/**
* Split a string <b>str</b> along all occurrences of <b>sep</b>,
* adding the split strings, in order, to <b>sl</b>.
*
* If <b>flags</b>&SPLIT_SKIP_SPACE is true, remove initial and
* trailing space from each entry.
* If <b>flags</b>&SPLIT_IGNORE_BLANK is true, remove any entries
* of length 0.
* If <b>flags</b>&SPLIT_STRIP_SPACE is true, strip spaces from each
* split string.
*
* If max>0, divide the string into no more than <b>max</b> pieces. If
* <b>sep</b> is NULL, split on any sequence of horizontal space.
*/
int
smartlist_split_string(smartlist_t *sl, const char *str, const char *sep,
int flags, int max)
{
const char *cp, *end, *next;
int n = 0;
tor_assert(sl);
tor_assert(str);
cp = str;
while (1) {
if (flags&SPLIT_SKIP_SPACE) {
while (TOR_ISSPACE(*cp)) ++cp;
}
if (max>0 && n == max-1) {
end = strchr(cp,'\0');
} else if (sep) {
end = strstr(cp,sep);
if (!end)
end = strchr(cp,'\0');
} else {
for (end = cp; *end && *end != '\t' && *end != ' '; ++end)
;
}
tor_assert(end);
if (!*end) {
next = NULL;
} else if (sep) {
next = end+strlen(sep);
} else {
next = end+1;
while (*next == '\t' || *next == ' ')
++next;
}
if (flags&SPLIT_SKIP_SPACE) {
while (end > cp && TOR_ISSPACE(*(end-1)))
--end;
}
if (end != cp || !(flags&SPLIT_IGNORE_BLANK)) {
char *string = tor_strndup(cp, end-cp);
if (flags&SPLIT_STRIP_SPACE)
tor_strstrip(string, " ");
smartlist_add(sl, string);
++n;
}
if (!next)
break;
cp = next;
}
return n;
}
/** Add an entry to the GeoIP table, parsing it from <b>line</b>. The
* format is as for geoip_load_file(). */
/*private*/ int
geoip_parse_entry(const char *line)
{
unsigned int low, high;
char b[3];
if (!geoip_countries)
init_geoip_countries();
if (!geoip_entries)
geoip_entries = smartlist_new();
while (TOR_ISSPACE(*line))
++line;
if (*line == '#')
return 0;
if (tor_sscanf(line,"%u,%u,%2s", &low, &high, b) == 3) {
geoip_add_entry(low, high, b);
return 0;
} else if (tor_sscanf(line,"\"%u\",\"%u\",\"%2s\",", &low, &high, b) == 3) {
geoip_add_entry(low, high, b);
return 0;
} else {
log_warn(LD_GENERAL, "Unable to parse line from GEOIP file: %s",
escaped(line));
return -1;
}
}
/** Return true iff we need to quote and escape the string <b>s</b> to encode
* it. */
static bool
needs_escape(const char *s, bool as_keyless_val)
{
if (as_keyless_val && *s == 0)
return true;
for (; *s; ++s) {
if (*s >= 127 || TOR_ISSPACE(*s) || ! TOR_ISPRINT(*s) ||
*s == '\'' || *s == '\"') {
return true;
}
}
return false;
}
/** Add an entry to the GeoIP table indicated by <b>family</b>,
* parsing it from <b>line</b>. The format is as for geoip_load_file(). */
/*private*/ int
geoip_parse_entry(const char *line, sa_family_t family)
{
tor_addr_t low_addr, high_addr;
char c[3];
char *country = NULL;
if (!geoip_countries)
init_geoip_countries();
if (family == AF_INET) {
if (!geoip_ipv4_entries)
geoip_ipv4_entries = smartlist_new();
} else if (family == AF_INET6) {
if (!geoip_ipv6_entries)
geoip_ipv6_entries = smartlist_new();
} else {
log_warn(LD_GENERAL, "Unsupported family: %d", family);
return -1;
}
while (TOR_ISSPACE(*line))
++line;
if (*line == '#')
return 0;
if (family == AF_INET) {
unsigned int low, high;
if (tor_sscanf(line,"%u,%u,%2s", &low, &high, c) == 3 ||
tor_sscanf(line,"\"%u\",\"%u\",\"%2s\",", &low, &high, c) == 3) {
tor_addr_from_ipv4h(&low_addr, low);
tor_addr_from_ipv4h(&high_addr, high);
} else
goto fail;
country = c;
} else { /* AF_INET6 */
char buf[512];
char *low_str, *high_str;
struct in6_addr low, high;
char *strtok_state;
strlcpy(buf, line, sizeof(buf));
low_str = tor_strtok_r(buf, ",", &strtok_state);
if (!low_str)
goto fail;
high_str = tor_strtok_r(NULL, ",", &strtok_state);
if (!high_str)
goto fail;
country = tor_strtok_r(NULL, "\n", &strtok_state);
if (!country)
goto fail;
if (strlen(country) != 2)
goto fail;
if (tor_inet_pton(AF_INET6, low_str, &low) <= 0)
goto fail;
tor_addr_from_in6(&low_addr, &low);
if (tor_inet_pton(AF_INET6, high_str, &high) <= 0)
goto fail;
tor_addr_from_in6(&high_addr, &high);
}
geoip_add_entry(&low_addr, &high_addr, country);
return 0;
fail:
log_warn(LD_GENERAL, "Unable to parse line from GEOIP %s file: %s",
family == AF_INET ? "IPv4" : "IPv6", escaped(line));
return -1;
}
/**
* Helper function to parse out a line in the measured bandwidth file
* into a measured_bw_line_t output structure.
*
* If <b>line_is_after_headers</b> is true, then if we encounter an incomplete
* bw line, return -1 and warn, since we are after the headers and we should
* only parse bw lines. Return 0 otherwise.
*
* If <b>line_is_after_headers</b> is false then it means that we are not past
* the header block yet. If we encounter an incomplete bw line, return -1 but
* don't warn since there could be additional header lines coming. If we
* encounter a proper bw line, return 0 (and we got past the headers).
*
* If the line contains "vote=0", stop parsing it, and return -1, so that the
* line is ignored during voting.
*/
STATIC int
measured_bw_line_parse(measured_bw_line_t *out, const char *orig_line,
int line_is_after_headers)
{
char *line = tor_strdup(orig_line);
char *cp = line;
int got_bw = 0;
int got_node_id = 0;
char *strtok_state; /* lame sauce d'jour */
if (strlen(line) == 0) {
log_warn(LD_DIRSERV, "Empty line in bandwidth file");
tor_free(line);
return -1;
}
/* Remove end of line character, so that is not part of the token */
if (line[strlen(line) - 1] == '\n') {
line[strlen(line) - 1] = '\0';
}
cp = tor_strtok_r(cp, " \t", &strtok_state);
if (!cp) {
log_warn(LD_DIRSERV, "Invalid line in bandwidth file: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
if (orig_line[strlen(orig_line)-1] != '\n') {
log_warn(LD_DIRSERV, "Incomplete line in bandwidth file: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
do {
// If the line contains vote=0, ignore it.
if (strcmpstart(cp, "vote=0") == 0) {
log_debug(LD_DIRSERV, "Ignoring bandwidth file line that contains "
"vote=0: %s",escaped(orig_line));
tor_free(line);
return -1;
} else if (strcmpstart(cp, "bw=") == 0) {
int parse_ok = 0;
char *endptr;
if (got_bw) {
log_warn(LD_DIRSERV, "Double bw= in bandwidth file line: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
cp+=strlen("bw=");
out->bw_kb = tor_parse_long(cp, 10, 0, LONG_MAX, &parse_ok, &endptr);
if (!parse_ok || (*endptr && !TOR_ISSPACE(*endptr))) {
log_warn(LD_DIRSERV, "Invalid bandwidth in bandwidth file line: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
got_bw=1;
} else if (strcmpstart(cp, "node_id=$") == 0) {
if (got_node_id) {
log_warn(LD_DIRSERV, "Double node_id= in bandwidth file line: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
cp+=strlen("node_id=$");
if (strlen(cp) != HEX_DIGEST_LEN ||
base16_decode(out->node_id, DIGEST_LEN,
cp, HEX_DIGEST_LEN) != DIGEST_LEN) {
log_warn(LD_DIRSERV, "Invalid node_id in bandwidth file line: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
strlcpy(out->node_hex, cp, sizeof(out->node_hex));
got_node_id=1;
}
} while ((cp = tor_strtok_r(NULL, " \t", &strtok_state)));
if (got_bw && got_node_id) {
tor_free(line);
return 0;
} else if (line_is_after_headers == 0) {
/* There could be additional header lines, therefore do not give warnings
* but returns -1 since it's not a complete bw line. */
log_debug(LD_DIRSERV, "Missing bw or node_id in bandwidth file line: %s",
escaped(orig_line));
tor_free(line);
return -1;
} else {
log_warn(LD_DIRSERV, "Incomplete line in bandwidth file: %s",
escaped(orig_line));
tor_free(line);
return -1;
}
}
