Compare commits

..

3 Commits

Author SHA1 Message Date
Xavier Roche
75fd5ae90f Extract the duplicated stats-refresh loop tick into hts_loop_tick()
The wait-for-socket macro body (stats refresh + loop callback) also
existed open-coded at six more sites across htsname.c and htsparse.c,
differing only in the slot index passed to the callback and the abort
action. Collapse all of them onto a shared hts_loop_tick(); each caller
keeps its own abort path.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Signed-off-by: Xavier Roche <roche@httrack.com>
2026-07-04 14:55:52 +02:00
Xavier Roche
1e7744865f Fold the twin wait-for-socket macros into one shared function
URLSAVENAME_WAIT_FOR_AVAILABLE_SOCKET (htsname.c) and
WAIT_FOR_AVAILABLE_SOCKET (htsparse.h) were byte-for-byte duplicates,
both carrying the #481 checkmirror break. Replace them with a single
hts_wait_available_socket() in htscore.c; the callback-abort path still
returns -1 from the callers.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Signed-off-by: Xavier Roche <roche@httrack.com>
2026-07-04 14:51:15 +02:00
Xavier Roche
dfafe28002 Enforce the -E time limit inside the transfer wait cycle (#482)
-E was only evaluated at per-link boundaries, so a slow or throttling
server starved the check for minutes, and the smooth stop it finally
requested drained the remaining transfers at server pace with no bound.
back_wait now checks the deadline every cycle and, once a short grace
period expires, aborts the in-flight HTTP transfers like the -T timeout
path does (FTP slots stay with their owning thread). back_checkmirror's
0 return, previously dead, now carries the hard stop.

Signed-off-by: Xavier Roche <roche@httrack.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-07-04 13:13:33 +02:00
5 changed files with 69 additions and 169 deletions

View File

@@ -3371,6 +3371,41 @@ int back_pluggable_sockets_strict(struct_back * sback, httrackp * opt) {
return n;
}
/* One engine-loop tick: refresh the transfer stats and run the loop callback
for slot b (-1 = none). HTS_FALSE = the callback requested an abort. */
hts_boolean hts_loop_tick(struct_back *sback, httrackp *opt, int b, int ptr) {
engine_stats();
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
return RUN_CALLBACK7(
opt, loop, sback->lnk, sback->count, b, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)
? HTS_TRUE
: HTS_FALSE;
}
/* Single implementation of the historical WAIT_FOR_AVAILABLE_SOCKET macros. */
hts_boolean hts_wait_available_socket(struct_back *sback, httrackp *opt,
cache_back *cache, int ptr) {
const int prev = opt->state._hts_in_html_parsing;
while (back_pluggable_sockets_strict(sback, opt) <= 0) {
opt->state._hts_in_html_parsing = 6;
back_wait(sback, opt, cache, 0);
/* time limit (-E) exceeded: stop waiting for a socket (#481) */
if (!back_checkmirror(opt))
break;
if (!hts_loop_tick(sback, opt, -1, ptr))
return HTS_FALSE;
}
opt->state._hts_in_html_parsing = prev;
return HTS_TRUE;
}
int back_pluggable_sockets(struct_back * sback, httrackp * opt) {
int n;

View File

@@ -432,6 +432,15 @@ int back_pluggable_sockets(struct_back * sback, httrackp * opt);
int back_pluggable_sockets_strict(struct_back * sback, httrackp * opt);
/* One engine-loop tick: refresh the transfer stats and run the loop callback
for slot b (-1 = none). HTS_FALSE = the callback requested an abort. */
hts_boolean hts_loop_tick(struct_back *sback, httrackp *opt, int b, int ptr);
/* Wait until a test socket can be plugged, pumping transfers, stats and the
loop callback; gives up past the -E deadline. HTS_FALSE = callback abort. */
hts_boolean hts_wait_available_socket(struct_back *sback, httrackp *opt,
cache_back *cache, int ptr);
/* Randomized inter-file pause target in [min_ms,max_ms] (#185), derived from a
timestamp seed so it is stable within one gap and rerolls per launch. */
int hts_pause_target_ms(TStamp seed, int min_ms, int max_ms);

View File

@@ -74,37 +74,6 @@ static const char *hts_tbdev[] = {
""
};
#define URLSAVENAME_WAIT_FOR_AVAILABLE_SOCKET() \
do { \
int prev = opt->state._hts_in_html_parsing; \
while (back_pluggable_sockets_strict(sback, opt) <= 0) { \
opt->state._hts_in_html_parsing = 6; \
/* Wait .. */ \
back_wait(sback, opt, cache, 0); \
/* time limit (-E) exceeded: stop waiting for a socket (#481) */ \
if (!back_checkmirror(opt)) \
break; \
/* Transfer rate */ \
engine_stats(); \
/* Refresh various stats */ \
HTS_STAT.stat_nsocket = back_nsoc(sback); \
HTS_STAT.stat_errors = fspc(opt, NULL, "error"); \
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning"); \
HTS_STAT.stat_infos = fspc(opt, NULL, "info"); \
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr); \
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback); \
/* Check */ \
{ \
if (!RUN_CALLBACK7( \
opt, loop, sback->lnk, sback->count, -1, ptr, opt->lien_tot, \
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)) { \
return -1; \
} \
} \
} \
opt->state._hts_in_html_parsing = prev; \
} while (0)
/* Strip all // */
static void cleanDoubleSlash(char *s) {
int i, j;
@@ -658,11 +627,10 @@ int url_savename(lien_adrfilsave *const afs,
int has_been_moved = 0;
lien_adrfil current;
/* Ensure we don't use too many sockets by using a "testing" one
If we have only 1 simultaneous connection authorized, wait for pending download
Wait for an available slot
/* Wait for an available test slot, honoring the connection limits
*/
URLSAVENAME_WAIT_FOR_AVAILABLE_SOCKET();
if (!hts_wait_available_socket(sback, opt, cache, ptr))
return -1;
/* Rock'in */
current.adr[0] = current.fil[0] = '\0';
@@ -692,24 +660,11 @@ int url_savename(lien_adrfilsave *const afs,
if (ptr >= 0) {
back_fillmax(sback, opt, cache, ptr, numero_passe);
}
// on est obligé d'appeler le shell pour le refresh..
// Transfer rate
engine_stats();
// Refresh various stats
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
if (!RUN_CALLBACK7
(opt, loop, sback->lnk, sback->count, b, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart),
&HTS_STAT)) {
if (!hts_loop_tick(sback, opt, b, ptr)) {
return -1;
} else if (opt->state._hts_cancel || !back_checkmirror(opt)) { // cancel 2 ou 1 (cancel parsing)
} else if (opt->state._hts_cancel ||
!back_checkmirror(
opt)) { // cancel level 2 or 1 (cancel parsing)
back_delete(opt, cache, sback, b); // cancel test
stop_looping = 1;
}
@@ -774,8 +729,9 @@ int url_savename(lien_adrfilsave *const afs,
"Loop with HEAD request (during prefetch) at %s%s",
current.adr, current.fil);
}
// Ajouter
URLSAVENAME_WAIT_FOR_AVAILABLE_SOCKET();
if (!hts_wait_available_socket(sback, opt,
cache, ptr))
return -1;
if (back_add(sback, opt, cache, moved.adr, moved.fil, methode, referer_adr, referer_fil, 1) != -1) { // OK
hts_log_print(opt, LOG_DEBUG,
"(during prefetch) %s (%d) to link %s at %s%s",

View File

@@ -3399,20 +3399,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
back_wait(sback, opt, cache, HTS_STAT.stat_timestart);
back_fillmax(sback, opt, cache, ptr, numero_passe);
// Transfer rate
engine_stats();
// Refresh various stats
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
if (!RUN_CALLBACK7
(opt, loop, sback->lnk, sback->count, 0, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)) {
if (!hts_loop_tick(sback, opt, 0, ptr)) {
hts_log_print(opt, LOG_ERROR, "Exit requested by shell or user");
*stre->exit_xh_ = 1; // exit requested
XH_uninit;
@@ -3423,7 +3410,6 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
nofollow = 1; // moins violent
opt->state._hts_cancel = 0;
}
}
// refresh the backing system each 2 seconds
if (engine_stats()) {
@@ -3960,22 +3946,8 @@ void hts_mirror_process_user_interaction(htsmoduleStruct * str,
{
back_wait(sback, opt, cache, HTS_STAT.stat_timestart);
// Transfer rate
engine_stats();
// Refresh various stats
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
b = 0;
if (!RUN_CALLBACK7
(opt, loop, sback->lnk, sback->count, b, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)
|| !back_checkmirror(opt)) {
if (!hts_loop_tick(sback, opt, b, ptr) || !back_checkmirror(opt)) {
hts_log_print(opt, LOG_ERROR, "Exit requested by shell or user");
*stre->exit_xh_ = 1; // exit requested
XH_uninit;
@@ -4081,20 +4053,7 @@ void hts_mirror_process_user_interaction(htsmoduleStruct * str,
if (!back_checkmirror(opt))
break;
// Transfer rate
engine_stats();
// Refresh various stats
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
if (!RUN_CALLBACK7
(opt, loop, sback->lnk, sback->count, b, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)) {
if (!hts_loop_tick(sback, opt, b, ptr)) {
hts_log_print(opt, LOG_ERROR, "Exit requested by shell or user");
*stre->exit_xh_ = 1; // exit requested
XH_uninit;
@@ -4281,26 +4240,12 @@ int hts_mirror_wait_for_next_file(htsmoduleStruct * str,
freet(s);
}
// Transfer rate
engine_stats();
// Refresh various stats
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
if (!RUN_CALLBACK7
(opt, loop, sback->lnk, sback->count, b, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)) {
if (!hts_loop_tick(sback, opt, b, ptr)) {
hts_log_print(opt, LOG_ERROR, "Exit requested by shell or user");
*stre->exit_xh_ = 1; // exit requested
XH_uninit;
return 0;
}
}
#if HTS_POLL
@@ -4533,10 +4478,9 @@ int hts_wait_delayed(htsmoduleStruct * str, lien_adrfilsave *afs,
IS_DELAYED_EXT(afs->save) && continue_loop && loops < 7; loops++) {
continue_loop = 0;
/*
Wait for an available slot
*/
WAIT_FOR_AVAILABLE_SOCKET();
/* Wait for an available slot */
if (!hts_wait_available_socket(sback, opt, cache, ptr))
return -1;
/* We can lookup directly in the cache to speedup this mess */
if (opt->delayed_cached) {
@@ -4682,29 +4626,14 @@ int hts_wait_delayed(htsmoduleStruct * str, lien_adrfilsave *afs,
if (ptr >= 0) {
back_fillmax(sback, opt, cache, ptr, numero_passe);
}
// on est obligé d'appeler le shell pour le refresh..
{
// Transfer rate
engine_stats();
// Refresh various stats
HTS_STAT.stat_nsocket = back_nsoc(sback);
HTS_STAT.stat_errors = fspc(opt, NULL, "error");
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning");
HTS_STAT.stat_infos = fspc(opt, NULL, "info");
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr);
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback);
if (!RUN_CALLBACK7
(opt, loop, sback->lnk, sback->count, b, ptr, opt->lien_tot,
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)) {
back_set_unlocked(sback, b);
return -1;
} else if (opt->state._hts_cancel || !back_checkmirror(opt)) { // cancel 2 ou 1 (cancel parsing)
back_delete(opt, cache, sback, b); // cancel test
break;
}
if (!hts_loop_tick(sback, opt, b, ptr)) {
back_set_unlocked(sback, b);
return -1;
} else if (opt->state._hts_cancel ||
!back_checkmirror(
opt)) { // cancel level 2 or 1 (cancel parsing)
back_delete(opt, cache, sback, b); // cancel test
break;
}
} while (
/* dns/connect/request */

View File

@@ -175,33 +175,4 @@ int hts_wait_delayed(htsmoduleStruct * str, lien_adrfilsave *afs,
/* Apply changes */ \
* str->ptr_ = ptr
#define WAIT_FOR_AVAILABLE_SOCKET() \
do { \
int prev = opt->state._hts_in_html_parsing; \
while (back_pluggable_sockets_strict(sback, opt) <= 0) { \
opt->state._hts_in_html_parsing = 6; \
/* Wait .. */ \
back_wait(sback, opt, cache, 0); \
/* time limit (-E) exceeded: stop waiting for a socket (#481) */ \
if (!back_checkmirror(opt)) \
break; \
/* Transfer rate */ \
engine_stats(); \
/* Refresh various stats */ \
HTS_STAT.stat_nsocket = back_nsoc(sback); \
HTS_STAT.stat_errors = fspc(opt, NULL, "error"); \
HTS_STAT.stat_warnings = fspc(opt, NULL, "warning"); \
HTS_STAT.stat_infos = fspc(opt, NULL, "info"); \
HTS_STAT.nbk = backlinks_done(sback, opt->liens, opt->lien_tot, ptr); \
HTS_STAT.nb = back_transferred(HTS_STAT.stat_bytes, sback); \
/* Check */ \
if (!RUN_CALLBACK7( \
opt, loop, sback->lnk, sback->count, -1, ptr, opt->lien_tot, \
(int) (time_local() - HTS_STAT.stat_timestart), &HTS_STAT)) { \
return -1; \
} \
} \
opt->state._hts_in_html_parsing = prev; \
} while (0)
#endif