Scanner and parser use different tables for constants

Moreover, each function being parsed has its own table.

The code is cleaner when each table is used for one specific purpose:
The scanner uses its table to anchor and unify strings, mapping strings
to themselves; the parser uses it to reuse constants in the code,
mapping constants to their indices in the constant table. A different
table for each task avoids false collisions.

ltable.c

@@ -962,7 +962,7 @@ lu_byte luaH_getint (Table *t, lua_Integer key, TValue *res) {
 */
 const TValue *luaH_Hgetshortstr (Table *t, TString *key) {
   Node *n = hashstr(t, key);
-  lua_assert(key->tt == LUA_VSHRSTR);
+  lua_assert(strisshr(key));
   for (;;) {  /* check whether 'key' is somewhere in the chain */
     if (keyisshrstr(n) && eqshrstr(keystrval(n), key))
       return gval(n);  /* that's it */
@@ -997,15 +997,6 @@ lu_byte luaH_getstr (Table *t, TString *key, TValue *res) {
 }
 
 
-TString *luaH_getstrkey (Table *t, TString *key) {
-  const TValue *o = Hgetstr(t, key);
-  if (!isabstkey(o))  /* string already present? */
-    return keystrval(nodefromval(o));  /* get saved copy */
-  else
-    return NULL;
-}
-
-
 /*
 ** main search function
 */