Bug fix: regexp should not treat \n as |

Remove a "feature" in the built-in regexp, where
a newline in the pattern was treated as alternation, like |.

Signed-off-by: Steve Bennett <steveb@workware.net.au>

2 files changed, 7 insertions, 4 deletions

diff --git a/jimregexp.c b/jimregexp.c
index 66a1ba3..6c4e7bc 100644
--- a/jimregexp.c
+++ b/jimregexp.c
@@ -43,7 +43,7 @@
  *   on 16 October 2010, to remove static state and add better Tcl ARE compatibility.
  *   This includes counted repetitions, UTF-8 support, character classes,
  *   shorthand character classes, increased number of parentheses to 100,
- *   backslash escape sequences.
+ *   backslash escape sequences. It also removes \n as an alternative to |.
  *
  * Beware that some of this code is subtly aware of the way operator
  * precedence is structured in regular expressions.  Serious changes in
@@ -342,7 +342,7 @@ static int *reg(regex_t *preg, int paren /* Parenthesized? */, int *flagp )
 	if (!(flags&HASWIDTH))
 		*flagp &= ~HASWIDTH;
 	*flagp |= flags&SPSTART;
-	while (*preg->regparse == '|' || *preg->regparse == '\n') {
+	while (*preg->regparse == '|') {
 		preg->regparse++;
 		br = regbranch(preg, &flags);
 		if (br == NULL)
@@ -395,7 +395,7 @@ static int *regbranch(regex_t *preg, int *flagp )
 	ret = regnode(preg, BRANCH);
 	chain = NULL;
 	while (*preg->regparse != '\0' && *preg->regparse != ')' &&
-	       *preg->regparse != '\n' && *preg->regparse != '|') {
+	       *preg->regparse != '|') {
 		latest = regpiece(preg, &flags);
 		if (latest == NULL)
 			return(NULL);
@@ -852,7 +852,6 @@ static int *regatom(regex_t *preg, int *flagp)
 		break;
 	case '\0':
 	case '|':
-	case '\n':
 	case ')':
 		preg->err = REG_ERR_INTERNAL;
 		return NULL;	/* Supposed to be caught earlier. */
diff --git a/tests/regexp.test b/tests/regexp.test
index a198fc8..885e44c 100644
--- a/tests/regexp.test
+++ b/tests/regexp.test
@@ -640,6 +640,10 @@ test regexp-21.13 {multiple matches handle newlines} {
     regexp -all -inline -indices -line -- ^ "a\nb\nc"
 } {{0 -1} {2 1} {4 3}}
 
+test regexp-21.14 {Literal newline in pattern} {
+    regexp -all -inline "\n(\[ \t\]+)" "\n\t\t# This  is  a test"
+} "{\n\t\t} {\t\t}"
+
 test regexp-22.1 {effect of caching} jim {
 
     set filedata {BEGIN_TABLE ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;; END_TABLE}