int mget_cookie_load_public_suffixes(const char *fname)
{
PUBLIC_SUFFIX suffix, *suffixp;
FILE *fp;
int nsuffixes = 0;
char *buf = NULL, *linep, *p;
size_t bufsize = 0;
ssize_t buflen;
// as of 02.11.2012, the list at http://publicsuffix.org/list/ contains ~6000 rules
// and 40 exceptions.
if (!suffixes)
suffixes = mget_vector_create(6*1024, -2, (int(*)(const void *, const void *))suffix_compare);
if (!suffix_exceptions)
suffix_exceptions = mget_vector_create(64, -2, (int(*)(const void *, const void *))suffix_compare);
if ((fp = fopen(fname, "r"))) {
while ((buflen = mget_getline(&buf, &bufsize, fp)) >= 0) {
linep = buf;
while (isspace(*linep)) linep++; // ignore leading whitespace
if (!*linep) continue; // skip empty lines
if (*linep == '/' && linep[1] == '/')
continue; // skip comments
// parse suffix rule
for (p = linep; *linep && !isspace(*linep);) linep++;
*linep = 0;
if (*p == '!') {
// add to exceptions
suffix_init(&suffix, p + 1, linep - p - 1);
suffixp = mget_vector_get(suffix_exceptions, mget_vector_add(suffix_exceptions, &suffix, sizeof(suffix)));
} else {
suffix_init(&suffix, p, linep - p);
suffixp = mget_vector_get(suffixes, mget_vector_add(suffixes, &suffix, sizeof(suffix)));
}
if (suffixp)
suffixp->label = suffixp->label_buf; // set label to changed address
nsuffixes++;;
}
xfree(buf);
fclose(fp);
mget_vector_sort(suffix_exceptions);
mget_vector_sort(suffixes);
} else
error_printf(_("Failed to open public suffix file '%s'\n"), fname);
return nsuffixes;
}
static void _rss_get_url(void *context, int flags, const char *dir, const char *attr, const char *val, size_t len, size_t pos G_GNUC_MGET_UNUSED)
{
struct rss_context *ctx = context;
mget_string_t url;
if (!val || !len)
return;
url.p = NULL;
if ((flags & XML_FLG_ATTRIBUTE)) {
if (!mget_strcasecmp_ascii(attr, "url") || !mget_strcasecmp_ascii(attr, "href")
|| !mget_strcasecmp_ascii(attr, "src") || !mget_strcasecmp_ascii(attr, "domain")
|| !mget_strcasecmp_ascii(attr, "xmlns") || !mget_strncasecmp_ascii(attr, "xmlns:", 6))
{
for (;len && isspace(*val); val++, len--); // skip leading spaces
for (;len && isspace(val[len - 1]); len--); // skip trailing spaces
url.p = val;
url.len = len;
if (!ctx->urls)
ctx->urls = mget_vector_create(32, -2, NULL);
mget_vector_add(ctx->urls, &url, sizeof(url));
}
}
else if ((flags & XML_FLG_CONTENT)) {
const char *elem = strrchr(dir, '/');
if (elem) {
elem++;
if (!mget_strcasecmp_ascii(elem, "guid") || !mget_strcasecmp_ascii(elem, "link")
|| !mget_strcasecmp_ascii(elem, "comments") || !mget_strcasecmp_ascii(elem, "docs"))
{
for (;len && isspace(*val); val++, len--); // skip leading spaces
for (;len && isspace(val[len - 1]); len--); // skip trailing spaces
// debug_printf("#2 %02X %s %s '%.*s' %zd\n", flags, dir, attr, (int) len, val, len);
url.p = val;
url.len = len;
if (!ctx->urls)
ctx->urls = mget_vector_create(32, -2, NULL);
mget_vector_add(ctx->urls, &url, sizeof(url));
}
}
}
}
void mget_cookie_store_cookie(MGET_COOKIE *cookie)
{
MGET_COOKIE *old;
int pos;
debug_printf("got cookie %s=%s\n", cookie->name, cookie->value);
if (!cookie->normalized)
return;
pthread_mutex_lock(&cookies_mutex);
if (!cookies) {
cookies = mget_vector_create(128, -2, (int(*)(const void *, const void *))compare_cookie);
old = NULL;
} else
old = mget_vector_get(cookies, pos = mget_vector_find(cookies, cookie));
if (old) {
debug_printf("replace old cookie %s=%s\n", cookie->name, cookie->value);
cookie->creation = old->creation;
mget_cookie_free_cookie(old);
mget_vector_replace(cookies, cookie, sizeof(*cookie), pos);
} else {
debug_printf("store new cookie %s=%s\n", cookie->name, cookie->value);
mget_vector_insert_sorted(cookies, cookie, sizeof(*cookie));
}
pthread_mutex_unlock(&cookies_mutex);
}
mget_cookie_db_t *mget_cookie_db_init(mget_cookie_db_t *cookie_db)
{
if (!cookie_db)
cookie_db = xmalloc(sizeof(mget_cookie_db_t));
memset(cookie_db, 0, sizeof(*cookie_db));
cookie_db->cookies = mget_vector_create(32, -2, (int(*)(const void *, const void *))_compare_cookie);
mget_vector_set_destructor(cookie_db->cookies, (void(*)(void *))mget_cookie_deinit);
mget_thread_mutex_init(&cookie_db->mutex);
#ifdef WITH_LIBPSL
cookie_db->psl = (psl_ctx_t *)psl_builtin();
#endif
return cookie_db;
}
static void test_vector(void)
{
struct ENTRY
*tmp,
txt_sorted[5] = { {""}, {"four"}, {"one"}, {"three"}, {"two"} },
*txt[countof(txt_sorted)];
mget_vector_t
*v = mget_vector_create(2, -2, (int(*)(const void *, const void *))compare_txt);
unsigned
it;
int
n;
// copy
for (it = 0; it < countof(txt); it++)
txt[it] = &txt_sorted[it];
// shuffle txt
for (it = 0; it < countof(txt); it++) {
n = rand() % countof(txt);
tmp = txt[n];
txt[n] = txt[it];
txt[it] = tmp;
}
for (it = 0; it < countof(txt); it++) {
mget_vector_insert_sorted(v, txt[it], sizeof(struct ENTRY));
}
for (it = 0; it < countof(txt); it++) {
struct ENTRY *e = mget_vector_get(v, it);
if (!strcmp(e->txt,txt_sorted[it].txt))
ok++;
else
failed++;
}
mget_vector_free(&v);
}
void job_create_parts(JOB *job)
{
PART part;
mget_metalink_t *metalink;
ssize_t fsize;
int it;
if (!job || !(metalink = job->metalink))
return;
memset(&part, 0, sizeof(PART));
// create space to hold enough parts
if (!job->parts)
job->parts = mget_vector_create(mget_vector_size(metalink->pieces), 4, NULL);
else
mget_vector_clear(job->parts);
fsize = metalink->size;
for (it = 0; it < mget_vector_size(metalink->pieces); it++) {
mget_metalink_piece_t *piece = mget_vector_get(metalink->pieces, it);
if (fsize >= piece->length) {
part.length = piece->length;
} else {
part.length = fsize;
}
part.id = it + 1;
mget_vector_add(job->parts, &part, sizeof(PART));
part.position += part.length;
fsize -= piece->length;
}
}
static void test_stringmap(void)
{
mget_stringmap_t *m;
char key[128], value[128], *val;
int run, it;
size_t valuesize;
// the initial size of 16 forces the internal reshashing function to be called twice
m = mget_stringmap_create(16);
for (run = 0; run < 2; run++) {
if (run) {
mget_stringmap_clear(m);
mget_stringmap_sethashfunc(m, hash_txt);
}
for (it = 0; it < 26; it++) {
sprintf(key, "http://www.example.com/subdir/%d.html", it);
valuesize = sprintf(value, "%d.html", it);
if (mget_stringmap_put(m, key, value, valuesize + 1)) {
failed++;
info_printf("stringmap_put(%s) returns unexpected old value\n", key);
} else ok++;
}
if ((it = mget_stringmap_size(m)) != 26) {
failed++;
info_printf("stringmap_size() returned %d (expected %d)\n", it, 26);
} else ok++;
// now, look up every single entry
for (it = 0; it < 26; it++) {
sprintf(key, "http://www.example.com/subdir/%d.html", it);
sprintf(value, "%d.html", it);
if (!(val = mget_stringmap_get(m, key))) {
failed++;
info_printf("stringmap_get(%s) didn't find entry\n", key);
} else if (strcmp(val, value)) {
failed++;
info_printf("stringmap_get(%s) found '%s' (expected '%s')\n", key, val, value);
} else ok++;
}
mget_stringmap_clear(m);
if ((it = mget_stringmap_size(m)) != 0) {
failed++;
info_printf("stringmap_size() returned %d (expected 0)\n", it);
} else ok++;
for (it = 0; it < 26; it++) {
sprintf(key, "http://www.example.com/subdir/%d.html", it);
valuesize = sprintf(value, "%d.html", it);
if (mget_stringmap_put(m, key, value, valuesize + 1)) {
failed++;
info_printf("stringmap_put(%s) returns unexpected old value\n", key);
} else ok++;
}
if ((it = mget_stringmap_size(m)) != 26) {
failed++;
info_printf("stringmap_size() returned %d (expected %d)\n", it, 26);
} else ok++;
// now, remove every single entry
for (it = 0; it < 26; it++) {
sprintf(key, "http://www.example.com/subdir/%d.html", it);
sprintf(value, "%d.html", it);
mget_stringmap_remove(m, key);
}
if ((it = mget_stringmap_size(m)) != 0) {
failed++;
info_printf("stringmap_size() returned %d (expected 0)\n", it);
} else ok++;
for (it = 0; it < 26; it++) {
sprintf(key, "http://www.example.com/subdir/%d.html", it);
valuesize = sprintf(value, "%d.html", it);
if (mget_stringmap_put(m, key, value, valuesize + 1)) {
failed++;
info_printf("stringmap_put(%s) returns unexpected old value\n", key);
} else ok++;
}
if ((it = mget_stringmap_size(m)) != 26) {
failed++;
info_printf("stringmap_size() returned %d (expected %d)\n", it, 26);
} else ok++;
}
// testing alloc/free in stringmap/hashmap
mget_stringmap_clear(m);
mget_stringmap_put(m, "thekey", NULL, 0) ? failed++ : ok++;
mget_stringmap_put(m, "thekey", NULL, 0) ? ok++ : failed++;
mget_stringmap_put(m, "thekey", "thevalue", 9) ? ok++ : failed++;
mget_stringmap_put(m, "thekey", "thevalue", 9) ? ok++ : failed++;
mget_stringmap_put(m, "thekey", NULL, 0) ? ok++ : failed++;
// testing key/value identity alloc/free in stringmap/hashmap
mget_stringmap_clear(m);
mget_stringmap_put(m, "thekey", NULL, 0) ? failed++ : ok++;
mget_stringmap_put(m, "thekey", NULL, 0) ? ok++ : failed++;
mget_stringmap_put(m, "thekey", "thevalue", 9) ? ok++ : failed++;
mget_stringmap_put(m, "thekey", NULL, 0) ? ok++ : failed++;
mget_stringmap_free(&m);
mget_http_challenge_t challenge;
mget_http_parse_challenge("Basic realm=\"test realm\"", &challenge);
mget_http_free_challenge(&challenge);
mget_vector_t *challenges;
challenges = mget_vector_create(2, 2, NULL);
mget_vector_set_destructor(challenges, (void(*)(void *))mget_http_free_challenge);
mget_http_parse_challenge("Basic realm=\"test realm\"", &challenge);
mget_vector_add(challenges, &challenge, sizeof(challenge));
mget_http_free_challenges(&challenges);
char *response_text = strdup(
"HTTP/1.1 401 Authorization Required\r\n"\
"Date: Sun, 23 Dec 2012 21:03:45 GMT\r\n"\
"Server: Apache/2.2.22 (Debian)\r\n"\
"WWW-Authenticate: Digest realm=\"therealm\", nonce=\"Ip6MaovRBAA=c4af733c51270698260f5d357724c2cbce20fa3d\", algorithm=MD5, domain=\"/prot_digest_md5\", qop=\"auth\"\r\n"\
"Vary: Accept-Encoding\r\n"\
"Content-Length: 476\r\n"\
"Keep-Alive: timeout=5, max=99\r\n"\
"Connection: Keep-Alive\r\n"\
"Content-Type: text/html; charset=iso-8859-1\r\n\r\n");
mget_iri_t *iri = mget_iri_parse("http://localhost/prot_digest_md5/", NULL);
mget_http_request_t *req = mget_http_create_request(iri, "GET");
mget_http_response_t *resp = mget_http_parse_response_header(response_text);
mget_http_add_credentials(req, mget_vector_get(resp->challenges, 0), "tim", "123");
// for (it=0;it<vec_size(req->lines);it++) {
// info_printf("%s\n", (char *)vec_get(req->lines, it));
// }
mget_http_free_response(&resp);
mget_http_free_request(&req);
mget_iri_free(&iri);
xfree(response_text);
// Authorization: Digest username="******", realm="therealm", nonce="Ip6MaovRBAA=c4af733c51270698260f5d357724c2cbce20fa3d", uri="/prot_digest_md5/", response="a99e2012d507a73dd46eb044d3f4641c", qop=auth, nc=00000001, cnonce="3d20faa1"
}
static void test_parse_challenge(void)
{
static const struct test_data {
const char *
input;
const char *
scheme[3];
} test_data[] = {
{ // simplebasic
"Basic realm=\"foo\"",
{ "Basic", NULL }
},
{ // simplebasicucase
"BASIC REALM=\"foo\"",
{ "Basic", NULL }
},
{ // simplebasicucase
"Basic , realm=\"foo\"",
{ "Basic", NULL }
},
{ //
"Basic realm=\"test realm\"",
{ "Basic", NULL }
},
{ //
"Basic realm=\"test-äöÜ\"",
{ "Basic", NULL }
},
{ //
"Basic realm=\"basic\", Newauth realm=\"newauth\"",
{ "Basic", "Newauth", NULL }
},
};
mget_vector_t *challenges;
mget_http_challenge_t *challenge;
// Testcases found here http://greenbytes.de/tech/tc/httpauth/
challenges = mget_vector_create(2, 2, NULL);
mget_vector_set_destructor(challenges, (void(*)(void *))mget_http_free_challenge);
for (unsigned it = 0; it < countof(test_data); it++) {
const struct test_data *t = &test_data[it];
mget_http_parse_challenges(t->input, challenges);
for (unsigned nchal = 0; nchal < countof(test_data[0].scheme) && t->scheme[nchal]; nchal++) {
challenge = mget_vector_get(challenges, nchal);
if (!t->scheme[nchal]) {
if (challenge) {
failed++;
info_printf("Failed [%u]: mget_http_parse_challenges(%s) found %d challenges (expected %d)\n", it, t->input, mget_vector_size(challenges), nchal);
}
break;
}
if (!challenge) {
failed++;
info_printf("Failed [%u]: mget_http_parse_challenges(%s) did not find enough challenges\n", it, t->input);
break;
}
if (!mget_strcasecmp_ascii(challenge->auth_scheme, t->scheme[nchal])) {
ok++;
} else {
failed++;
info_printf("Failed [%u]: mget_http_parse_challenges(%s) -> '%s' (expected '%s')\n", it, t->input, challenge->auth_scheme, t->scheme[nchal]);
}
}
mget_vector_clear(challenges);
}
mget_http_free_challenges(&challenges);
}
int job_validate_file(JOB *job)
{
PART part;
mget_metalink_t *metalink;
off_t fsize;
int fd, rc = -1, it;
struct stat st;
if (!job || !(metalink = job->metalink))
return 0;
memset(&part, 0, sizeof(PART));
// create space to hold enough parts
if (!job->parts)
job->parts = mget_vector_create(mget_vector_size(metalink->pieces), 4, NULL);
else
mget_vector_clear(job->parts);
fsize = metalink->size;
if (mget_vector_size(metalink->hashes) == 0) {
// multipart non-metalink download: do not clobber if file has expected size
if (stat(metalink->name, &st) == 0 && st.st_size == fsize) {
return 1; // we are done
}
}
// truncate file if needed
if (stat(metalink->name, &st) == 0 && st.st_size > fsize) {
if (truncate(metalink->name, fsize) == -1)
error_printf(_("Failed to truncate %s\n from %llu to %llu bytes\n"),
metalink->name, (unsigned long long)st.st_size, (unsigned long long)fsize);
}
if ((fd = open(metalink->name, O_RDONLY)) != -1) {
// file exists, check which piece is invalid and requeue it
for (it = 0; errno != EINTR && it < mget_vector_size(metalink->hashes); it++) {
mget_metalink_hash_t *hash = mget_vector_get(metalink->hashes, it);
if ((rc = check_file_fd(hash, fd)) == -1)
continue; // hash type not available, try next
break;
}
if (rc == 1) {
info_printf(_("Checksum OK for '%s'\n"), metalink->name);
return 1; // we are done
}
else if (rc == -1) {
// failed to check file, continue as if file is ok
info_printf(_("Failed to build checksum, assuming file to be OK\n"));
return 1; // we are done
} else
info_printf(_("Bad checksum for '%s'\n"), metalink->name);
// if (vec_size(metalink->pieces) < 1)
// return;
for (it = 0; errno != EINTR && it < mget_vector_size(metalink->pieces); it++) {
mget_metalink_piece_t *piece = mget_vector_get(metalink->pieces, it);
mget_metalink_hash_t *hash = &piece->hash;
if (fsize >= piece->length) {
part.length = piece->length;
} else {
part.length = (size_t)fsize;
}
part.id = it + 1;
if ((rc = check_piece_hash(hash, fd, part.position, part.length)) != 1) {
info_printf(_("Piece %d/%d not OK - requeuing\n"), it + 1, mget_vector_size(metalink->pieces));
mget_vector_add(job->parts, &part, sizeof(PART));
debug_printf(" need to download %llu bytes from pos=%llu\n",
(unsigned long long)part.length, (unsigned long long)part.position);
}
part.position += part.length;
fsize -= piece->length;
}
close(fd);
} else {
for (it = 0; it < mget_vector_size(metalink->pieces); it++) {
mget_metalink_piece_t *piece = mget_vector_get(metalink->pieces, it);
if (fsize >= piece->length) {
part.length = piece->length;
} else {
part.length = fsize;
}
part.id = it + 1;
mget_vector_add(job->parts, &part, sizeof(PART));
part.position += part.length;
fsize -= piece->length;
}
}
return 0;
}
ROBOTS *mget_robots_parse(const char *data)
{
ROBOTS *robots;
ROBOTS_PATH path;
int collect = 0;
const char *p;
if (!data || !*data)
return NULL;
robots = xcalloc(1, sizeof (ROBOTS));
do {
if (collect < 2 && !strncasecmp(data, "User-agent:", 11)) {
if (!collect) {
for (data += 11; *data == ' ' || *data == '\t'; data++);
if (!strncasecmp(data, "mget", 4)) {
collect = 1;
}
else if (*data == '*') {
collect = 1;
}
} else
collect = 2;
}
else if (collect == 1 && !strncasecmp(data, "Disallow:", 9)) {
for (data += 9; *data == ' ' || *data == '\t'; data++);
if (*data == '\r' || *data == '\n' || !*data) {
// all allowed
mget_vector_free(&robots->paths);
collect = 2;
} else {
if (!robots->paths) {
robots->paths = mget_vector_create(32, -2, NULL);
mget_vector_set_destructor(robots->paths, (void(*)(void *))_free_path);
}
for (p = data; !isspace(*p); p++);
path.len = p - data;
path.path = strndup(data, path.len);
mget_vector_add(robots->paths, &path, sizeof(path));
}
}
else if (!strncasecmp(data, "Sitemap:", 8)) {
for (data += 8; *data==' ' || *data == '\t'; data++);
for (p = data; !isspace(*p); p++);
if (!robots->sitemaps)
robots->sitemaps = mget_vector_create(4, -2, NULL);
mget_vector_add_noalloc(robots->sitemaps, strndup(data, p - data));
}
if ((data = strchr(data, '\n')))
data++; // point to next line
} while (data && *data);
/*
for (int it = 0; it < mget_vector_size(robots->paths); it++) {
ROBOTS_PATH *path = mget_vector_get(robots->paths, it);
info_printf("path '%s'\n", path->path);
}
for (int it = 0; it < mget_vector_size(robots->sitemaps); it++) {
const char *sitemap = mget_vector_get(robots->sitemaps, it);
info_printf("sitemap '%s'\n", sitemap);
}
*/
return robots;
}
