void crawler::crawl(std::string &path) {
std::string normalized_path = vmprobe::path::normalize(path);
file_index already_seen_files;
_crawl(normalized_path, already_seen_files);
}
static void
_test_empty(void)
{
gchar *basedir = g_strdup_printf("%s/try-%u", global_base, _next_id());
if (0 != g_mkdir_with_parents(basedir, 0700))
g_error("Try failure [%s] : (%d) %s", basedir, errno, strerror(errno));
_crawl(basedir);
g_free(basedir);
}
static void
_test_simple(void)
{
static struct citizen_s citizens[] = {
{"/", "", 0400},
{"/", "plop", 0600},
{"/", "plip", 0600},
{"/", "plup", 0600},
{"/", "plap", 0600},
{NULL, NULL, 0}
};
gchar *basedir = g_strdup_printf("%s/try-%u", global_base, _next_id());
if (0 != g_mkdir_with_parents(basedir, 0700))
g_error("Try failure [%s] : (%d) %s", basedir, errno, strerror(errno));
_populate(basedir, citizens);
_crawl(basedir);
g_free(basedir);
}
void crawler::_crawl(std::string &path_std_string, file_index &already_seen_files) {
const char *path = path_std_string.c_str();
struct ::stat sb;
int res = follow_symlinks ? stat(path, &sb) : lstat(path, &sb);
if (res) {
warning("unable to stat %s (%s)", path, strerror(errno));
return;
}
if (S_ISLNK(sb.st_mode)) {
warning("not following symbolic link %s", path);
return;
}
if (skip_duplicate_hardlinks && sb.st_nlink > 1) {
file_index::iterator dev = already_seen_files.find(sb.st_dev);
// Haven't seen this device before, initialize it
if (dev == already_seen_files.end()) {
already_seen_files.emplace(sb.st_dev, std::unordered_set<ino_t> {sb.st_ino});
}
// Already seen this inode before
else if (!dev->second.insert(sb.st_ino).second) {
warning("skipping duplicate hardlink %s", path);
return;
}
}
if (S_ISDIR(sb.st_mode)) {
if (curr_crawl_depth == max_crawl_depth) {
warning("maximum directory crawl depth reached: %s", path);
return;
}
num_dirs++;
retry_opendir:
DIR *dirp = opendir(path);
if (dirp == NULL) {
if (errno == ENFILE || errno == EMFILE) {
increment_nofile_rlimit();
goto retry_opendir;
}
warning("unable to opendir %s (%s), skipping", path, strerror(errno));
return;
}
struct dirent *de;
std::vector<std::string> dir_entries;
while((de = readdir(dirp)) != NULL) {
if (strcmp(de->d_name, ".") == 0 || strcmp(de->d_name, "..") == 0) continue;
dir_entries.emplace_back(de->d_name);
}
if (closedir(dirp)) fatal("unable to closedir %s (%s)", path, strerror(errno));
std::sort(dir_entries.begin(), dir_entries.end());
for (auto &s : dir_entries) {
std::string npath = path_std_string + std::string("/") + s;
curr_crawl_depth++;
_crawl(npath, already_seen_files);
curr_crawl_depth--;
}
} else if (S_ISREG(sb.st_mode)) {
num_files++;
// FIXME: catch exceptions thrown by the callback
file_handler(path_std_string, sb);
} else {
warning("skipping non-regular file: %s", path);
}
}
