constify character property tables

* enc/jis/props.kwd: constify character property tables of JIS
  based encodings by perfect hash.
* enc/euc_jp.c, enc/shift_jis.c: use character property functions.

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46039 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

1 files changed, 52 insertions, 0 deletions

diff --git a/enc/jis/props.src b/enc/jis/props.src
new file mode 100644
index 0000000000..b2ddde50c9
--- /dev/null
+++ b/enc/jis/props.src
@@ -0,0 +1,52 @@
+%{/* -*- c -*- */
+#define GPERF_DOWNCASE 1
+#define GPERF_CASE_STRNCMP 1
+
+static inline int
+gperf_case_strncmp(const char *s1, const char *s2, unsigned int n)
+{
+    const UChar *str = (const UChar *)s1;
+    const UChar *s = (const UChar *)s2;
+    return onigenc_with_ascii_strnicmp(ONIG_ENCODING_ASCII, str, str + n, s, n);
+}
+
+enum onigenc_jis_ctype {
+    onigenc_jis_min = ONIGENC_MAX_STD_CTYPE,
+    onigenc_jis_hiragana,
+    onigenc_jis_katakana,
+    onigenc_jis_han,
+    onigenc_jis_latin,
+    onigenc_jis_greek,
+    onigenc_jis_cyrillic,
+    onigenc_jis_max
+};
+
+enum {PropertyListNum = onigenc_jis_max - onigenc_jis_min - 1};
+
+static const OnigCodePoint* PropertyList[PropertyListNum] = {
+    CR_Hiragana,
+    CR_Katakana,
+    CR_Han,
+    CR_Latin,
+    CR_Greek,
+    CR_Cyrillic,
+};
+
+struct enc_property {
+    signed char name;
+    unsigned char ctype;
+};
+
+static const struct enc_property *onig_jis_property(const char *str, unsigned int len);
+%}
+
+struct enc_property;
+
+%%
+hiragana, onigenc_jis_hiragana
+katakana, onigenc_jis_katakana
+han,      onigenc_jis_han
+latin,    onigenc_jis_latin
+greek,    onigenc_jis_greek
+cyrillic, onigenc_jis_cyrillic
+%%