Merge origin/master into warnfix-dead

# Conflicts:
#	tests/Makefile.am

.github/workflows/windows-build.yml

@@ -225,8 +225,9 @@ jobs:
 
           # Every gate here exits 77, so an all-skipped suite would report green having
           # tested nothing: pin the skips, and floor the passes in case the glob empties.
-          # footer-overflow skips on Windows (needs a path past MAX_PATH); crange pending #581.
-          expected_skips=" 01_engine-footer-overflow.test 48_local-crange-memresume.test 71_local-crange-repaircache.test"
+          # footer-overflow skips on Windows (needs a path past MAX_PATH); crange pending #581;
+          # webdav-mime needs a reapable background listener, which MSYS cannot give it.
+          expected_skips=" 01_engine-footer-overflow.test 48_local-crange-memresume.test 71_local-crange-repaircache.test 79_local-proxytrack-webdav-mime.test"
           [ "$pass" -ge 90 ] || { echo "::error::only $pass tests passed ($skip skipped)"; exit 1; }
           [ "$skipped" = "$expected_skips" ] || { echo "::error::unexpected skips:$skipped"; exit 1; }
           [ "$fail" -eq 0 ] || { echo "::error::failing:$failed"; exit 1; }

src/htsback.c

@@ -541,7 +541,7 @@ static int create_back_tmpfile(httrackp *opt, lien_back *const back,
                                const char *ext) {
   // do not use tempnam() but a regular filename
   back->tmpfile_buffer[0] = '\0';
-  if (back->url_sav != NULL && back->url_sav[0] != '\0') {
+  if (back->url_sav[0] != '\0') {
     snprintf(back->tmpfile_buffer, sizeof(back->tmpfile_buffer), "%s.%s",
              back->url_sav, ext);
     back->tmpfile = back->tmpfile_buffer;
@@ -887,8 +887,7 @@ int back_finalize(httrackp * opt, cache_back * cache, struct_back * sback,
               HTS_STAT.stat_bytes += back[p].r.size;
               HTS_STAT.stat_files++;
               hts_log_print(opt, LOG_TRACE, "added file %s%s => %s",
-                            back[p].url_adr, back[p].url_fil,
-                            back[p].url_sav != NULL ? back[p].url_sav : "");
+                            back[p].url_adr, back[p].url_fil, back[p].url_sav);
             }
             if ((!back[p].r.notmodified) && (opt->is_update)) {
               HTS_STAT.stat_updated_files++;    // page modifiée
@@ -2302,26 +2301,24 @@ int back_add(struct_back *sback, httrackp *opt, cache_back *cache,
               && slot_can_be_finalized(opt, &back[i]);
             int may_serialize = slot_can_be_cached_on_disk(&back[i]);
 
-            hts_log_print(opt, LOG_DEBUG,
-                          "back[%03d]: may_clean=%d, may_finalize_disk=%d, may_serialize=%d:"
-                          LF "\t"
-                          "finalized(%d), status(%d), locked(%d), delayed(%d), test(%d), "
-                          LF "\t"
-                          "statuscode(%d), size(%d), is_write(%d), may_hypertext(%d), "
-                          LF "\t" "contenttype(%s), url(%s%s), save(%s)", i,
-                          may_clean, may_finalize, may_serialize,
-                          back[i].finalized, back[i].status, back[i].locked,
-                          IS_DELAYED_EXT(back[i].url_sav), back[i].testmode,
-                          back[i].r.statuscode, (int) back[i].r.size,
-                          back[i].r.is_write, may_be_hypertext_mime(opt,
-                                                                    back[i].r.
-                                                                    contenttype,
-                                                                    back[i].
-                                                                    url_fil),
-                          /* */
-                          back[i].r.contenttype, back[i].url_adr,
-                          back[i].url_fil,
-                          back[i].url_sav ? back[i].url_sav : "<null>");
+            hts_log_print(
+                opt, LOG_DEBUG,
+                "back[%03d]: may_clean=%d, may_finalize_disk=%d, "
+                "may_serialize=%d:" LF "\t"
+                "finalized(%d), status(%d), locked(%d), delayed(%d), "
+                "test(%d), " LF "\t"
+                "statuscode(%d), size(%d), is_write(%d), may_hypertext(%d), " LF
+                "\t"
+                "contenttype(%s), url(%s%s), save(%s)",
+                i, may_clean, may_finalize, may_serialize, back[i].finalized,
+                back[i].status, back[i].locked, IS_DELAYED_EXT(back[i].url_sav),
+                back[i].testmode, back[i].r.statuscode, (int) back[i].r.size,
+                back[i].r.is_write,
+                may_be_hypertext_mime(opt, back[i].r.contenttype,
+                                      back[i].url_fil),
+                /* */
+                back[i].r.contenttype, back[i].url_adr, back[i].url_fil,
+                back[i].url_sav);
           }
         }
       }
@@ -2857,7 +2854,8 @@ void back_wait(struct_back * sback, httrackp * opt, cache_back * cache,
               // new session
               back[i].r.ssl_con = SSL_new(openssl_ctx);
               if (back[i].r.ssl_con) {
-                const char* hostname = jump_protocol_const(back[i].url_adr);
+                /* non-const twin: the OpenSSL macro casts the qualifier away */
+                char *hostname = jump_protocol(back[i].url_adr);
                 // some servers expect the hostname on the clienthello (SNI TLS extension)
                 SSL_set_tlsext_host_name(back[i].r.ssl_con, hostname);
                 SSL_clear(back[i].r.ssl_con);

src/htscache.c

@@ -629,8 +629,8 @@ static htsblk cache_readex_new(httrackp * opt, cache_back * cache,
                 // File exists on disk with declared cache name (this is expected!)
                 if (fexist_utf8(fconv(catbuff, sizeof(catbuff), previous_save))) {       // un fichier existe déja
                   // Expected size ?
-                  const size_t fsize =
-                    fsize_utf8(fconv(catbuff, sizeof(catbuff), previous_save));
+                  const LLint fsize = fsize_utf8(
+                      fconv(catbuff, sizeof(catbuff), previous_save));
                   if (fsize == r.size) {
                     // Target name is the previous name, and the file looks good: nothing to do!
                     if (strcmp(previous_save, target_save) == 0) {
@@ -666,7 +666,8 @@ static htsblk cache_readex_new(httrackp * opt, cache_back * cache,
                 // Suppose a broken mirror, with a file being renamed: OK
                 else if (fexist_utf8(fconv(catbuff, sizeof(catbuff), target_save))) {
                   // Expected size ?
-                  const size_t fsize = fsize_utf8(fconv(catbuff, sizeof(catbuff), target_save));
+                  const LLint fsize =
+                      fsize_utf8(fconv(catbuff, sizeof(catbuff), target_save));
 
                   if (fsize == r.size) {
                     // So far so good
@@ -1440,7 +1441,7 @@ int cache_brstr(char *adr, char *s, size_t s_size) {
 
 /* binput bounded to a NUL-terminated buffer: refuse to start a read at or
    past `end`, so a prior over-advance can't walk a cache-index parse OOB. */
-int cache_binput(char *adr, const char *end, char *s, int max) {
+int cache_binput(const char *adr, const char *end, char *s, int max) {
   if (adr >= end) {
     s[0] = '\0';
     return 0;

src/htscache.h

@@ -93,7 +93,7 @@ void cache_rstr(FILE *fp, char *s, size_t s_size);
 char *cache_rstr_addr(FILE * fp);
 int cache_brstr(char *adr, char *s, size_t s_size);
 /* binput over a NUL-terminated buffer, bounded: no read starts at/past end. */
-int cache_binput(char *adr, const char *end, char *s, int max);
+int cache_binput(const char *adr, const char *end, char *s, int max);
 int cache_brint(char *adr, int *i);
 void cache_rint(FILE * fp, int *i);
 void cache_rLLint(FILE * fp, LLint * i);

src/htscore.c

@@ -696,17 +696,19 @@ int httpmirror(char *url1, httrackp * opt) {
   // copier adresse(s) dans liste des adresses
   {
     char *a = url1;
-    int primary_len = 8192;
-
-    if (StringNotEmpty(opt->filelist)) {
-      primary_len += max(0, fsize_utf8(StringBuff(opt->filelist)) * 2);
-    }
-    primary_len += (int) strlen(url1) * 2;
+    const LLint list_sz = StringNotEmpty(opt->filelist)
+                              ? fsize_utf8(StringBuff(opt->filelist))
+                              : 0;
+    /* two bytes reserved per list byte; -1 makes an undoublable size refused */
+    const LLint list_room =
+        list_sz > 0 ? (list_sz <= INT64_MAX / 2 ? list_sz * 2 : -1) : 0;
+    const size_t primary_len =
+        llint_grow_size_t(8192 + strlen(url1) * 2, list_room, 0);
 
     // création de la première page, qui contient les liens de base à scanner
     // c'est plus propre et plus logique que d'entrer à la main les liens dans la pile
     // on bénéficie ainsi des vérifications et des tests du robot pour les liens "primaires"
-    primary = (char *) malloct(primary_len);
+    primary = primary_len != (size_t) -1 ? (char *) malloct(primary_len) : NULL;
     if (!primary) {
       printf("PANIC! : Not enough memory [%d]\n", __LINE__);
       XH_extuninit;
@@ -887,7 +889,7 @@ int httpmirror(char *url1, httrackp * opt) {
       }
 
       if (filelist_buff != NULL) {
-        int filelist_ptr = 0;
+        size_t filelist_ptr = 0;
         int n = 0;
         char BIGSTK line[HTS_URLMAXSIZE * 2];
 

src/htscoremain.c

@@ -145,7 +145,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
   int argv_url = -1;            // ==0 : utiliser cache et doit.log
   char *argv_firsturl = NULL;   // utilisé pour nommage par défaut
   char *url = NULL;             // URLS séparées par un espace
-  int url_sz = 65535;
+  size_t url_sz = 65535;
 
   // the parametres
   int httrack_logmode = 3;      // ONE log file
@@ -224,21 +224,22 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
 
   /* create x_argvblk buffer for transformed command line */
   {
-    int current_size = 0;
-    int size;
+    size_t current_size = 0;
+    const LLint size = fsize("config");
+    size_t blk_size;
     int na;
 
     for(na = 0; na < argc; na++)
-      current_size += (int) (strlen(argv[na]) + 1);
-    if ((size = fsize("config")) > 0)
-      current_size += size;
-    x_argvblk = (char *) malloct(current_size + 32768);
+      current_size += strlen(argv[na]) + 1;
+    /* a huge file named "config" must saturate, not wrap, the capacity */
+    blk_size = llint_grow_size_t(current_size, size > 0 ? size : 0, 32768);
+    x_argvblk = blk_size != (size_t) -1 ? (char *) malloct(blk_size) : NULL;
     if (x_argvblk == NULL) {
       HTS_PANIC_PRINTF("Error, not enough memory");
       htsmain_free();
       return -1;
     }
-    x_argvblk_size = (size_t) (current_size + 32768);
+    x_argvblk_size = blk_size;
     x_argvblk[0] = '\0';
     x_ptr = 0;
 
@@ -1456,20 +1457,29 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
                     FILE *fp = FOPEN(argv[na], "rb");
 
                     if (fp != NULL) {
-                      int cl = (int) strlen(url);
+                      size_t cl = strlen(url);
+                      const size_t fzs = llint_to_size_t(fz);
+                      const size_t capa = llint_grow_size_t(cl, fz, 8192);
 
-                      ensureUrlCapacity(url, url_sz, cl + fz + 8192);
+                      if (capa == (size_t) -1) {
+                        fclose(fp);
+                        HTS_PANIC_PRINTF("File url list too large");
+                        htsmain_free();
+                        return -1;
+                      }
+                      ensureUrlCapacity(url, url_sz, capa);
                       if (cl > 0) {     /* don't stick! (3.43) */
                         url[cl] = ' ';
                         cl++;
                       }
-                      if (fread(url + cl, 1, fz, fp) != fz) {
+                      if (fread(url + cl, 1, fzs, fp) != fzs) {
+                        fclose(fp);
                         HTS_PANIC_PRINTF("File url list could not be read");
                         htsmain_free();
                         return -1;
                       }
                       fclose(fp);
-                      *(url + cl + fz) = '\0';
+                      *(url + cl + fzs) = '\0';
                     }
                   }
                 }
@@ -2296,8 +2306,8 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
 
       } else {                  // URL/filters
         char catbuff[CATBUFF_SIZE];
-        const int urlSize = (int) strlen(argv[na]);
-        const int capa = (int) (strlen(url) + urlSize + 32);
+        const size_t urlSize = strlen(argv[na]);
+        const size_t capa = strlen(url) + urlSize + 32;
 
         assertf(urlSize < HTS_URLMAXSIZE);
         if (urlSize < HTS_URLMAXSIZE) {

src/htsglobal.h

@@ -336,16 +336,20 @@ typedef int INTsys;
 #endif
 
 /* Socket-handle type. An unsigned integer wide enough for a Windows SOCKET;
-   a plain int file descriptor on POSIX. */
+   a plain int file descriptor on POSIX. T_SOCP is its printf conversion,
+   '%' included: unsigned __int64 on Win64 must not be printed with "%d". */
 #ifdef _WIN32
 #if defined(_WIN64)
 
 typedef unsigned __int64 T_SOC;
+#define T_SOCP "%" PRIu64
 #else
 typedef unsigned __int32 T_SOC;
+#define T_SOCP "%" PRIu32
 #endif
 #else
 typedef int T_SOC;
+#define T_SOCP "%d"
 #endif
 
 /* Buffer size for a printed network address (IPv4 or IPv6, NUL included). */

src/htshelp.c

@@ -156,9 +156,7 @@ void help_wizard(httrackp * opt) {
   char *a;
 
   //
-  if (urls == NULL || mainpath == NULL || projname == NULL || stropt == NULL
-      || stropt2 == NULL || strwild == NULL || cmd == NULL || str == NULL
-      || argv == NULL) {
+  if (buffers == NULL) {
     fprintf(stderr, "* memory exhausted in %s, line %d\n", __FILE__, __LINE__);
     return;
   }
@@ -251,6 +249,7 @@ void help_wizard(httrackp * opt) {
       strcatbuff(stropt2, "--update ");
       break;
     case 0:
+      freet(buffers);
       return;
       break;
     }
@@ -315,8 +314,10 @@ void help_wizard(httrackp * opt) {
       fflush(stdout);
       linput(stdin, str, 250);
       if (strnotempty(str)) {
-        if (!((str[0] == 'y') || (str[0] == 'Y')))
+        if (!((str[0] == 'y') || (str[0] == 'Y'))) {
+          freet(buffers);
           return;
+        }
       }
       printf("\n");
 
@@ -340,7 +341,7 @@ void help_wizard(httrackp * opt) {
   }
 
   /* Free buffers */
-  free(buffers);
+  freet(buffers);
 #undef urls
 #undef mainpath
 #undef projname

src/htslib.c

@@ -704,18 +704,16 @@ T_SOC http_xfopen(httrackp *opt, int mode, int treat, int waitconnect,
   /* Check for errors */
   if (soc == INVALID_SOCKET) {
     if (retour) {
-      if (retour->msg) {
-        if (!strnotempty(retour->msg)) {
+      if (!strnotempty(retour->msg)) {
 #ifdef _WIN32
-          int last_errno = WSAGetLastError();
+        int last_errno = WSAGetLastError();
 
-          sprintf(retour->msg, "Connect error: %s", strerror(last_errno));
+        sprintf(retour->msg, "Connect error: %s", strerror(last_errno));
 #else
-          int last_errno = errno;
+        int last_errno = errno;
 
-          sprintf(retour->msg, "Connect error: %s", strerror(last_errno));
+        sprintf(retour->msg, "Connect error: %s", strerror(last_errno));
 #endif
-        }
       }
     }
   }
@@ -2207,7 +2205,7 @@ T_SOC newhttp_addr(httrackp *opt, const char *_iadr, htsblk *retour, int port,
 #if DEBUG
       printf("erreur gethostbyname\n");
 #endif
-      if (retour && retour->msg) {
+      if (retour != NULL) {
 #ifdef _WIN32
         snprintf(retour->msg, sizeof(retour->msg),
                  "Unable to get server's address: %s", error);
@@ -2238,7 +2236,7 @@ T_SOC newhttp_addr(httrackp *opt, const char *_iadr, htsblk *retour, int port,
     DEBUG_W("socket()=%d\n" _(int) soc);
 #endif
     if (soc == INVALID_SOCKET) {
-      if (retour && retour->msg) {
+      if (retour != NULL) {
 #ifdef _WIN32
         int last_errno = WSAGetLastError();
 
@@ -2262,17 +2260,8 @@ T_SOC newhttp_addr(httrackp *opt, const char *_iadr, htsblk *retour, int port,
                              &bind_addr, &error) == NULL
           || bind(soc, &SOCaddr_sockaddr(bind_addr), 
                   SOCaddr_size(bind_addr)) != 0) {
-        if (retour && retour->msg) {
-#ifdef _WIN32
-          snprintf(retour->msg, sizeof(retour->msg),
-                   "Unable to bind the specificied server address: %s",
-                   error);
-#else
-          snprintf(retour->msg, sizeof(retour->msg),
-                   "Unable to bind the specificied server address: %s",
-                   error);
-#endif
-        }
+        snprintf(retour->msg, sizeof(retour->msg),
+                 "Unable to bind the specificied server address: %s", error);
         deletesoc(soc);
         return INVALID_SOCKET;
       }
@@ -2319,7 +2308,7 @@ T_SOC newhttp_addr(httrackp *opt, const char *_iadr, htsblk *retour, int port,
 #if HDEBUG
         printf("unable to connect!\n");
 #endif
-        if (retour != NULL && retour->msg) {
+        if (retour != NULL) {
 #ifdef _WIN32
           const int last_errno = WSAGetLastError();
 
@@ -2598,13 +2587,15 @@ void deletesoc(T_SOC soc) {
     if (closesocket(soc) != 0) {
       int err = WSAGetLastError();
 
-      fprintf(stderr, "* error closing socket %d: %s\n", soc, strerror(err));
+      fprintf(stderr, "* error closing socket " T_SOCP ": %s\n", soc,
+              strerror(err));
     }
 #else
     if (close(soc) != 0) {
       const int err = errno;
 
-      fprintf(stderr, "* error closing socket %d: %s\n", soc, strerror(err));
+      fprintf(stderr, "* error closing socket " T_SOCP ": %s\n", soc,
+              strerror(err));
     }
 #endif
 #if HTS_WIDE_DEBUG
@@ -3017,7 +3008,7 @@ int finput(T_SOC fd, char *s, int max) {
 }
 
 // Like linput, but in memory (optimized)
-int binput(char *buff, char *s, int max) {
+int binput(const char *buff, char *s, int max) {
   int count = 0;
   int destCount = 0;
 

src/htslib.h

@@ -262,7 +262,7 @@ HTS_INLINE void time_rfc822_local(char *s, struct tm *A);
 
 HTS_INLINE int sendc(htsblk * r, const char *s);
 int finput(T_SOC fd, char *s, int max);
-int binput(char *buff, char *s, int max);
+int binput(const char *buff, char *s, int max);
 int linput(FILE * fp, char *s, int max);
 int linputsoc(T_SOC soc, char *s, int max);
 int linputsoc_t(T_SOC soc, char *s, int max, int timeout);
@@ -607,6 +607,21 @@ static HTS_UNUSED size_t llint_to_size_t(LLint o) {
   }
 }
 
+/* Capacity for @p used bytes plus @p extra more plus @p slack spare;
+   (size_t) -1 if the total exceeds (size_t) -2 or @p extra is negative
+   (llint_to_size_t() would map that to a huge valid-looking size). */
+static HTS_UNUSED size_t llint_grow_size_t(size_t used, LLint extra,
+                                           size_t slack) {
+  const size_t max = (size_t) -2; /* (size_t) -1 is the error value */
+  const size_t e = extra >= 0 ? llint_to_size_t(extra) : (size_t) -1;
+
+  if (e == (size_t) -1 || used > max || slack > max - used ||
+      e > max - used - slack) {
+    return (size_t) -1;
+  }
+  return used + e + slack;
+}
+
 /* dirent() compatibility */
 #ifdef _WIN32
 /* Holds a UTF-8 d_name: MAX_PATH (260) UTF-16 units expand to <=3 bytes each.

src/htsselftest.c

@@ -2023,6 +2023,77 @@ static int st_fsize(httrackp *opt, int argc, char **argv) {
   return rc;
 }
 
+/* 4GB+100KB wraps to ~108KB through an int, and needs 33 unsigned bits. A
+   macro, not a static const: MSVC's C mode (/TC) rejects a const object
+   used inside another object's static initializer below (C2099). */
+#define HTS_ST_GROWSIZE_OVER32 (4LL * 1024 * 1024 * 1024 + 100 * 1024)
+
+/* llint_grow_size_t() sizes the buffer holding a whole -%S list file: the
+   result must be the exact 64-bit sum or a clean refusal, never a short one. */
+static int st_growsize(httrackp *opt, int argc, char **argv) {
+  enum { REFUSE, ACCEPT, WIDTH };
+
+  static const struct {
+    size_t used;
+    LLint extra;
+    size_t slack;
+    int want;
+  } cases[] = {
+      {0, 0, 0, ACCEPT},
+      {10, 100, 8192, ACCEPT},
+      {(size_t) -2 - 8, 4, 4, ACCEPT}, /* exact fit, no room to spare */
+      {(size_t) -2, 0, 0, ACCEPT},     /* largest representable capacity */
+      {0, -1, 0, REFUSE},              /* fsize() failure */
+      /* -1 already maps to SIZE_MAX; only this exercises the negative guard */
+      {0, -4096, 0, REFUSE},
+      {(size_t) -1, 1, 0, REFUSE},
+      {(size_t) -2, 0, 1, REFUSE},     /* slack alone overruns */
+      {(size_t) -1 - 8, 4, 4, REFUSE}, /* total would be the error value */
+      {(size_t) -1 - 8, 4, 8, REFUSE},
+      {0, HTS_ST_GROWSIZE_OVER32, 8192,
+       WIDTH}, /* 32-bit size_t can't hold these */
+      {10, HTS_ST_GROWSIZE_OVER32, 8192, WIDTH},
+  };
+
+  size_t k;
+  int rc = 0;
+
+  (void) opt;
+  (void) argc;
+  (void) argv;
+  for (k = 0; k < sizeof(cases) / sizeof(cases[0]); k++) {
+    const size_t used = cases[k].used, slack = cases[k].slack;
+    const LLint extra = cases[k].extra;
+    const size_t got = llint_grow_size_t(used, extra, slack);
+    const hts_boolean refused = got == (size_t) -1 ? HTS_TRUE : HTS_FALSE;
+    const hts_boolean exact =
+        !refused && extra >= 0 && got - used - slack == (size_t) extra;
+    hts_boolean ok;
+
+    switch (cases[k].want) {
+    case ACCEPT:
+      ok = exact;
+      break;
+    case REFUSE:
+      ok = refused;
+      break;
+    default:
+      ok = sizeof(size_t) >= sizeof(LLint) ? exact : refused;
+      break;
+    }
+    if (!ok) {
+      fprintf(stderr,
+              "growsize: grow(" LLintP ", " LLintP ", " LLintP ") = " LLintP
+              " (want %s)\n",
+              (LLint) used, extra, (LLint) slack, (LLint) got,
+              cases[k].want == REFUSE ? "refusal" : "exact sum");
+      rc = 1;
+    }
+  }
+  printf("growsize self-test %s\n", rc == 0 ? "OK" : "FAILED");
+  return rc;
+}
+
 static int st_savename(httrackp *opt, int argc, char **argv) {
   lien_adrfilsave afs;
   cache_back cache;
@@ -2417,17 +2488,20 @@ static int st_cookies(httrackp *opt, int argc, char **argv) {
     static t_cookie ck2;
     htsblk r;
     char host[600];
+    char line[64]; /* treathead NUL-cuts the header in place: never a literal */
 
     memset(&r, 0, sizeof(r));
     memset(host, 'a', sizeof(host) - 1);
     host[sizeof(host) - 1] = '\0';
     ck2.max_len = (int) sizeof(ck2.data);
     ck2.data[0] = '\0';
-    treathead(&ck2, host, "/", &r, "Set-Cookie: SID=1; path=/");
+    strcpybuff(line, "Set-Cookie: SID=1; path=/");
+    treathead(&ck2, host, "/", &r, line);
     if (strnotempty(ck2.data)) // oversize-host cookie was not dropped
       err = 1;
     /* control: a normal host still yields a cookie through treathead */
-    treathead(&ck2, dom, "/", &r, "Set-Cookie: SID=1; path=/");
+    strcpybuff(line, "Set-Cookie: SID=1; path=/");
+    treathead(&ck2, dom, "/", &r, line);
     if (strstr(ck2.data, "SID") == NULL) // guard wrongly dropped a valid cookie
       err = 1;
   }
@@ -2917,7 +2991,7 @@ static int ae_write_packed(const char *path, int windowBits,
     deflateEnd(&strm);
     return 1;
   }
-  strm.next_in = (Bytef *) src;
+  strm.next_in = (const Bytef *) src;
   strm.avail_in = (uInt) len;
   do {
     size_t n;
@@ -4868,6 +4942,8 @@ static const struct selftest_entry {
     {"sniff", "<content-type> <hex:..|text>", "MIME magic consistency",
      st_sniff},
     {"fsize", "<dir>", "file size past the 2GB signed-32-bit wrap", st_fsize},
+    {"growsize", "", "buffer capacity for a 64-bit file size (no int wrap)",
+     st_growsize},
     {"cache", "<dir>", "cache read/write round-trip self-test", st_cache},
     {"cacheindex", "", "cache-index (.ndx) parse must stay in bounds",
      st_cacheindex},

src/htsserver.c

@@ -147,7 +147,8 @@ HTS_UNUSED static int LANG_LIST(const char *path, char *buffer, size_t size);
 // 0- Init the URL catcher with standard port
 
 // smallserver_init(&port,&return_host);
-T_SOC smallserver_init_std(int *port_prox, char *adr_prox, int defaultPort) {
+T_SOC smallserver_init_std(int *port_prox, char *adr_prox, int defaultPort,
+                           const char *bindAddr) {
   T_SOC soc;
 
   if (defaultPort <= 0) {
@@ -160,12 +161,12 @@ T_SOC smallserver_init_std(int *port_prox, char *adr_prox, int defaultPort) {
     int i = 0;
 
     do {
-      soc = smallserver_init(&try_to_listen_to[i], adr_prox);
+      soc = smallserver_init(&try_to_listen_to[i], adr_prox, bindAddr);
       *port_prox = try_to_listen_to[i];
       i++;
     } while((soc == INVALID_SOCKET) && (try_to_listen_to[i] >= 0));
   } else {
-    soc = smallserver_init(&defaultPort, adr_prox);
+    soc = smallserver_init(&defaultPort, adr_prox, bindAddr);
     *port_prox = defaultPort;
   }
   return soc;
@@ -243,9 +244,10 @@ static int my_gethostname(char *h_loc, size_t size) {
 }
 
 // smallserver_init(&port,&return_host);
-T_SOC smallserver_init(int *port, char *adr) {
+T_SOC smallserver_init(int *port, char *adr, const char *bindAddr) {
   T_SOC soc = INVALID_SOCKET;
   char h_loc[256 + 2];
+  SOCaddr server;
 
   commandRunning = commandEnd = commandReturn = commandReturnSet =
     commandEndRequested = 0;
@@ -256,25 +258,23 @@ T_SOC smallserver_init(int *port, char *adr) {
     free(commandReturnCmdl);
   commandReturnCmdl = NULL;
 
-  if (my_gethostname(h_loc, 256) == 0) {   // host name
-    SOCaddr server;
+  SOCaddr_initany(server);
+  if (bindAddr != NULL && *bindAddr != '\0') {
+    /* advertise the bound address, else the URL we print is unreachable */
+    if (strlen(bindAddr) >= sizeof(h_loc) || !gethost(bindAddr, &server)) {
+      return INVALID_SOCKET;
+    }
+    strcpybuff(h_loc, bindAddr);
+  } else if (my_gethostname(h_loc, 256) != 0) { // host name
+    return INVALID_SOCKET;
+  }
 
-    SOCaddr_initany(server);
-    if ((soc =
-         (T_SOC) socket(SOCaddr_sinfamily(server), SOCK_STREAM,
-                        0)) != INVALID_SOCKET) {
-      SOCaddr_initport(server, *port);
-      if (bind(soc, &SOCaddr_sockaddr(server), SOCaddr_size(server)) == 0) {
-        if (listen(soc, 10) >= 0) {
-          strcpy(adr, h_loc);
-        } else {
-#ifdef _WIN32
-          closesocket(soc);
-#else
-          close(soc);
-#endif
-          soc = INVALID_SOCKET;
-        }
+  if ((soc = (T_SOC) socket(SOCaddr_sinfamily(server), SOCK_STREAM, 0)) !=
+      INVALID_SOCKET) {
+    SOCaddr_initport(server, *port);
+    if (bind(soc, &SOCaddr_sockaddr(server), SOCaddr_size(server)) == 0) {
+      if (listen(soc, 10) >= 0) {
+        strcpy(adr, h_loc);
       } else {
 #ifdef _WIN32
         closesocket(soc);
@@ -283,6 +283,13 @@ T_SOC smallserver_init(int *port, char *adr) {
 #endif
         soc = INVALID_SOCKET;
       }
+    } else {
+#ifdef _WIN32
+      closesocket(soc);
+#else
+      close(soc);
+#endif
+      soc = INVALID_SOCKET;
     }
   }
   return soc;
@@ -311,6 +318,51 @@ typedef struct {
   error_redirect = "/server/error.html"; \
 } while(0)
 
+/* Longest "sid" value worth unescaping: the expected one is an md5 hex digest,
+   so anything near this is already invalid and is rejected unread. */
+#define SID_VALUE_MAX 64
+
+/** Does the urlencoded request body present the expected session id?
+    True only if at least one "sid" field is present and every occurrence
+    matches, so it holds whichever one a later last-write-wins parse keeps.
+    Non-destructive: it runs before the body is tokenized in place. */
+static hts_boolean body_sid_is_valid(const char *body, const char *expected) {
+  const char *s = body;
+  hts_boolean seen = HTS_FALSE;
+
+  while (s != NULL && *s != '\0') {
+    const char *const amp = strchr(s, '&');
+    const char *const eq = strchr(s, '=');
+
+    if (eq != NULL && (amp == NULL || eq < amp) && (size_t) (eq - s) == 3 &&
+        strncmp(s, "sid", 3) == 0) {
+      const size_t len = amp != NULL ? (size_t) (amp - eq - 1) : strlen(eq + 1);
+      hts_boolean match = HTS_FALSE;
+
+      if (len < SID_VALUE_MAX) {
+        char raw[SID_VALUE_MAX];
+        String value = STRING_EMPTY;
+
+        memcpy(raw, eq + 1, len);
+        raw[len] = '\0';
+        unescapehttp(raw, &value);
+        /* StringBuff is NULL until written, so an empty value lands here. */
+        if (StringBuff(value) != NULL &&
+            strcmp(StringBuff(value), expected) == 0) {
+          match = HTS_TRUE;
+        }
+        StringFree(value);
+      }
+      if (!match) {
+        return HTS_FALSE;
+      }
+      seen = HTS_TRUE;
+    }
+    s = amp != NULL ? amp + 1 : NULL;
+  }
+  return seen;
+}
+
 int smallserver(T_SOC soc, char *url, char *method, char *data, char *path) {
   int timeout = 30;
   int retour = 0;
@@ -395,6 +447,7 @@ int smallserver(T_SOC soc, char *url, char *method, char *data, char *path) {
     T_SOC soc_c;
     LLint length = 0;
     const char *error_redirect = NULL;
+    hts_boolean denied = HTS_FALSE;
 
     line[0] = '\0';
     buffer[0] = '\0';
@@ -506,6 +559,22 @@ int smallserver(T_SOC soc, char *url, char *method, char *data, char *path) {
         }
       }
 
+      /* Authenticate the body before parsing it: every field it carries is
+         written straight into the global key store below, "command" included,
+         and that one reaches the engine. Checking afterwards cannot work — the
+         damage is already done, and the pre-seeded "sid" above would compare
+         equal to itself for a request that simply omits the field. */
+      if (meth && buffer[0]) {
+        intptr_t expected = 0;
+
+        if (!coucal_readptr(NewLangList, "_sid", &expected) ||
+            !body_sid_is_valid(buffer, (const char *) expected)) {
+          buffer[0] = '\0';
+          meth = 0;
+          denied = HTS_TRUE;
+        }
+      }
+
       /* check variables */
       if (meth && buffer[0]) {
         char *s = buffer;
@@ -526,20 +595,6 @@ int smallserver(T_SOC soc, char *url, char *method, char *data, char *path) {
         }
       }
 
-      /* Error check */
-      {
-        intptr_t adr = 0;
-        intptr_t adr2 = 0;
-
-        if (coucal_readptr(NewLangList, "sid", &adr)) {
-          if (coucal_readptr(NewLangList, "_sid", &adr2)) {
-            if (strcmp((char *) adr, (char *) adr2) != 0) {
-              meth = 0;
-            }
-          }
-        }
-      }
-
       /* Check variables (internal) */
       if (meth) {
         int doLoad = 0;
@@ -903,13 +958,11 @@ int smallserver(T_SOC soc, char *url, char *method, char *data, char *path) {
                 }
               }
               StringMemcat(headers, redir, strlen(redir));
-              {
-                char tmp[256];
-
-                if (strlen(file) < sizeof(tmp) - 32) {
-                  sprintf(tmp, "Location: %s\r\n", newfile);
-                  StringMemcat(headers, tmp, strlen(tmp));
-                }
+              /* client-supplied: a CR/LF here would split the response */
+              if (newfile[strcspn(newfile, "\r\n")] == '\0') {
+                StringCat(headers, "Location: ");
+                StringCat(headers, newfile);
+                StringCat(headers, "\r\n");
               }
               coucal_write(NewLangList, "redirect", (intptr_t) NULL);
             } else if (is_html(file)) {
@@ -1368,6 +1421,11 @@ int smallserver(T_SOC soc, char *url, char *method, char *data, char *path) {
             StringCat(output, error);
           }
         }
+      } else if (denied) {
+        StringCat(headers, "HTTP/1.0 403 Forbidden\r\n"
+                           "Server: httrack small server\r\n"
+                           "Content-type: text/html\r\n");
+        StringCat(output, "Missing or invalid session id.\r\n");
       } else {
 #ifdef _DEBUG
         char error_hdr[] =

src/htsserver.h

@@ -43,8 +43,11 @@ Please visit our Website: http://www.httrack.com
 
 // Fonctions
 void socinput(T_SOC soc, char *s, int max);
-T_SOC smallserver_init_std(int *port_prox, char *adr_prox, int defaultPort);
-T_SOC smallserver_init(int *port, char *adr);
+/* Listen on bindAddr, or every interface if NULL/empty; adr (>= 258 bytes) gets
+   the address to advertise. INVALID_SOCKET on error. */
+T_SOC smallserver_init_std(int *port_prox, char *adr_prox, int defaultPort,
+                           const char *bindAddr);
+T_SOC smallserver_init(int *port, char *adr, const char *bindAddr);
 int smallserver(T_SOC soc, char *url, char *method, char *data, char *path);
 
 #define CATCH_RESPONSE \

src/htstools.c

@@ -1318,19 +1318,19 @@ HTSEXT_API hts_boolean hts_findnext(find_handle find) {
   if (find) {
 #ifdef _WIN32
     if ((FindNextFileA(find->handle, &find->hdata)))
-      return 1;
+      return HTS_TRUE;
 #else
     char catbuff[CATBUFF_SIZE];
 
     memset(&(find->filestat), 0, sizeof(find->filestat));
     if ((find->dirp = readdir(find->hdir)))
-      if (find->dirp->d_name)
-        if (!STAT
-            (concat(catbuff, sizeof(catbuff), find->path, find->dirp->d_name), &find->filestat))
-          return 1;
+      if (!STAT(
+              concat(catbuff, sizeof(catbuff), find->path, find->dirp->d_name),
+              &find->filestat))
+        return HTS_TRUE;
 #endif
   }
-  return 0;
+  return HTS_FALSE;
 }
 
 HTSEXT_API int hts_findclose(find_handle find) {

src/htsweb.c

@@ -88,7 +88,7 @@ Please visit our Website: http://www.httrack.com
 
 static htsmutex refreshMutex = HTSMUTEX_INIT;
 
-static int help_server(char *dest_path, int defaultPort);
+static int help_server(char *dest_path, int defaultPort, const char *bindAddr);
 extern int commandRunning;
 extern int commandEnd;
 extern int commandReturn;
@@ -153,6 +153,8 @@ int main(int argc, char *argv[]) {
   int ret = 0;
   int defaultPort = 0;
   int parentPid = 0;
+  /* loopback by default: the handler trusts its input; --bind widens it */
+  const char *bindAddr = "127.0.0.1";
 
   printf("Initializing the server..\n");
 
@@ -179,7 +181,8 @@ int main(int argc, char *argv[]) {
   if (argc < 2 || (argc % 2) != 0) {
     fprintf(stderr, "** Warning: use the webhttrack frontend if available\n");
     fprintf(stderr,
-            "usage: %s [--port <port>] [--ppid parent-pid] <path-to-html-root-dir> [key value [key value]..]\n",
+            "usage: %s [--port <port>] [--bind <address>] [--ppid parent-pid] "
+            "<path-to-html-root-dir> [key value [key value]..]\n",
             argv[0]);
     fprintf(stderr, "example: %s /usr/share/httrack/\n", argv[0]);
     return 1;
@@ -267,6 +270,14 @@ int main(int argc, char *argv[]) {
         fprintf(stderr, "couldn't set the port number to %s\n", argv[i + 1]);
         return -1;
       }
+    } else if (strcmp(argv[i], "--bind") == 0 && i + 1 < argc) {
+      /* empty would fall back to every interface, silently undoing the default
+       */
+      if (!strnotempty(argv[i + 1])) {
+        fprintf(stderr, "--bind needs an address\n");
+        return -1;
+      }
+      bindAddr = argv[i + 1];
     } else if (strcmp(argv[i], "--ppid") == 0 && i + 1 < argc) {
       if (sscanf(argv[i + 1], "%u", &parentPid) != 1) {
         fprintf(stderr, "couldn't set the parent PID to %s\n", argv[i + 1]);
@@ -293,7 +304,7 @@ int main(int argc, char *argv[]) {
   }
 
   /* launch */
-  ret = help_server(argv[1], defaultPort);
+  ret = help_server(argv[1], defaultPort, bindAddr);
 
   htsthread_wait_n(background_threads - 1);
   hts_uninit();
@@ -427,11 +438,11 @@ static int webhttrack_runmain(httrackp * opt, int argc, char **argv) {
   return ret;
 }
 
-static int help_server(char *dest_path, int defaultPort) {
+static int help_server(char *dest_path, int defaultPort, const char *bindAddr) {
   int returncode = 0;
   char adr_prox[HTS_URLMAXSIZE * 2];
   int port_prox;
-  T_SOC soc = smallserver_init_std(&port_prox, adr_prox, defaultPort);
+  T_SOC soc = smallserver_init_std(&port_prox, adr_prox, defaultPort, bindAddr);
 
   if (soc != INVALID_SOCKET) {
     char url[HTS_URLMAXSIZE * 2];

src/libhttrack.vcxproj

@@ -59,9 +59,10 @@
   <ItemDefinitionGroup>
     <ClCompile>
       <!-- LIBHTTRACK_EXPORTS turns HTSEXT_API into __declspec(dllexport); ZLIB_DLL
-           imports zlib from its DLL. HTS_USEBROTLI/HTS_USEZSTD enable the br and
-           zstd content codings. Windows 7 floor, matching WinHTTrack. -->
-      <PreprocessorDefinitions>WIN32;_WINDOWS;_MBCS;_USRDLL;LIBHTTRACK_EXPORTS;ZLIB_DLL;HTS_USEBROTLI=1;HTS_USEZSTD=1;WINVER=0x0601;_WIN32_WINNT=0x0601;_CRT_SECURE_NO_WARNINGS;_CRT_NONSTDC_NO_DEPRECATE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
+           imports zlib from its DLL, ZLIB_CONST makes its next_in const as the
+           autotools build does. HTS_USEBROTLI/HTS_USEZSTD enable the br and zstd
+           content codings. Windows 7 floor, matching WinHTTrack. -->
+      <PreprocessorDefinitions>WIN32;_WINDOWS;_MBCS;_USRDLL;LIBHTTRACK_EXPORTS;ZLIB_DLL;ZLIB_CONST;HTS_USEBROTLI=1;HTS_USEZSTD=1;WINVER=0x0601;_WIN32_WINNT=0x0601;_CRT_SECURE_NO_WARNINGS;_CRT_NONSTDC_NO_DEPRECATE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
       <AdditionalIncludeDirectories>$(MSBuildThisFileDirectory);$(MSBuildThisFileDirectory)coucal;%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
       <WarningLevel>Level3</WarningLevel>
       <MultiProcessorCompilation>true</MultiProcessorCompilation>

src/proxy/proxytrack.c

@@ -771,7 +771,7 @@ static PT_Element proxytrack_process_DAV_Request(PT_Indexes indexes,
                   PT_ReadIndex(indexes, StringBuff(itemUrl) + 1, FETCH_HEADERS);
                 if (file != NULL && file->statuscode == HTTP_OK) {
                   size = file->size;
-                  if (file->lastmodified) {
+                  if (file->lastmodified[0] != '\0') {
                     timestamp = get_time_rfc822(file->lastmodified);
                   }
                   if (timestamp == (time_t) 0) {
@@ -785,7 +785,7 @@ static PT_Element proxytrack_process_DAV_Request(PT_Indexes indexes,
                     }
                     timestamp = timestampRep;
                   }
-                  if (file->contenttype) {
+                  if (file->contenttype[0] != '\0') {
                     mimeType = file->contenttype;
                   }
                 }

tests/01_engine-growsize.test

@@ -0,0 +1,25 @@
+#!/bin/bash
+#
+# Buffer capacity for a 64-bit file size ('httrack -#test=growsize'): a -%S list
+# file past 4GB must size its buffer exactly or be refused, never wrap.
+
+set -euo pipefail
+
+out=$(httrack -#test=growsize)
+echo "$out"
+test "$out" == "growsize self-test OK"
+
+tmp=$(mktemp -d "${TMPDIR:-/tmp}/httrack_growsize.XXXXXX")
+trap 'rm -rf "$tmp"' EXIT HUP INT QUIT PIPE TERM
+
+echo '<html><body>hi</body></html>' >"$tmp/index.html"
+printf -- '-*/zzmarker*\n' >"$tmp/rules.txt"
+
+# the rules file lands in the URL/filter string, echoed back by the banner
+run=$(httrack -O "$tmp/out" --quiet -n "-%S" "$tmp/rules.txt" \
+    "file://$tmp/index.html" 2>&1) || true
+printf '%s\n' "$run" | grep -q 'zzmarker' || {
+    echo "FAIL: -%S rules file was not loaded"
+    printf '%s\n' "$run"
+    exit 1
+}

tests/68_webhttrack-outdir-charset.test

@@ -60,13 +60,20 @@ test -n "${url:-}" || fail "htsserver did not start: $(cat "${srvlog}")"
 
 # Post a "start" whose -O dir is 'café' in the form's ISO-8859-1 charset.
 "${python}" - "${url}" "${work}" <<'PY'
-import sys, urllib.parse, urllib.request
+import re, sys, urllib.parse, urllib.request
 
 url, work = sys.argv[1], sys.argv[2]
 outdir = work + "/caf\xe9"  # 'café' as the single byte the browser would send
 # Port 1 refuses at once: only the decoded -O dir matters, not the fetch.
 cmd = "httrack --quiet --robots=0 http://127.0.0.1:1/x.html -O " + outdir
-fields = [("path", work), ("projname", "proj"), ("command_do", "start"),
+# The body is refused without the session id the server renders into each form.
+# Note the wizard page is under /server/; a bare /step4.html is the doc page.
+page = urllib.request.urlopen(url + "server/step4.html", timeout=20).read()
+m = re.search(rb'name="sid" value="([0-9a-f]+)"', page)
+if not m:
+    raise SystemExit("no session id in server/step4.html")
+fields = [("sid", m.group(1).decode()),
+          ("path", work), ("projname", "proj"), ("command_do", "start"),
           ("winprofile", "x"), ("command", cmd)]
 body = "&".join("%s=%s" % (k, urllib.parse.quote(v, safe="", encoding="latin-1"))
                 for k, v in fields)

tests/77_webhttrack-redirect.test

@@ -0,0 +1,185 @@
+#!/bin/bash
+#
+# htsserver's POST redirect: the Location header is built from a client-supplied
+# value, and the listen address is loopback unless --bind widens it.
+
+set -euo pipefail
+
+testdir=$(cd "$(dirname "$0")" && pwd)
+distdir=${top_srcdir:-$(cd "${testdir}/.." && pwd)}
+distdir=$(cd "${distdir}" && pwd)
+# run_with_timeout/kill_tree: timeout(1) is absent on macOS
+# shellcheck source=tests/testlib.sh
+. "${testdir}/testlib.sh"
+
+fail() {
+    echo "FAIL: $*" >&2
+    exit 1
+}
+
+command -v htsserver >/dev/null || fail "no htsserver in PATH"
+command -v python3 >/dev/null || {
+    echo "python3 not found; skipping" >&2
+    exit 77
+}
+
+log=$(mktemp)
+# start() runs inside a command substitution, so its $! never reaches this
+# shell. Take the pid from the server's own announcement instead: a missed kill
+# leaves an orphan holding the CI job open long after the suite has passed.
+srvpid() { sed -n 's/^PID=//p' "${log}" 2>/dev/null | head -1; }
+cleanup() {
+    stop
+    rm -f "${log}"
+}
+trap cleanup EXIT HUP INT QUIT PIPE TERM
+
+freeport() {
+    python3 -c 'import socket
+s = socket.socket()
+s.bind(("127.0.0.1", 0))
+print(s.getsockname()[1])
+s.close()'
+}
+
+# Start htsserver, echo the announced URL. Extra args are passed through.
+start() {
+    local port url
+    port=$(freeport)
+    : >"${log}"
+    (
+        trap '' TERM TTOU
+        exec htsserver "${distdir}/" --port "${port}" "$@" >"${log}" 2>&1
+    ) &
+    for _ in $(seq 1 40); do
+        url=$(sed -n 's/^URL=//p' "${log}" 2>/dev/null) && test -n "${url}" && break
+        sleep 0.25
+    done
+    test -n "${url:-}" || fail "htsserver did not come up: $(cat "${log}")"
+    echo "${url}"
+}
+
+stop() {
+    local pid
+    pid=$(srvpid)
+    test -z "${pid}" || kill -9 "${pid}" 2>/dev/null || true
+}
+
+# The session id gates the request body, so a POST has to carry one. The server
+# renders it into every form, which is where a browser picks it up too.
+scrape_sid() {
+    python3 -c 'import socket, sys
+s = socket.create_connection(("127.0.0.1", int(sys.argv[1])), 10)
+s.settimeout(20)
+s.sendall(b"GET /server/index.html HTTP/1.0\r\nHost: 127.0.0.1\r\n\r\n")
+out = b""
+while True:
+    b = s.recv(65536)
+    if not b:
+        break
+    out += b
+s.close()
+sys.stdout.write(out.decode("latin-1"))' "$1" |
+        sed -n 's/.*name="sid" value="\([0-9a-f]*\)".*/\1/p' | head -1
+}
+
+# POST redirect=$1 to 127.0.0.1:$2 with session id $3, print the raw headers.
+post_redirect() {
+    python3 -c 'import socket, sys, urllib.parse
+body = ("sid=" + sys.argv[3] + "&redirect="
+        + urllib.parse.quote(sys.argv[1], safe=""))
+req = ("POST / HTTP/1.0\r\nHost: 127.0.0.1\r\n"
+       "Content-type: application/x-www-form-urlencoded\r\n"
+       "Content-length: %d\r\n\r\n%s" % (len(body), body))
+s = socket.create_connection(("127.0.0.1", int(sys.argv[2])), 10)
+s.settimeout(20)
+s.sendall(req.encode())
+out = b""
+# stop at the end of the header block: the server need not close the connection,
+# and an unbounded recv() hung the macOS runner for over an hour.
+while b"\r\n\r\n" not in out:
+    b = s.recv(65536)
+    if not b:
+        break
+    out += b
+s.close()
+sys.stdout.write(out.split(b"\r\n\r\n")[0].decode("latin-1"))' "$1" "$2" "$3"
+}
+
+portof() { echo "${1##*:}" | tr -d /; }
+
+# "free"/"inuse"/"unusable": can $1 still be bound on 127.0.0.2? A wildcard
+# listener takes the port on every address, a loopback-only one does not, so
+# this discriminates where the announced URL cannot.
+probe_alias() {
+    python3 -c 'import socket, sys
+s = socket.socket()
+try:
+    s.bind(("127.0.0.2", int(sys.argv[1])))
+except OSError as e:
+    print("unusable" if e.errno in (99, 49, 10049) else "inuse")
+else:
+    print("free")
+finally:
+    s.close()' "$1"
+}
+
+# Non-repeating and not a round number: a truncation, a reorder or a dropped
+# interior byte all stay visible.
+long=$(python3 -c 'print("".join(chr(33 + i % 90) for i in range(4097)))')
+url=$(start)
+port=$(portof "${url}")
+sid=$(scrape_sid "${port}")
+test "${#sid}" -eq 32 || fail "did not scrape a 32-hex sid (got '${sid}')"
+resp=$(post_redirect "${long}" "${port}" "${sid}") ||
+    fail "no response to an oversized redirect value"
+stop
+loc=$(printf '%s' "${resp}" | sed -n 's/^Location: //p' | tr -d '\r')
+test "${loc}" = "${long}" ||
+    fail "oversized redirect not echoed whole (got ${#loc} bytes, want ${#long})"
+
+# CR/LF must suppress the header outright. Sanitising it instead would keep the
+# grep for an injected header quiet while still emitting a mangled Location.
+url=$(start)
+port=$(portof "${url}")
+sid=$(scrape_sid "${port}")
+resp=$(post_redirect '/foo
+X-Injected: pwned' "${port}" "${sid}") || fail "no response to a CRLF redirect value"
+stop
+printf '%s' "${resp}" | grep -qi 'X-Injected' &&
+    fail "CRLF in the redirect value reached the response"
+test "$(printf '%s' "${resp}" | grep -ci '^Location:')" -eq 0 ||
+    fail "CRLF redirect still emitted a Location header"
+test "$(printf '%s' "${resp}" | grep -c '^HTTP/1\.')" -eq 1 ||
+    fail "CRLF redirect did not yield exactly one status line"
+
+# Loopback unless asked otherwise. Assert the socket, not the announcement.
+url=$(start)
+port=$(portof "${url}")
+alias_state=$(probe_alias "${port}")
+stop
+if test "${alias_state}" = unusable; then
+    echo "no 127.0.0.2 alias; skipping the listen-address assertions" >&2
+else
+    test "${alias_state}" = free ||
+        fail "default listen address is not loopback-only (127.0.0.2:${port} taken)"
+    case "${url}" in
+    http://127.0.0.1:*) ;;
+    *) fail "default announcement is not loopback: ${url}" ;;
+    esac
+
+    url=$(start --bind 0.0.0.0)
+    port=$(portof "${url}")
+    alias_state=$(probe_alias "${port}")
+    stop
+    test "${alias_state}" = inuse ||
+        fail "--bind 0.0.0.0 did not take the wildcard (127.0.0.2:${port} ${alias_state})"
+fi
+
+# An empty --bind must not quietly fall back to every interface. Bounded: if it
+# were accepted the server would listen instead of exiting.
+run_with_timeout 15 htsserver "${distdir}/" --bind "" >"${log}" 2>&1 || true
+grep -q -- "--bind needs an address" "${log}" ||
+    fail "empty --bind not refused: $(cat "${log}")"
+
+echo "PASS"

tests/78_webhttrack-sid.test

@@ -0,0 +1,142 @@
+#!/bin/bash
+#
+# htsserver's session id must gate the request body: every field in it is
+# written into the global key store, and "command" from there reaches the engine.
+
+set -euo pipefail
+
+testdir=$(cd "$(dirname "$0")" && pwd)
+distdir=${top_srcdir:-$(cd "${testdir}/.." && pwd)}
+distdir=$(cd "${distdir}" && pwd)
+
+fail() {
+    echo "FAIL: $*" >&2
+    exit 1
+}
+
+command -v htsserver >/dev/null || fail "no htsserver in PATH"
+command -v python3 >/dev/null || {
+    echo "python3 not found; skipping" >&2
+    exit 77
+}
+
+srv=
+log=$(mktemp)
+cleanup() {
+    test -z "${srv}" || kill -9 "${srv}" 2>/dev/null || true
+    rm -f "${log}"
+}
+trap cleanup EXIT HUP INT QUIT PIPE TERM
+
+freeport() {
+    python3 -c 'import socket
+s = socket.socket()
+s.bind(("127.0.0.1", 0))
+print(s.getsockname()[1])
+s.close()'
+}
+
+# Echo the announced URL. Runs in a command substitution, so it is a
+# subshell and cannot export the pid: the caller reads it back with srvpid.
+start() {
+    local port url
+    port=$(freeport)
+    : >"${log}"
+    (
+        trap '' TERM TTOU
+        exec htsserver "${distdir}/" --port "${port}" >"${log}" 2>&1
+    ) &
+    for _ in $(seq 1 40); do
+        url=$(sed -n 's/^URL=//p' "${log}" 2>/dev/null) && test -n "${url}" && break
+        sleep 0.25
+    done
+    test -n "${url:-}" || fail "htsserver did not come up: $(cat "${log}")"
+    echo "${url}"
+}
+
+# The server reports its own pid; the aliveness assertions below hang off it.
+srvpid() { sed -n 's/^PID=//p' "${log}" | head -1; }
+
+alive() { kill -0 "$1" 2>/dev/null; }
+
+portof() { echo "${1##*:}" | tr -d /; }
+
+# Raw request to 127.0.0.1:$1; $2 is the body ("" for a GET of the $3 page,
+# default index). Prints the reply.
+request() {
+    python3 -c 'import socket, sys
+port, body = int(sys.argv[1]), sys.argv[2]
+page = sys.argv[3] if len(sys.argv) > 3 else "index"
+if body:
+    req = ("POST / HTTP/1.0\r\nHost: 127.0.0.1\r\n"
+           "Content-type: application/x-www-form-urlencoded\r\n"
+           "Content-length: %d\r\n\r\n%s" % (len(body), body))
+else:
+    req = ("GET /server/%s.html HTTP/1.0\r\nHost: 127.0.0.1\r\n\r\n" % page)
+s = socket.create_connection(("127.0.0.1", port), 10)
+s.settimeout(30)
+s.sendall(req.encode())
+out = b""
+while True:
+    b = s.recv(65536)
+    if not b:
+        break
+    out += b
+s.close()
+sys.stdout.write(out.decode("latin-1"))' "$1" "$2" ${3+"$3"}
+}
+
+# The server hands the sid to any client in every form; scraping it is the
+# legitimate flow, and it is what makes the accept case below meaningful.
+scrape_sid() {
+    request "$1" "" | sed -n 's/.*name="sid" value="\([0-9a-f]*\)".*/\1/p' | head -1
+}
+
+url=$(start)
+port=$(portof "${url}")
+srv=$(srvpid)
+test -n "${srv}" || fail "htsserver did not report its pid"
+alive "${srv}" || fail "htsserver is not running"
+
+sid=$(scrape_sid "${port}")
+# Control: without a real sid every assertion below would pass vacuously.
+test "${#sid}" -eq 32 || fail "did not scrape a 32-hex sid from the page (got '${sid}')"
+
+# Accept: the legitimate flow still works. "redirect" is the cheapest field
+# with a reply that is visible in the headers.
+resp=$(request "${port}" "sid=${sid}&redirect=/accepted")
+printf '%s' "${resp}" | grep -q '^Location: /accepted' ||
+    fail "a body carrying the correct sid was refused"
+
+# Refuse: missing and wrong. A missing one is the regression under test — the
+# expected value used to be pre-seeded into the compared key, so omitting the
+# field compared equal to itself.
+for bad in "redirect=/nosid" "sid=&redirect=/empty" \
+    "sid=00000000000000000000000000000000&redirect=/wrong"; do
+    resp=$(request "${port}" "${bad}")
+    printf '%s' "${resp}" | grep -q '^Location:' &&
+        fail "body accepted without a valid sid: ${bad}"
+    # A refusal has to be a well-formed reply, not a headerless fragment: the
+    # release build used to emit only Content-length, which reads as a protocol
+    # error to any client and hides the reason.
+    printf '%s' "${resp}" | grep -q '^HTTP/1\.0 403 ' ||
+        fail "refusal was not a 403: ${bad}"
+done
+
+# Suppressing the reply is not the same as refusing the write: the body is
+# applied to one global key store that later requests render from, and the
+# command dispatcher reads it from there. Probe the store itself — step3
+# interpolates ${projname} into its title — rather than the refused reply.
+title() { request "$1" "" step3; }
+
+request "${port}" "projname=UNAUTHWRITE" >/dev/null 2>&1 || true
+title "${port}" | grep -q 'UNAUTHWRITE' &&
+    fail "a body without a valid sid was written to the key store"
+
+# Paired accept case: without it the assertion above passes even if the server
+# simply ignores every body, which would prove nothing.
+request "${port}" "sid=${sid}&projname=AUTHWRITE" >/dev/null 2>&1 || true
+title "${port}" | grep -q 'AUTHWRITE' ||
+    fail "a body carrying the correct sid was not written to the key store"
+
+echo "PASS"

tests/79_local-proxytrack-webdav-mime.test

@@ -0,0 +1,90 @@
+#!/bin/bash
+#
+# proxytrack's WebDAV PROPFIND fallback must behave identically whether or not
+# the cache entry supplies Content-Type and Last-Modified.
+
+set -euo pipefail
+
+: "${top_srcdir:=..}"
+# shellcheck source=tests/testlib.sh
+. "$top_srcdir/tests/testlib.sh"
+
+python=$(find_python) || {
+    echo "python3 missing, skipping"
+    exit 77
+}
+command -v curl >/dev/null 2>&1 || {
+    echo "curl missing, skipping"
+    exit 77
+}
+# First test to run proxytrack as a live server, and MSYS cannot reap a native
+# listener: the orphan wedged the whole Windows suite for 48 minutes (#595).
+if is_windows; then
+    echo "windows: cannot reap a backgrounded proxytrack, skipping"
+    exit 77
+fi
+
+dir=$(mktemp -d)
+ptpid=
+cleanup() {
+    stop_server "$ptpid"
+    rm -rf "$dir"
+}
+trap cleanup EXIT
+
+# Neither header is present, so file->contenttype and ->lastmodified stay "".
+printf 'HTTP/1.1 200 OK\r\nContent-Length: 5\r\n\r\n' >"$dir/hdr"
+printf 'hello' >"$dir/body"
+alen=$(($(wc -c <"$dir/hdr") + $(wc -c <"$dir/body")))
+{
+    printf 'filedesc://t.arc 0.0.0.0 20250101000000 text/plain 200 - - 0 t.arc 9\n'
+    printf '2 0 test\n'
+    printf '\n\n'
+    printf 'http://example.com/page.html 0.0.0.0 20250101000000 text/html 200 - - 0 t.arc %d\n' "$alen"
+    cat "$dir/hdr" "$dir/body"
+} >"$dir/in.arc"
+
+freeport() {
+    "$python" -c 'import socket; s=socket.socket(); s.bind(("127.0.0.1",0)); print(s.getsockname()[1]); s.close()'
+}
+proxyport=$(freeport)
+icpport=$(freeport)
+
+proxytrack "127.0.0.1:$proxyport" "127.0.0.1:$icpport" "$dir/in.arc" >"$dir/pt.log" 2>&1 &
+ptpid=$!
+waited=0
+until grep -qE "HTTP Proxy installed on|Unable to (initialize a temporary server|create the server)" "$dir/pt.log"; do
+    kill -0 "$ptpid" 2>/dev/null || {
+        echo "FAIL: proxytrack exited before listening"
+        cat "$dir/pt.log"
+        exit 1
+    }
+    test "$waited" -lt 50 || {
+        echo "FAIL: proxytrack never announced its listen port"
+        exit 1
+    }
+    sleep 0.1
+    waited=$((waited + 1))
+done
+grep -q "HTTP Proxy installed on" "$dir/pt.log" || {
+    echo "FAIL: proxytrack failed to bind"
+    cat "$dir/pt.log"
+    exit 1
+}
+
+# --max-time: an unbounded read here would wedge the runner rather than fail
+curl -s --max-time 30 -X PROPFIND -H "Depth: 1" \
+    "http://127.0.0.1:$proxyport/webdav/example.com/" >"$dir/resp.xml"
+
+grep -q '<getcontenttype>application/octet-stream</getcontenttype>' "$dir/resp.xml" || {
+    echo "FAIL: missing Content-Type did not fall back to application/octet-stream"
+    cat "$dir/resp.xml"
+    exit 1
+}
+grep -q '<getlastmodified>Wed, 01 Jan 2025 00:00:00 GMT</getlastmodified>' "$dir/resp.xml" || {
+    echo "FAIL: missing Last-Modified did not fall back to the index timestamp"
+    cat "$dir/resp.xml"
+    exit 1
+}
+
+echo "OK: WebDAV listing falls back correctly for an entry with no Content-Type/Last-Modified"

tests/Makefile.am

@@ -64,6 +64,7 @@ TESTS = \
 	01_engine-reconcile.test \
 	01_engine-expandhome.test \
 	01_engine-fsize.test \
+	01_engine-growsize.test \
 	01_engine-redirect.test \
 	01_engine-longpath-io.test \
 	01_engine-mirror-io.test \
@@ -171,6 +172,9 @@ TESTS = \
 	74_local-warc-wacz.test \
 	74_local-warc-verbatim.test \
 	75_engine-longpath-posix.test \
-	76_cli-resize.test
+	76_cli-resize.test \
+	77_webhttrack-redirect.test \
+	78_webhttrack-sid.test \
+	79_local-proxytrack-webdav-mime.test
 
 CLEANFILES = check-network_sh.cache