From c756bd444c42dffbd979ab5797ad62e6c0adf30f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E4=BA=B6=E7=88=B6=20=28taan=C2=B2=20fu=C2=B2=29?=
 <tanxpyox@gmail.com>
Date: Wed, 24 Sep 2025 01:12:19 -0400
Subject: [PATCH 1/4] Update(CI): Streamline jyupting checker

Resolves #202
Delegates responsibility to upstream Python checker
---
 .ci/checker.cpp                   | 113 ------------------------------
 .ci/verify.php                    |  87 -----------------------
 .github/workflows/deploy-test.yml |  25 -------
 3 files changed, 225 deletions(-)
 delete mode 100644 .ci/checker.cpp
 delete mode 100644 .ci/verify.php

diff --git a/.ci/checker.cpp b/.ci/checker.cpp
deleted file mode 100644
index a3dbcd4a..00000000
--- a/.ci/checker.cpp
+++ /dev/null
@@ -1,113 +0,0 @@
-#include <iostream>
-#include <fstream>
-#include <sstream>
-#include <string>
-#include <boost/regex.hpp>
-#include <algorithm>
-#include <cstdio>
-#define JYUT phonemes[0]
-#define INITIAL phonemes[1]
-#define STEM phonemes[2]
-#define FINAL phonemes[3]
-#define TONE phonemes[4]
-using namespace boost;
-
-std::string word, line, dict;
-smatch phonemes;
-
-regex expr("(gw?|kw?|ng?|sh?|[bpmfdtlhzcjw])?(aa?|eo?|oe?|[iu]|yu|ng|m)([iumptk]|ng?)?([1-6])");
-
-// CHANGE THIS -- a function to check if a jyutping word is valid 
-bool valid_jyutping(std::string a)
-{
-	// Has a jyutping-like structure -- catches most invalid spellings ('ch-', 'eu-', etc)
-	//      If cannot be decomposed into exactly 4 groups, treat as invalid
-	if (!regex_match(a, phonemes, expr)) return false;
-	
-	/* Note: after the first regex_match, the jyutping string will be decomposed into phonemes acc to the following:
-		phonemes[0] = JYUT = full jyutping string
-		--------------------------------
-		phonemes[1] = INITIAL = initial (empty if none)
-		phonemes[2] = STEM = stem (aa / a / e / i / o / u / eo / oe / yu / ng* / m*)
-		phonemes[3] = FINAL = final (empty if none; including -i and -u; thus -aai will be treated as -aa + -i)
-		phonemes[4] = TONE = tone 
-
-		*Only applies to /ng/ and /m/
-	*/
-	
-	/* ------ ADDITIONAL CHECKING RULES ------ */
-	
-	// eo followed by t / i / n
-	if (STEM == "eo" && !(FINAL == "t" || FINAL == "i" || FINAL == "n")) return false;
-	
-	// i / yu must begin with an initial, but allow ik1 and ik6 (hiccups)
-	if (JYUT == "ik1" || JYUT == "ik6") return true;
-	else if (INITIAL == "" && (STEM == "i" || STEM == "yu")) return false;
-
-	// u must begin with an initial, except for /ung/ and /uk/
-	if (INITIAL == "" && STEM == "u") return (FINAL == "ng" || FINAL == "k");
-	
-	// Add more rules here...
-
-	// catch invalid matches of /m/ and /ng/
-	if ((STEM == "m" || STEM == "ng") && (INITIAL != "" && INITIAL != "h")) return false;
-	if ((STEM == "m" || STEM == "ng") && FINAL != "") return false;
-	
-	return true;
-}
-
-//Strips down a jyutping string into words and passes to valid_jyutping() for checking
-bool process(std::string a)
-{
-	std::istringstream sin(a);
-	while(sin >> word) {
-		if (!valid_jyutping(word)) return false;
-	}
-	return true;
-}
-
-//Processes file and keeps track of discovered errors
-int parse() 
-{
-	int count_invalid = 0;
-	int line_no = 0;
-	
-	// Discard lines until '...' (i.e. end of header block)
-	while (true) {
-		getline(std::cin, line);
-		line_no++;
-		if (line == "...") break;
-	} 
-	
-	// Real checking
-	while(getline(std::cin, line)) {
-		line_no++;
-		int temp = line.find('\t');
-		std::string jyut_string = line.substr(temp+1, line.find('\t', temp + 1) - temp);
-		if (!process(jyut_string)) {
-			std::cerr << "[L" << line_no << "] Invalid Jyutping detected: " << jyut_string << std::endl;
-			count_invalid++;
-		} 
-	}
-
-	return count_invalid;
-}
-
-// Wrapper for command line interface
-int main (int argc, char** argv) 
-{
-	std::ios_base::sync_with_stdio(false);
-	std::cin.tie(NULL);
-	
-	if (argc == 2) {
-		dict = argv[1];
-		std::cout << "Loading " << dict << std::endl;
-	} else {
-		std::cerr << "Invalid input format" << std::endl;
-		std::cerr << "Use ./checker input_file_name.yaml" << std::endl;
-		return 1;
-	}
-	
-	std::freopen(dict.c_str(), "r", stdin); // redirect file to stdin stream
-	return parse();
-}
diff --git a/.ci/verify.php b/.ci/verify.php
deleted file mode 100644
index f92f10e5..00000000
--- a/.ci/verify.php
+++ /dev/null
@@ -1,87 +0,0 @@
-<?php
-
-echo '<pre>';
-$starttime = microtime(true);
-
-$fp = fopen('./jyut6ping3.dict.yaml', 'r');
-$errorcnt = 0;
-$i = 0;
-while($line = fgets($fp)) {
-	if (trim($line) === '' || $line[0] === '#') goto next;
-	if (strpos($line, "\t") === false) goto next;
-
-	list($char, $val) = explode("\t", trim($line), 3);
-
-	$syllable = strtok($val, " ");
-	while ($syllable !== false) {
-		$tone = substr($syllable, -1);
-		if ($tone !== '1' && $tone !== '2' && $tone !== '3' && $tone !== '4' && $tone !== '5' && $tone !== '6') {
-			goto error;
-		}
-
-		$syllable = substr($syllable, 0, -1);
-
-		if ($syllable == 'm' || $syllable == 'ng' || $syllable == 'hm' || $syllable == 'hng') goto success;
-
-		$syllable = strtr($syllable, [
-			'q' => 'qqqq', 'v' => 'qqqq',
-
-			'aang' => 'v', 'aan' => 'v', 'aam' => 'v',
-			'aak' => 'v', 'aat' => 'v', 'aap' => 'v',
-			'aai' => 'v', 'aau' => 'v', 'aa' => 'v',
-
-			'ang' => 'v', 'an' => 'v', 'am' => 'v',
-			'ak' => 'v', 'at' => 'v', 'ap' => 'v',
-			'ai' => 'v', 'au' => 'v', 'a' => 'v',
-
-			'em' => 'v', 'eng' => 'v', 'en' => 'v',
-			'ep' => 'v', 'ek' => 'v', 'et' => 'v',
-			'ei' => 'v', 'eu' => 'v', 'e' => 'v',
-
-			'ing' => 'v', 'in' => 'v', 'im' => 'v',
-			'ik' => 'v', 'it' => 'v', 'ip' => 'v',
-			'iu' => 'v', 'i' => 'v',
-
-			'oi' => 'v', 'ou' => 'v',
-			'ong' => 'v', 'on' => 'v', 'ot' => 'v', 'ok' => 'v', 'om' => 'v',
-			'o' => 'v', 
-
-			'eoi' => 'v', 'eon' => 'v', 'eot' => 'v', 
-			'oeng' => 'v', 'oet' => 'v', 'oek' => 'v', 'oe' => 'v', 
-			'yun' => 'v', 'yut' => 'v', 'yu' => 'v',
-			
-			'ui' => 'v', 'ung' => 'v', 'un' => 'v', 'ut' => 'v', 'uk' => 'v', 
-			'u' => 'v', 
-		]);
-
-		$syllable = strtr($syllable, [
-			'gw' => 'q', 'kw' => 'q',
-			'b' => 'q', 'p' => 'q', 'm' => 'q', 'f' => 'q',
-			'd' => 'q', 't' => 'q', 'n' => 'q', 'l' => 'q',
-			'g' => 'q', 'k' => 'q', 'ng' => 'q', 'h' => 'q',
-			'z' => 'q', 'c' => 'q', 's' => 'q',
-			'j' => 'q', 'w' => 'q'
-		]);
-
-		if ($syllable !== 'qv' && $syllable !== 'v') {
-			echo $syllable . " ";
-			goto error;
-		}
-		
-
-		success:
-		$syllable = strtok(" ");
-		continue;
-
-		error:
-		echo "Invalid Jyutping detected on line $i: $line" . "\r\n";
-		$errorcnt++;
-		$syllable = strtok(" ");
-	}
-
-	next:
-	$i++;
-}
-
-echo "Time used: " . (microtime(true) - $starttime) . ' s';
-exit(min($errorcnt, 1));
diff --git a/.github/workflows/deploy-test.yml b/.github/workflows/deploy-test.yml
index ffea1029..b3f327e5 100644
--- a/.github/workflows/deploy-test.yml
+++ b/.github/workflows/deploy-test.yml
@@ -39,28 +39,3 @@ jobs:
         run: |
           cat log.tmp
           exit $(cat log.tmp | grep -c ^[EW]) 
-
-  check_jyutping_cpp:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - run: |
-          echo "Installing prerequisites..."
-          sudo apt-get install libboost-all-dev
-      - run: |
-          echo "Building checker from source..."
-          export LD_LIBRARY_PATH="/usr/lib/boost/lib"
-          g++ .ci/checker.cpp -o checker.o -Ofast -I/usr/include/boost -L/usr/lib/boost/lib -lboost_regex
-          chmod u+x ./checker.o
-      - run: |
-          echo "Checking jyut6ping3.dict.yaml"
-          time ./checker.o jyut6ping3.dict.yaml
-
-  check_jyutping_php:
-    runs-on: ubuntu-latest
-    steps:
-      - name: Checkout
-        uses: actions/checkout@v4
-      - run: |
-          php .ci/verify.php 

From dc17381c16e227274139c4f5b830b834e086f70e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E4=BA=B6=E7=88=B6=20=28taan=C2=B2=20fu=C2=B2=29?=
 <tanxpyox@gmail.com>
Date: Wed, 24 Sep 2025 01:25:37 -0400
Subject: [PATCH 2/4] Update(schema): change simplifier priority

Addresses #189
---
 jyut6ping3.schema.yaml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/jyut6ping3.schema.yaml b/jyut6ping3.schema.yaml
index 94a03bd2..b907cca6 100644
--- a/jyut6ping3.schema.yaml
+++ b/jyut6ping3.schema.yaml
@@ -78,9 +78,9 @@ engine:
     - table_translator@cangjie5
   filters:
     - simplifier@emoji_cantonese_suggestion
+    - simplifier
     - simplifier@variants_hk
     - simplifier@trad_tw
-    - simplifier
     - uniquifier
     - reverse_lookup_filter@jyut6ping3_reverse_lookup
 

From 23287e6473acba1fa3b5cf832b0942e85221e874 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E4=BA=B6=E7=88=B6=20=28taan=C2=B2=20fu=C2=B2=29?=
 <tanxpyox@gmail.com>
Date: Wed, 24 Sep 2025 08:48:17 -0400
Subject: [PATCH 3/4] Fix(CI): pull request refs

Old version always fetches commits from the master branch; fixed it so that it will now install files from the checked out branch (e.g. from a pull request).
---
 .github/workflows/deploy-test.yml | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/.github/workflows/deploy-test.yml b/.github/workflows/deploy-test.yml
index b3f327e5..3d336803 100644
--- a/.github/workflows/deploy-test.yml
+++ b/.github/workflows/deploy-test.yml
@@ -22,15 +22,20 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout
-        uses: actions/checkout@v4
+        uses: actions/checkout@v5
       - name: Install rime engine
         run: |
           sudo apt-get install ibus-rime -y
-      - name: Install rime-cantonese files
+      - name: Install rime-cantonese files from checked out branch
         run: |
           chmod u+x ./.ci/*
           export rime_dir=~/.config/ibus/rime
-          ./.ci/install-schema.sh
+          curl -fsSL https://git.io/rime-install | bash -s -- :preset emoji CanCLID/rime-loengfan custom:set:config=default,key=installed_from,value=rime-cantonese custom:clear_schema_list custom:add:schema=jyut6ping3 custom:add:schema=cangjie5 custom:add:schema=stroke custom:add:schema=luna_pinyin lotem/rime-octagram-data lotem/rime-octagram-data@hant lotem/rime-octagram-data:customize:schema=jyut6ping3,model=hant
+          cp ./*.{txt,yaml} $rime_dir 
+          cp ./opencc/* $rime_dir
+
+# Use cp ./** in lieu of ./.ci/install_schema.sh to ensure that action is running on checked out branch (e.g. PR refs)
+
       - name: Compile
         run: |
           chmod u+wx ~/.config/ibus/rime/*

From 611506b4ea228f71f6142c99ad39f8e88aa13d80 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E4=BA=B6=E7=88=B6=20=28taan=C2=B2=20fu=C2=B2=29?=
 <tanxpyox@gmail.com>
Date: Wed, 24 Sep 2025 09:07:50 -0400
Subject: [PATCH 4/4] Temporarily reinstate cpp checker for transition

---
 .ci/checker.cpp                   | 109 ++++++++++++++++++++++++++++++
 .github/workflows/deploy-test.yml |  17 +++++
 2 files changed, 126 insertions(+)
 create mode 100644 .ci/checker.cpp

diff --git a/.ci/checker.cpp b/.ci/checker.cpp
new file mode 100644
index 00000000..a3889d47
--- /dev/null
+++ b/.ci/checker.cpp
@@ -0,0 +1,109 @@
+#include <iostream>
+#include <fstream>
+#include <sstream>
+#include <string>
+#include <boost/regex.hpp>
+#include <algorithm>
+#include <cstdio>
+#define JYUT phonemes[0]
+#define INITIAL phonemes[1]
+#define STEM phonemes[2]
+#define FINAL phonemes[3]
+#define TONE phonemes[4]
+using namespace boost;
+
+std::string word, line, dict;
+smatch phonemes;
+
+regex expr("(gw?|kw?|ng?|sh?|[bpmfdtlhzcjw])?(aa?|eo?|oe?|[iu]|yu|ng|m)([iumptk]|ng?)?([1-6])");
+
+// CHANGE THIS -- a function to check if a jyutping word is valid 
+bool valid_jyutping(std::string a) {
+	// Has a jyutping-like structure -- catches most invalid spellings ('ch-', 'eu-', etc)
+	//      If cannot be decomposed into exactly 4 groups, treat as invalid
+	if (!regex_match(a, phonemes, expr)) return false;
+	
+	/* Note: after the first regex_match, the jyutping string will be decomposed into phonemes acc to the following:
+		phonemes[0] = JYUT = full jyutping string
+		--------------------------------
+		phonemes[1] = INITIAL = initial (empty if none)
+		phonemes[2] = STEM = stem (aa / a / e / i / o / u / eo / oe / yu / ng* / m*)
+		phonemes[3] = FINAL = final (empty if none; including -i and -u; thus -aai will be treated as -aa + -i)
+		phonemes[4] = TONE = tone 
+
+		*Only applies to /ng/ and /m/
+	*/
+	
+	/* ------ ADDITIONAL CHECKING RULES ------ */
+	
+	// eo followed by t / i / n
+	if (STEM == "eo" && !(FINAL == "t" || FINAL == "i" || FINAL == "n")) return false;
+	
+	// i / yu must begin with an initial, but allow ik1 and ik6 (hiccups)
+	if (JYUT == "ik1" || JYUT == "ik6") return true;
+	else if (INITIAL == "" && (STEM == "i" || STEM == "yu")) return false;
+
+	// u must begin with an initial, except for /ung/ and /uk/
+	if (INITIAL == "" && STEM == "u") return (FINAL == "ng" || FINAL == "k");
+	
+	// Add more rules here...
+
+	// catch invalid matches of /m/ and /ng/
+	if ((STEM == "m" || STEM == "ng") && (INITIAL != "" && INITIAL != "h")) return false;
+	if ((STEM == "m" || STEM == "ng") && FINAL != "") return false;
+	
+	return true;
+}
+
+//Strips down a jyutping string into words and passes to valid_jyutping() for checking
+bool process(std::string a) {
+	std::istringstream sin(a);
+	while(sin >> word) {
+		if (!valid_jyutping(word)) return false;
+	}
+	return true;
+}
+
+//Processes file and keeps track of discovered errors
+int parse()  {
+	int count_invalid = 0;
+	int line_no = 0;
+	
+	// Discard lines until '...' (i.e. end of header block)
+	while (true) {
+		getline(std::cin, line);
+		line_no++;
+		if (line == "...") break;
+	} 
+	
+	// Real checking
+	while(getline(std::cin, line)) {
+		line_no++;
+		int temp = line.find('\t');
+		std::string jyut_string = line.substr(temp+1, line.find('\t', temp + 1) - temp);
+		if (!process(jyut_string)) {
+			std::cerr << "[L" << line_no << "] Invalid Jyutping detected: " << jyut_string << std::endl;
+			count_invalid++;
+		} 
+	}
+
+	return count_invalid;
+}
+
+// Wrapper for command line interface
+int main (int argc, char** argv) {
+	std::ios_base::sync_with_stdio(false);
+	std::cin.tie(NULL);
+	
+	if (argc == 2) {
+		dict = argv[1];
+		std::cout << "Loading " << dict << std::endl;
+	} else {
+		std::cerr << "Invalid input format" << std::endl;
+		std::cerr << "Use ./checker input_file_name.yaml" << std::endl;
+		return 1;
+	}
+	
+	std::freopen(dict.c_str(), "r", stdin); // redirect file to stdin stream
+	return parse();
+}
\ No newline at end of file
diff --git a/.github/workflows/deploy-test.yml b/.github/workflows/deploy-test.yml
index 3d336803..83832190 100644
--- a/.github/workflows/deploy-test.yml
+++ b/.github/workflows/deploy-test.yml
@@ -44,3 +44,20 @@ jobs:
         run: |
           cat log.tmp
           exit $(cat log.tmp | grep -c ^[EW]) 
+
+  check_jyutping_cpp:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - run: |
+          echo "Installing prerequisites..."
+          sudo apt-get install libboost-all-dev
+      - run: |
+          echo "Building checker from source..."
+          export LD_LIBRARY_PATH="/usr/lib/boost/lib"
+          g++ .ci/checker.cpp -o checker.o -Ofast -I/usr/include/boost -L/usr/lib/boost/lib -lboost_regex
+          chmod u+x ./checker.o
+      - run: |
+          echo "Checking jyut6ping3.dict.yaml"
+          time ./checker.o jyut6ping3.dict.yaml
\ No newline at end of file