Make source code character checking smarter.

3 files changed, 74 insertions, 12 deletions

diff --git a/ChangeLog b/ChangeLog
index 35e08c91..3d8affd7 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2014-07-10         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* awkgram.y (check_for_bad): New routine to do the fatal message,
+	with smarter checking.
+	(nextc): Call it as appropriate.
+
 2014-07-03         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* awkgram.y (nextc): Add bool check_for_bad parameter to check
diff --git a/awkgram.c b/awkgram.c
index 94ca1313..711c2e37 100644
--- a/awkgram.c
+++ b/awkgram.c
@@ -5130,6 +5130,34 @@ tokexpand()
 	return tok;
 }
 
+/* check_bad_char --- fatal if c isn't allowed in gawk source code */
+
+/*
+ * The error message was inspired by someone who decided to put
+ * a physical \0 byte into the source code to see what would
+ * happen and then filed a bug report about it.  Sigh.
+ */
+
+static void
+check_bad_char(int c)
+{
+	/* allow escapes. needed for autoconf. bleah. */
+	switch (c) {
+	case '\a':
+	case '\b':
+	case '\f':
+	case '\n':
+	case '\r':
+	case '\t':
+		return;
+	default:
+		break;
+	}
+
+	if (iscntrl(c) && ! isspace(c))
+		fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), c);
+}
+
 /* nextc --- get the next input character */
 
 #if MBS_SUPPORT
@@ -5186,8 +5214,8 @@ again:
 				0 : work_ring_idx + 1;
 			cur_char_ring[work_ring_idx] = 0;
 		}
-		if (check_for_bad && iscntrl(*lexptr) && ! isspace(*lexptr))
-			fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), *lexptr);
+		if (check_for_bad)
+			check_bad_char(*lexptr);
 
 		return (int) (unsigned char) *lexptr++;
 	} else {
@@ -5195,8 +5223,8 @@ again:
 			if (lexeof)
 				return END_FILE;
 			if (lexptr && lexptr < lexend) {
-				if (check_for_bad && iscntrl(*lexptr) && ! isspace(*lexptr))
-					fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), *lexptr);
+				if (check_for_bad)
+					check_bad_char(*lexptr);
 				return ((int) (unsigned char) *lexptr++);
 			}
 		} while (get_src_buf());
@@ -5213,8 +5241,8 @@ nextc(bool check_for_bad)
 		if (lexeof)
 			return END_FILE;
 		if (lexptr && lexptr < lexend) {
-			if (check_for_bad && iscntrl(*lexptr) && ! isspace(*lexptr))
-				fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), *lexptr);
+			if (check_for_bad)
+				check_bad_char(*lexptr);
 			return ((int) (unsigned char) *lexptr++);
 		}
 	} while (get_src_buf());
diff --git a/awkgram.y b/awkgram.y
index b512584c..1a199469 100644
--- a/awkgram.y
+++ b/awkgram.y
@@ -2791,6 +2791,34 @@ tokexpand()
 	return tok;
 }
 
+/* check_bad_char --- fatal if c isn't allowed in gawk source code */
+
+/*
+ * The error message was inspired by someone who decided to put
+ * a physical \0 byte into the source code to see what would
+ * happen and then filed a bug report about it.  Sigh.
+ */
+
+static void
+check_bad_char(int c)
+{
+	/* allow escapes. needed for autoconf. bleah. */
+	switch (c) {
+	case '\a':
+	case '\b':
+	case '\f':
+	case '\n':
+	case '\r':
+	case '\t':
+		return;
+	default:
+		break;
+	}
+
+	if (iscntrl(c) && ! isspace(c))
+		fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), c);
+}
+
 /* nextc --- get the next input character */
 
 #if MBS_SUPPORT
@@ -2847,8 +2875,8 @@ again:
 				0 : work_ring_idx + 1;
 			cur_char_ring[work_ring_idx] = 0;
 		}
-		if (check_for_bad && iscntrl(*lexptr) && ! isspace(*lexptr))
-			fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), *lexptr);
+		if (check_for_bad)
+			check_bad_char(*lexptr);
 
 		return (int) (unsigned char) *lexptr++;
 	} else {
@@ -2856,8 +2884,8 @@ again:
 			if (lexeof)
 				return END_FILE;
 			if (lexptr && lexptr < lexend) {
-				if (check_for_bad && iscntrl(*lexptr) && ! isspace(*lexptr))
-					fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), *lexptr);
+				if (check_for_bad)
+					check_bad_char(*lexptr);
 				return ((int) (unsigned char) *lexptr++);
 			}
 		} while (get_src_buf());
@@ -2874,8 +2902,8 @@ nextc(bool check_for_bad)
 		if (lexeof)
 			return END_FILE;
 		if (lexptr && lexptr < lexend) {
-			if (check_for_bad && iscntrl(*lexptr) && ! isspace(*lexptr))
-				fatal(_("PEBKAC error: invalid character '\\%03o' in source code"), *lexptr);
+			if (check_for_bad)
+				check_bad_char(*lexptr);
 			return ((int) (unsigned char) *lexptr++);
 		}
 	} while (get_src_buf());