[ruby/prism] EUC-KR encodings

https://github.com/ruby/prism/commit/ba5218385a

4 files changed, 89 insertions, 1 deletions

diff --git a/prism/enc/pm_encoding.h b/prism/enc/pm_encoding.h
index e81ecad25b..3bea57fbf9 100644
--- a/prism/enc/pm_encoding.h
+++ b/prism/enc/pm_encoding.h
@@ -169,7 +169,10 @@ extern pm_encoding_t pm_encoding_cp950;
 extern pm_encoding_t pm_encoding_euc_jp;
 extern pm_encoding_t pm_encoding_euc_jp_ms;
 extern pm_encoding_t pm_encoding_euc_jis_2004;
+extern pm_encoding_t pm_encoding_euc_kr;
+extern pm_encoding_t pm_encoding_gb12345;
 extern pm_encoding_t pm_encoding_gb1988;
+extern pm_encoding_t pm_encoding_gb2312;
 extern pm_encoding_t pm_encoding_gbk;
 extern pm_encoding_t pm_encoding_ibm437;
 extern pm_encoding_t pm_encoding_ibm720;
diff --git a/prism/enc/pm_euc_jp.c b/prism/enc/pm_euc_jp.c
index 9bee6a1292..a33ba4f0a5 100644
--- a/prism/enc/pm_euc_jp.c
+++ b/prism/enc/pm_euc_jp.c
@@ -97,3 +97,81 @@ pm_encoding_t pm_encoding_cp51932 = {
     .isupper_char = pm_encoding_euc_jp_isupper_char,
     .multibyte = true
 };
+
+static size_t
+pm_encoding_euc_kr_char_width(const uint8_t *b, ptrdiff_t n) {
+    // These are the single byte characters.
+    if (*b < 0x80) {
+        return 1;
+    }
+
+    // These are the double byte characters.
+    if (
+        (n > 1) &&
+        (
+            (b[0] >= 0xA1 && b[0] <= 0xFE) &&
+            (b[1] >= 0xA1 && b[1] <= 0xFE)
+        )
+    ) {
+        return 2;
+    }
+
+    return 0;
+}
+
+static size_t
+pm_encoding_euc_kr_alpha_char(const uint8_t *b, ptrdiff_t n) {
+    if (pm_encoding_euc_kr_char_width(b, n) == 1) {
+        return pm_encoding_ascii_alpha_char(b, n);
+    } else {
+        return 0;
+    }
+}
+
+static size_t
+pm_encoding_euc_kr_alnum_char(const uint8_t *b, ptrdiff_t n) {
+    if (pm_encoding_euc_kr_char_width(b, n) == 1) {
+        return pm_encoding_ascii_alnum_char(b, n);
+    } else {
+        return 0;
+    }
+}
+
+static bool
+pm_encoding_euc_kr_isupper_char(const uint8_t *b, ptrdiff_t n) {
+    if (pm_encoding_euc_kr_char_width(b, n) == 1) {
+        return pm_encoding_ascii_isupper_char(b, n);
+    } else {
+        return 0;
+    }
+}
+
+/** EUC-KR encoding */
+pm_encoding_t pm_encoding_euc_kr = {
+    .name = "EUC-KR",
+    .char_width = pm_encoding_euc_kr_char_width,
+    .alnum_char = pm_encoding_euc_kr_alnum_char,
+    .alpha_char = pm_encoding_euc_kr_alpha_char,
+    .isupper_char = pm_encoding_euc_kr_isupper_char,
+    .multibyte = true
+};
+
+/** GB2312 encoding */
+pm_encoding_t pm_encoding_gb2312 = {
+    .name = "GB2312",
+    .char_width = pm_encoding_euc_kr_char_width,
+    .alnum_char = pm_encoding_euc_kr_alnum_char,
+    .alpha_char = pm_encoding_euc_kr_alpha_char,
+    .isupper_char = pm_encoding_euc_kr_isupper_char,
+    .multibyte = true
+};
+
+/** GB12345 encoding */
+pm_encoding_t pm_encoding_gb12345 = {
+    .name = "GB12345",
+    .char_width = pm_encoding_euc_kr_char_width,
+    .alnum_char = pm_encoding_euc_kr_alnum_char,
+    .alpha_char = pm_encoding_euc_kr_alpha_char,
+    .isupper_char = pm_encoding_euc_kr_isupper_char,
+    .multibyte = true
+};
diff --git a/prism/prism.c b/prism/prism.c
index 4679ebb4ed..e18244ddf4 100644
--- a/prism/prism.c
+++ b/prism/prism.c
@@ -6250,14 +6250,18 @@ parser_lex_magic_comment_encoding_value(pm_parser_t *parser, const uint8_t *star
                 ENCODING2("EUC-JP", "eucJP", pm_encoding_euc_jp);
                 ENCODING2("eucJP-ms", "euc-jp-ms", pm_encoding_euc_jp_ms);
                 ENCODING2("EUC-JIS-2004", "EUC-JISX0213", pm_encoding_euc_jis_2004);
+                ENCODING2("EUC-KR", "eucKR", pm_encoding_euc_kr);
+                ENCODING2("EUC-CN", "eucCN", pm_encoding_gb2312);
                 ENCODING1("external", pm_encoding_utf_8);
                 break;
             case 'F': case 'f':
                 ENCODING1("filesystem", pm_encoding_utf_8);
                 break;
             case 'G': case 'g':
-                ENCODING1("GB1988", pm_encoding_gb1988);
                 ENCODING1("GBK", pm_encoding_gbk);
+                ENCODING1("GB12345", pm_encoding_gb12345);
+                ENCODING1("GB1988", pm_encoding_gb1988);
+                ENCODING1("GB2312", pm_encoding_gb2312);
                 break;
             case 'I': case 'i':
                 ENCODING1("IBM437", pm_encoding_ibm437);
diff --git a/test/prism/encoding_test.rb b/test/prism/encoding_test.rb
index b206ab20e5..eddb0b294d 100644
--- a/test/prism/encoding_test.rb
+++ b/test/prism/encoding_test.rb
@@ -74,7 +74,10 @@ module Prism
       Encoding::Big5_UAO =>      codepoints_2bytes,
       Encoding::CP949 =>         codepoints_2bytes,
       Encoding::CP950 =>         codepoints_2bytes,
+      Encoding::EUC_KR =>        codepoints_2bytes,
       Encoding::GBK =>           codepoints_2bytes,
+      Encoding::GB12345 =>       codepoints_2bytes,
+      Encoding::GB2312 =>        codepoints_2bytes,
       Encoding::MACJAPANESE =>   codepoints_2bytes,
       Encoding::Shift_JIS =>     codepoints_2bytes,
       Encoding::SJIS_DoCoMo =>   codepoints_2bytes,