[19_2] Move cjk predicates to lolly::data

Kernel/Types/analyze.cpp

@@ -86,44 +86,6 @@ is_numeric (string s) {
   return true;
 }
 
-bool
-is_cjk_unified_ideographs (string s) {
-  int n= N (s);
-  for (int i= 0; i < n; i++)
-    if (s[i] == '<' && i + 1 < n && s[i + 1] == '#') {
-      int start= i + 2;
-      i        = i + 2;
-      while (i < n && s[i] != '>')
-        i++;
-      string r= s (start, i);
-      if ("4E00" <= r && r <= "9FBF") continue;
-      else return false;
-    }
-    else {
-      return false;
-    }
-  return true;
-}
-
-bool
-has_cjk_unified_ideographs (string s) {
-  int n= N (s);
-  for (int i= 0; i < n; i++)
-    if (s[i] == '<' && i + 1 < n && s[i + 1] == '#') {
-      int start= i + 2;
-      i        = i + 2;
-      while (i < n && s[i] != '>')
-        i++;
-      string r= s (start, i);
-      if ("4E00" <= r && r <= "9FBF") return true;
-      else continue;
-    }
-    else {
-      continue;
-    }
-  return false;
-}
-
 /******************************************************************************
  * Changing cases
  ******************************************************************************/

Kernel/Types/analyze.hpp

@@ -125,28 +125,6 @@ bool is_iso_alpha (string s);
  */
 bool is_numeric (string s);
 
-/**
- * @brief Checks if a string contains only CJK Unified Ideographs.
- *
- * @param s The string to check.
- * @return True if all characters in the string are CJK Unified Ideographs,
- * otherwise false.
- * @note This function expects the CJK Unified Ideographs to be in a specific
- * encoded format.
- */
-bool is_cjk_unified_ideographs (string s);
-
-/**
- * @brief Checks if a string contains any CJK Unified Ideographs.
- *
- * @param s The string to check.
- * @return True if the string contains at least one CJK Unified Ideograph,
- * otherwise false.
- * @note This function expects the CJK Unified Ideographs to be in a specific
- * encoded format.
- */
-bool has_cjk_unified_ideographs (string s);
-
 /**
  * @brief Converts a lowercase character to uppercase.
  *

lolly/data/unicode.cpp

@@ -30,5 +30,44 @@ unicode_get_range (int code) {
   else if (code >= 0x1d400 && code <= 0x1d7ff) return "mathletters";
   else return "";
 }
+
+bool
+is_cjk_unified_ideographs (string s) {
+  int n= N (s);
+  for (int i= 0; i < n; i++)
+    if (s[i] == '<' && i + 1 < n && s[i + 1] == '#') {
+      int start= i + 2;
+      i        = i + 2;
+      while (i < n && s[i] != '>')
+        i++;
+      string r= s (start, i);
+      if ("4E00" <= r && r <= "9FBF") continue;
+      else return false;
+    }
+    else {
+      return false;
+    }
+  return true;
+}
+
+bool
+has_cjk_unified_ideographs (string s) {
+  int n= N (s);
+  for (int i= 0; i < n; i++)
+    if (s[i] == '<' && i + 1 < n && s[i + 1] == '#') {
+      int start= i + 2;
+      i        = i + 2;
+      while (i < n && s[i] != '>')
+        i++;
+      string r= s (start, i);
+      if ("4E00" <= r && r <= "9FBF") return true;
+      else continue;
+    }
+    else {
+      continue;
+    }
+  return false;
+}
+
 } // namespace data
 } // namespace lolly

lolly/data/unicode.hpp

@@ -16,5 +16,27 @@
 namespace lolly {
 namespace data {
 string unicode_get_range (int code);
-}
+
+/**
+ * @brief Checks if a string contains only CJK Unified Ideographs.
+ *
+ * @param s The string to check.
+ * @return True if all characters in the string are CJK Unified Ideographs,
+ * otherwise false.
+ * @note This function expects the CJK Unified Ideographs to be in a specific
+ * encoded format.
+ */
+bool is_cjk_unified_ideographs (string s);
+
+/**
+ * @brief Checks if a string contains any CJK Unified Ideographs.
+ *
+ * @param s The string to check.
+ * @return True if the string contains at least one CJK Unified Ideograph,
+ * otherwise false.
+ * @note This function expects the CJK Unified Ideographs to be in a specific
+ * encoded format.
+ */
+bool has_cjk_unified_ideographs (string s);
+} // namespace data
 } // namespace lolly

tests/Kernel/Types/analyze_test.cpp

@@ -75,13 +75,6 @@ TEST_CASE ("as_hex") {
   }
 }
 
-TEST_CASE ("cjk_unified_ideographs") {
-  CHECK (is_cjk_unified_ideographs ("<#4E2D>"));
-  CHECK (has_cjk_unified_ideographs ("<#4E2D>"));
-  CHECK (has_cjk_unified_ideographs ("bib-<#4E2D>"));
-  CHECK (!is_cjk_unified_ideographs ("bib-<#4E2D>"));
-}
-
 TEST_CASE ("test locase all") {
   CHECK_EQ (locase_all (string ("true")) == string ("true"), true);
   CHECK_EQ (locase_all (string ("TRue")) == string ("true"), true);

tests/lolly/data/unicode_test.cpp

@@ -12,10 +12,19 @@
 #include "a_tbox_main.cpp"
 #include "lolly/data/unicode.hpp"
 
+using lolly::data::has_cjk_unified_ideographs;
+using lolly::data::is_cjk_unified_ideographs;
 using lolly::data::unicode_get_range;
 
 TEST_CASE ("unicode_get_range") {
   string_eq (unicode_get_range ((int) 'a'), "ascii");
   string_eq (unicode_get_range (0x2460), "enclosed_alphanumerics"); // ①
   string_eq (unicode_get_range (0x24ff), "enclosed_alphanumerics"); // ⓿
 }
+
+TEST_CASE ("cjk_unified_ideographs") {
+  CHECK (is_cjk_unified_ideographs ("<#4E2D>"));
+  CHECK (has_cjk_unified_ideographs ("<#4E2D>"));
+  CHECK (has_cjk_unified_ideographs ("bib-<#4E2D>"));
+  CHECK (!is_cjk_unified_ideographs ("bib-<#4E2D>"));
+}