From: Neil Brown <neilb@suse.de>
Date: Wed, 24 Aug 2011 11:39:47 +0000 (+1000)
Subject: Clean up split.c
X-Git-Tag: v0.9~60
X-Git-Url: http://git.neil.brown.name/?a=commitdiff_plain;h=a584162354dfc1f9e70139f3f1f1708d54c81ffa;p=wiggle.git

Clean up split.c

Add/fix comments, remove unused function parameter.

Signed-off-by: NeilBrown <neilb@suse.de>
---

diff --git a/split.c b/split.c
index f21c642..0562310 100644
--- a/split.c
+++ b/split.c
@@ -23,21 +23,18 @@
  */
 
 /*
- * split a stream into words or line
- * When splitting into words we can either be approximate or precise.
- *  Precise mode includes every char in a word.
- *  Approximate mode excluses white-space words and might unite some special chars
+ * Split a stream into words or line
  *
- * In general, a word is one of:
+ * A word is one of:
  *    string of [A-Za-z0-9_]
  *    or string of [ \t]
- *    or single char.
+ *    or single char (i.e. punctuation and newlines).
  *
  * A line is any string that ends with \n
  *
  * As a special case to allow proper aligning of multiple chunks
- * in a patch, a word starting \0 will include 5 chars and a newline
- *
+ * in a patch, a word starting \0 will include 20 chars with a newline
+ * second from the end.
  *
  * We make two passes through the stream.
  * Firstly we count the number of item so an array can be allocated,
@@ -53,7 +50,8 @@
 
 #include "hash.h"
 
-static int split_internal(char *start, char *end, int type, struct elmnt *list, int reverse)
+static int split_internal(char *start, char *end, int type,
+			  struct elmnt *list)
 {
 	int cnt = 0;
 
@@ -63,49 +61,48 @@ static int split_internal(char *start, char *end, int type, struct elmnt *list,
 		if (*cp == '\0' && cp+19 < end && cp[18] == '\n') {
 			/* special word */
 			cp += 20;
-		} else switch (type) {
-		case ByLine:
-			while (cp < end && *cp != '\n')
-				cp++;
-			if (cp < end)
-				cp++;
-			break;
-		case ByWord:
-		case ApproxWord:
-			if (isalnum(*cp) || *cp == '_') {
-				do
+		} else
+			switch (type) {
+			case ByLine:
+				while (cp < end && *cp != '\n')
 					cp++;
-				while (cp < end && (isalnum(*cp)  || *cp == '_'));
-			} else if (*cp == ' ' || *cp == '\t') {
-				do
+				if (cp < end)
 					cp++;
-				while (cp < end && (*cp == ' ' || *cp == '\t'));
-			} else
-				cp++;
-			break;
-		}
-		if (type != ApproxWord || *start == '\0' ||
-		    (isalnum(*start) || *start == '_')) {
-			if (list) {
-				if (reverse)
-					list--;
-				list->start = start;
-				list->len = cp-start;
-				if (*start)
-					list->hash = hash_mem(start, list->len, BITS_PER_LONG);
-				else
-					list->hash = atoi(start+1);
-				if (!reverse)
-					list++;
+				break;
+			case ByWord:
+				if (isalnum(*cp) || *cp == '_') {
+					do
+						cp++;
+					while (cp < end
+					       && (isalnum(*cp)
+						   || *cp == '_'));
+				} else if (*cp == ' ' || *cp == '\t') {
+					do
+						cp++;
+					while (cp < end
+					       && (*cp == ' '
+						   || *cp == '\t'));
+				} else
+					cp++;
+				break;
 			}
-			cnt++;
+		if (list) {
+			list->start = start;
+			list->len = cp-start;
+			if (*start)
+				list->hash = hash_mem(start, list->len,
+						      BITS_PER_LONG);
+			else
+				list->hash = atoi(start+1);
+			list++;
 		}
+		cnt++;
 		start = cp;
 	}
 	return cnt;
 }
 
-struct file split_stream(struct stream s, int type, int reverse)
+struct file split_stream(struct stream s, int type)
 {
 	int cnt;
 	struct file f;
@@ -115,10 +112,9 @@ struct file split_stream(struct stream s, int type, int reverse)
 	end = s.body+s.len;
 	c = s.body;
 
-	cnt = split_internal(c, end, type, NULL, reverse);
-/*	fprintf(stderr, "cnt %d\n", cnt);*/
+	cnt = split_internal(c, end, type, NULL);
 	f.list = malloc(cnt*sizeof(struct elmnt));
 
-	f.elcnt = split_internal(c, end, type, f.list + reverse*cnt, reverse);
+	f.elcnt = split_internal(c, end, type, f.list);
 	return f;
 }
diff --git a/vpatch.c b/vpatch.c
index 26b4756..eb3fd02 100644
--- a/vpatch.c
+++ b/vpatch.c
@@ -1115,9 +1115,9 @@ static void merge_window(struct plist *p, FILE *f, int reverse)
 			sm = load_file(p->file);
 		}
 	}
-	fm = split_stream(sm, ByWord, 0);
-	fb = split_stream(sb, ByWord, 0);
-	fa = split_stream(sa, ByWord, 0);
+	fm = split_stream(sm, ByWord);
+	fb = split_stream(sb, ByWord);
+	fa = split_stream(sa, ByWord);
 
 	if (ch)
 		csl1 = pdiff(fm, fb, ch);
@@ -2014,9 +2014,9 @@ static void calc_one(struct plist *pl, FILE *f, int reverse)
 		struct file ff, fp1, fp2;
 		struct csl *csl1, *csl2;
 		struct ci ci;
-		ff = split_stream(sf, ByWord, 0);
-		fp1 = split_stream(s1, ByWord, 0);
-		fp2 = split_stream(s2, ByWord, 0);
+		ff = split_stream(sf, ByWord);
+		fp1 = split_stream(s1, ByWord);
+		fp2 = split_stream(s2, ByWord);
 		if (pl->chunks)
 			csl1 = pdiff(ff, fp1, pl->chunks);
 		else
diff --git a/wiggle.c b/wiggle.c
index 4810e9c..d9b26b6 100644
--- a/wiggle.c
+++ b/wiggle.c
@@ -430,8 +430,8 @@ int main(int argc, char *argv[])
 		}
 		if (obj == 'l') {
 			int a, b;
-			fl[0] = split_stream(flist[0], ByLine, 0);
-			fl[1] = split_stream(flist[1], ByLine, 0);
+			fl[0] = split_stream(flist[0], ByLine);
+			fl[1] = split_stream(flist[1], ByLine);
 			if (chunks2 && !chunks1)
 				csl1 = pdiff(fl[0], fl[1], chunks2);
 			else
@@ -476,8 +476,8 @@ int main(int argc, char *argv[])
 		} else {
 			int a, b;
 			int sol = 1; /* start of line */
-			fl[0] = split_stream(flist[0], ByWord, 0);
-			fl[1] = split_stream(flist[1], ByWord, 0);
+			fl[0] = split_stream(flist[0], ByWord);
+			fl[1] = split_stream(flist[1], ByWord);
 			if (chunks2 && !chunks1)
 				csl1 = pdiff(fl[0], fl[1], chunks2);
 			else
@@ -692,13 +692,13 @@ int main(int argc, char *argv[])
 		}
 
 		if (obj == 'l') {
-			fl[0] = split_stream(flist[0], ByLine, 0);
-			fl[1] = split_stream(flist[1], ByLine, 0);
-			fl[2] = split_stream(flist[2], ByLine, 0);
+			fl[0] = split_stream(flist[0], ByLine);
+			fl[1] = split_stream(flist[1], ByLine);
+			fl[2] = split_stream(flist[2], ByLine);
 		} else {
-			fl[0] = split_stream(flist[0], ByWord, 0);
-			fl[1] = split_stream(flist[1], ByWord, 0);
-			fl[2] = split_stream(flist[2], ByWord, 0);
+			fl[0] = split_stream(flist[0], ByWord);
+			fl[1] = split_stream(flist[1], ByWord);
+			fl[2] = split_stream(flist[2], ByWord);
 		}
 		if (chunks2 && !chunks1)
 			csl1 = pdiff(fl[0], fl[1], chunks2);
diff --git a/wiggle.h b/wiggle.h
index 1006ed3..a58918a 100644
--- a/wiggle.h
+++ b/wiggle.h
@@ -108,7 +108,7 @@ extern struct stream load_file(char *name);
 extern int split_patch(struct stream, struct stream*, struct stream*);
 extern int split_merge(struct stream, struct stream*, struct stream*,
 		       struct stream*);
-extern struct file split_stream(struct stream s, int type, int reverse);
+extern struct file split_stream(struct stream s, int type);
 extern struct csl *pdiff(struct file a, struct file b, int chunks);
 extern struct csl *diff(struct file a, struct file b);
 extern struct csl *diff_partial(struct file a, struct file b,
@@ -158,4 +158,3 @@ extern void cleanlist(struct file a, struct file b, struct csl *list);
 
 #define	ByLine	0
 #define	ByWord	1
-#define	ApproxWord 2