@matheusbsbs / matheus-bagliano / commits / 3a85dc7d53
Fork 0

is_ntfs_dotgit(): speed it up

Previously, this function was written without focusing on speed, intending to make reviewing the code as easy as possible, to avoid any bugs in this critical code. Turns out: we can do much better on both accounts. With this patch, we make it as fast as this developer can make it go: - We avoid the call to `is_dir_sep()` and make all the character comparisons explicit. - We avoid the cost of calling `strncasecmp()` and unroll the test for `.git` and `git~1`, not even using `tolower()` because it is faster to compare against two constant values. - We look for `.git` and `.git~1` first thing, and return early if not found. - We also avoid calling a separate function for detecting chains of spaces and periods. Each of these improvements has a noticeable impact on the speed of `is_ntfs_dotgit()`. Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>

Johannes Schindelin committed Sep 6, 2019 at 21:09 UTC 3a85dc7d534fc2d410ddc0c771c963b20d1b4857
1 file changed +30 -25
path.c
+30 -25
@@ -1288,20 +1288,6 @@ int daemon_avoid_alias(const char *p)
1288 }
1289 }
1290
1291 -static int only_spaces_and_periods(const char *path, size_t len, size_t skip)
1292 -{
1293 - if (len < skip)
1294 - return 0;
1295 - len -= skip;
1296 - path += skip;
1297 - while (len-- > 0) {
1298 - char c = *(path++);
1299 - if (c != ' ' && c != '.')
1300 - return 0;
1301 - }
1302 - return 1;
1303 -}
1304 -
1291 /*
1292 * On NTFS, we need to be careful to disallow certain synonyms of the `.git/`
1293 * directory:
@@ -1341,19 +1327,38 @@ static int only_spaces_and_periods(const char *path, size_t len, size_t skip)
1327 */
1328 int is_ntfs_dotgit(const char *name)
1329 {
1344 - size_t len;
1330 + char c;
1331
1346 - for (len = 0; ; len++)
1347 - if (!name[len] || name[len] == '\\' || is_dir_sep(name[len]) ||
1348 - name[len] == ':') {
1349 - if (only_spaces_and_periods(name, len, 4) &&
1350 - !strncasecmp(name, ".git", 4))
1351 - return 1;
1352 - if (only_spaces_and_periods(name, len, 5) &&
1353 - !strncasecmp(name, "git~1", 5))
1354 - return 1;
1332 + /*
1333 + * Note that when we don't find `.git` or `git~1` we end up with `name`
1334 + * advanced partway through the string. That's okay, though, as we
1335 + * return immediately in those cases, without looking at `name` any
1336 + * further.
1337 + */
1338 + c = *(name++);
1339 + if (c == '.') {
1340 + /* .git */
1341 + if (((c = *(name++)) != 'g' && c != 'G') ||
1342 + ((c = *(name++)) != 'i' && c != 'I') ||
1343 + ((c = *(name++)) != 't' && c != 'T'))
1344 return 0;
1356 - }
1345 + } else if (c == 'g' || c == 'G') {
1346 + /* git ~1 */
1347 + if (((c = *(name++)) != 'i' && c != 'I') ||
1348 + ((c = *(name++)) != 't' && c != 'T') ||
1349 + *(name++) != '~' ||
1350 + *(name++) != '1')
1351 + return 0;
1352 + } else
1353 + return 0;
1354 +
1355 + for (;;) {
1356 + c = *(name++);
1357 + if (!c || c == '\\' || c == '/' || c == ':')
1358 + return 1;
1359 + if (c != '.' && c != ' ')
1360 + return 0;
1361 + }
1362 }
1363
1364 static int is_ntfs_dot_generic(const char *name,