Fix up \cX for 5.14

Throughout 5.13 there was temporary code to deprecate and forbid
certain values of X following a \c in qq strings.  This patch fixes
this to the final 5.14 semantics.

These are:

1) a utf8 non-ASCII character will croak.  This is the same
behavior as pre-5.13, but it gives a correct error message, rather than
the malformed utf8 message previously.

2) \c{ and \cX where X is above ASCII will generate a deprecated
message.  The intent is to remove these capabilities in 5.16.  The
original agreement was to croak on above ASCII, but that does violate
our stability policy, so I'm deprecating it instead.

3) A non-deprecated warning is generated for all other \cX; this is the
same as throughout the 5.13 series.

I did not have the tuits to use \c{} as I had planned in 5.14, but \N{}
can be used instead.

diff --git a/dquote_static.c b/dquote_static.c
index b5a2ccc..e23ec46 100644 (file)
--- a/dquote_static.c
+++ b/dquote_static.c
@@ -39,26 +39,34 @@ S_regcurly(pTHX_ register const char *s)
 /* XXX Add documentation after final interface and behavior is decided */
 /* May want to show context for error, so would pass Perl_bslash_c(pTHX_ const char* current, const char* start, const bool output_warning)
     U8 source = *current;
-
-    May want to add eg, WARN_REGEX
 */
 
 STATIC char
-S_grok_bslash_c(pTHX_ const char source, const bool output_warning)
+S_grok_bslash_c(pTHX_ const char source, const bool utf8, const bool output_warning)
 {
 
     U8 result;
 
-    if (! isASCII(source)) {
-       Perl_croak(aTHX_ "Character following \"\\c\" must be ASCII");
+    if (utf8) {
+       /* Trying to deprecate non-ASCII usages.  This construct has never
+        * worked for a utf8 variant.  So, even though are accepting non-ASCII
+        * Latin1 in 5.14, no need to make them work under utf8 */
+       if (! isASCII(source)) {
+           Perl_croak(aTHX_ "Character following \"\\c\" must be ASCII");
+       }
     }
 
     result = toCTRL(source);
-    if (! isCNTRL(result)) {
+    if (! isASCII(source)) {
+           Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_SYNTAX),
+                           "Character following \"\\c\" must be ASCII");
+    }
+    else if (! isCNTRL(result) && output_warning) {
        if (source == '{') {
-           Perl_croak(aTHX_ "It is proposed that \"\\c{\" no longer be valid. It has historically evaluated to\n \";\".  If you disagree with this proposal, send email to perl5-porters@perl.org\nOtherwise, or in the meantime, you can work around this failure by changing\n\"\\c{\" to \";\"");
+           Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_SYNTAX),
+                           "\"\\c{\" is deprecated and is more clearly written as \";\"");
        }
-       else if (output_warning) {
+       else {
            U8 clearer[3];
            U8 i = 0;
            if (! isALNUM(result)) {
@@ -67,8 +75,8 @@ S_grok_bslash_c(pTHX_ const char source, const bool output_warning)
            clearer[i++] = result;
            clearer[i++] = '\0';
 
-           Perl_ck_warner_d(aTHX_ packWARN(WARN_DEPRECATED),
-                           "\"\\c%c\" more clearly written simply as \"%s\"",
+           Perl_ck_warner(aTHX_ packWARN(WARN_SYNTAX),
+                           "\"\\c%c\" is more clearly written simply as \"%s\"",
                            source,
                            clearer);
        }

diff --git a/embed.fnc b/embed.fnc
index b311bb2..233d1a4 100644 (file)
--- a/embed.fnc
+++ b/embed.fnc
@@ -657,7 +657,7 @@ p   |OP*    |localize       |NN OP *o|I32 lex
 ApdR   |I32    |looks_like_number|NN SV *const sv
 Apd    |UV     |grok_bin       |NN const char* start|NN STRLEN* len_p|NN I32* flags|NULLOK NV *result
 #ifdef PERL_IN_DQUOTE_STATIC_C
-EXMsR  |char   |grok_bslash_c  |const char source|const bool output_warning
+EXMsR  |char   |grok_bslash_c  |const char source|const bool utf8|const bool output_warning
 EXMsR  |bool   |grok_bslash_o  |NN const char* s|NN UV* uv|NN STRLEN* len|NN const char** error_msg|const bool output_warning
 #endif
 Apd    |UV     |grok_hex       |NN const char* start|NN STRLEN* len_p|NN I32* flags|NULLOK NV *result

diff --git a/embed.h b/embed.h
index 532a91c..b9cd903 100644 (file)
--- a/embed.h
+++ b/embed.h
@@ -859,7 +859,7 @@
 #    endif
 #  endif
 #  if defined(PERL_IN_DQUOTE_STATIC_C)
-#define grok_bslash_c(a,b)     S_grok_bslash_c(aTHX_ a,b)
+#define grok_bslash_c(a,b,c)   S_grok_bslash_c(aTHX_ a,b,c)
 #define grok_bslash_o(a,b,c,d,e)       S_grok_bslash_o(aTHX_ a,b,c,d,e)
 #define regcurly(a)            S_regcurly(aTHX_ a)
 #  endif

diff --git a/pod/perldiag.pod b/pod/perldiag.pod
index bf22e1e..3d35b1c 100644 (file)
--- a/pod/perldiag.pod
+++ b/pod/perldiag.pod
@@ -1258,7 +1258,12 @@ Perhaps you need to copy the value to a temporary, and repeat that.
 
 =item Character following "\c" must be ASCII
 
-(F) In C<\cI<X>>, I<X> must be an ASCII character.
+(F|W deprecated, syntax) In C<\cI<X>>, I<X> must be an ASCII character.
+It is planned to make this fatal in all instances in Perl 5.16.  In the
+cases where it isn't fatal, the character this evaluates to is
+derived by exclusive or'ing the code point of this character with 0x40.
+
+Note that non-alphabetic ASCII characters are discouraged here as well.
 
 =item Character in 'C' format wrapped in pack
 
@@ -1498,12 +1503,20 @@ valid magic number.
 you have also specified an explicit size for the string.  See
 L<perlfunc/pack>.
 
-=item "\c%c" more clearly written simply as "%s"
+=item "\c{" is deprecated and is more clearly written as ";"
+
+(D deprecated, syntax) The C<\cI<X>> construct is intended to be a way
+to specify non-printable characters.  You used it with a "{" which
+evaluates to ";", which is printable.  It is planned to remove the
+ability to specify a semi-colon this way in Perl 5.16.  Just use a
+semi-colon or a backslash-semi-colon without the "\c".
+
+=item "\c%c" is more clearly written simply as "%s"
 
-(D deprecated) The C<\cI<X>> construct is intended to be a way to specify
+(W syntax) The C<\cI<X>> construct is intended to be a way to specify
 non-printable characters.  You used it for a printable one, which is better
 written as simply itself, perhaps preceded by a backslash for non-word
-characters.  This message may not remain as Deprecated beyond 5.13.
+characters.
 
 =item Deep recursion on subroutine "%s"
 

diff --git a/pod/perlop.pod b/pod/perlop.pod
index eb71b89..e11fff2 100644 (file)
--- a/pod/perlop.pod
+++ b/pod/perlop.pod
@@ -1114,7 +1114,8 @@ L<perlebcdic/OPERATOR DIFFERENCES> for the complete list of what these
 sequences mean on both ASCII and EBCDIC platforms.
 
 Use of any other character following the "c" besides those listed above is
-discouraged, and may become deprecated or forbidden.  What happens for those
+discouraged, and some are deprecated with the intention of removing
+those in Perl 5.16.  What happens for any of these
 other characters currently though, is that the value is derived by inverting
 the 7th bit (0x40).
 

diff --git a/proto.h b/proto.h
index 0272be9..061cd06 100644 (file)
--- a/proto.h
+++ b/proto.h
@@ -5182,7 +5182,7 @@ STATIC I32        S_do_trans_simple_utf8(pTHX_ SV * const sv)
 
 #endif
 #if defined(PERL_IN_DQUOTE_STATIC_C)
-STATIC char    S_grok_bslash_c(pTHX_ const char source, const bool output_warning)
+STATIC char    S_grok_bslash_c(pTHX_ const char source, const bool utf8, const bool output_warning)
                        __attribute__warn_unused_result__;
 
 STATIC bool    S_grok_bslash_o(pTHX_ const char* s, UV* uv, STRLEN* len, const char** error_msg, const bool output_warning)

diff --git a/regcomp.c b/regcomp.c
index d986c32..43fa7f0 100644 (file)
--- a/regcomp.c
+++ b/regcomp.c
@@ -8522,7 +8522,7 @@ tryagain:
                        break;
                    case 'c':
                        p++;
-                       ender = grok_bslash_c(*p++, SIZE_ONLY);
+                       ender = grok_bslash_c(*p++, UTF, SIZE_ONLY);
                        break;
                    case '0': case '1': case '2': case '3':case '4':
                    case '5': case '6': case '7': case '8':case '9':
@@ -9287,7 +9287,7 @@ parseit:
                    goto recode_encoding;
                break;
            case 'c':
-               value = grok_bslash_c(*RExC_parse++, SIZE_ONLY);
+               value = grok_bslash_c(*RExC_parse++, UTF, SIZE_ONLY);
                break;
            case '0': case '1': case '2': case '3': case '4':
            case '5': case '6': case '7':

diff --git a/t/lib/warnings/toke b/t/lib/warnings/toke
index f37c36e..425b613 100644 (file)
--- a/t/lib/warnings/toke
+++ b/t/lib/warnings/toke
@@ -940,3 +940,19 @@ Use of ?PATTERN? without explicit operator is deprecated at - line 4.
 yes
 no
 no
+########
+# toke.c
+use warnings;
+my $a = "\c{ack}";
+$a = "\c,";
+$a = "\c`";
+no warnings 'syntax';
+$a = "\c{ack}";
+$a = "\c,";
+$a = "\c`";
+no warnings 'deprecated';
+EXPECT
+"\c{" is deprecated and is more clearly written as ";" at - line 3.
+"\c," is more clearly written simply as "l" at - line 4.
+"\c`" is more clearly written simply as "\ " at - line 5.
+"\c{" is deprecated and is more clearly written as ";" at - line 7.

diff --git a/t/op/qq.t b/t/op/qq.t
index 65e3e82..3d1bfac 100644 (file)
--- a/t/op/qq.t
+++ b/t/op/qq.t
@@ -5,7 +5,7 @@ BEGIN {
     @INC = '../lib';
 }
 
-print q(1..30
+print q(1..28
 );
 
 # This is() function is written to avoid ""
@@ -68,26 +68,6 @@ is ("a\o{120}b", "a" . chr(0x50) . "b");
 is ("a\o{400}b", "a" . chr(0x100) . "b");
 is ("a\o{1000}b", "a" . chr(0x200) . "b");
 
-# These kludged tests should change when we remove the temporary fatal error
-# in util.c for "\c{".  And, the warning there should probably not be
-# deprecated; See [perl #75138].
-# BE SURE TO remove the message from the __DATA__ section of porting/diag.t,
-# and to verify the messages in util.c are adequately covered in
-# perldiag.pod, and to revise the explanatory wording that is there now.
-my $value = eval '"\c{ACK}"';
-if ($^V lt v5.13.0 || $^V ge v5.14.0) {
-    is ($@, "");
-    is ($value, ";ACK}");
-}
-elsif ($@ ne "") {  # 5.13 series, expect the eval to fail, so pass it.
-    is ("1", "1");  # This .t only has 'is' at its disposal
-    is ("1", "1");
-}
-else {  # Something wrong; someone has removed the failure in util.c
-    is ("Should fail for 5.13 until fix test", "0");
-    is ("1", "1");
-}
-
 # This caused a memory fault
 no warnings "utf8";
 is ("abc", eval qq[qq\x{8000_0000}abc\x{8000_0000}])

diff --git a/t/porting/diag.t b/t/porting/diag.t
index 1bedf73..f001d94 100644 (file)
--- a/t/porting/diag.t
+++ b/t/porting/diag.t
@@ -432,7 +432,6 @@ Invalid type '%c' in pack
 Invalid type '%c' in %s
 Invalid type '%c' in unpack
 Invalid type ',' in %s
-It is proposed that "\c{" no longer be valid. It has historically evaluated to  ";".  If you disagree with this proposal, send email to perl5-porters@perl.org Otherwise, or in the meantime, you can work around this failure by changing "\c{" to ";"
 'j' not supported on this platform
 'J' not supported on this platform
 leaving effective gid failed

diff --git a/t/re/re_tests b/t/re/re_tests
index bc4a7fc..c80667f 100644 (file)
--- a/t/re/re_tests
+++ b/t/re/re_tests
@@ -1451,8 +1451,8 @@ abc\N{def -       c       -       \\N{NAME} must be resolved by the lexer
 # Verify works in single quotish context; regex compiler delivers slightly different msg
 # \N{U+BEEF.BEAD} succeeds here, because can't completely hide it from the outside.
 \N{U+0xBEEF}   -       c       -       Invalid hexadecimal number
-\c`    -       c       -       \"\\c`\" more clearly written simply as \"\\ \"
-\c1    -       c       -       \"\\c1\" more clearly written simply as \"q\"
+\c`    -       c       -       \"\\c`\" is more clearly written simply as \"\\ \"
+\c1    -       c       -       \"\\c1\" is more clearly written simply as \"q\"
 \cA    \001    y       $&      \1
 
 \400   \x{100} y       $&      \x{100}

diff --git a/toke.c b/toke.c
index 17152e5..8dadf61 100644 (file)
--- a/toke.c
+++ b/toke.c
@@ -3363,7 +3363,7 @@ S_scan_const(pTHX_ char *start)
            case 'c':
                s++;
                if (s < send) {
-                   *d++ = grok_bslash_c(*s++, 1);
+                   *d++ = grok_bslash_c(*s++, has_utf8, 1);
                }
                else {
                    yyerror("Missing control char name in \\c");