git wrapper: DWIM mistyped commands
This patch introduces a modified Damerau-Levenshtein algorithm into Git's code base, and uses it with the following penalties to show some similar commands when an unknown command was encountered: swap = 0, insertion = 1, substitution = 2, deletion = 4 A typical output would now look like this: $ git sm git: 'sm' is not a git-command. See 'git --help'. Did you mean one of these? am rm The cut-off is at similarity rating 6, which was empirically determined to give sensible results. As a convenience, if there is only one candidate, Git continues under the assumption that the user mistyped it. Example: $ git reabse WARNING: You called a Git program named 'reabse', which does not exist. Continuing under the assumption that you meant 'rebase' [...] Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de> Signed-off-by: Alex Riesen <raa.lkml@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Johannes Schindelin committed
Aug 31, 2008 at 15:50 UTC
8af84dadb142f7321ff0ce8690385e99da8ede2f
7 files changed
+133
-4
Makefile
+2
@@ -358,6 +358,7 @@ LIB_H += graph.h
358
LIB_H += grep.h
359
LIB_H += hash.h
360
LIB_H += help.h
361
+LIB_H += levenshtein.h
362
LIB_H += list-objects.h
363
LIB_H += ll-merge.h
364
LIB_H += log-tree.h
@@ -433,6 +434,7 @@ LIB_OBJS += hash.o
434
LIB_OBJS += help.o
435
LIB_OBJS += ident.o
436
LIB_OBJS += interpolate.o
437
+LIB_OBJS += levenshtein.o
438
LIB_OBJS += list-objects.o
439
LIB_OBJS += ll-merge.o
440
LIB_OBJS += lockfile.o
builtin.h
+1
-1
@@ -11,7 +11,7 @@ extern const char git_usage_string[];
11
extern const char git_more_info_string[];
12
13
extern void list_common_cmds_help(void);
14
-extern void help_unknown_cmd(const char *cmd);
14
+extern const char *help_unknown_cmd(const char *cmd);
15
extern void prune_packed_objects(int);
16
extern int read_line_with_nul(char *buf, int size, FILE *file);
17
extern int fmt_merge_msg(int merge_summary, struct strbuf *in,
git.c
+3
-1
@@ -499,7 +499,9 @@ int main(int argc, const char **argv)
499
cmd, argv[0]);
500
exit(1);
501
}
502
- help_unknown_cmd(cmd);
502
+ argv[0] = help_unknown_cmd(cmd);
503
+ handle_internal_command(argc, argv);
504
+ execv_dashed_external(argv);
505
}
506
507
fprintf(stderr, "Failed to run command '%s': %s\n",
help.c
+71
-1
@@ -1,6 +1,7 @@
1
#include "cache.h"
2
#include "builtin.h"
3
#include "exec_cmd.h"
4
+#include "levenshtein.h"
5
#include "help.h"
6
7
/* most GUI terminals set COLUMNS (although some don't export it) */
@@ -37,6 +38,16 @@ void add_cmdname(struct cmdnames *cmds, const char *name, int len)
38
cmds->names[cmds->cnt++] = ent;
39
}
40
41
+static void clean_cmdnames(struct cmdnames *cmds)
42
+{
43
+ int i;
44
+ for (i = 0; i < cmds->cnt; ++i)
45
+ free(cmds->names[i]);
46
+ free(cmds->names);
47
+ cmds->cnt = 0;
48
+ cmds->alloc = 0;
49
+}
50
+
51
static int cmdname_compare(const void *a_, const void *b_)
52
{
53
struct cmdname *a = *(struct cmdname **)a_;
@@ -257,9 +268,68 @@ int is_in_cmdlist(struct cmdnames *c, const char *s)
268
return 0;
269
}
270
260
-void help_unknown_cmd(const char *cmd)
271
+static int levenshtein_compare(const void *p1, const void *p2)
272
+{
273
+ const struct cmdname *const *c1 = p1, *const *c2 = p2;
274
+ const char *s1 = (*c1)->name, *s2 = (*c2)->name;
275
+ int l1 = (*c1)->len;
276
+ int l2 = (*c2)->len;
277
+ return l1 != l2 ? l1 - l2 : strcmp(s1, s2);
278
+}
279
+
280
+const char *help_unknown_cmd(const char *cmd)
281
{
282
+ int i, n, best_similarity = 0;
283
+ struct cmdnames main_cmds, other_cmds;
284
+
285
+ memset(&main_cmds, 0, sizeof(main_cmds));
286
+ memset(&other_cmds, 0, sizeof(main_cmds));
287
+
288
+ load_command_list("git-", &main_cmds, &other_cmds);
289
+
290
+ ALLOC_GROW(main_cmds.names, main_cmds.cnt + other_cmds.cnt,
291
+ main_cmds.alloc);
292
+ memcpy(main_cmds.names + main_cmds.cnt, other_cmds.names,
293
+ other_cmds.cnt * sizeof(other_cmds.names[0]));
294
+ main_cmds.cnt += other_cmds.cnt;
295
+ free(other_cmds.names);
296
+
297
+ /* This reuses cmdname->len for similarity index */
298
+ for (i = 0; i < main_cmds.cnt; ++i)
299
+ main_cmds.names[i]->len =
300
+ levenshtein(cmd, main_cmds.names[i]->name, 0, 2, 1, 4);
301
+
302
+ qsort(main_cmds.names, main_cmds.cnt,
303
+ sizeof(*main_cmds.names), levenshtein_compare);
304
+
305
+ if (!main_cmds.cnt)
306
+ die ("Uh oh. Your system reports no Git commands at all.");
307
+
308
+ best_similarity = main_cmds.names[0]->len;
309
+ n = 1;
310
+ while (n < main_cmds.cnt && best_similarity == main_cmds.names[n]->len)
311
+ ++n;
312
+ if (n == 1) {
313
+ const char *assumed = main_cmds.names[0]->name;
314
+ main_cmds.names[0] = NULL;
315
+ clean_cmdnames(&main_cmds);
316
+ fprintf(stderr, "WARNING: You called a Git program named '%s', "
317
+ "which does not exist.\n"
318
+ "Continuing under the assumption that you meant '%s'\n",
319
+ cmd, assumed);
320
+ return assumed;
321
+ }
322
+
323
fprintf(stderr, "git: '%s' is not a git-command. See 'git --help'.\n", cmd);
324
+
325
+ if (best_similarity < 6) {
326
+ fprintf(stderr, "\nDid you mean %s?\n",
327
+ n < 2 ? "this": "one of these");
328
+
329
+ for (i = 0; i < n; i++)
330
+ fprintf(stderr, "\t%s\n", main_cmds.names[i]->name);
331
+ }
332
+
333
exit(1);
334
}
335
help.h
+1
-1
@@ -5,7 +5,7 @@ struct cmdnames {
5
int alloc;
6
int cnt;
7
struct cmdname {
8
- size_t len;
8
+ size_t len; /* also used for similarity index in help.c */
9
char name[FLEX_ARRAY];
10
} **names;
11
};
levenshtein.c
new
+47
@@ -0,0 +1,47 @@
1
+#include "cache.h"
2
+#include "levenshtein.h"
3
+
4
+int levenshtein(const char *string1, const char *string2,
5
+ int w, int s, int a, int d)
6
+{
7
+ int len1 = strlen(string1), len2 = strlen(string2);
8
+ int *row0 = xmalloc(sizeof(int) * (len2 + 1));
9
+ int *row1 = xmalloc(sizeof(int) * (len2 + 1));
10
+ int *row2 = xmalloc(sizeof(int) * (len2 + 1));
11
+ int i, j;
12
+
13
+ for (j = 0; j <= len2; j++)
14
+ row1[j] = j * a;
15
+ for (i = 0; i < len1; i++) {
16
+ int *dummy;
17
+
18
+ row2[0] = (i + 1) * d;
19
+ for (j = 0; j < len2; j++) {
20
+ /* substitution */
21
+ row2[j + 1] = row1[j] + s * (string1[i] != string2[j]);
22
+ /* swap */
23
+ if (i > 0 && j > 0 && string1[i - 1] == string2[j] &&
24
+ string1[i] == string2[j - 1] &&
25
+ row2[j + 1] > row0[j - 1] + w)
26
+ row2[j + 1] = row0[j - 1] + w;
27
+ /* deletion */
28
+ if (j + 1 < len2 && row2[j + 1] > row1[j + 1] + d)
29
+ row2[j + 1] = row1[j + 1] + d;
30
+ /* insertion */
31
+ if (row2[j + 1] > row2[j] + a)
32
+ row2[j + 1] = row2[j] + a;
33
+ }
34
+
35
+ dummy = row0;
36
+ row0 = row1;
37
+ row1 = row2;
38
+ row2 = dummy;
39
+ }
40
+
41
+ i = row1[len2];
42
+ free(row0);
43
+ free(row1);
44
+ free(row2);
45
+
46
+ return i;
47
+}
levenshtein.h
new
+8
@@ -0,0 +1,8 @@
1
+#ifndef LEVENSHTEIN_H
2
+#define LEVENSHTEIN_H
3
+
4
+int levenshtein(const char *string1, const char *string2,
5
+ int swap_penalty, int substition_penalty,
6
+ int insertion_penalty, int deletion_penalty);
7
+
8
+#endif