Diff of /trunk/mkinitrd-magellan/busybox/editors/diff.c

Parent Directory | Revision Log | Patch

	revision 983 by niro, Fri Apr 24 18:33:46 2009 UTC	revision 984 by niro, Sun May 30 11:32:42 2010 UTC
#	Line 2	Line 2
2	/*	/*
3	* Mini diff implementation for busybox, adapted from OpenBSD diff.	* Mini diff implementation for busybox, adapted from OpenBSD diff.
4	*	*
5		* Copyright (C) 2010 by Matheus Izvekov <mizvekov@gmail.com>
6	* Copyright (C) 2006 by Robert Sullivan <cogito.ergo.cogito@hotmail.com>	* Copyright (C) 2006 by Robert Sullivan <cogito.ergo.cogito@hotmail.com>
7	* Copyright (c) 2003 Todd C. Miller <Todd.Miller@courtesan.com>	* Copyright (c) 2003 Todd C. Miller <Todd.Miller@courtesan.com>
8	*	*
#	Line 12	Line 13
13	* Licensed under GPLv2 or later, see file LICENSE in this tarball for details.	* Licensed under GPLv2 or later, see file LICENSE in this tarball for details.
14	*/	*/
15
	#include "libbb.h"
	// #define FSIZE_MAX 32768
	/* NOINLINEs added to prevent gcc from merging too much into diffreg()
	* (it bites more than it can (efficiently) chew). */
	/*
	* Output flags
	*/
	enum {
	/* Print a header/footer between files */
	/* D_HEADER = 1, - unused */
	/* Treat file as empty (/dev/null) */
	D_EMPTY1 = 2 * ENABLE_FEATURE_DIFF_DIR,
	D_EMPTY2 = 4 * ENABLE_FEATURE_DIFF_DIR,
	};
16	/*	/*
17	* Status values for print_status() and diffreg() return values	* The following code uses an algorithm due to Harold Stone,
18	* Guide:	* which finds a pair of longest identical subsequences in
19	* D_SAME - files are the same	* the two files.
20	* D_DIFFER - files differ	*
21	* D_BINARY - binary files differ	* The major goal is to generate the match vector J.
22	* D_COMMON - subdirectory common to both dirs	* J[i] is the index of the line in file1 corresponding
23	* D_ONLY - file only exists in one dir	* to line i in file0. J[i] = 0 if there is no
24	* D_ISDIR1 - path1 a dir, path2 a file	* such line in file1.
25	* D_ISDIR2 - path1 a file, path2 a dir	*
26	* D_ERROR - error occurred	* Lines are hashed so as to work in core. All potential
27	* D_SKIPPED1 - skipped path1 as it is a special file	* matches are located by sorting the lines of each file
28	* D_SKIPPED2 - skipped path2 as it is a special file	* on the hash (called "value"). In particular, this
29		* collects the equivalence classes in file1 together.
30		* Subroutine equiv replaces the value of each line in
31		* file0 by the index of the first element of its
32		* matching equivalence in (the reordered) file1.
33		* To save space equiv squeezes file1 into a single
34		* array member in which the equivalence classes
35		* are simply concatenated, except that their first
36		* members are flagged by changing sign.
37		*
38		* Next the indices that point into member are unsorted into
39		* array class according to the original order of file0.
40		*
41		* The cleverness lies in routine stone. This marches
42		* through the lines of file0, developing a vector klist
43		* of "k-candidates". At step i a k-candidate is a matched
44		* pair of lines x,y (x in file0, y in file1) such that
45		* there is a common subsequence of length k
46		* between the first i lines of file0 and the first y
47		* lines of file1, but there is no such subsequence for
48		* any smaller y. x is the earliest possible mate to y
49		* that occurs in such a subsequence.
50		*
51		* Whenever any of the members of the equivalence class of
52		* lines in file1 matable to a line in file0 has serial number
53		* less than the y of some k-candidate, that k-candidate
54		* with the smallest such y is replaced. The new
55		* k-candidate is chained (via pred) to the current
56		* k-1 candidate so that the actual subsequence can
57		* be recovered. When a member has serial number greater
58		* that the y of all k-candidates, the klist is extended.
59		* At the end, the longest subsequence is pulled out
60		* and placed in the array J by unravel
61		*
62		* With J in hand, the matches there recorded are
63		* checked against reality to assure that no spurious
64		* matches have crept in due to hashing. If they have,
65		* they are broken, and "jackpot" is recorded--a harmless
66		* matter except that a true match for a spuriously
67		* mated line may now be unnecessarily reported as a change.
68		*
69		* Much of the complexity of the program comes simply
70		* from trying to minimize core utilization and
71		* maximize the range of doable problems by dynamically
72		* allocating what is needed and reusing what is not.
73		* The core requirements for problems larger than somewhat
74		* are (in words) 2*length(file0) + length(file1) +
75		* 3*(number of k-candidates installed), typically about
76		* 6n words for files of length n.
77	*/	*/
	#define D_SAME          0
	#define D_DIFFER        (1 << 0)
	#define D_BINARY        (1 << 1)
	#define D_COMMON        (1 << 2)
	/*#define D_ONLY        (1 << 3) - unused */
	#define D_ISDIR1        (1 << 4)
	#define D_ISDIR2        (1 << 5)
	#define D_ERROR         (1 << 6)
	#define D_SKIPPED1      (1 << 7)
	#define D_SKIPPED2      (1 << 8)
	/* Command line options */
	#define FLAG_a  (1 << 0)
	#define FLAG_b  (1 << 1)
	#define FLAG_d  (1 << 2)
	#define FLAG_i  (1 << 3)
	#define FLAG_L  (1 << 4)
	#define FLAG_N  (1 << 5)
	#define FLAG_q  (1 << 6)
	#define FLAG_r  (1 << 7)
	#define FLAG_s  (1 << 8)
	#define FLAG_S  (1 << 9)
	#define FLAG_t  (1 << 10)
	#define FLAG_T  (1 << 11)
	#define FLAG_U  (1 << 12)
	#define FLAG_w  (1 << 13)
78
79		#include "libbb.h"
80
81	struct cand {	#if 0
82	int x;	//#define dbg_error_msg(...) bb_error_msg(__VA_ARGS__)
83	int y;	#else
84	int pred;	#define dbg_error_msg(...) ((void)0)
85	};	#endif
86
87	struct line {	enum {                  /* print_status() and diffreg() return values */
88	int serial;	STATUS_SAME,    /* files are the same */
89	int value;	STATUS_DIFFER,  /* files differ */
90		STATUS_BINARY,  /* binary files differ */
91	};	};
92
93	/*	enum {                  /* Commandline flags */
94	* The following struct is used to record change information	FLAG_a,
95	* doing a "context" or "unified" diff.  (see routine "change" to	FLAG_b,
96	* understand the highly mnemonic field names)	FLAG_d,
97	*/	FLAG_i,
98	struct context_vec {	FLAG_L,         /* never used, handled by getopt32 */
99	int a;          /* start line in old file */	FLAG_N,
100	int b;          /* end line in old file */	FLAG_q,
101	int c;          /* start line in new file */	FLAG_r,
102	int d;          /* end line in new file */	FLAG_s,
103		FLAG_S,         /* never used, handled by getopt32 */
104		FLAG_t,
105		FLAG_T,
106		FLAG_U,         /* never used, handled by getopt32 */
107		FLAG_w,
108		FLAG_u,         /* ignored, this is the default */
109		FLAG_p,         /* not implemented */
110		FLAG_B,
111		FLAG_E,         /* not implemented */
112	};	};
113		#define FLAG(x) (1 << FLAG_##x)
114
115		/* We cache file position to avoid excessive seeking */
116	#define g_read_buf bb_common_bufsiz1	typedef struct FILE_and_pos_t {
117		FILE *ft_fp;
118		off_t ft_pos;
119		} FILE_and_pos_t;
120
121	struct globals {	struct globals {
	bool anychange;
122	smallint exit_status;	smallint exit_status;
123	int opt_U_context;	int opt_U_context;
124	size_t max_context;     /* size of context_vec_start */	char *label[2];
125	USE_FEATURE_DIFF_DIR(int dl_count;)	struct stat stb[2];
	USE_FEATURE_DIFF_DIR(char **dl;)
	char *opt_S_start;
	const char *label1;
	const char *label2;
	int *J;                 /* will be overlaid on class */
	int clen;
	int pref, suff;         /* length of prefix and suffix */
	int nlen[2];
	int slen[2];
	int clistlen;           /* the length of clist */
	struct cand *clist;     /* merely a free storage pot for candidates */
	long *ixnew;            /* will be overlaid on nfile[1] */
	long *ixold;            /* will be overlaid on klist */
	struct line *nfile[2];
	struct line *sfile[2];  /* shortened by pruning common prefix/suffix */
	struct context_vec *context_vec_start;
	struct context_vec *context_vec_end;
	struct context_vec *context_vec_ptr;
	char *tempname1, *tempname2;
	struct stat stb1, stb2;
126	};	};
127	#define G (*ptr_to_globals)	#define G (*ptr_to_globals)
	#define anychange          (G.anychange         )
128	#define exit_status        (G.exit_status       )	#define exit_status        (G.exit_status       )
129	#define opt_U_context      (G.opt_U_context     )	#define opt_U_context      (G.opt_U_context     )
130	#define max_context        (G.max_context       )	#define label              (G.label             )
131	#define dl_count           (G.dl_count          )	#define stb                (G.stb               )
	#define dl                 (G.dl                )
	#define opt_S_start        (G.opt_S_start       )
	#define label1             (G.label1            )
	#define label2             (G.label2            )
	#define J                  (G.J                 )
	#define clen               (G.clen              )
	#define pref               (G.pref              )
	#define suff               (G.suff              )
	#define nlen               (G.nlen              )
	#define slen               (G.slen              )
	#define clistlen           (G.clistlen          )
	#define clist              (G.clist             )
	#define ixnew              (G.ixnew             )
	#define ixold              (G.ixold             )
	#define nfile              (G.nfile             )
	#define sfile              (G.sfile             )
	#define context_vec_start  (G.context_vec_start )
	#define context_vec_end    (G.context_vec_end   )
	#define context_vec_ptr    (G.context_vec_ptr   )
	#define stb1               (G.stb1              )
	#define stb2               (G.stb2              )
	#define tempname1          (G.tempname1         )
	#define tempname2          (G.tempname2         )
132	#define INIT_G() do { \	#define INIT_G() do { \
133	SET_PTR_TO_GLOBALS(xzalloc(sizeof(G))); \	SET_PTR_TO_GLOBALS(xzalloc(sizeof(G))); \
134	opt_U_context = 3; \	opt_U_context = 3; \
	max_context = 64; \
135	} while (0)	} while (0)
136
137		typedef int token_t;
138
139	#if ENABLE_FEATURE_DIFF_DIR	enum {
140	static void print_only(const char *path, const char *entry)	/* Public */
141	{	TOK_EMPTY = 1 << 9,  /* Line fully processed, you can proceed to the next */
142	printf("Only in %s: %s\n", path, entry);	TOK_EOF   = 1 << 10, /* File ended */
143	}	/* Private (Only to be used by read_token() */
144	#endif	TOK_EOL   = 1 << 11, /* we saw EOL (sticky) */
145		TOK_SPACE = 1 << 12, /* used -b code, means we are skipping spaces */
146		SHIFT_EOF = (sizeof(token_t)*8 - 8) - 1,
147		CHAR_MASK = 0x1ff,   /* 8th bit is used to distinguish EOF from 0xff */
148		};
149
150		/* Restores full EOF from one 8th bit: */
151		//#define TOK2CHAR(t) (((t) << SHIFT_EOF) >> SHIFT_EOF)
152		/* We don't really need the above, we only need to have EOF != any_real_char: */
153		#define TOK2CHAR(t) ((t) & CHAR_MASK)
154
155	static void print_status(int val, char *_path1, char *_path2)	static void seek_ft(FILE_and_pos_t *ft, off_t pos)
156	{	{
157	/*const char *const _entry = entry ? entry : "";*/	if (ft->ft_pos != pos) {
158	/*char *const _path1 = entry ? concat_path_file(path1, _entry) : path1;*/	ft->ft_pos = pos;
159	/*char *const _path2 = entry ? concat_path_file(path2, _entry) : path2;*/	fseeko(ft->ft_fp, pos, SEEK_SET);
	switch (val) {
	/* case D_ONLY:
	print_only(path1, entry);
	break;
	*/
	case D_COMMON:
	printf("Common subdirectories: %s and %s\n", _path1, _path2);
	break;
	case D_BINARY:
	printf("Binary files %s and %s differ\n", _path1, _path2);
	break;
	case D_DIFFER:
	if (option_mask32 & FLAG_q)
	printf("Files %s and %s differ\n", _path1, _path2);
	break;
	case D_SAME:
	if (option_mask32 & FLAG_s)
	printf("Files %s and %s are identical\n", _path1, _path2);
	break;
	case D_ISDIR1:
	printf("File %s is a %s while file %s is a %s\n",
	_path1, "directory", _path2, "regular file");
	break;
	case D_ISDIR2:
	printf("File %s is a %s while file %s is a %s\n",
	_path1, "regular file", _path2, "directory");
	break;
	case D_SKIPPED1:
	printf("File %s is not a regular file or directory and was skipped\n",
	_path1);
	break;
	case D_SKIPPED2:
	printf("File %s is not a regular file or directory and was skipped\n",
	_path2);
	break;
	}
	/*
	if (entry) {
	free(_path1);
	free(_path2);
160	}	}
	*/
161	}	}
162
163		/* Reads tokens from given fp, handling -b and -w flags
164	/* Read line, return its nonzero hash. Return 0 if EOF.	* The user must reset tok every line start
	*
	* Hash function taken from Robert Sedgewick, Algorithms in C, 3d ed., p 578.
165	*/	*/
166	static ALWAYS_INLINE int fiddle_sum(int sum, int t)	static int read_token(FILE_and_pos_t *ft, token_t tok)
	{
	return sum * 127 + t;
	}
	static int readhash(FILE *fp)
167	{	{
168	int i, t, space;	tok |= TOK_EMPTY;
169	int sum;	while (!(tok & TOK_EOL)) {
170		bool is_space;
171		int t;
172
173		t = fgetc(ft->ft_fp);
174		if (t != EOF)
175		ft->ft_pos++;
176		is_space = (t == EOF || isspace(t));
177
178		/* If t == EOF (-1), set both TOK_EOF and TOK_EOL */
179		tok |= (t & (TOK_EOF + TOK_EOL));
180		/* Only EOL? */
181		if (t == '\n')
182		tok |= TOK_EOL;
183
184	sum = 1;	if (option_mask32 & FLAG(i)) /* Handcoded tolower() */
185	space = 0;	t = (t >= 'A' && t <= 'Z') ? t - ('A' - 'a') : t;
	i = 0;
	if (!(option_mask32 & (FLAG_b | FLAG_w))) {
	while ((t = getc(fp)) != '\n') {
	if (t == EOF) {
	if (i == 0)
	return 0;
	break;
	}
	sum = fiddle_sum(sum, t);
	i = 1;
	}
	} else {
	while (1) {
	switch (t = getc(fp)) {
	case '\t':
	case '\r':
	case '\v':
	case '\f':
	case ' ':
	space = 1;
	continue;
	default:
	if (space && !(option_mask32 & FLAG_w)) {
	i = 1;
	space = 0;
	}
	sum = fiddle_sum(sum, t);
	i = 1;
	continue;
	case EOF:
	if (i == 0)
	return 0;
	/* FALLTHROUGH */
	case '\n':
	break;
	}
	break;
	}
	}
	/*
	* There is a remote possibility that we end up with a zero sum.
	* Zero is used as an EOF marker, so return 1 instead.
	*/
	return (sum == 0 ? 1 : sum);
	}
186
187		if ((option_mask32 & FLAG(w)) && is_space)
188		continue;
189
190	/* Our diff implementation is using seek.	/* Trim char value to low 9 bits */
191	* When we meet non-seekable file, we must make a temp copy.	t &= CHAR_MASK;
	*/
	static char *make_temp(FILE *f, struct stat *sb)
	{
	char *name;
	int fd;
192
193	if (S_ISREG(sb->st_mode) || S_ISBLK(sb->st_mode))	if (option_mask32 & FLAG(b)) {
194	return NULL;	/* Was prev char whitespace? */
195	name = xstrdup("/tmp/difXXXXXX");	if (tok & TOK_SPACE) { /* yes */
196	fd = mkstemp(name);	if (is_space) /* this one too, ignore it */
197	if (fd < 0)	continue;
198	bb_perror_msg_and_die("mkstemp");	tok &= ~TOK_SPACE;
199	if (bb_copyfd_eof(fileno(f), fd) < 0) {	} else if (is_space) {
200	clean_up:	/* 1st whitespace char.
201	unlink(name);	* Set TOK_SPACE and replace char by ' ' */
202	xfunc_die(); /* error message is printed by bb_copyfd_eof */	t = TOK_SPACE + ' ';
203	}	}
204	fstat(fd, sb);	}
205	close(fd);	/* Clear EMPTY */
206	if (freopen(name, "r+", f) == NULL) {	tok &= ~(TOK_EMPTY + CHAR_MASK);
207	bb_perror_msg("freopen");	/* Assign char value (low 9 bits) and maybe set TOK_SPACE */
208	goto clean_up;	tok |= t;
209		break;
210	}	}
211	return name;	#if 0
212		bb_error_msg("fp:%p tok:%x '%c'%s%s%s%s", fp, tok, tok & 0xff
213		, tok & TOK_EOF ? " EOF" : ""
214		, tok & TOK_EOL ? " EOL" : ""
215		, tok & TOK_EMPTY ? " EMPTY" : ""
216		, tok & TOK_SPACE ? " SPACE" : ""
217		);
218		#endif
219		return tok;
220	}	}
221
222		struct cand {
223		int x;
224		int y;
225		int pred;
226		};
227
228	/*	static int search(const int *c, int k, int y, const struct cand *list)
	* Check to see if the given files differ.
	* Returns 0 if they are the same, 1 if different, and -1 on error.
	*/
	static NOINLINE int files_differ(FILE *f1, FILE *f2)
229	{	{
230	size_t i, j;	int i, j;
231
232	/* Prevent making copies for "/dev/null" (too common) */	if (list[c[k]].y < y) /* quick look for typical case */
233	/* Deal with input from pipes etc */	return k + 1;
234	tempname1 = make_temp(f1, &stb1);
235	tempname2 = make_temp(f2, &stb2);	for (i = 0, j = k + 1;;) {
236	if (stb1.st_size != stb2.st_size) {	const int l = (i + j) >> 1;
237	return 1;	if (l > i) {
238	}	const int t = list[c[l]].y;
239	while (1) {	if (t > y)
240	i = fread(g_read_buf,                    1, COMMON_BUFSIZE/2, f1);	j = l;
241	j = fread(g_read_buf + COMMON_BUFSIZE/2, 1, COMMON_BUFSIZE/2, f2);	else if (t < y)
242	if (i != j)	i = l;
243	return 1;	else
244	if (i == 0)	return l;
245	return (ferror(f1) || ferror(f2)) ? -1 : 0;	} else
246	if (memcmp(g_read_buf,	return l + 1;
	g_read_buf + COMMON_BUFSIZE/2, i) != 0)
	return 1;
247	}	}
248	}	}
249
250		static unsigned isqrt(unsigned n)
	static void prepare(int i, FILE *fp /*, off_t filesize*/)
251	{	{
252	struct line *p;	unsigned x = 1;
253	int h;	while (1) {
254	size_t j, sz;	const unsigned y = x;
255		x = ((n / x) + x) >> 1;
256	rewind(fp);	if (x <= (y + 1) && x >= (y - 1))
257		return x;
	/*sz = (filesize <= FSIZE_MAX ? filesize : FSIZE_MAX) / 25;*/
	/*if (sz < 100)*/
	sz = 100;
	p = xmalloc((sz + 3) * sizeof(p[0]));
	j = 0;
	while ((h = readhash(fp)) != 0) { /* while not EOF */
	if (j == sz) {
	sz = sz * 3 / 2;
	p = xrealloc(p, (sz + 3) * sizeof(p[0]));
	}
	p[++j].value = h;
258	}	}
	nlen[i] = j;
	nfile[i] = p;
259	}	}
260
261		static void stone(const int *a, int n, const int *b, int *J, int pref)
262	static void prune(void)	{
263	{	const unsigned isq = isqrt(n);
264	int i, j;	const unsigned bound =
265		(option_mask32 & FLAG(d)) ? UINT_MAX : MAX(256, isq);
266	for (pref = 0; pref < nlen[0] && pref < nlen[1] &&	int clen = 1;
267	nfile[0][pref + 1].value == nfile[1][pref + 1].value; pref++)	int clistlen = 100;
268	continue;	int k = 0;
269	for (suff = 0; suff < nlen[0] - pref && suff < nlen[1] - pref &&	struct cand *clist = xzalloc(clistlen * sizeof(clist[0]));
270	nfile[0][nlen[0] - suff].value == nfile[1][nlen[1] - suff].value;	struct cand cand;
271	suff++)	struct cand *q;
272	continue;	int *klist = xzalloc((n + 2) * sizeof(klist[0]));
273	for (j = 0; j < 2; j++) {	/*clist[0] = (struct cand){0}; - xzalloc did it */
274	sfile[j] = nfile[j] + pref;	/*klist[0] = 0; */
275	slen[j] = nlen[j] - pref - suff;
276	for (i = 0; i <= slen[j]; i++)	for (cand.x = 1; cand.x <= n; cand.x++) {
277	sfile[j][i].serial = i;	int j = a[cand.x], oldl = 0;
278		unsigned numtries = 0;
279		if (j == 0)
280		continue;
281		cand.y = -b[j];
282		cand.pred = klist[0];
283		do {
284		int l, tc;
285		if (cand.y <= clist[cand.pred].y)
286		continue;
287		l = search(klist, k, cand.y, clist);
288		if (l != oldl + 1)
289		cand.pred = klist[l - 1];
290		if (l <= k && clist[klist[l]].y <= cand.y)
291		continue;
292		if (clen == clistlen) {
293		clistlen = clistlen * 11 / 10;
294		clist = xrealloc(clist, clistlen * sizeof(clist[0]));
295		}
296		clist[clen] = cand;
297		tc = klist[l];
298		klist[l] = clen++;
299		if (l <= k) {
300		cand.pred = tc;
301		oldl = l;
302		numtries++;
303		} else {
304		k++;
305		break;
306		}
307		} while ((cand.y = b[++j]) > 0 && numtries < bound);
308	}	}
309		/* Unravel */
310		for (q = clist + klist[k]; q->y; q = clist + q->pred)
311		J[q->x + pref] = q->y + pref;
312		free(klist);
313		free(clist);
314	}	}
315
316		struct line {
317		/* 'serial' is not used in the begining, so we reuse it
318		* to store line offsets, thus reducing memory pressure
319		*/
320		union {
321		unsigned serial;
322		off_t offset;
323		};
324		unsigned value;
325		};
326
327	static void equiv(struct line *a, int n, struct line *b, int m, int *c)	static void equiv(struct line *a, int n, struct line *b, int m, int *c)
328	{	{
329	int i, j;	int i = 1, j = 1;
330
	i = j = 1;
331	while (i <= n && j <= m) {	while (i <= n && j <= m) {
332	if (a[i].value < b[j].value)	if (a[i].value < b[j].value)
333	a[i++].value = 0;	a[i++].value = 0;
#	Line 413  static void equiv(struct line *a, int n,	Line 350  static void equiv(struct line *a, int n,
350	c[j] = -1;	c[j] = -1;
351	}	}
352
353		static void unsort(const struct line *f, int l, int *b)
	static int isqrt(int n)
	{
	int y, x;
	if (n == 0)
	return 0;
	x = 1;
	do {
	y = x;
	x = n / x;
	x += y;
	x /= 2;
	} while ((x - y) > 1 || (x - y) < -1);
	return x;
	}
	static int newcand(int x, int y, int pred)
	{
	struct cand *q;
	if (clen == clistlen) {
	clistlen = clistlen * 11 / 10;
	clist = xrealloc(clist, clistlen * sizeof(struct cand));
	}
	q = clist + clen;
	q->x = x;
	q->y = y;
	q->pred = pred;
	return clen++;
	}
	static int search(int *c, int k, int y)
354	{	{
	int i, j, l, t;
	if (clist[c[k]].y < y) /* quick look for typical case */
	return k + 1;
	i = 0;
	j = k + 1;
	while (1) {
	l = i + j;
	if ((l >>= 1) <= i)
	break;
	t = clist[c[l]].y;
	if (t > y)
	j = l;
	else if (t < y)
	i = l;
	else
	return l;
	}
	return l + 1;
	}
	static int stone(int *a, int n, int *b, int *c)
	{
	int i, k, y, j, l;
	int oldc, tc, oldl;
	unsigned int numtries;
	#if ENABLE_FEATURE_DIFF_MINIMAL
	const unsigned int bound =
	(option_mask32 & FLAG_d) ? UINT_MAX : MAX(256, isqrt(n));
	#else
	const unsigned int bound = MAX(256, isqrt(n));
	#endif
	k = 0;
	c[0] = newcand(0, 0, 0);
	for (i = 1; i <= n; i++) {
	j = a[i];
	if (j == 0)
	continue;
	y = -b[j];
	oldl = 0;
	oldc = c[0];
	numtries = 0;
	do {
	if (y <= clist[oldc].y)
	continue;
	l = search(c, k, y);
	if (l != oldl + 1)
	oldc = c[l - 1];
	if (l <= k) {
	if (clist[c[l]].y <= y)
	continue;
	tc = c[l];
	c[l] = newcand(i, y, oldc);
	oldc = tc;
	oldl = l;
	numtries++;
	} else {
	c[l] = newcand(i, y, oldc);
	k++;
	break;
	}
	} while ((y = b[++j]) > 0 && numtries < bound);
	}
	return k;
	}
	static void unravel(int p)
	{
	struct cand *q;
355	int i;	int i;
356		int *a = xmalloc((l + 1) * sizeof(a[0]));
	for (i = 0; i <= nlen[0]; i++)
	J[i] = i <= pref ? i : i > nlen[0] - suff ? i + nlen[1] - nlen[0] : 0;
	for (q = clist + p; q->y != 0; q = clist + q->pred)
	J[q->x + pref] = q->y + pref;
	}
	static void unsort(struct line *f, int l, int *b)
	{
	int *a, i;
	a = xmalloc((l + 1) * sizeof(int));
357	for (i = 1; i <= l; i++)	for (i = 1; i <= l; i++)
358	a[f[i].serial] = f[i].value;	a[f[i].serial] = f[i].value;
359	for (i = 1; i <= l; i++)	for (i = 1; i <= l; i++)
#	Line 543  static void unsort(struct line *f, int l	Line 361  static void unsort(struct line *f, int l
361	free(a);	free(a);
362	}	}
363
364		static int line_compar(const void *a, const void *b)
	static int skipline(FILE *f)
365	{	{
366	int i, c;	#define l0 ((const struct line*)a)
367		#define l1 ((const struct line*)b)
368	for (i = 1; (c = getc(f)) != '\n' && c != EOF; i++)	int r = l0->value - l1->value;
369	continue;	if (r)
370	return i;	return r;
371		return l0->serial - l1->serial;
372		#undef l0
373		#undef l1
374	}	}
375
376		static void fetch(FILE_and_pos_t *ft, const off_t *ix, int a, int b, int ch)
	/*
	* Check does double duty:
	*  1.  ferret out any fortuitous correspondences due
	*      to confounding by hashing (which result in "jackpot")
	*  2.  collect random access indexes to the two files
	*/
	static NOINLINE void check(FILE *f1, FILE *f2)
377	{	{
378	int i, j, jackpot, c, d;	int i, j, col;
	long ctold, ctnew;
	rewind(f1);
	rewind(f2);
	j = 1;
	ixold[0] = ixnew[0] = 0;
	jackpot = 0;
	ctold = ctnew = 0;
	for (i = 1; i <= nlen[0]; i++) {
	if (J[i] == 0) {
	ixold[i] = ctold += skipline(f1);
	continue;
	}
	while (j < J[i]) {
	ixnew[j] = ctnew += skipline(f2);
	j++;
	}
	if (option_mask32 & (FLAG_b | FLAG_w | FLAG_i)) {
	while (1) {
	c = getc(f1);
	d = getc(f2);
	/*
	* GNU diff ignores a missing newline
	* in one file if bflag || wflag.
	*/
	if ((option_mask32 & (FLAG_b | FLAG_w))
	&& ((c == EOF && d == '\n') || (c == '\n' && d == EOF))
	) {
	break;
	}
	ctold++;
	ctnew++;
	if ((option_mask32 & FLAG_b) && isspace(c) && isspace(d)) {
	do {
	if (c == '\n')
	break;
	ctold++;
	c = getc(f1);
	} while (isspace(c));
	do {
	if (d == '\n')
	break;
	ctnew++;
	d = getc(f2);
	} while (isspace(d));
	} else if (option_mask32 & FLAG_w) {
	while (isspace(c) && c != '\n') {
	c = getc(f1);
	ctold++;
	}
	while (isspace(d) && d != '\n') {
	d = getc(f2);
	ctnew++;
	}
	}
	if (c != d) {
	jackpot++;
	J[i] = 0;
	if (c != '\n' && c != EOF)
	ctold += skipline(f1);
	if (d != '\n' && c != EOF)
	ctnew += skipline(f2);
	break;
	}
	if (c == '\n' || c == EOF)
	break;
	}
	} else {
	while (1) {
	ctold++;
	ctnew++;
	c = getc(f1);
	d = getc(f2);
	if (c != d) {
	J[i] = 0;
	if (c != '\n' && c != EOF)
	ctold += skipline(f1);
	/* was buggy? "if (d != '\n' && c != EOF)" */
	if (d != '\n' && d != EOF)
	ctnew += skipline(f2);
	break;
	}
	if (c == '\n' || c == EOF)
	break;
	}
	}
	ixold[i] = ctold;
	ixnew[j] = ctnew;
	j++;
	}
	for (; j <= nlen[1]; j++)
	ixnew[j] = ctnew += skipline(f2);
	}
	/* shellsort CACM #201 */
	static void sort(struct line *a, int n)
	{
	struct line *ai, *aim, w;
	int j, m = 0, k;
	if (n == 0)
	return;
	for (j = 1; j <= n; j *= 2)
	m = 2 * j - 1;
	for (m /= 2; m != 0; m /= 2) {
	k = n - m;
	for (j = 1; j <= k; j++) {
	for (ai = &a[j]; ai > a; ai -= m) {
	aim = &ai[m];
	if (aim < ai)
	break; /* wraparound */
	if (aim->value > ai[0].value
	|| (aim->value == ai[0].value && aim->serial > ai[0].serial)
	) {
	break;
	}
	w.value = ai[0].value;
	ai[0].value = aim->value;
	aim->value = w.value;
	w.serial = ai[0].serial;
	ai[0].serial = aim->serial;
	aim->serial = w.serial;
	}
	}
	}
	}
	static void uni_range(int a, int b)
	{
	if (a < b)
	printf("%d,%d", a, b - a + 1);
	else if (a == b)
	printf("%d", b);
	else
	printf("%d,0", b);
	}
	static void fetch(long *f, int a, int b, FILE *lb, int ch)
	{
	int i, j, c, lastc, col, nc;
	if (a > b)
	return;
379	for (i = a; i <= b; i++) {	for (i = a; i <= b; i++) {
380	fseek(lb, f[i - 1], SEEK_SET);	seek_ft(ft, ix[i - 1]);
381	nc = f[i] - f[i - 1];	putchar(ch);
382	if (ch != '\0') {	if (option_mask32 & FLAG(T))
383	putchar(ch);	putchar('\t');
384	if (option_mask32 & FLAG_T)	for (j = 0, col = 0; j < ix[i] - ix[i - 1]; j++) {
385	putchar('\t');	int c = fgetc(ft->ft_fp);
	}
	col = 0;
	for (j = 0, lastc = '\0'; j < nc; j++, lastc = c) {
	c = getc(lb);
386	if (c == EOF) {	if (c == EOF) {
387	printf("\n\\ No newline at end of file\n");	printf("\n\\ No newline at end of file\n");
388	return;	return;
389	}	}
390	if (c == '\t' && (option_mask32 & FLAG_t)) {	ft->ft_pos++;
391	do {	if (c == '\t' && (option_mask32 & FLAG(t)))
392	putchar(' ');	do putchar(' '); while (++col & 7);
393	} while (++col & 7);	else {
	} else {
394	putchar(c);	putchar(c);
395	col++;	col++;
396	}	}
#	Line 736  static void fetch(long *f, int a, int b,	Line 398  static void fetch(long *f, int a, int b,
398	}	}
399	}	}
400
401		/* Creates the match vector J, where J[i] is the index
402	#if ENABLE_FEATURE_DIFF_BINARY	* of the line in the new file corresponding to the line i
403	static int asciifile(FILE *f)	* in the old file. Lines start at 1 instead of 0, that value
404		* being used instead to denote no corresponding line.
405		* This vector is dynamically allocated and must be freed by the caller.
406		*
407		* * fp is an input parameter, where fp[0] and fp[1] are the open
408		*   old file and new file respectively.
409		* * nlen is an output variable, where nlen[0] and nlen[1]
410		*   gets the number of lines in the old and new file respectively.
411		* * ix is an output variable, where ix[0] and ix[1] gets
412		*   assigned dynamically allocated vectors of the offsets of the lines
413		*   of the old and new file respectively. These must be freed by the caller.
414		*/
415		static NOINLINE int *create_J(FILE_and_pos_t ft[2], int nlen[2], off_t *ix[2])
416	{	{
417	int i, cnt;	int *J, slen[2], *class, *member;
418		struct line *nfile[2], *sfile[2];
419	if (option_mask32 & FLAG_a)	int pref = 0, suff = 0, i, j, delta;
420	return 1;
421	rewind(f);	/* Lines of both files are hashed, and in the process
422	cnt = fread(g_read_buf, 1, COMMON_BUFSIZE, f);	* their offsets are stored in the array ix[fileno]
423	for (i = 0; i < cnt; i++) {	* where fileno == 0 points to the old file, and
424	if (!isprint(g_read_buf[i])	* fileno == 1 points to the new one.
425	&& !isspace(g_read_buf[i])	*/
426	) {	for (i = 0; i < 2; i++) {
427	return 0;	unsigned hash;
428		token_t tok;
429		size_t sz = 100;
430		nfile[i] = xmalloc((sz + 3) * sizeof(nfile[i][0]));
431		/* ft gets here without the correct position, cant use seek_ft */
432		ft[i].ft_pos = 0;
433		fseeko(ft[i].ft_fp, 0, SEEK_SET);
434
435		nlen[i] = 0;
436		/* We could zalloc nfile, but then zalloc starts showing in gprof at ~1% */
437		nfile[i][0].offset = 0;
438		goto start; /* saves code */
439		while (1) {
440		tok = read_token(&ft[i], tok);
441		if (!(tok & TOK_EMPTY)) {
442		/* Hash algorithm taken from Robert Sedgewick, Algorithms in C, 3d ed., p 578. */
443		/*hash = hash * 128 - hash + TOK2CHAR(tok);
444		* gcc insists on optimizing above to "hash * 127 + ...", thus... */
445		unsigned o = hash - TOK2CHAR(tok);
446		hash = hash * 128 - o; /* we want SPEED here */
447		continue;
448		}
449		if (nlen[i]++ == sz) {
450		sz = sz * 3 / 2;
451		nfile[i] = xrealloc(nfile[i], (sz + 3) * sizeof(nfile[i][0]));
452		}
453		/* line_compar needs hashes fit into positive int */
454		nfile[i][nlen[i]].value = hash & INT_MAX;
455		/* like ftello(ft[i].ft_fp) but faster (avoids lseek syscall) */
456		nfile[i][nlen[i]].offset = ft[i].ft_pos;
457		if (tok & TOK_EOF) {
458		/* EOF counts as a token, so we have to adjust it here */
459		nfile[i][nlen[i]].offset++;
460		break;
461		}
462		start:
463		hash = tok = 0;
464	}	}
465		/* Exclude lone EOF line from the end of the file, to make fetch()'s job easier */
466		if (nfile[i][nlen[i]].offset - nfile[i][nlen[i] - 1].offset == 1)
467		nlen[i]--;
468		/* Now we copy the line offsets into ix */
469		ix[i] = xmalloc((nlen[i] + 2) * sizeof(ix[i][0]));
470		for (j = 0; j < nlen[i] + 1; j++)
471		ix[i][j] = nfile[i][j].offset;
472		}
473
474		/* length of prefix and suffix is calculated */
475		for (; pref < nlen[0] && pref < nlen[1] &&
476		nfile[0][pref + 1].value == nfile[1][pref + 1].value;
477		pref++);
478		for (; suff < nlen[0] - pref && suff < nlen[1] - pref &&
479		nfile[0][nlen[0] - suff].value == nfile[1][nlen[1] - suff].value;
480		suff++);
481		/* Arrays are pruned by the suffix and prefix lenght,
482		* the result being sorted and stored in sfile[fileno],
483		* and their sizes are stored in slen[fileno]
484		*/
485		for (j = 0; j < 2; j++) {
486		sfile[j] = nfile[j] + pref;
487		slen[j] = nlen[j] - pref - suff;
488		for (i = 0; i <= slen[j]; i++)
489		sfile[j][i].serial = i;
490		qsort(sfile[j] + 1, slen[j], sizeof(*sfile[j]), line_compar);
491	}	}
492	return 1;	/* nfile arrays are reused to reduce memory pressure
493	}	* The #if zeroed out section performs the same task as the
494	#else	* one in the #else section.
495	#define asciifile(f) 1	* Peak memory usage is higher, but one array copy is avoided
496	#endif	* by not using unsort()
	/* dump accumulated "unified" diff changes */
	static void dump_unified_vec(FILE *f1, FILE *f2)
	{
	struct context_vec *cvp = context_vec_start;
	int lowa, upb, lowc, upd;
	int a, b, c, d;
	char ch;
	if (context_vec_start > context_vec_ptr)
	return;
	b = d = 0; /* gcc */
	lowa = MAX(1, cvp->a - opt_U_context);
	upb = MIN(nlen[0], context_vec_ptr->b + opt_U_context);
	lowc = MAX(1, cvp->c - opt_U_context);
	upd = MIN(nlen[1], context_vec_ptr->d + opt_U_context);
	printf("@@ -");
	uni_range(lowa, upb);
	printf(" +");
	uni_range(lowc, upd);
	printf(" @@\n");
	/*
	* Output changes in "unified" diff format--the old and new lines
	* are printed together.
497	*/	*/
	for (; cvp <= context_vec_ptr; cvp++) {
	a = cvp->a;
	b = cvp->b;
	c = cvp->c;
	d = cvp->d;
	/*
	* c: both new and old changes
	* d: only changes in the old file
	* a: only changes in the new file
	*/
	if (a <= b && c <= d)
	ch = 'c';
	else
	ch = (a <= b) ? 'd' : 'a';
498	#if 0	#if 0
499	switch (ch) {	member = xmalloc((slen[1] + 2) * sizeof(member[0]));
500	case 'c':	equiv(sfile[0], slen[0], sfile[1], slen[1], member);
501	// fetch() seeks!	free(nfile[1]);
502	fetch(ixold, lowa, a - 1, f1, ' ');
503	fetch(ixold, a, b, f1, '-');	class = xmalloc((slen[0] + 1) * sizeof(class[0]));
504	fetch(ixnew, c, d, f2, '+');	for (i = 1; i <= slen[0]; i++) /* Unsorting */
505	break;	class[sfile[0][i].serial] = sfile[0][i].value;
506	case 'd':	free(nfile[0]);
	fetch(ixold, lowa, a - 1, f1, ' ');
	fetch(ixold, a, b, f1, '-');
	break;
	case 'a':
	fetch(ixnew, lowc, c - 1, f2, ' ');
	fetch(ixnew, c, d, f2, '+');
	break;
	}
507	#else	#else
508	if (ch == 'c' || ch == 'd') {	member = (int *)nfile[1];
509	fetch(ixold, lowa, a - 1, f1, ' ');	equiv(sfile[0], slen[0], sfile[1], slen[1], member);
510	fetch(ixold, a, b, f1, '-');	member = xrealloc(member, (slen[1] + 2) * sizeof(member[0]));
511	}
512	if (ch == 'a')	class = (int *)nfile[0];
513	fetch(ixnew, lowc, c - 1, f2, ' ');	unsort(sfile[0], slen[0], (int *)nfile[0]);
514	if (ch == 'c' || ch == 'a')	class = xrealloc(class, (slen[0] + 2) * sizeof(class[0]));
	fetch(ixnew, c, d, f2, '+');
515	#endif	#endif
516	lowa = b + 1;	J = xmalloc((nlen[0] + 2) * sizeof(J[0]));
517	lowc = d + 1;	/* The elements of J which fall inside the prefix and suffix regions
518	}	* are marked as unchanged, while the ones which fall outside
519	fetch(ixnew, d + 1, upd, f2, ' ');	* are initialized with 0 (no matches), so that function stone can
520		* then assign them their right values
521		*/
522		for (i = 0, delta = nlen[1] - nlen[0]; i <= nlen[0]; i++)
523		J[i] = i <= pref            ?  i :
524		i > (nlen[0] - suff) ? (i + delta) : 0;
525		/* Here the magic is performed */
526		stone(class, slen[0], member, J, pref);
527		J[nlen[0] + 1] = nlen[1] + 1;
528
529	context_vec_ptr = context_vec_start - 1;	free(class);
530	}	free(member);
531
532		/* Both files are rescanned, in an effort to find any lines
533		* which, due to limitations intrinsic to any hashing algorithm,
534		* are different but ended up confounded as the same
535		*/
536		for (i = 1; i <= nlen[0]; i++) {
537		if (!J[i])
538		continue;
539
540	static void print_header(const char *file1, const char *file2)	seek_ft(&ft[0], ix[0][i - 1]);
541	{	seek_ft(&ft[1], ix[1][J[i] - 1]);
	if (label1)
	printf("--- %s\n", label1);
	else
	printf("--- %s\t%s", file1, ctime(&stb1.st_mtime));
	if (label2)
	printf("+++ %s\n", label2);
	else
	printf("+++ %s\t%s", file2, ctime(&stb2.st_mtime));
	}
542
543		for (j = J[i]; i <= nlen[0] && J[i] == j; i++, j++) {
544		token_t tok0 = 0, tok1 = 0;
545		do {
546		tok0 = read_token(&ft[0], tok0);
547		tok1 = read_token(&ft[1], tok1);
548
549	/*	if (((tok0 ^ tok1) & TOK_EMPTY) != 0 /* one is empty (not both) */
550	* Indicate that there is a difference between lines a and b of the from file	|| (!(tok0 & TOK_EMPTY) && TOK2CHAR(tok0) != TOK2CHAR(tok1))
551	* to get to lines c to d of the to file.  If a is greater than b then there	) {
552	* are no lines in the from file involved and this means that there were	J[i] = 0; /* Break the correspondence */
553	* lines appended (beginning at b).  If c is greater than d then there are	}
554	* lines missing from the to file.	} while (!(tok0 & tok1 & TOK_EMPTY));
555	*/	}
	static void change(char *file1, FILE *f1, char *file2, FILE *f2,
	int a, int b, int c, int d)
	{
	if ((a > b && c > d) || (option_mask32 & FLAG_q)) {
	anychange = 1;
	return;
556	}	}
557
558	/*	return J;
	* Allocate change records as needed.
	*/
	if (context_vec_ptr == context_vec_end - 1) {
	ptrdiff_t offset = context_vec_ptr - context_vec_start;
	max_context <<= 1;
	context_vec_start = xrealloc(context_vec_start,
	max_context * sizeof(struct context_vec));
	context_vec_end = context_vec_start + max_context;
	context_vec_ptr = context_vec_start + offset;
	}
	if (anychange == 0) {
	/*
	* Print the context/unidiff header first time through.
	*/
	print_header(file1, file2);
	} else if (a > context_vec_ptr->b + (2 * opt_U_context) + 1
	&& c > context_vec_ptr->d + (2 * opt_U_context) + 1
	) {
	/*
	* If this change is more than 'context' lines from the
	* previous change, dump the record and reset it.
	*/
	// dump_unified_vec() seeks!
	dump_unified_vec(f1, f2);
	}
	context_vec_ptr++;
	context_vec_ptr->a = a;
	context_vec_ptr->b = b;
	context_vec_ptr->c = c;
	context_vec_ptr->d = d;
	anychange = 1;
559	}	}
560
561		static bool diff(FILE* fp[2], char *file[2])
	static void output(char *file1, FILE *f1, char *file2, FILE *f2)
562	{	{
563	/* Note that j0 and j1 can't be used as they are defined in math.h.	int nlen[2];
564	* This also allows the rather amusing variable 'j00'... */	off_t *ix[2];
565	int m, i0, i1, j00, j01;	FILE_and_pos_t ft[2];
566		typedef struct { int a, b; } vec_t[2];
567	rewind(f1);	vec_t *vec = NULL;
568	rewind(f2);	int i = 1, j, k, idx = -1;
569	m = nlen[0];	bool anychange = false;
570	J[0] = 0;	int *J;
571	J[m + 1] = nlen[1] + 1;
572	for (i0 = 1; i0 <= m; i0 = i1 + 1) {	ft[0].ft_fp = fp[0];
573	while (i0 <= m && J[i0] == J[i0 - 1] + 1)	ft[1].ft_fp = fp[1];
574	i0++;	/* note that ft[i].ft_pos is unintitalized, create_J()
575	j00 = J[i0 - 1] + 1;	* must not assume otherwise */
576	i1 = i0 - 1;	J = create_J(ft, nlen, ix);
	while (i1 < m && J[i1 + 1] == 0)
	i1++;
	j01 = J[i1 + 1] - 1;
	J[i1] = j01;
	// change() seeks!
	change(file1, f1, file2, f2, i0, i1, j00, j01);
	}
	if (m == 0) {
	// change() seeks!
	change(file1, f1, file2, f2, 1, 0, 1, nlen[1]);
	}
	if (anychange != 0 && !(option_mask32 & FLAG_q)) {
	// dump_unified_vec() seeks!
	dump_unified_vec(f1, f2);
	}
	}
577
578	/*	do {
579	* The following code uses an algorithm due to Harold Stone,	bool nonempty = false;
	* which finds a pair of longest identical subsequences in
	* the two files.
	*
	* The major goal is to generate the match vector J.
	* J[i] is the index of the line in file1 corresponding
	* to line i in file0. J[i] = 0 if there is no
	* such line in file1.
	*
	* Lines are hashed so as to work in core. All potential
	* matches are located by sorting the lines of each file
	* on the hash (called "value"). In particular, this
	* collects the equivalence classes in file1 together.
	* Subroutine equiv replaces the value of each line in
	* file0 by the index of the first element of its
	* matching equivalence in (the reordered) file1.
	* To save space equiv squeezes file1 into a single
	* array member in which the equivalence classes
	* are simply concatenated, except that their first
	* members are flagged by changing sign.
	*
	* Next the indices that point into member are unsorted into
	* array class according to the original order of file0.
	*
	* The cleverness lies in routine stone. This marches
	* through the lines of file0, developing a vector klist
	* of "k-candidates". At step i a k-candidate is a matched
	* pair of lines x,y (x in file0, y in file1) such that
	* there is a common subsequence of length k
	* between the first i lines of file0 and the first y
	* lines of file1, but there is no such subsequence for
	* any smaller y. x is the earliest possible mate to y
	* that occurs in such a subsequence.
	*
	* Whenever any of the members of the equivalence class of
	* lines in file1 matable to a line in file0 has serial number
	* less than the y of some k-candidate, that k-candidate
	* with the smallest such y is replaced. The new
	* k-candidate is chained (via pred) to the current
	* k-1 candidate so that the actual subsequence can
	* be recovered. When a member has serial number greater
	* that the y of all k-candidates, the klist is extended.
	* At the end, the longest subsequence is pulled out
	* and placed in the array J by unravel
	*
	* With J in hand, the matches there recorded are
	* checked against reality to assure that no spurious
	* matches have crept in due to hashing. If they have,
	* they are broken, and "jackpot" is recorded--a harmless
	* matter except that a true match for a spuriously
	* mated line may now be unnecessarily reported as a change.
	*
	* Much of the complexity of the program comes simply
	* from trying to minimize core utilization and
	* maximize the range of doable problems by dynamically
	* allocating what is needed and reusing what is not.
	* The core requirements for problems larger than somewhat
	* are (in words) 2*length(file0) + length(file1) +
	* 3*(number of k-candidates installed), typically about
	* 6n words for files of length n.
	*/
	/* NB: files can be not REGular. The only sure thing that they
	* are not both DIRectories. */
	static unsigned diffreg(char *file1, char *file2, int flags)
	{
	int *member;     /* will be overlaid on nfile[1] */
	int *class;      /* will be overlaid on nfile[0] */
	int *klist;      /* will be overlaid on nfile[0] after class */
	FILE *f1;
	FILE *f2;
	unsigned rval;
	int i;
580
581	anychange = 0;	while (1) {
582	context_vec_ptr = context_vec_start - 1;	vec_t v;
	tempname1 = tempname2 = NULL;
	/* Is any of them a directory? Then it's simple */
	if (S_ISDIR(stb1.st_mode) != S_ISDIR(stb2.st_mode))
	return (S_ISDIR(stb1.st_mode) ? D_ISDIR1 : D_ISDIR2);
	/* None of them are directories */
	rval = D_SAME;
	if (flags & D_EMPTY1)
	/* can't be stdin, but xfopen_stdin() is smaller code */
	f1 = xfopen_stdin(bb_dev_null);
	else
	f1 = xfopen_stdin(file1);
	if (flags & D_EMPTY2)
	f2 = xfopen_stdin(bb_dev_null);
	else
	f2 = xfopen_stdin(file2);
	/* NB: if D_EMPTY1/2 is set, other file is always a regular file,
	* not pipe/fifo/chardev/etc - D_EMPTY is used by "diff -r" only,
	* and it never diffs non-ordinary files in subdirs. */
	if (!(flags & (D_EMPTY1 | D_EMPTY2))) {
	/* Quick check whether they are different */
	/* NB: copies non-REG files to tempfiles and fills tempname1/2 */
	i = files_differ(f1, f2);
	if (i != 1) { /* not different? */
	if (i != 0) /* error? */
	exit_status |= 2;
	goto closem;
	}
	}
583
584	if (!asciifile(f1) || !asciifile(f2)) {	for (v[0].a = i; v[0].a <= nlen[0] && J[v[0].a] == J[v[0].a - 1] + 1; v[0].a++)
585	rval = D_BINARY;	continue;
586	exit_status |= 1;	v[1].a = J[v[0].a - 1] + 1;
	goto closem;
	}
587
588	// Rewind inside!	for (v[0].b = v[0].a - 1; v[0].b < nlen[0] && !J[v[0].b + 1]; v[0].b++)
589	prepare(0, f1 /*, stb1.st_size*/);	continue;
590	prepare(1, f2 /*, stb2.st_size*/);	v[1].b = J[v[0].b + 1] - 1;
591	prune();	/*
592	sort(sfile[0], slen[0]);	* Indicate that there is a difference between lines a and b of the 'from' file
593	sort(sfile[1], slen[1]);	* to get to lines c to d of the 'to' file. If a is greater than b then there
594		* are no lines in the 'from' file involved and this means that there were
595		* lines appended (beginning at b).  If c is greater than d then there are
596		* lines missing from the 'to' file.
597		*/
598		if (v[0].a <= v[0].b || v[1].a <= v[1].b) {
599		/*
600		* If this change is more than 'context' lines from the
601		* previous change, dump the record and reset it.
602		*/
603		int ct = (2 * opt_U_context) + 1;
604		if (idx >= 0
605		&& v[0].a > vec[idx][0].b + ct
606		&& v[1].a > vec[idx][1].b + ct
607		) {
608		break;
609		}
610
611	member = (int *) nfile[1];	for (j = 0; j < 2; j++)
612	equiv(sfile[0], slen[0], sfile[1], slen[1], member);	for (k = v[j].a; k < v[j].b; k++)
613	//TODO: xrealloc_vector?	nonempty |= (ix[j][k+1] - ix[j][k] != 1);
	member = xrealloc(member, (slen[1] + 2) * sizeof(int));
614
615	class = (int *) nfile[0];	vec = xrealloc_vector(vec, 6, ++idx);
616	unsort(sfile[0], slen[0], class);	memcpy(vec[idx], v, sizeof(v));
617	class = xrealloc(class, (slen[0] + 2) * sizeof(int));	}
	klist = xmalloc((slen[0] + 2) * sizeof(int));
	clen = 0;
	clistlen = 100;
	clist = xmalloc(clistlen * sizeof(struct cand));
	i = stone(class, slen[0], member, klist);
	free(member);
	free(class);
618
619	J = xrealloc(J, (nlen[0] + 2) * sizeof(int));	i = v[0].b + 1;
620	unravel(klist[i]);	if (i > nlen[0])
621	free(clist);	break;
622	free(klist);	J[v[0].b] = v[1].b;
623		}
624		if (idx < 0 || ((option_mask32 & FLAG(B)) && !nonempty))
625		goto cont;
626		if (!(option_mask32 & FLAG(q))) {
627		int lowa;
628		vec_t span, *cvp = vec;
629
630		if (!anychange) {
631		/* Print the context/unidiff header first time through */
632		printf("--- %s\n", label[0] ? label[0] : file[0]);
633		printf("+++ %s\n", label[1] ? label[1] : file[1]);
634		}
635
636	ixold = xrealloc(ixold, (nlen[0] + 2) * sizeof(long));	printf("@@");
637	ixnew = xrealloc(ixnew, (nlen[1] + 2) * sizeof(long));	for (j = 0; j < 2; j++) {
638	// Rewind inside!	int a = span[j].a = MAX(1, (*cvp)[j].a - opt_U_context);
639	check(f1, f2);	int b = span[j].b = MIN(nlen[j], vec[idx][j].b + opt_U_context);
	// Rewind inside!
	output(file1, f1, file2, f2);
640
641	closem:	printf(" %c%d", j ? '+' : '-', MIN(a, b));
642	if (anychange) {	if (a == b)
643	exit_status |= 1;	continue;
644	if (rval == D_SAME)	printf(",%d", (a < b) ? b - a + 1 : 0);
645	rval = D_DIFFER;	}
646	}	printf(" @@\n");
647	fclose_if_not_stdin(f1);	/*
648	fclose_if_not_stdin(f2);	* Output changes in "unified" diff format--the old and new lines
649	if (tempname1) {	* are printed together.
650	unlink(tempname1);	*/
651	free(tempname1);	for (lowa = span[0].a; ; lowa = (*cvp++)[0].b + 1) {
652	}	bool end = cvp > &vec[idx];
653	if (tempname2) {	fetch(&ft[0], ix[0], lowa, end ? span[0].b : (*cvp)[0].a - 1, ' ');
654	unlink(tempname2);	if (end)
655	free(tempname2);	break;
656		for (j = 0; j < 2; j++)
657		fetch(&ft[j], ix[j], (*cvp)[j].a, (*cvp)[j].b, j ? '+' : '-');
658		}
659		}
660		anychange = true;
661		cont:
662		idx = -1;
663		} while (i <= nlen[0]);
664
665		free(vec);
666		free(ix[0]);
667		free(ix[1]);
668		free(J);
669		return anychange;
670		}
671
672		static int diffreg(char *file[2])
673		{
674		FILE *fp[2] = { stdin, stdin };
675		bool binary = false, differ = false;
676		int status = STATUS_SAME, i;
677
678		for (i = 0; i < 2; i++) {
679		int fd = open_or_warn_stdin(file[i]);
680		if (fd == -1)
681		goto out;
682		/* Our diff implementation is using seek.
683		* When we meet non-seekable file, we must make a temp copy.
684		*/
685		if (lseek(fd, 0, SEEK_SET) == -1 && errno == ESPIPE) {
686		char name[] = "/tmp/difXXXXXX";
687		int fd_tmp = mkstemp(name);
688		if (fd_tmp < 0)
689		bb_perror_msg_and_die("mkstemp");
690		unlink(name);
691		if (bb_copyfd_eof(fd, fd_tmp) < 0)
692		xfunc_die();
693		if (fd) /* Prevents closing of stdin */
694		close(fd);
695		fd = fd_tmp;
696		}
697		fp[i] = fdopen(fd, "r");
698	}	}
	return rval;
	}
699
700		while (1) {
701	#if ENABLE_FEATURE_DIFF_DIR	const size_t sz = COMMON_BUFSIZE / 2;
702	static void do_diff(char *dir1, char *path1, char *dir2, char *path2)	char *const buf0 = bb_common_bufsiz1;
703	{	char *const buf1 = buf0 + sz;
704	int flags = 0; /*D_HEADER;*/	int j, k;
705	int val;	i = fread(buf0, 1, sz, fp[0]);
706	char *fullpath1 = NULL; /* if -N */	j = fread(buf1, 1, sz, fp[1]);
707	char *fullpath2 = NULL;	if (i != j) {
708		differ = true;
709	if (path1)	i = MIN(i, j);
	fullpath1 = concat_path_file(dir1, path1);
	if (path2)
	fullpath2 = concat_path_file(dir2, path2);
	if (!fullpath1 || stat(fullpath1, &stb1) != 0) {
	flags |= D_EMPTY1;
	memset(&stb1, 0, sizeof(stb1));
	if (path2) {
	free(fullpath1);
	fullpath1 = concat_path_file(dir1, path2);
710	}	}
711	}	if (i == 0)
712	if (!fullpath2 || stat(fullpath2, &stb2) != 0) {	break;
713	flags |= D_EMPTY2;	for (k = 0; k < i; k++) {
714	memset(&stb2, 0, sizeof(stb2));	if (!buf0[k] || !buf1[k])
715	stb2.st_mode = stb1.st_mode;	binary = true;
716	if (path1) {	if (buf0[k] != buf1[k])
717	free(fullpath2);	differ = true;
	fullpath2 = concat_path_file(dir2, path1);
718	}	}
719	}	}
720		if (differ) {
721		if (binary && !(option_mask32 & FLAG(a)))
722		status = STATUS_BINARY;
723		else if (diff(fp, file))
724		status = STATUS_DIFFER;
725		}
726		if (status != STATUS_SAME)
727		exit_status |= 1;
728		out:
729		fclose_if_not_stdin(fp[0]);
730		fclose_if_not_stdin(fp[1]);
731
732	if (stb1.st_mode == 0)	return status;
	stb1.st_mode = stb2.st_mode;
	if (S_ISDIR(stb1.st_mode) && S_ISDIR(stb2.st_mode)) {
	printf("Common subdirectories: %s and %s\n", fullpath1, fullpath2);
	goto ret;
	}
	if (!S_ISREG(stb1.st_mode) && !S_ISDIR(stb1.st_mode))
	val = D_SKIPPED1;
	else if (!S_ISREG(stb2.st_mode) && !S_ISDIR(stb2.st_mode))
	val = D_SKIPPED2;
	else {
	/* Both files are either REGular or DIRectories */
	val = diffreg(fullpath1, fullpath2, flags);
	}
	print_status(val, fullpath1, fullpath2 /*, NULL*/);
	ret:
	free(fullpath1);
	free(fullpath2);
733	}	}
	#endif
734
735		static void print_status(int status, char *path[2])
736		{
737		switch (status) {
738		case STATUS_BINARY:
739		case STATUS_DIFFER:
740		if ((option_mask32 & FLAG(q)) || status == STATUS_BINARY)
741		printf("Files %s and %s differ\n", path[0], path[1]);
742		break;
743		case STATUS_SAME:
744		if (option_mask32 & FLAG(s))
745		printf("Files %s and %s are identical\n", path[0], path[1]);
746		break;
747		}
748		}
749
750	#if ENABLE_FEATURE_DIFF_DIR	#if ENABLE_FEATURE_DIFF_DIR
751		struct dlist {
752		size_t len;
753		int s, e;
754		char **dl;
755		};
756
757	/* This function adds a filename to dl, the directory listing. */	/* This function adds a filename to dl, the directory listing. */
758	static int FAST_FUNC add_to_dirlist(const char *filename,	static int FAST_FUNC add_to_dirlist(const char *filename,
759	struct stat *sb UNUSED_PARAM,	struct stat *sb UNUSED_PARAM,
760	void *userdata,	void *userdata, int depth UNUSED_PARAM)
	int depth UNUSED_PARAM)
761	{	{
762	dl = xrealloc_vector(dl, 5, dl_count);	struct dlist *const l = userdata;
763	dl[dl_count] = xstrdup(filename + (int)(ptrdiff_t)userdata);	l->dl = xrealloc_vector(l->dl, 6, l->e);
764	dl_count++;	/* + 1 skips "/" after dirname */
765		l->dl[l->e] = xstrdup(filename + l->len + 1);
766		l->e++;
767	return TRUE;	return TRUE;
768	}	}
769
770		/* If recursion is not set, this function adds the directory
771	/* This returns a sorted directory listing. */	* to the list and prevents recursive_action from recursing into it.
772	static char **get_recursive_dirlist(char *path)	*/
773	{	static int FAST_FUNC skip_dir(const char *filename,
774	dl_count = 0;	struct stat *sb, void *userdata,
775	dl = xzalloc(sizeof(dl[0]));	int depth)
776		{
777	/* We need to trim root directory prefix.	if (!(option_mask32 & FLAG(r)) && depth) {
778	* Using void *userdata to specify its length,	add_to_dirlist(filename, sb, userdata, depth);
779	* add_to_dirlist will remove it. */	return SKIP;
	if (option_mask32 & FLAG_r) {
	recursive_action(path, ACTION_RECURSE|ACTION_FOLLOWLINKS,
	add_to_dirlist, /* file_action */
	NULL, /* dir_action */
	(void*)(ptrdiff_t)(strlen(path) + 1),
	0);
	} else {
	DIR *dp;
	struct dirent *ep;
	dp = warn_opendir(path);
	while ((ep = readdir(dp))) {
	if (!strcmp(ep->d_name, "..") || LONE_CHAR(ep->d_name, '.'))
	continue;
	add_to_dirlist(ep->d_name, NULL, (void*)(int)0, 0);
	}
	closedir(dp);
780	}	}
781		return TRUE;
	/* Sort dl alphabetically. */
	qsort_string_vector(dl, dl_count);
	dl[dl_count] = NULL;
	return dl;
782	}	}
783
784		static void diffdir(char *p[2], const char *s_start)
	static void diffdir(char *p1, char *p2)
785	{	{
786	char **dirlist1, **dirlist2;	struct dlist list[2];
787	char *dp1, *dp2;	int i;
	int pos;
	/* Check for trailing slashes. */
	dp1 = last_char_is(p1, '/');
	if (dp1 != NULL)
	*dp1 = '\0';
	dp2 = last_char_is(p2, '/');
	if (dp2 != NULL)
	*dp2 = '\0';
	/* Get directory listings for p1 and p2. */
	dirlist1 = get_recursive_dirlist(p1);
	dirlist2 = get_recursive_dirlist(p2);
	/* If -S was set, find the starting point. */
	if (opt_S_start) {
	while (*dirlist1 != NULL && strcmp(*dirlist1, opt_S_start) < 0)
	dirlist1++;
	while (*dirlist2 != NULL && strcmp(*dirlist2, opt_S_start) < 0)
	dirlist2++;
	if ((*dirlist1 == NULL) || (*dirlist2 == NULL))
	bb_error_msg(bb_msg_invalid_arg, "NULL", "-S");
	}
788
789		memset(&list, 0, sizeof(list));
790		for (i = 0; i < 2; i++) {
791		/*list[i].s = list[i].e = 0; - memset did it */
792		/*list[i].dl = NULL; */
793
794		/* We need to trim root directory prefix.
795		* Using list.len to specify its length,
796		* add_to_dirlist will remove it. */
797		list[i].len = strlen(p[i]);
798		recursive_action(p[i], ACTION_RECURSE | ACTION_FOLLOWLINKS,
799		add_to_dirlist, skip_dir, &list[i], 0);
800		/* Sort dl alphabetically.
801		* GNU diff does this ignoring any number of trailing dots.
802		* We don't, so for us dotted files almost always are
803		* first on the list.
804		*/
805		qsort_string_vector(list[i].dl, list[i].e);
806		/* If -S was set, find the starting point. */
807		if (!s_start)
808		continue;
809		while (list[i].s < list[i].e && strcmp(list[i].dl[list[i].s], s_start) < 0)
810		list[i].s++;
811		}
812	/* Now that both dirlist1 and dirlist2 contain sorted directory	/* Now that both dirlist1 and dirlist2 contain sorted directory
813	* listings, we can start to go through dirlist1. If both listings	* listings, we can start to go through dirlist1. If both listings
814	* contain the same file, then do a normal diff. Otherwise, behaviour	* contain the same file, then do a normal diff. Otherwise, behaviour
815	* is determined by whether the -N flag is set. */	* is determined by whether the -N flag is set. */
816	while (*dirlist1 != NULL || *dirlist2 != NULL) {	while (1) {
817	dp1 = *dirlist1;	char *dp[2];
818	dp2 = *dirlist2;	int pos;
819	pos = dp1 == NULL ? 1 : (dp2 == NULL ? -1 : strcmp(dp1, dp2));	int k;
820
821		dp[0] = list[0].s < list[0].e ? list[0].dl[list[0].s] : NULL;
822		dp[1] = list[1].s < list[1].e ? list[1].dl[list[1].s] : NULL;
823		if (!dp[0] && !dp[1])
824		break;
825		pos = !dp[0] ? 1 : (!dp[1] ? -1 : strcmp(dp[0], dp[1]));
826		k = pos > 0;
827		if (pos && !(option_mask32 & FLAG(N)))
828		printf("Only in %s: %s\n", p[k], dp[k]);
829		else {
830		char *fullpath[2], *path[2]; /* if -N */
831
832		for (i = 0; i < 2; i++) {
833		if (pos == 0 || i == k) {
834		path[i] = fullpath[i] = concat_path_file(p[i], dp[i]);
835		stat(fullpath[i], &stb[i]);
836		} else {
837		fullpath[i] = concat_path_file(p[i], dp[1 - i]);
838		path[i] = (char *)bb_dev_null;
839		}
840		}
841		if (pos)
842		stat(fullpath[k], &stb[1 - k]);
843
844		if (S_ISDIR(stb[0].st_mode) && S_ISDIR(stb[1].st_mode))
845		printf("Common subdirectories: %s and %s\n", fullpath[0], fullpath[1]);
846		else if (!S_ISREG(stb[0].st_mode) && !S_ISDIR(stb[0].st_mode))
847		printf("File %s is not a regular file or directory and was skipped\n", fullpath[0]);
848		else if (!S_ISREG(stb[1].st_mode) && !S_ISDIR(stb[1].st_mode))
849		printf("File %s is not a regular file or directory and was skipped\n", fullpath[1]);
850		else if (S_ISDIR(stb[0].st_mode) != S_ISDIR(stb[1].st_mode)) {
851		if (S_ISDIR(stb[0].st_mode))
852		printf("File %s is a %s while file %s is a %s\n", fullpath[0], "directory", fullpath[1], "regular file");
853		else
854		printf("File %s is a %s while file %s is a %s\n", fullpath[0], "regular file", fullpath[1], "directory");
855		} else
856		print_status(diffreg(path), fullpath);
857
858		free(fullpath[0]);
859		free(fullpath[1]);
860		}
861		free(dp[k]);
862		list[k].s++;
863	if (pos == 0) {	if (pos == 0) {
864	do_diff(p1, dp1, p2, dp2);	free(dp[1 - k]);
865	dirlist1++;	list[1 - k].s++;
	dirlist2++;
	} else if (pos < 0) {
	if (option_mask32 & FLAG_N)
	do_diff(p1, dp1, p2, NULL);
	else
	print_only(p1, dp1);
	dirlist1++;
	} else {
	if (option_mask32 & FLAG_N)
	do_diff(p1, NULL, p2, dp2);
	else
	print_only(p2, dp2);
	dirlist2++;
866	}	}
867	}	}
868		if (ENABLE_FEATURE_CLEAN_UP) {
869		free(list[0].dl);
870		free(list[1].dl);
871		}
872	}	}
873	#endif	#endif
874
875		#if ENABLE_FEATURE_DIFF_LONG_OPTIONS
876		static const char diff_longopts[] ALIGN1 =
877		"ignore-case\0"              No_argument       "i"
878		"ignore-tab-expansion\0"     No_argument       "E"
879		"ignore-space-change\0"      No_argument       "b"
880		"ignore-all-space\0"         No_argument       "w"
881		"ignore-blank-lines\0"       No_argument       "B"
882		"text\0"                     No_argument       "a"
883		"unified\0"                  Required_argument "U"
884		"label\0"                    Required_argument "L"
885		"show-c-function\0"          No_argument       "p"
886		"brief\0"                    No_argument       "q"
887		"expand-tabs\0"              No_argument       "t"
888		"initial-tab\0"              No_argument       "T"
889		"recursive\0"                No_argument       "r"
890		"new-file\0"                 No_argument       "N"
891		"report-identical-files\0"   No_argument       "s"
892		"starting-file\0"            Required_argument "S"
893		"minimal\0"                  No_argument       "d"
894		;
895		#endif
896
897	int diff_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;	int diff_main(int argc, char **argv) MAIN_EXTERNALLY_VISIBLE;
898	int diff_main(int argc UNUSED_PARAM, char **argv)	int diff_main(int argc UNUSED_PARAM, char **argv)
899	{	{
900	int gotstdin = 0;	int gotstdin = 0, i;
901	char *f1, *f2;	char *file[2], *s_start = NULL;
902	llist_t *L_arg = NULL;	llist_t *L_arg = NULL;
903
904	INIT_G();	INIT_G();
905
906	/* exactly 2 params; collect multiple -L <label>; -U N */	/* exactly 2 params; collect multiple -L <label>; -U N */
907	opt_complementary = "=2:L::U+";	opt_complementary = "=2:L::U+";
908	getopt32(argv, "abdiL:NqrsS:tTU:wu"	#if ENABLE_FEATURE_DIFF_LONG_OPTIONS
909	"p" /* ignored (for compatibility) */,	applet_long_options = diff_longopts;
910	&L_arg, &opt_S_start, &opt_U_context);	#endif
911	/*argc -= optind;*/	getopt32(argv, "abdiL:NqrsS:tTU:wupBE",
912		&L_arg, &s_start, &opt_U_context);
913	argv += optind;	argv += optind;
914	while (L_arg) {	while (L_arg)
915	if (label1 && label2)	label[!!label[0]] = llist_pop(&L_arg);
916	bb_show_usage();	xfunc_error_retval = 2;
917	if (label1) /* then label2 is NULL */	for (i = 0; i < 2; i++) {
918	label2 = label1;	file[i] = argv[i];
919	label1 = llist_pop(&L_arg);	/* Compat: "diff file name_which_doesnt_exist" exits with 2 */
920		if (LONE_DASH(file[i])) {
921		fstat(STDIN_FILENO, &stb[i]);
922		gotstdin++;
923		} else
924		xstat(file[i], &stb[i]);
925	}	}
926		xfunc_error_retval = 1;
927	/*	if (gotstdin && (S_ISDIR(stb[0].st_mode) || S_ISDIR(stb[1].st_mode)))
	* Do sanity checks, fill in stb1 and stb2 and call the appropriate
	* driver routine.  Both drivers use the contents of stb1 and stb2.
	*/
	f1 = argv[0];
	f2 = argv[1];
	if (LONE_DASH(f1)) {
	fstat(STDIN_FILENO, &stb1);
	gotstdin++;
	} else
	xstat(f1, &stb1);
	if (LONE_DASH(f2)) {
	fstat(STDIN_FILENO, &stb2);
	gotstdin++;
	} else
	xstat(f2, &stb2);
	if (gotstdin && (S_ISDIR(stb1.st_mode) || S_ISDIR(stb2.st_mode)))
928	bb_error_msg_and_die("can't compare stdin to a directory");	bb_error_msg_and_die("can't compare stdin to a directory");
929
930	if (S_ISDIR(stb1.st_mode) && S_ISDIR(stb2.st_mode)) {	if (S_ISDIR(stb[0].st_mode) && S_ISDIR(stb[1].st_mode)) {
931	#if ENABLE_FEATURE_DIFF_DIR	#if ENABLE_FEATURE_DIFF_DIR
932	diffdir(f1, f2);	diffdir(file, s_start);
	return exit_status;
933	#else	#else
934	bb_error_msg_and_die("no support for directory comparison");	bb_error_msg_and_die("no support for directory comparison");
935	#endif	#endif
936		} else {
937		bool dirfile = S_ISDIR(stb[0].st_mode) || S_ISDIR(stb[1].st_mode);
938		bool dir = S_ISDIR(stb[1].st_mode);
939		if (dirfile) {
940		const char *slash = strrchr(file[!dir], '/');
941		file[dir] = concat_path_file(file[dir], slash ? slash + 1 : file[!dir]);
942		xstat(file[dir], &stb[dir]);
943		}
944		/* diffreg can get non-regular files here */
945		print_status(gotstdin > 1 ? STATUS_SAME : diffreg(file), file);
946
947		if (dirfile)
948		free(file[dir]);
949	}	}
950
	if (S_ISDIR(stb1.st_mode)) { /* "diff dir file" */
	/* NB: "diff dir      dir2/dir3/file" must become
	*     "diff dir/file dir2/dir3/file" */
	char *slash = strrchr(f2, '/');
	f1 = concat_path_file(f1, slash ? slash + 1 : f2);
	xstat(f1, &stb1);
	}
	if (S_ISDIR(stb2.st_mode)) {
	char *slash = strrchr(f1, '/');
	f2 = concat_path_file(f2, slash ? slash + 1 : f1);
	xstat(f2, &stb2);
	}
	/* diffreg can get non-regular files here,
	* they are not both DIRestories */
	print_status((gotstdin > 1 ? D_SAME : diffreg(f1, f2, 0)),
	f1, f2 /*, NULL*/);
951	return exit_status;	return exit_status;
952	}	}

---

Colored Diff Long Colored Diff Full Colored Diff Unidiff Context Diff Side by Side

Legend: Removed from v.983   changed lines   Added in v.984