Improve unicode compile times

code/unicode/unicode_compile.py

@@ -38,8 +38,15 @@ with open('unicode.h', 'w') as header:
     header.write('#define Zl 26\n');
     header.write('#define Zp 27\n');
     header.write('#define Zs 28\n');
-    header.write('\n');
+    header.write(
-    header.write('#define UNI_TAB \\\n');
+'''
+struct _uni_elm {
+    wint_t code;
+    wint_t cat;
+    wint_t lower;
+    wint_t upper;
+} uni_codepoints[] = {
+''');
 
     with open('unicode_data.txt') as file:
         for line in file:
@@ -52,11 +59,11 @@ with open('unicode.h', 'w') as header:
                 lower = code
             if upper == '' or upper == '\n':
                 upper = code
-            header.write('    X(' + \
+            header.write('    {' + \
                 '0x' + code  + ', ' + \
                 cat          + ', ' + \
                 '0x' + lower + ', ' + \
-                '0x' + upper + ')\\\n');
+                '0x' + upper + '},\n');
 
-    header.write('\n');
+    header.write('};\n\n');
     header.close();

code/unicode/wctype.c

@@ -5,12 +5,7 @@
 #include "unicode.h"
 
 static inline int char_cat(wint_t wc) {
-    #define X(code, cat, l, u) case code: return cat;
+    return uni_codepoints[wc].cat;
-    switch(wc) {
-        UNI_TAB
-    }
-    #undef X
-    return -1;
 }
 
 int iswctype(wint_t wc, wctype_t desc) {
@@ -109,19 +104,9 @@ int iswxdigit(wint_t wc) {
 }
 
 wint_t towlower(wint_t wc) {
-    #define X(code, cat, l, u) case code: return l;
+    return uni_codepoints[wc].lower;
-    switch(wc) {
-        UNI_TAB
-    }
-    #undef X
-    return wc;
 }
 
 wint_t towupper(wint_t wc) {
-    #define X(code, cat, l, u) case code: return u;
+    return uni_codepoints[wc].upper;
-    switch(wc) {
-        UNI_TAB
-    }
-    #undef X
-    return wc;
 }