Add more backslash sequences: \b, \e, \f, \h, \r, \t, \v and negations.

perl6 · Oct 9, 2009 · 1f30c77 · 1f30c77
1 parent cef1481
commit 1f30c77
Show file tree

Hide file tree

Showing 4 changed files with 59 additions and 8 deletions.
diff --git a/STATUS b/STATUS
@@ -1,4 +1,4 @@
-2009-10-08:  
+2009-10-09:  
 
 At the moment, nqp-rx is configured to build an executable called
 "p6regex", which is a Perl 6 regular expression compiler for Parrot.
@@ -41,7 +41,7 @@ The key files for the p6regex compiler are:
     src/Regex/P6Regex/Actions.pm     # actions to create PAST from parse
 
 
-Things that work (2009-10-08, 22h23 UTC):
+Things that work (2009-10-09, 08h07 UTC):
 
 * bare literal strings
 * quantifiers  *, +, ?, *:, +:, ?:, *?, +?, ??, *!, +!, ?!
@@ -52,4 +52,6 @@ Things that work (2009-10-08, 22h23 UTC):
 * anchors ^, ^^, $, $$, <<, >>
 * backslash-quoted punctuation
 * #-comments (mostly)
-* obsolete backslash sequences \A \Z \z \Q \E
+* obsolete backslash sequences \A \Z \z \Q
+* \b, \B, \e, \E, \f, \F, \h, \H, \r, \R, \t, \T, \v, \V
+* enumerated character lists <[ab0..9]>
diff --git a/src/PAST/Compiler-Regex.pir b/src/PAST/Compiler-Regex.pir
@@ -477,16 +477,16 @@ character list.
 
     .local string charlist
     charlist = node[0]
+    charlist = self.'escape'(charlist)
     .local pmc negate, testop
     negate = node.'negate'()
     testop = self.'??!!'(negate, 'ge', 'lt')
 
-    ops.'push_pirop'('inline', charlist, negate, 'inline'=>'  # rx enumcharlist %0 negate=%1')
+    ops.'push_pirop'('inline', negate, 'inline'=>'  # rx enumcharlist negate=%0')
     ops.'push_pirop'('ge', pos, eos, fail)
     ops.'push_pirop'('sub', '$I10', pos, off)
     ops.'push_pirop'('substr', '$S10', tgt, '$I10', 1)
-    $S0 = self.'escape'(charlist)
-    ops.'push_pirop'('index', '$I11', $S0, '$S10')
+    ops.'push_pirop'('index', '$I11', charlist, '$S10')
     ops.'push_pirop'(testop, '$I11', 0, fail)
     ops.'push_pirop'('inc', pos)
     .return (ops)

diff --git a/src/Regex/P6Regex/Actions.pm b/src/Regex/P6Regex/Actions.pm
@@ -136,6 +136,49 @@ method backslash:sym<w>($/) {
     make $past;
 }
 
+method backslash:sym<b>($/) {
+    my $past := PAST::Regex.new( "\b", :pasttype('enumcharlist'),
+                    :negate($<sym> eq 'B'));
+    make $past;
+}
+
+method backslash:sym<e>($/) {
+    my $past := PAST::Regex.new( "\e", :pasttype('enumcharlist'),
+                    :negate($<sym> eq 'E'));
+    make $past;
+}
+
+method backslash:sym<f>($/) {
+    my $past := PAST::Regex.new( "\f", :pasttype('enumcharlist'), 
+                    :negate($<sym> eq 'F'));
+    make $past;
+}
+
+method backslash:sym<h>($/) {
+    my $past := PAST::Regex.new( "\x[09,20,a0,1680,180e,2000,2001,2002,2003,2004,2005,2006,2007,2008,2009,200a,202f,205f,3000]", :pasttype('enumcharlist'),
+                    :negate($<sym> eq 'H'));
+    make $past;
+}
+
+method backslash:sym<r>($/) {
+    my $past := PAST::Regex.new( "\r", :pasttype('enumcharlist'), 
+                    :negate($<sym> eq 'R'));
+    make $past;
+}
+
+method backslash:sym<t>($/) {
+    my $past := PAST::Regex.new( "\t", :pasttype('enumcharlist'), 
+                    :negate($<sym> eq 'T'));
+    make $past;
+}
+
+method backslash:sym<v>($/) {
+    my $past := PAST::Regex.new( "\x[0a,0b,0c,0d,85,2028,2029]", :pasttype('enumcharlist'),
+                    :negate($<sym> eq 'V'));
+    make $past;
+}
+
+
 method backslash:sym<misc>($/) {
     my $past := PAST::Regex.new( ~$/ , :pasttype('literal') );
     make $past;
@@ -173,6 +216,6 @@ method cclass_elem($/) {
         else { $str := $str ~ $_[0]; }
     }
     my $past := PAST::Regex.new( $str, :pasttype('enumcharlist') );
-    if $<sign> eq '-' { $past.negate(1); }
+    $past.negate( $<sign> eq '-' );
     make $past;
 }
diff --git a/src/Regex/P6Regex/Grammar.pm b/src/Regex/P6Regex/Grammar.pm
@@ -64,10 +64,16 @@ grammar Regex::P6Regex::Grammar is PCT::Grammar;
 
     # proto token backslash { <...> }
     token backslash:sym<w> { $<sym>:=[<[dswnDSWN]>] {*} }
+    token backslash:sym<b> { $<sym>:=[<[bB]>] {*} }
+    token backslash:sym<e> { $<sym>:=[<[eE]>] {*} }
+    token backslash:sym<f> { $<sym>:=[<[fF]>] {*} }
+    token backslash:sym<h> { $<sym>:=[<[hH]>] {*} }
+    token backslash:sym<r> { $<sym>:=[<[rR]>] {*} }
+    token backslash:sym<t> { $<sym>:=[<[tT]>] {*} }
+    token backslash:sym<v> { $<sym>:=[<[vV]>] {*} }
     token backslash:sym<A> { 'A' <.obs: '\\A as beginning-of-string matcher;^'> }
     token backslash:sym<z> { 'z' <.obs: '\\z as end-of-string matcher;$'> }
     token backslash:sym<Z> { 'Z' <.obs: '\\Z as end-of-string matcher;\\n?$'> }
-    token backslash:sym<E> { 'E' <.obs: '\\E as quotemeta;quotes or literal variable match'> }
     token backslash:sym<Q> { 'Q' <.obs: '\\Q as quotemeta;quotes or literal variable match'> }
     token backslash:sym<misc> { \W {*} }