From f835b53da4d7285c0ce00933ca9e84286517b3d4 Mon Sep 17 00:00:00 2001
From: Miles Johnson <milesj@users.noreply.github.com>
Date: Sat, 8 Jun 2024 17:59:28 -0700
Subject: [PATCH] fix: Improve regex patterns. (#175)

* Update regex.

* Go back.

* Add more tests.

* Fix.
---
 bin/generateRegex.mjs              | 11 +++++++++++
 package.json                       |  1 +
 packages/regex/tests/regex.test.ts | 31 +++++++++++++++++++++++++++---
 3 files changed, 40 insertions(+), 3 deletions(-)
 create mode 100644 bin/generateRegex.mjs

diff --git a/bin/generateRegex.mjs b/bin/generateRegex.mjs
new file mode 100644
index 00000000..4dfba8df
--- /dev/null
+++ b/bin/generateRegex.mjs
@@ -0,0 +1,11 @@
+#!/usr/bin/env node
+
+import { generateRegex } from 'emojibase-generator';
+
+function handleError(error) {
+	console.error(error.message, error.stack);
+}
+
+process.on('unhandledRejection', handleError);
+
+generateRegex().catch(handleError);
diff --git a/package.json b/package.json
index e0678755..17f461a8 100644
--- a/package.json
+++ b/package.json
@@ -13,6 +13,7 @@
     "generate": "yarn run generate:emojis && yarn run generate:shortcodes && node ./bin/buildFilesizeTable.mjs && yarn run generate:tests",
     "generate:emojis": "rm -rf ./cache/final && node ./bin/generateEmoji.mjs",
     "generate:po": "node ./bin/generatePo.mjs",
+    "generate:regex": "node ./bin/generateRegex.mjs",
     "generate:shortcodes": "node ./bin/generateShortcodes.mjs",
     "generate:tests": "cp ./packages/data/en/data.json ./packages/test-utils/test-data.json && cp ./packages/data/en/shortcodes/emojibase.json ./packages/test-utils/test-shortcodes.json && cp ./packages/data/en/messages.json ./packages/test-utils/test-messages.json",
     "generate:types": "node ./bin/generateTypes.mjs",
diff --git a/packages/regex/tests/regex.test.ts b/packages/regex/tests/regex.test.ts
index 560bc3bc..4968fd5f 100644
--- a/packages/regex/tests/regex.test.ts
+++ b/packages/regex/tests/regex.test.ts
@@ -72,7 +72,8 @@ describe('regex', () => {
 			tests.push({
 				type: 'emoji',
 				unicode: emoji.emoji,
-				pass: [...BASE_PATTERNS, 'emojiLoose', 'emojiCodepoint', 'emojiCodepointLoose'],
+				pass: [...BASE_PATTERNS, 'emoji', 'emojiLoose', 'emojiCodepoint', 'emojiCodepointLoose'],
+				// fail: ['text', 'textLoose', 'textCodepoint', 'textCodepointLoose'],
 			});
 		}
 
@@ -81,11 +82,12 @@ describe('regex', () => {
 			tests.push({
 				type: 'text',
 				unicode: emoji.text,
-				pass: [...BASE_PATTERNS, 'textLoose', 'textCodepoint', 'textCodepointLoose'],
+				pass: [...BASE_PATTERNS, 'text', 'textLoose', 'textCodepoint', 'textCodepointLoose'],
+				// fail: ['emoji', 'emojiLoose', 'emojiCodepoint', 'emojiCodepointLoose'],
 			});
 		}
 
-		tests.forEach(({ unicode, type, pass }) => {
+		tests.forEach(({ unicode, type, pass, fail }) => {
 			describe(`${VARIATION_DESCRIPTIONS[type]}`, () => {
 				pass.forEach((passType) => {
 					const pattern = PATTERNS[passType];
@@ -113,6 +115,17 @@ describe('regex', () => {
 						});
 					});
 				});
+
+				fail?.forEach((passType) => {
+					const pattern = PATTERNS[passType];
+
+					describe(`fails ${PATTERN_DESCRIPTIONS[passType]}`, () => {
+						it(`doesnt match unicode by itself for ${unicode}`, () => {
+							const match = unicode.match(pattern)!;
+							expect(match).toBeNull();
+						});
+					});
+				});
 			});
 		});
 
@@ -194,4 +207,16 @@ describe('regex', () => {
 			});
 		});
 	});
+
+	it('manual cases', () => {
+		expect('👍🏻'.match(PATTERNS.combo)).not.toBeNull();
+		expect('👍🏻'.match(PATTERNS.emoji)).not.toBeNull();
+		expect('👍🏻'.match(PATTERNS.emojiLoose)).not.toBeNull();
+		// expect('👍🏻'.match(PATTERNS.text)).toBeNull();
+
+		expect('👍'.match(PATTERNS.combo)).not.toBeNull();
+		expect('👍'.match(PATTERNS.emoji)).toBeNull();
+		expect('👍'.match(PATTERNS.emojiLoose)).not.toBeNull();
+		// expect('👍'.match(PATTERNS.text)).toBeNull();
+	});
 });