Improve consistency of the substitution command further.

When the opening square bracket ('[') is abused as the delimiter, the regular
expression contains a bracket expression, and the bracket expression contains
another opening square bracket (sick! - i mean, sic!), then do not require
escaping that innermost bracket and treat a preceding backslash as a literal
backslash character, in accordance with POSIX:

   $ printf 'x[x\\x\n' | sed 's[\[[][R[g'
  xRx\x
   $ printf 'x[x\\x\n' | sed 's[\[\[][R[g'
  xRxRx

While here, also make the implementation more readable and insert
some comments.

Joint work with martijn@ (started during g2k18) and OK martijn@.

1 files changed, 32 insertions, 20 deletions

diff --git a/usr.bin/sed/compile.c b/usr.bin/sed/compile.c
index f15dd9f9303..64e31cd9b44 100644
--- a/usr.bin/sed/compile.c
+++ b/usr.bin/sed/compile.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: compile.c,v 1.48 2018/07/09 09:43:54 schwarze Exp $	*/
+/*	$OpenBSD: compile.c,v 1.49 2018/08/14 18:10:09 schwarze Exp $	*/
 
 /*-
  * Copyright (c) 1992 Diomidis Spinellis.
@@ -362,28 +362,40 @@ compile_delimited(char *p, char *d)
 		error(COMPILE, "\\ can not be used as a string delimiter");
 	else if (c == '\n')
 		error(COMPILE, "newline can not be used as a string delimiter");
-	while (*p) {
-		if (*p == '[' && *p != c) {
-			if ((d = compile_ccl(&p, d)) == NULL)
-				error(COMPILE, "unbalanced brackets ([])");
-			continue;
-		} else if (*p == '\\' && p[1] == c) {
-			p++;
-		} else if (*p == '\\' && p[1] == '[') {
-			*d++ = *p++;
-		} else if (*p == '\\' && p[1] == 'n') {
-			*d++ = '\n';
-			p += 2;
-			continue;
-		} else if (*p == '\\' && p[1] == '\\') {
-			*d++ = *p++;
-		} else if (*p == c) {
+
+	while (p[0]) {
+		/* Unescaped delimiter: We are done. */
+		if (p[0] == c) {
 			*d = '\0';
-			return (p + 1);
+			return p + 1;
 		}
-		*d++ = *p++;
+		if (p[0] == '\\') {
+			/* Escaped delimiter: Skip the backslash. */
+			if (p[1] == c) {
+				p++;
+			} else {
+				/* Backslash-n: Match linefeed. */
+				if (p[1] == 'n') {
+					*d++ = '\n';
+					p += 2;
+				/* Other escapes remain unchanged. */
+				} else {
+					*d++ = *p++;
+					*d++ = *p++;
+				}
+				continue;
+			}
+		}
+		if (p[0] != '[')
+			*d++ = *p++;
+		/*
+		 * Bracket expression:
+		 * It may contain the delimiter without escaping.
+		 */
+		else if ((d = compile_ccl(&p, d)) == NULL)
+			error(COMPILE, "unbalanced brackets ([])");
 	}
-	return (NULL);
+	return NULL;
 }