From: NeilBrown <neil@brown.name>
Date: Sat, 29 Aug 2020 08:15:52 +0000 (+1000)
Subject: Introduce --non-space option
X-Git-Tag: v1.3~14
X-Git-Url: http://git.neil.brown.name/?a=commitdiff_plain;h=b78945cad0c9901a2b7b2792343cca5bca575187;p=wiggle.git

Introduce --non-space option

This can significantly reduce the number of words by treating
punctuation as part of the surrounding word, rather than as single-char
words.
Fewer words can mean much faster comparisons.

Signed-off-by: NeilBrown <neil@brown.name>
---

diff --git a/ReadMe.c b/ReadMe.c
index 05d8ec2..847e697 100644
--- a/ReadMe.c
+++ b/ReadMe.c
@@ -58,9 +58,10 @@ struct option long_options[] = {
 	{"no-ignore",	0, 0, 'i'},
 	{"show-wiggles",0, 0, 'W'},
 	{"ignore-blanks",0,0, 'b'},
-	{"no-backup",   0, 0, NO_BACKUP },
-	{"self-test",   0, 0, SELF_TEST},
+	{"no-backup",	0, 0, NO_BACKUP },
+	{"self-test",	0, 0, SELF_TEST},
 	{"report-wiggles", 0, 0, REPORT_WIGGLES},
+	{"non-space",	0, 0, NON_SPACE},
 	{0, 0, 0, 0}
 };
 
@@ -84,6 +85,7 @@ char Help[] =  "\n"
 "\n"
 "   --words     -w    : word-wise diff and merge.\n"
 "   --lines     -l    : line-wise diff and merge.\n"
+"   --non-space       : words are separated by spaces.\n"
 "\n"
 "   --patch     -p    : treat last file as a patch file.\n"
 "   -1  -2  -3        : select which component of patch or merge to use.\n"
diff --git a/split.c b/split.c
index 27f9b7e..a9d66b0 100644
--- a/split.c
+++ b/split.c
@@ -58,7 +58,7 @@ static int split_internal(char *start, char *end, int type,
 		char *cp2;
 		int prefix = 0;
 
-		if (type == (ByWord | IgnoreBlanks))
+		if ((type & ByWord) && (type & IgnoreBlanks))
 			while (cp < end &&
 			       (*cp == ' ' || *cp == '\t')) {
 				prefix++;
@@ -79,24 +79,28 @@ static int split_internal(char *start, char *end, int type,
 					cp++;
 				break;
 			case ByWord:
-				if (isalnum(*cp) || *cp == '_') {
+				if (*cp == ' ' || *cp == '\t') {
 					do
 						cp++;
 					while (cp < end
-					       && (isalnum(*cp)
-						   || *cp == '_'));
-				} else if (*cp == ' ' || *cp == '\t') {
+					       && (*cp == ' '
+						   || *cp == '\t'));
+				} else if ((type & WholeWord) ||
+					   isalnum(*cp) || *cp == '_') {
 					do
 						cp++;
 					while (cp < end
-					       && (*cp == ' '
-						   || *cp == '\t'));
+					       && (((type & WholeWord)
+						    && *cp != ' ' && *cp != '\t'
+						    && *cp != '\n')
+						   || isalnum(*cp)
+						   || *cp == '_'));
 				} else
 					cp++;
 				break;
 			}
 		cp2 = cp;
-		if (type == (ByWord | IgnoreBlanks) &&
+		if ((type & ByWord) && (type & IgnoreBlanks) &&
 		    *start && *start != '\n')
 			while (cp2 < end &&
 			       (*cp2 == ' ' || *cp2 == '\t' || *cp2 == '\n')) {
diff --git a/wiggle.1 b/wiggle.1
index f14166f..89047cb 100644
--- a/wiggle.1
+++ b/wiggle.1
@@ -144,6 +144,11 @@ patch to be wiggled in to place.
 Request that all operations and display be word based.  This is the
 default for the "diff" function.
 .TP
+.BR \-\-non\-space
+Request that words be defined as sequences of non-white-space.  Without
+this flag words are sequences of alphanumerics or single non-white-space
+characters.
+.TP
 .BR \-l ", " \-\-lines
 Request that all operations and display be line based.
 .TP
diff --git a/wiggle.c b/wiggle.c
index aef2d75..2347ffe 100644
--- a/wiggle.c
+++ b/wiggle.c
@@ -759,6 +759,10 @@ int main(int argc, char *argv[])
 				Cmd, mode, opt);
 			exit(2);
 
+		case NON_SPACE:
+			ignore_blanks |= WholeWord;
+			continue;
+
 		case 'w':
 		case 'l':
 			if (obj == 0 || obj == opt) {
@@ -784,7 +788,7 @@ int main(int argc, char *argv[])
 			continue;
 
 		case 'b':
-			ignore_blanks = IgnoreBlanks;
+			ignore_blanks |= IgnoreBlanks;
 			continue;
 
 		case 'i':
diff --git a/wiggle.h b/wiggle.h
index 1befdd5..dfde4b7 100644
--- a/wiggle.h
+++ b/wiggle.h
@@ -199,8 +199,9 @@ extern char short_options[];
 extern struct option long_options[];
 enum other_options {
 	SELF_TEST = 300,
-	REPORT_WIGGLES = 301,
+	REPORT_WIGGLES,
 	NO_BACKUP,
+	NON_SPACE,
 };
 extern char Usage[];
 extern char Help[];
@@ -216,4 +217,5 @@ enum {
 	ByWord = 1,
 	ByMask = 3,
 	IgnoreBlanks = 8, /* 'or'ed in */
+	WholeWord = 16,
 };