unexpand: fix incorrect expansion, add test for it

function                                             old     new   delta
expand_main                                          676     656     -20

coreutils/expand.c

@@ -68,51 +68,39 @@ static void expand(FILE *file, int tab_size, unsigned opt)
 static void unexpand(FILE *file, unsigned tab_size, unsigned opt)
 {
 	char *line;
-	char *ptr;
-	int convert;
-	int pos;
-	int i = 0;
-	unsigned column = 0;
 
 	while ((line = xmalloc_fgets(file)) != NULL) {
-		convert = 1;
+		char *ptr = line;
-		pos = 0;
+		unsigned column = 0;
-		ptr = line;
+
-		while (*line) {
+		while (*ptr) {
-			while ((*line == ' ' || *line == '\t') && convert) {
+			unsigned n;
-				pos += (*line == ' ') ? 1 : tab_size;
+
-				line++;
+			while (*ptr == ' ') {
 				column++;
-				if ((opt & OPT_ALL) && column == tab_size) {
+				ptr++;
-					column = 0;
-					goto put_tab;
-				}
 			}
-			if (pos) {
+			if (*ptr == '\t') {
-				i = pos / tab_size;
+				column += tab_size - (column % tab_size);
-				if (i) {
+				ptr++;
-					for (; i > 0; i--) {
+				continue;
- put_tab:
-						bb_putchar('\t');
-					}
-				} else {
-					for (i = pos % tab_size; i > 0; i--) {
-						bb_putchar(' ');
-					}
-				}
-				pos = 0;
-			} else {
-				if (opt & OPT_INITIAL) {
-					convert = 0;
-				}
-				if (opt & OPT_ALL) {
-					column++;
-				}
-				bb_putchar(*line);
-				line++;
 			}
+
+			n = column / tab_size;
+			column = column % tab_size;
+			while (n--)
+				putchar('\t');
+
+			if ((opt & OPT_INITIAL) && ptr != line) {
+				printf("%*s%s", column, "", ptr);
+				break;
+			}
+			n = strcspn(ptr, "\t ");
+			printf("%*s%.*s", column, "", n, ptr);
+			ptr += n;
+			column = (column + n) % tab_size;
 		}
-		free(ptr);
+		free(line);
 	}
 }
 #endif

testsuite/unexpand.tests

@@ -0,0 +1,30 @@
+#!/bin/sh
+# Copyright 2008 by Denys Vlasenko
+# Licensed under GPL v2, see file LICENSE for details.
+
+. testing.sh
+
+# testing "test name" "options" "expected result" "file input" "stdin"
+
+testing "unexpand case 1" "unexpand" \
+	"\t12345678\n" "" "        12345678\n" \
+
+testing "unexpand case 2" "unexpand" \
+	"\t 12345678\n" "" "         12345678\n" \
+
+testing "unexpand case 3" "unexpand" \
+	"\t  12345678\n" "" "          12345678\n" \
+
+testing "unexpand case 4" "unexpand" \
+	"\t12345678\n" "" "       \t12345678\n" \
+
+testing "unexpand case 5" "unexpand" \
+	"\t12345678\n" "" "      \t12345678\n" \
+
+testing "unexpand case 6" "unexpand" \
+	"\t12345678\n" "" "     \t12345678\n" \
+
+testing "unexpand case 7" "unexpand" \
+	"123\t 45678\n" "" "123 \t 45678\n" \
+
+exit $FAILCOUNT