EncodedRegex, RegexSequence } from '../types';
  */
 export function negativeLookbehind(sequence: RegexSequence): EncodedRegex {
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: `(?<!${encode(sequence).pattern})`,
   };
 }
diff --git a/src/constructs/quantifiers.ts b/src/constructs/quantifiers.ts
index 3473a1e..f261b4c 100644
--- a/src/constructs/quantifiers.ts
+++ b/src/constructs/quantifiers.ts
@@ -15,7 +15,7 @@ export interface QuantifierOptions {
 export function zeroOrMore(sequence: RegexSequence, options?: QuantifierOptions): EncodedRegex {
   const elements = ensureElements(sequence);
   return {
-    precedence: 'sequence',
+    precedence: 1,
     pattern: `${encodeAtomic(elements)}*${options?.greedy === false ? '?' : ''}`,
   };
 }
@@ -29,7 +29,7 @@ export function zeroOrMore(sequence: RegexSequence, options?: QuantifierOptions)
 export function oneOrMore(sequence: RegexSequence, options?: QuantifierOptions): EncodedRegex {
   const elements = ensureElements(sequence);
   return {
-    precedence: 'sequence',
+    precedence: 1,
     pattern: `${encodeAtomic(elements)}+${options?.greedy === false ? '?' : ''}`,
   };
 }
@@ -43,7 +43,7 @@ export function oneOrMore(sequence: RegexSequence, options?: QuantifierOptions):
 export function optional(sequence: RegexSequence, options?: QuantifierOptions): EncodedRegex {
   const elements = ensureElements(sequence);
   return {
-    precedence: 'sequence',
+    precedence: 1,
     pattern: `${encodeAtomic(elements)}?${options?.greedy === false ? '?' : ''}`,
   };
 }
diff --git a/src/constructs/repeat.ts b/src/constructs/repeat.ts
index b445ded..62769b5 100644
--- a/src/constructs/repeat.ts
+++ b/src/constructs/repeat.ts
@@ -22,13 +22,13 @@ export function repeat(sequence: RegexSequence, options: RepeatOptions): Encoded
 
   if (typeof options === 'number') {
     return {
-      precedence: 'sequence',
+      precedence: 1,
       pattern: `${encodeAtomic(elements)}{${options}}`,
     };
   }
 
   return {
-    precedence: 'sequence',
+    precedence: 1,
     pattern: `${encodeAtomic(elements)}{${options.min},${options?.max ?? ''}}${
       options.greedy === false ? '?' : ''
     }`,
diff --git a/src/constructs/unicode.ts b/src/constructs/unicode.ts
index 9fd17a6..b729db1 100644
--- a/src/constructs/unicode.ts
+++ b/src/constructs/unicode.ts
@@ -23,7 +23,7 @@ export function unicodeChar(codePoint: number): CharacterEscape {
       : `\\u{${codePoint.toString(16)}}`; // 1-6 digit hex (requires unicode-aware mode)
 
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: escape,
     elements: [escape],
   };
@@ -50,7 +50,7 @@ export function unicodeProperty(property: string, value?: string): CharacterEsca
   const escape = `\\p{${property}${value ? `=${value}` : ''}}`;
 
   return {
-    precedence: 'atom',
+    precedence: 0,
     pattern: escape,
     elements: [escape],
   };
diff --git a/src/encoder.ts b/src/encoder.ts
index c2aaeda..5efe84f 100644
--- a/src/encoder.ts
+++ b/src/encoder.ts
@@ -10,16 +10,14 @@ export function encode(sequence: RegexSequence): EncodedRegex {
   }
 
   return {
-    precedence: 'sequence',
-    pattern: encoded
-      .map((n) => (n.precedence === 'disjunction' ? encodeAtomic(n) : n.pattern))
-      .join(''),
+    precedence: 1,
+    pattern: encoded.map((n) => (n.precedence === 2 ? encodeAtomic(n) : n.pattern)).join(''),
   };
 }
 
 export function encodeAtomic(sequence: RegexSequence): string {
   const encoded = encode(sequence);
-  return encoded.precedence === 'atom' ? encoded.pattern : `(?:${encoded.pattern})`;
+  return encoded.precedence === 0 ? encoded.pattern : `(?:${encoded.pattern})`;
 }
 
 function encodeElement(element: RegexElement): EncodedRegex {
@@ -51,7 +49,7 @@ function encodeText(text: string): EncodedRegex {
 
   return {
     // Optimize for single character case
-    precedence: text.length === 1 ? 'atom' : 'sequence',
+    precedence: text.length === 1 ? 0 : 1,
     pattern: escapeText(text),
   };
 }
@@ -61,7 +59,7 @@ function encodeRegExp(regexp: RegExp): EncodedRegex {
 
   return {
     // Encode at safe precedence
-    precedence: isAtomicPattern(pattern) ? 'atom' : 'disjunction',
+    precedence: isAtomicPattern(pattern) ? 0 : 2,
     pattern,
   };
 }
diff --git a/src/types.ts b/src/types.ts
index 4a4b056..61dd143 100644
--- a/src/types.ts
+++ b/src/types.ts
@@ -28,7 +28,7 @@ export interface EncodedRegex {
 /**
  * Precedence of given regex pattern.
  */
-export type EncodePrecedence = 'atom' | 'sequence' | 'disjunction';
+export type EncodePrecedence = 0 | 1 | 2;
 
 /**
  * Regex patter that can be encoded by calling the `encode` method.