This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Extract common code to an inline function
authorKarl Williamson <public@khwilliamson.com>
Sun, 10 Feb 2013 04:23:30 +0000 (21:23 -0700)
committerKarl Williamson <public@khwilliamson.com>
Thu, 29 Aug 2013 15:55:50 +0000 (09:55 -0600)
This fairly short paradigm is repeated in several places; a later commit
will improve it.

embed.fnc
embed.h
inline.h
pp_pack.c
proto.h
sv.c
toke.c
utf8.c

index bebed86..7747453 100644 (file)
--- a/embed.fnc
+++ b/embed.fnc
@@ -2316,6 +2316,8 @@ iR        |bool   |is_utf8_common |NN const U8 *const p|NN SV **swash|NN const char * con
 sR     |SV*    |swatch_get     |NN SV* swash|UV start|UV span
 #endif
 
+AiMn   |void   |append_utf8_from_native_byte|const U8 byte|NN U8** dest
+
 Apd    |void   |sv_setsv_flags |NN SV *dstr|NULLOK SV *sstr|const I32 flags
 Apd    |void   |sv_catpvn_flags|NN SV *const dstr|NN const char *sstr|const STRLEN len \
                                |const I32 flags
diff --git a/embed.h b/embed.h
index 1329937..9d31874 100644 (file)
--- a/embed.h
+++ b/embed.h
@@ -41,6 +41,7 @@
 #define _to_utf8_upper_flags(a,b,c,d,e)        Perl__to_utf8_upper_flags(aTHX_ a,b,c,d,e)
 #define amagic_call(a,b,c,d)   Perl_amagic_call(aTHX_ a,b,c,d)
 #define amagic_deref_call(a,b) Perl_amagic_deref_call(aTHX_ a,b)
+#define append_utf8_from_native_byte   S_append_utf8_from_native_byte
 #define apply_attrs_string(a,b,c,d)    Perl_apply_attrs_string(aTHX_ a,b,c,d)
 #define atfork_lock            Perl_atfork_lock
 #define atfork_unlock          Perl_atfork_unlock
index 066edf1..63a5e06 100644 (file)
--- a/inline.h
+++ b/inline.h
@@ -203,8 +203,26 @@ S_croak_memory_wrap(void)
 
 /* ------------------------------- utf8.h ------------------------------- */
 
-/* These exist only to replace the macros they formerly were so that their use
- * can be deprecated */
+PERL_STATIC_INLINE void
+S_append_utf8_from_native_byte(const U8 byte, U8** dest)
+{
+    /* Takes an input 'byte' (Latin1 or EBCDIC) and appends it to the UTF-8
+     * encoded string at '*dest', updating '*dest' to include it */
+
+    const U8 uv = NATIVE_TO_LATIN1(byte);
+
+    PERL_ARGS_ASSERT_APPEND_UTF8_FROM_NATIVE_BYTE;
+
+    if (UNI_IS_INVARIANT(uv))
+        *(*dest)++ = UNI_TO_NATIVE(uv);
+    else {
+        *(*dest)++ = UTF8_EIGHT_BIT_HI(uv);
+        *(*dest)++ = UTF8_EIGHT_BIT_LO(uv);
+    }
+}
+
+/* These two exist only to replace the macros they formerly were so that their
+ * use can be deprecated */
 
 PERL_STATIC_INLINE bool
 S_isIDFIRST_lazy(pTHX_ const char* p)
index 3cfc03c..39f862e 100644 (file)
--- a/pp_pack.c
+++ b/pp_pack.c
@@ -357,24 +357,12 @@ S_bytes_to_uni(const U8 *start, STRLEN len, char *dest, const bool needs_swap) {
     if (UNLIKELY(needs_swap)) {
         const U8 *p = start + len;
         while (p-- > start) {
-            const UV uv = NATIVE_TO_ASCII(*p);
-            if (UNI_IS_INVARIANT(uv))
-                *dest++ = (char)(U8)UTF_TO_NATIVE(uv);
-            else {
-                *dest++ = (char)(U8)UTF8_EIGHT_BIT_HI(uv);
-                *dest++ = (char)(U8)UTF8_EIGHT_BIT_LO(uv);
-            }
+            append_utf8_from_native_byte(*p, (U8 **) & dest);
         }
     } else {
         const U8 * const end = start + len;
         while (start < end) {
-            const UV uv = NATIVE_TO_ASCII(*start);
-            if (UNI_IS_INVARIANT(uv))
-                *dest++ = (char)(U8)UTF_TO_NATIVE(uv);
-            else {
-                *dest++ = (char)(U8)UTF8_EIGHT_BIT_HI(uv);
-                *dest++ = (char)(U8)UTF8_EIGHT_BIT_LO(uv);
-            }
+            append_utf8_from_native_byte(*start, (U8 **) & dest);
             start++;
         }
     }
diff --git a/proto.h b/proto.h
index 30fcba5..0da17ec 100644 (file)
--- a/proto.h
+++ b/proto.h
@@ -112,6 +112,11 @@ PERL_CALLCONV SV * Perl_amagic_deref_call(pTHX_ SV *ref, int method)
        assert(ref)
 
 PERL_CALLCONV bool     Perl_amagic_is_enabled(pTHX_ int method);
+PERL_STATIC_INLINE void        S_append_utf8_from_native_byte(const U8 byte, U8** dest)
+                       __attribute__nonnull__(2);
+#define PERL_ARGS_ASSERT_APPEND_UTF8_FROM_NATIVE_BYTE  \
+       assert(dest)
+
 PERL_CALLCONV I32      Perl_apply(pTHX_ I32 type, SV** mark, SV** sp)
                        __attribute__nonnull__(pTHX_2)
                        __attribute__nonnull__(pTHX_3);
diff --git a/sv.c b/sv.c
index 3945ab9..8ba0505 100644 (file)
--- a/sv.c
+++ b/sv.c
@@ -3422,13 +3422,8 @@ must_be_utf8:
                }
 
                while (t < e) {
-                   const UV uv = NATIVE8_TO_UNI(*t++);
-                   if (UNI_IS_INVARIANT(uv))
-                       *d++ = (U8)UNI_TO_NATIVE(uv);
-                   else {
-                       *d++ = (U8)UTF8_EIGHT_BIT_HI(uv);
-                       *d++ = (U8)UTF8_EIGHT_BIT_LO(uv);
-                   }
+                    append_utf8_from_native_byte(*t, &d);
+                    t++;
                }
                *d = '\0';
                SvPV_free(sv); /* No longer using pre-existing string */
@@ -5200,13 +5195,8 @@ Perl_sv_catpvn_flags(pTHX_ SV *const dsv, const char *sstr, const STRLEN slen, c
        d = (U8 *)SvPVX(dsv) + dlen;
 
        while (sstr < send) {
-           const UV uv = NATIVE_TO_ASCII((U8)*sstr++);
-           if (UNI_IS_INVARIANT(uv))
-               *d++ = (U8)UTF_TO_NATIVE(uv);
-           else {
-               *d++ = (U8)UTF8_EIGHT_BIT_HI(uv);
-               *d++ = (U8)UTF8_EIGHT_BIT_LO(uv);
-           }
+            append_utf8_from_native_byte(*sstr, &d);
+           sstr++;
        }
        SvCUR_set(dsv, d-(const U8 *)SvPVX(dsv));
     }
diff --git a/toke.c b/toke.c
index 96b0274..3ce12e2 100644 (file)
--- a/toke.c
+++ b/toke.c
@@ -3176,13 +3176,7 @@ S_scan_const(pTHX_ char *start)
                    for (i = min; i <= max; i++)
 #ifdef EBCDIC
                         if (has_utf8) {
-                            const U8 ch = (U8)NATIVE_TO_UTF(i);
-                            if (UNI_IS_INVARIANT(ch))
-                                *d++ = (U8)i;
-                            else {
-                                *d++ = (U8)UTF8_EIGHT_BIT_HI(ch);
-                                *d++ = (U8)UTF8_EIGHT_BIT_LO(ch);
-                            }
+                            append_utf8_from_native_byte(i, &d);
                         }
                         else
 #endif
diff --git a/utf8.c b/utf8.c
index 1bdad1b..b445a2e 100644 (file)
--- a/utf8.c
+++ b/utf8.c
@@ -1376,13 +1376,8 @@ Perl_bytes_to_utf8(pTHX_ const U8 *s, STRLEN *len)
     dst = d;
 
     while (s < send) {
-        const UV uv = NATIVE_TO_ASCII(*s++);
-        if (UNI_IS_INVARIANT(uv))
-            *d++ = (U8)UTF_TO_NATIVE(uv);
-        else {
-            *d++ = (U8)UTF8_EIGHT_BIT_HI(uv);
-            *d++ = (U8)UTF8_EIGHT_BIT_LO(uv);
-        }
+        append_utf8_from_native_byte(*s, &d);
+        s++;
     }
     *d = '\0';
     *len = d-dst;