Bound htscoremain.c pointer-destination buffer writes (batch 13)

Continues the htssafe.h pointer-destination migration in the CLI parser
(hts_main_internal). All sites write into a bare char*.

* The cmdl_add()/cmdl_ins() macros build argv entries into the x_argvblk block
  (malloc'd as the command-line size + 32768). Thread the block's total size
  (recorded in a new x_argvblk_size) and bound the copy with strlcpybuff. The
  remaining room is computed by a cmdl_room() helper that yields 0 once the block
  is exhausted (alias expansion or doit.log insertion can outrun the 32768 slack)
  so the copy aborts cleanly instead of the size_t subtraction wrapping to a huge
  unbounded value.
* The in-place argv rewrites each write no more than the slot already holds, so
  they are bounded by strlen(dest)+1 (provably sufficient): the "(none)" ->
  "\"\"" replacement, the two quote-strip copies (tempo is argv[na] minus its
  surrounding quotes), and the "--catchurl" -> "-#P" rewrite. The "--clean"/
  "--tide" empty rewrite becomes a direct argv[i][1]='\0'.
* Guard the quote-strip's tempo[strlen(tempo)-1] read: a lone '"' argument left
  tempo empty and read tempo[-1] (out of bounds). It now takes the existing
  missing-quote error path.
* The URL accumulator append uses strlcatbuff against the tracked url_sz.

These are macros/locals inside hts_main_internal, so not -#7 unit-testable;
cmdl_add runs on every invocation (covered by the whole suite). New
01_engine-cmdline.test cases exercise the quote-strip rewrite as the sole URL (a
quoted URL is mirrored; dangling- and lone-quote arguments are refused cleanly,
never a crash).

htscoremain.c pointer-destination warnings: 10 -> 0.

Signed-off-by: Xavier Roche <roche@httrack.com>

configure

@@ -3685,7 +3685,9 @@ fi
 
 
 
-VERSION_INFO="2:49:0"
+# 3:0:0: htsblk layout changed (contenttype/charset/contentencoding widened to
+# 128), an incompatible ABI break, so bump current and reset revision/age.
+VERSION_INFO="3:0:0"
 
 { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking whether to enable maintainer-specific portions of Makefiles" >&5
 printf %s "checking whether to enable maintainer-specific portions of Makefiles... " >&6; }

configure.ac

@@ -29,7 +29,9 @@ AC_CONFIG_SRCDIR(src/httrack.c)
 AC_CONFIG_MACRO_DIR([m4])
 AC_CONFIG_HEADERS(config.h)
 AM_INIT_AUTOMAKE([subdir-objects])
-VERSION_INFO="2:49:0"
+# 3:0:0: htsblk layout changed (contenttype/charset/contentencoding widened to
+# 128), an incompatible ABI break, so bump current and reset revision/age.
+VERSION_INFO="3:0:0"
 AM_MAINTAINER_MODE
 AC_USE_SYSTEM_EXTENSIONS
 

src/htsback.c

@@ -3584,8 +3584,9 @@ void back_wait(struct_back * sback, httrackp * opt, cache_back * cache,
                         back[i].r.is_file = 1;
                         back[i].r.totalsize = back[i].r.size =
                           fsize_utf8(back[i].url_sav);
-                        get_httptype(opt, back[i].r.contenttype,
-                                     back[i].url_sav, 1);
+                        get_httptype_sized(opt, back[i].r.contenttype,
+                                           sizeof(back[i].r.contenttype),
+                                           back[i].url_sav, 1);
                         hts_log_print(opt, LOG_DEBUG,
                                       "Not-modified status without cache guessed: %s%s",
                                       back[i].url_adr, back[i].url_fil);

src/htscore.c

@@ -1734,7 +1734,7 @@ int httpmirror(char *url1, httrackp * opt) {
                 {
                   char buff[256];
 
-                  guess_httptype(opt, buff, urlfil());
+                  guess_httptype_sized(opt, buff, sizeof(buff), urlfil());
                   if (strcmp(buff, "image/gif") == 0)
                     create_gif_warning = 1;
                 }
@@ -3150,7 +3150,7 @@ static void postprocess_file(httrackp * opt, const char *save, const char *adr,
           /* CID */
           make_content_id(adr, fil, cid, sizeof(cid));
 
-          guess_httptype(opt, mimebuff, save);
+          guess_httptype_sized(opt, mimebuff, sizeof(mimebuff), save);
           fprintf(opt->state.mimefp, "--%s\r\n",
                   StringBuff(opt->state.mimemid));
           /*if (first)
@@ -3862,7 +3862,8 @@ int htsAddLink(htsmoduleStruct * str, char *link) {
             opt->savename_83 = b;
             if (r != -1 && !forbidden_url) {
               if (savename()) {
-                if (lienrelatif(tempo, afs.save, savename()) == 0) {
+                if (lienrelatif(tempo, sizeof(tempo), afs.save, savename()) ==
+                    0) {
                   hts_log_print(opt, LOG_DEBUG,
                                 "(module): relative link at %s build with %s and %s: %s",
                                 afs.af.adr, afs.save, savename(), tempo);

src/htscoremain.c

@@ -69,23 +69,29 @@ Please visit our Website: http://www.httrack.com
 /* Resolver */
 extern int IPV6_resolver;
 
-// Add a command in the argc/argv
-#define cmdl_add(token,argc,argv,buff,ptr) \
-  argv[argc]=(buff+ptr); \
-  strcpybuff(argv[argc],token); \
-  ptr += (int) (strlen(argv[argc])+2); \
+/* Remaining room in the argv block; 0 once it is exhausted (alias expansion or
+   doit.log insertion can outrun the +32768 slack), so the copy aborts cleanly
+   instead of the subtraction wrapping to a huge unbounded size. */
+#define cmdl_room(bufsize, ptr)                                                \
+  ((ptr) < (size_t) (bufsize) ? (size_t) (bufsize) - (ptr) : 0)
+
+// Add a command in the argc/argv (buff has total capacity bufsize)
+#define cmdl_add(token, argc, argv, buff, bufsize, ptr)                        \
+  argv[argc] = (buff + ptr);                                                   \
+  strlcpybuff(argv[argc], token, cmdl_room(bufsize, ptr));                     \
+  ptr += (int) (strlen(argv[argc]) + 2);                                       \
   argc++
 
-// Insert a command in the argc/argv
-#define cmdl_ins(token,argc,argv,buff,ptr) \
-  { \
-  int i; \
-  for(i=argc;i>0;i--)\
-  argv[i]=argv[i-1];\
-  } \
-  argv[0]=(buff+ptr); \
-  strcpybuff(argv[0],token); \
-  ptr += (int) (strlen(argv[0])+2); \
+// Insert a command in the argc/argv (buff has total capacity bufsize)
+#define cmdl_ins(token, argc, argv, buff, bufsize, ptr)                        \
+  {                                                                            \
+    int i;                                                                     \
+    for (i = argc; i > 0; i--)                                                 \
+      argv[i] = argv[i - 1];                                                   \
+  }                                                                            \
+  argv[0] = (buff + ptr);                                                      \
+  strlcpybuff(argv[0], token, cmdl_room(bufsize, ptr));                        \
+  ptr += (int) (strlen(argv[0]) + 2);                                          \
   argc++
 
 #define htsmain_free() do { \
@@ -285,6 +291,196 @@ static void basic_selftests(void) {
       assertf(end == NULL && strcmp(tok, "a\\") == 0);
     }
   }
+  // fil_normalized(): canonicalizes a URL path. Query arguments are sorted
+  // alphabetically (by the text after each '?'/'&') and the query is rebuilt
+  // through a bounded builder; outside the query, "//" collapses to "/".
+  // Regression for that builder.
+  {
+    char norm[256];
+
+    assertf(strcmp(fil_normalized("/p?b=2&a=1&c=3", norm), "/p?a=1&b=2&c=3") ==
+            0);
+    assertf(strcmp(fil_normalized("/a//b", norm), "/a/b") == 0);
+    // "//" is collapsed only before the query; inside the query it is kept
+    assertf(strcmp(fil_normalized("/a//b?x=c//d", norm), "/a/b?x=c//d") == 0);
+  }
+  // give_mimext(): mime type -> file extension, bounded into the caller buffer.
+  // Returns 1 when an extension was written, 0 otherwise.
+  {
+    char ext[16];
+
+    assertf(give_mimext(ext, sizeof(ext), "image/gif") == 1);
+    assertf(strcmp(ext, "gif") == 0);
+    assertf(give_mimext(ext, sizeof(ext), "text/html") == 1);
+    assertf(strcmp(ext, "html") == 0);
+    assertf(give_mimext(ext, sizeof(ext), "no/such-mime-type") == 0);
+    assertf(ext[0] == '\0');
+  }
+  // convtolower(): lower-cases into the caller buffer (bounded by its size).
+  {
+    char low[64];
+
+    assertf(strcmp(convtolower(low, sizeof(low), "ABC/Def.HTML"),
+                   "abc/def.html") == 0);
+  }
+  // cut_path(): splits a path into directory (with trailing '/') and basename,
+  // each bounded by its buffer size.
+  {
+    char path[256];
+    char pname[256];
+
+    {
+      char full[] = "/dir/sub/file.html";
+
+      cut_path(full, path, sizeof(path), pname, sizeof(pname));
+      assertf(strcmp(path, "/dir/sub/") == 0);
+      assertf(strcmp(pname, "file.html") == 0);
+    }
+    { // a trailing slash is trimmed before the split
+      char full[] = "/dir/sub/";
+
+      cut_path(full, path, sizeof(path), pname, sizeof(pname));
+      assertf(strcmp(path, "/dir/") == 0);
+      assertf(strcmp(pname, "sub") == 0);
+    }
+    { // a path of length <= 1 yields empty results
+      char full[] = "/";
+
+      cut_path(full, path, sizeof(path), pname, sizeof(pname));
+      assertf(path[0] == '\0' && pname[0] == '\0');
+    }
+  }
+  // get_httptype_sized(): a long MIME type (Office OOXML reaches 73 chars) is
+  // written whole into a contenttype-sized buffer; returns 1 on a match, 0 when
+  // flag==0 and nothing matched. Regression for the old contenttype[64]
+  // overflow.
+  {
+    httrackp *opt = hts_create_opt();
+    htsblk r; // write into the real struct field, not a stand-in
+
+    assertf(opt != NULL);
+    // a long MIME (Office OOXML reaches 73 chars) must fit htsblk.contenttype
+    // whole: a [64] field would make this bounded copy abort.
+    assertf(get_httptype_sized(opt, r.contenttype, sizeof(r.contenttype),
+                               "deck.pptx", 0) == 1);
+    assertf(strcmp(r.contenttype,
+                   "application/vnd.openxmlformats-officedocument."
+                   "presentationml.presentation") == 0);
+    assertf(get_httptype_sized(opt, r.contenttype, sizeof(r.contenttype),
+                               "x.gif", 0) == 1);
+    assertf(strcmp(r.contenttype, "image/gif") == 0);
+    // no extension and flag==0: nothing written, returns 0
+    assertf(get_httptype_sized(opt, r.contenttype, sizeof(r.contenttype),
+                               "noextfile", 0) == 0);
+    assertf(r.contenttype[0] == '\0');
+    // no extension and flag==1: octet-stream fallback, returns 1
+    assertf(get_httptype_sized(opt, r.contenttype, sizeof(r.contenttype),
+                               "noextfile", 1) == 1);
+    assertf(strcmp(r.contenttype, "application/octet-stream") == 0);
+    // a user --assume rule with an empty value matches but writes nothing:
+    // get_userhttptype returns 1 with the buffer empty, so get_httptype_sized
+    // must still report 0 (callers test the return like the old
+    // strnotempty(s)).
+    StringCopy(opt->mimedefs, "\ncgi=\n");
+    assertf(get_httptype_sized(opt, r.contenttype, sizeof(r.contenttype),
+                               "/x.cgi", 0) == 0);
+    assertf(r.contenttype[0] == '\0');
+    StringCopy(opt->mimedefs, "\ncgi=text/html\n");
+    assertf(get_httptype_sized(opt, r.contenttype, sizeof(r.contenttype),
+                               "/x.cgi", 0) == 1);
+    assertf(strcmp(r.contenttype, "text/html") == 0);
+    hts_free_opt(opt);
+  }
+  // adr_normalized_sized(): bounded host normalization (passthrough when
+  // already normal).
+  {
+    char n[HTS_URLMAXSIZE];
+
+    assertf(strcmp(adr_normalized_sized("example.com", n, sizeof(n)),
+                   "example.com") == 0);
+  }
+  // standard_name(): builds "<name><md5?>.<ext>" into a bounded buffer. The md5
+  // is appended (4 chars) only when the URL has a query string (see url_md5),
+  // so test both; pin the structure (name + ext, lengths), not the md5 chars.
+  {
+    char b[HTS_URLMAXSIZE * 2];
+    const char *nom = "index.html"; // name part
+    const char *dot = nom + 5;      // points at ".html"
+    size_t len;
+
+    // no query -> no md5: "index" + ".html"
+    standard_name(b, sizeof(b), dot, nom, "http://example.com/index.html", 0);
+    assertf(strcmp(b, "index.html") == 0);
+    // query -> 4 md5 chars between name and ext: "index" + md5(4) + ".html"
+    standard_name(b, sizeof(b), dot, nom, "http://example.com/index.html?v=1",
+                  0);
+    len = strlen(b);
+    assertf(len == 5 + 4 + 5);
+    assertf(strncmp(b, "index", 5) == 0);
+    assertf(strcmp(b + len - 5, ".html") == 0);
+    // short names: name kept (<=8), the extension is clamped to 3 -> ".htm"
+    standard_name(b, sizeof(b), dot, nom, "http://example.com/index.html?v=1",
+                  1);
+    len = strlen(b);
+    assertf(len == 5 + 4 + 4);
+    assertf(strcmp(b + len - 4, ".htm") == 0);
+    // short names with a >8-char name: the name is clamped to 8 ("indexpag")
+    {
+      const char *lnom = "indexpage.html";
+      const char *ldot = lnom + 9; // points at ".html"
+
+      standard_name(b, sizeof(b), ldot, lnom,
+                    "http://example.com/indexpage.html?v=1", 1);
+      len = strlen(b);
+      assertf(len == 8 + 4 + 4);
+      assertf(strncmp(b, "indexpag", 8) == 0);
+      assertf(strcmp(b + len - 4, ".htm") == 0);
+    }
+  }
+  // longfile_to_83(): single-name 8-3 (mode 1) / ISO9660 (mode 2) conversion;
+  // uppercases, clamps the name (8 / 31) and the extension (3). It rewrites
+  // 'save' in place, so pass a mutable array.
+  {
+    char n83[256];
+
+    {
+      char save[] = "longfilename.html";
+
+      longfile_to_83(1, n83, sizeof(n83), save); // 8-3: name->8, ext->3
+      assertf(strcmp(n83, "LONGFILE.HTM") == 0);
+    }
+    {
+      char save[] = "longfilename.html";
+
+      longfile_to_83(2, n83, sizeof(n83), save); // ISO9660: name->31, ext->3
+      assertf(strcmp(n83, "LONGFILENAME.HTM") == 0);
+    }
+    { // sanitization: leading '.'->'_', interior dots
+      char save[] = ".a b.c.d e"; // collapse to '_', spaces/specials -> '_'
+                                  // (only the last dot stays as the separator)
+      longfile_to_83(1, n83, sizeof(n83), save);
+      assertf(strcmp(n83, "_A_B_C.D_E") == 0);
+    }
+  }
+  // long_to_83(): per-segment 8-3 conversion of a whole path.
+  {
+    char n83[HTS_URLMAXSIZE * 2];
+    char save[] = "dir/longfilename.html";
+
+    long_to_83(1, n83, sizeof(n83), save);
+    assertf(strcmp(n83, "DIR/LONGFILE.HTM") == 0);
+  }
+  // lienrelatif(): relative path from the directory of curr_fil to link.
+  {
+    char s[HTS_URLMAXSIZE * 2];
+
+    // same directory -> just the basename
+    assertf(lienrelatif(s, sizeof(s), "dir/page.html", "dir/index.html") == 0);
+    assertf(strcmp(s, "page.html") == 0);
+    // link one level up -> a "../" prefix
+    assertf(lienrelatif(s, sizeof(s), "a.html", "dir/index.html") == 0);
+    assertf(strcmp(s, "../a.html") == 0);
+  }
 }
 
 /* Self-tests for the htssafe.h bounded string ops (driven by httrack -#8).
@@ -402,6 +598,7 @@ HTSEXT_API int hts_main2(int argc, char **argv, httrackp * opt) {
 static int hts_main_internal(int argc, char **argv, httrackp * opt) {
   char **x_argv = NULL;         // Patch pour argv et argc: en cas de récupération de ligne de commande
   char *x_argvblk = NULL;       // (reprise ou update)
+  size_t x_argvblk_size = 0;    // total capacity of x_argvblk
   int x_ptr = 0;                // offset
 
   //
@@ -479,7 +676,8 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
         *a = ' ';
       /* equivalent to "empty parameter" */
       if ((strcmp(argv[na], HTS_NOPARAM) == 0) || (strcmp(argv[na], HTS_NOPARAM2) == 0))        // (none)
-        strcpybuff(argv[na], "\"\"");
+        /* replacing "(none)"/"\"(none)\"" with "\"\"" always fits in place */
+        strlcpybuff(argv[na], "\"\"", strlen(argv[na]) + 1);
       if (strncmp(argv[na], "-&", 2) == 0)
         argv[na][1] = '%';
     }
@@ -501,6 +699,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
       htsmain_free();
       return -1;
     }
+    x_argvblk_size = (size_t) (current_size + 32768);
     x_argvblk[0] = '\0';
     x_ptr = 0;
 
@@ -522,7 +721,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
     //
     argv_url = 0;               /* pour comptage */
     //
-    cmdl_add(argv[0], x_argc, x_argv, x_argvblk, x_ptr);
+    cmdl_add(argv[0], x_argc, x_argv, x_argvblk, x_argvblk_size, x_ptr);
     na = 1;                     /* commencer après nom_prg */
     while(na < argc) {
       int result = 1;
@@ -543,9 +742,10 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
         }
 
         /* Copier */
-        cmdl_add(tmp_argv[0], x_argc, x_argv, x_argvblk, x_ptr);
+        cmdl_add(tmp_argv[0], x_argc, x_argv, x_argvblk, x_argvblk_size, x_ptr);
         if (tmp_argc > 1) {
-          cmdl_add(tmp_argv[1], x_argc, x_argv, x_argvblk, x_ptr);
+          cmdl_add(tmp_argv[1], x_argc, x_argv, x_argvblk, x_argvblk_size,
+                   x_ptr);
         }
 
         /* Compter URLs et détecter -i,-q.. */
@@ -617,7 +817,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
           char BIGSTK tempo[HTS_CDLMAXSIZE];
 
           strcpybuff(tempo, argv[na] + 1);
-          if (tempo[strlen(tempo) - 1] != '"') {
+          if (tempo[0] == '\0' || tempo[strlen(tempo) - 1] != '"') {
             char BIGSTK s[HTS_CDLMAXSIZE];
 
             sprintf(s, "Missing quote in %s", argv[na]);
@@ -626,7 +826,9 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
             return -1;
           }
           tempo[strlen(tempo) - 1] = '\0';
-          strcpybuff(argv[na], tempo);
+          /* tempo is argv[na] minus its surrounding quotes, so it fits in place
+           */
+          strlcpybuff(argv[na], tempo, strlen(argv[na]) + 1);
         }
 
         if (cmdl_opt(argv[na])) {       // option
@@ -791,7 +993,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
           if (strnotempty(lastp)) {
             insert_after_argc = argc - insert_after;
             cmdl_ins(lastp, insert_after_argc, (argv + insert_after), x_argvblk,
-                     x_ptr);
+                     x_argvblk_size, x_ptr);
             argc = insert_after_argc + insert_after;
             insert_after++;
           }
@@ -911,7 +1113,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
       if (argv[i][0] == '-') {
         if (argv[i][1] == '-') {        // --xxx
           if ((strfield2(argv[i] + 2, "clean")) || (strfield2(argv[i] + 2, "tide"))) {  // nettoyer
-            strcpybuff(argv[i] + 1, "");
+            argv[i][1] = '\0';
             if (fexist
                 (fconcat
                  (OPT_GET_BUFF(opt), OPT_GET_BUFF_SIZE(opt), StringBuff(opt->path_log), "hts-log.txt")))
@@ -1020,7 +1222,8 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
             //
           } else if (strfield2(argv[i] + 2, "catchurl")) {      // capture d'URL via proxy temporaire!
             argv_url = 1;       // forcer a passer les parametres
-            strcpybuff(argv[i] + 1, "#P");
+            /* argv[i] is "--catchurl"; "#P" fits after its first char */
+            strlcpybuff(argv[i] + 1, "#P", strlen(argv[i] + 1) + 1);
             //
           } else if (strfield2(argv[i] + 2, "updatehttrack")) {
 #ifdef _WIN32
@@ -1348,7 +1551,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
         char BIGSTK tempo[HTS_CDLMAXSIZE + 256];
 
         strcpybuff(tempo, argv[na] + 1);
-        if (tempo[strlen(tempo) - 1] != '"') {
+        if (tempo[0] == '\0' || tempo[strlen(tempo) - 1] != '"') {
           char s[HTS_CDLMAXSIZE + 256];
 
           sprintf(s, "Missing quote in %s", argv[na]);
@@ -1357,7 +1560,9 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
           return -1;
         }
         tempo[strlen(tempo) - 1] = '\0';
-        strcpybuff(argv[na], tempo);
+        /* tempo is argv[na] minus its surrounding quotes, so it fits in place
+         */
+        strlcpybuff(argv[na], tempo, strlen(argv[na]) + 1);
       }
 
       if (cmdl_opt(argv[na])) { // option
@@ -2598,15 +2803,12 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
                   // initialiser mimedefs
                   //get_userhttptype(opt,1,opt->mimedefs,NULL);
                   // check
-                  mime[0] = '\0';
-                  get_httptype(opt, mime, argv[na + 1], 0);
-                  if (mime[0] != '\0') {
+                  if (get_httptype_sized(opt, mime, sizeof(mime), argv[na + 1],
+                                         0)) {
                     char ext[256];
 
                     printf("%s is '%s'\n", argv[na + 1], mime);
-                    ext[0] = '\0';
-                    give_mimext(ext, mime);
-                    if (ext[0]) {
+                    if (give_mimext(ext, sizeof(ext), mime)) {
                       printf("and its local type is '.%s'\n", ext);
                     }
                   } else {
@@ -3019,7 +3221,7 @@ static int hts_main_internal(int argc, char **argv, httrackp * opt) {
         if (urlSize < HTS_URLMAXSIZE) {
           ensureUrlCapacity(url, url_sz, capa);
           if (strnotempty(url))
-            strcatbuff(url, " ");       // espace de séparation
+            strlcatbuff(url, " ", url_sz); // separator space
           append_escape_spc_url(unescape_http_unharm(catbuff, sizeof(catbuff), argv[na], 1), url, url_sz);
         }
       }                         // if argv=- etc. 

src/htsglobal.h

@@ -197,10 +197,13 @@ Please visit our Website: http://www.httrack.com
 
 #endif
 
-/* Taille max d'une URL */
+/* Max URL length */
 #define HTS_URLMAXSIZE 1024
-/* Taille max ligne de commande (>=HTS_URLMAXSIZE*2) */
+/* Max command-line length (>=HTS_URLMAXSIZE*2) */
 #define HTS_CDLMAXSIZE 1024
+/* MIME-type buffer contract (htsblk.contenttype/charset/contentencoding); holds
+   the longest registered MIME type, the Office OOXML ones reaching 73 chars */
+#define HTS_MIMETYPE_SIZE 128
 
 /* Copyright (C) 1998 Xavier Roche and other contributors */
 #define HTTRACK_AFF_AUTHORS "[XR&CO'2014]"
@@ -250,6 +253,22 @@ Please visit our Website: http://www.httrack.com
 #endif
 #endif
 
+/**
+ * Mark a function deprecated, with a message pointing at the replacement.
+ * Placed before the declaration so both the GCC/Clang attribute and the MSVC
+ * __declspec sit in a position both accept. Degrades to nothing elsewhere.
+ */
+#if defined(__GNUC__) &&                                                       \
+    (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 5))
+#define HTS_DEPRECATED(msg) __attribute__((deprecated(msg)))
+#elif defined(__GNUC__)
+#define HTS_DEPRECATED(msg) __attribute__((deprecated))
+#elif defined(_MSC_VER) && (_MSC_VER >= 1400)
+#define HTS_DEPRECATED(msg) __declspec(deprecated(msg))
+#else
+#define HTS_DEPRECATED(msg)
+#endif
+
 #ifndef HTS_LONGLONG
 #ifdef HTS_NO_64_BIT
 #define HTS_LONGLONG 0

src/htshash.c

@@ -76,7 +76,7 @@ static coucal_key key_duphandler(void *arg, coucal_key_const name) {
 /* Key sav hashes are using case-insensitive version */
 static coucal_hashkeys key_sav_hashes(void *arg, coucal_key_const key) {
   hash_struct *const hash = (hash_struct*) arg;
-  convtolower(hash->catbuff, (const char*) key);
+  convtolower(hash->catbuff, sizeof(hash->catbuff), (const char *) key);
   return coucal_hash_string(hash->catbuff);
 }
 

src/htsjava.c

@@ -472,9 +472,8 @@ static int tris(httrackp * opt, char *buffer) {
   {
     char type[256];
 
-    type[0] = '\0';
-    get_httptype(opt, type, buffer, 0);
-    if (strnotempty(type))      // type reconnu!
+    if (get_httptype_sized(opt, type, sizeof(type), buffer,
+                           0)) // recognized type
       return 1;
     // ajout RX 05/2001
     else if (is_dyntype(get_ext(catbuff, sizeof(catbuff), buffer)))      // asp,cgi...

src/htslib.c

@@ -754,7 +754,8 @@ T_SOC http_xfopen(httrackp * opt, int mode, int treat, int waitconnect,
       if (soc != INVALID_SOCKET) {
         retour->statuscode = HTTP_OK;   // OK
         strcpybuff(retour->msg, "OK");
-        guess_httptype(opt, retour->contenttype, fil);
+        guess_httptype_sized(opt, retour->contenttype,
+                             sizeof(retour->contenttype), fil);
       } else if (strnotempty(retour->msg) == 0)
         strcpybuff(retour->msg, "Unable to open local file");
       return soc;               // renvoyer
@@ -1530,8 +1531,9 @@ void treathead(t_cookie * cookie, const char *adr, const char *fil, htsblk * ret
       if (retour->location) {
         while(is_realspace(*(rcvd + p)))
           p++;                  // sauter espaces
-        if ((int) strlen(rcvd + p) < HTS_URLMAXSIZE)    // pas trop long?
-          strcpybuff(retour->location, rcvd + p);
+        if ((int) strlen(rcvd + p) < HTS_URLMAXSIZE) // not too long?
+          /* location aliases location_buffer[HTS_URLMAXSIZE * 2] */
+          strlcpybuff(retour->location, rcvd + p, HTS_URLMAXSIZE * 2);
         else                    // erreur.. ignorer
           retour->location[0] = '\0';
       }
@@ -3444,16 +3446,17 @@ HTSEXT_API char *fil_normalized(const char *source, char *dest) {
     /* Replace query by sorted query */
     copyBuff = malloct(qLen + 1);
     assertf(copyBuff != NULL);
-    copyBuff[0] = '\0';
-    for(i = 0; i < ampargs; i++) {
-      if (i == 0)
-        strcatbuff(copyBuff, "?");
-      else
-        strcatbuff(copyBuff, "&");
-      strcatbuff(copyBuff, amps[i] + 1);
+    {
+      htsbuff cb = htsbuff_ptr(copyBuff, qLen + 1);
+
+      for (i = 0; i < ampargs; i++) {
+        htsbuff_cat(&cb, i == 0 ? "?" : "&");
+        htsbuff_cat(&cb, amps[i] + 1);
+      }
+      assertf(cb.len == qLen);
     }
-    assertf(strlen(copyBuff) == qLen);
-    strcpybuff(query, copyBuff);
+    /* query points into dest where the original qLen-byte query was */
+    strlcpybuff(query, copyBuff, qLen + 1);
 
     /* Cleanup */
     freet(amps);
@@ -3464,12 +3467,19 @@ HTSEXT_API char *fil_normalized(const char *source, char *dest) {
 }
 
 #define endwith(a) ( (len >= (sizeof(a)-1)) ? ( strncmp(dest, a+len-(sizeof(a)-1), sizeof(a)-1) == 0 ) : 0 );
-HTSEXT_API char *adr_normalized(const char *source, char *dest) {
+HTSEXT_API char *adr_normalized_sized(const char *source, char *dest,
+                                      size_t destsize) {
   /* not yet too aggressive (no com<->net<->org checkings) */
-  strcpybuff(dest, jump_normalized_const(source));
+  strlcpybuff(dest, jump_normalized_const(source), destsize);
   return dest;
 }
 
+// deprecated variant; kept for ABI compatibility. Bounds to the implicit
+// contract the old callers relied on (an HTS_URLMAXSIZE*2 URL buffer).
+HTSEXT_API char *adr_normalized(const char *source, char *dest) {
+  return adr_normalized_sized(source, dest, HTS_URLMAXSIZE * 2);
+}
+
 #undef endwith
 
 // find port (:80) or NULL if not found
@@ -3894,9 +3904,9 @@ HTSEXT_API size_t escape_for_html_print_full(const char *const s, char *const de
 
 #undef ADD_CHAR
 
-// conversion minuscules, avec buffer
-char *convtolower(char *catbuff, const char *a) {
-  strcpybuff(catbuff, a);
+// lower-case conversion into caller buffer (capacity catbuffsize)
+char *convtolower(char *catbuff, size_t catbuffsize, const char *a) {
+  strlcpybuff(catbuff, a, catbuffsize);
   hts_lowcase(catbuff);         // lower case
   return catbuff;
 }
@@ -3919,22 +3929,34 @@ void hts_replace(char *s, char from, char to) {
   }
 }
 
-// deviner type d'un fichier local..
-// ex: fil="toto.gif" -> s="image/gif"
-void guess_httptype(httrackp * opt, char *s, const char *fil) {
-  get_httptype(opt, s, fil, 1);
+// guess a local file's mime type (e.g. fil="toto.gif" -> s="image/gif")
+// returns 1 if a type was written to s, 0 otherwise
+int guess_httptype_sized(httrackp *opt, char *s, size_t ssize,
+                         const char *fil) {
+  return get_httptype_sized(opt, s, ssize, fil, 1);
 }
 
-// idem
-// flag: 1 si toujours renvoyer un type
-HTSEXT_API void get_httptype(httrackp * opt, char *s, const char *fil, int flag) {
-  // userdef overrides get_httptype
+// deprecated variant; kept for ABI compatibility. Bounds to the implicit
+// contract the old callers relied on (a contenttype-sized buffer).
+void guess_httptype(httrackp * opt, char *s, const char *fil) {
+  (void) get_httptype_sized(opt, s, HTS_MIMETYPE_SIZE, fil, 1);
+}
+
+// write the mime type for fil into s (capacity ssize)
+// flag: 1 to always return a type (the "application/..." / octet-stream
+// fallback) returns 1 if a type was written to s, 0 otherwise
+HTSEXT_API int get_httptype_sized(httrackp *opt, char *s, size_t ssize,
+                                  const char *fil, int flag) {
+  // userdef overrides get_httptype (a rule with an empty value, e.g. "--assume
+  // cgi=", matches but writes nothing: report it as "no type" like the old
+  // code, whose callers tested strnotempty(s))
   if (get_userhttptype(opt, s, fil)) {
-    return;
+    return s[0] != '\0';
   }
   // regular tests
   if (ishtml(opt, fil) == 1) {
-    strcpybuff(s, "text/html");
+    strlcpybuff(s, "text/html", ssize);
+    return 1;
   } else {
     /* Check html -> text/html */
     const char *a = fil + strlen(fil) - 1;
@@ -3947,21 +3969,33 @@ HTSEXT_API void get_httptype(httrackp * opt, char *s, const char *fil, int flag)
       a++;
       while(strnotempty(hts_mime[j][1])) {
         if (strfield2(hts_mime[j][1], a)) {
-          if (hts_mime[j][0][0] != '*') {       // Une correspondance existe
-            strcpybuff(s, hts_mime[j][0]);
-            return;
+          if (hts_mime[j][0][0] != '*') { // a match exists
+            strlcpybuff(s, hts_mime[j][0], ssize);
+            return 1;
           }
         }
         j++;
       }
 
-      if (flag)
-        sprintf(s, "application/%s", a);
+      if (flag) {
+        snprintf(s, ssize, "application/%s", a);
+        return 1;
+      }
     } else {
-      if (flag)
-        strcpybuff(s, "application/octet-stream");
+      if (flag) {
+        strlcpybuff(s, "application/octet-stream", ssize);
+        return 1;
+      }
     }
   }
+  return 0;
+}
+
+// deprecated variant; kept for ABI compatibility. Bounds to the implicit
+// contract the old callers relied on (a contenttype-sized buffer).
+HTSEXT_API void get_httptype(httrackp *opt, char *s, const char *fil,
+                             int flag) {
+  (void) get_httptype_sized(opt, s, HTS_MIMETYPE_SIZE, fil, flag);
 }
 
 // get type of fil (php)
@@ -4071,17 +4105,17 @@ int get_userhttptype(httrackp * opt, char *s, const char *fil) {
   return 0;
 }
 
-// renvoyer extesion d'un type mime..
-// ex: "image/gif" -> gif
-void give_mimext(char *s, const char *st) {
+// give the file extension for a mime type (e.g. "image/gif" -> "gif")
+// returns 1 if an extension was found (and written to s), 0 otherwise
+int give_mimext(char *s, size_t ssize, const char *st) {
   int ok = 0;
   int j = 0;
 
   s[0] = '\0';
   while((!ok) && (strnotempty(hts_mime[j][1]))) {
     if (strfield2(hts_mime[j][0], st)) {
-      if (hts_mime[j][1][0] != '*') {   // Une correspondance existe
-        strcpybuff(s, hts_mime[j][1]);
+      if (hts_mime[j][1][0] != '*') { // a match exists
+        strlcpybuff(s, hts_mime[j][1], ssize);
         ok = 1;
       }
     }
@@ -4102,12 +4136,13 @@ void give_mimext(char *s, const char *st) {
     if (a) {
       if ((int) strlen(a) >= 1) {
         if ((int) strlen(a) <= 4) {
-          strcpybuff(s, a);
+          strlcpybuff(s, a, ssize);
           ok = 1;
         }
       }
     }
   }
+  return ok;
 }
 
 // extension connue?..
@@ -4205,9 +4240,8 @@ int may_bogus_multiple(httrackp * opt, const char *mime, const char *filename) {
     if (strfield2(hts_mime_bogus_multiple[j], mime)) {  /* found mime type in suspicious list */
       char ext[64];
 
-      ext[0] = '\0';
-      give_mimext(ext, mime);
-      if (ext[0] != 0) {        /* we have an extension for that */
+      if (give_mimext(ext, sizeof(ext),
+                      mime)) { /* we have an extension for that */
         const size_t ext_size = strlen(ext);
         const char *file = strrchr(filename, '/');      /* fetch terminal filename */
 
@@ -4930,7 +4964,8 @@ void hts_freeall(void) {
 
 // cut path and project name
 // patch also initial path
-void cut_path(char *fullpath, char *path, char *pname) {
+void cut_path(char *fullpath, char *path, size_t path_size, char *pname,
+              size_t pname_size) {
   path[0] = pname[0] = '\0';
   if (strnotempty(fullpath)) {
     if ((fullpath[strlen(fullpath) - 1] == '/')
@@ -4946,8 +4981,8 @@ void cut_path(char *fullpath, char *path, char *pname) {
         a--;
       if (*a == '/')
         a++;
-      strcpybuff(pname, a);
-      strncatbuff(path, fullpath, (int) (a - fullpath));
+      strlcpybuff(pname, a, pname_size);
+      strlncatbuff(path, fullpath, path_size, (size_t) (a - fullpath));
     }
   }
 }

src/htslib.h

@@ -252,7 +252,7 @@ int ishtml_ext(const char *a);
 int ishttperror(int err);
 
 int get_userhttptype(httrackp * opt, char *s, const char *fil);
-void give_mimext(char *s, const char *st);
+int give_mimext(char *s, size_t ssize, const char *st);
 
 int may_bogus_multiple(httrackp * opt, const char *mime, const char *filename);
 int may_unknown2(httrackp * opt, const char *mime, const char *filename);
@@ -264,7 +264,7 @@ void code64(unsigned char *a, int size_a, unsigned char *b, int crlf);
 
 #define copychar(catbuff,a) concat(catbuff,(a),NULL)
 
-char *convtolower(char *catbuff, const char *a);
+char *convtolower(char *catbuff, size_t catbuffsize, const char *a);
 void hts_lowcase(char *s);
 void hts_replace(char *s, char from, char to);
 int multipleStringMatch(const char *s, const char *match);
@@ -276,7 +276,8 @@ void fprintfio(FILE * fp, const char *buff, const char *prefix);
 int sig_ignore_flag(int setflag);  // flag ignore
 #endif
 
-void cut_path(char *fullpath, char *path, char *pname);
+void cut_path(char *fullpath, char *path, size_t path_size, char *pname,
+              size_t pname_size);
 int fexist(const char *s);
 int fexist_utf8(const char *s);
 
@@ -499,7 +500,8 @@ HTS_STATIC int is_hypertext_mime(httrackp * opt, const char *mime,
     char guessed[256];
 
     guessed[0] = '\0';
-    guess_httptype(opt, guessed, file);
+    if (!guess_httptype_sized(opt, guessed, sizeof(guessed), file))
+      return 0;
     return is_hypertext_mime__(guessed);
   }
   return 0;
@@ -514,7 +516,8 @@ HTS_STATIC int may_be_hypertext_mime(httrackp * opt, const char *mime,
     char guessed[256];
 
     guessed[0] = '\0';
-    guess_httptype(opt, guessed, file);
+    if (!guess_httptype_sized(opt, guessed, sizeof(guessed), file))
+      return 0;
     return may_be_hypertext_mime__(guessed);
   }
   return 0;
@@ -529,7 +532,8 @@ HTS_STATIC int compare_mime(httrackp * opt, const char *mime, const char *file,
     char guessed[256];
 
     guessed[0] = '\0';
-    guess_httptype(opt, guessed, file);
+    if (!guess_httptype_sized(opt, guessed, sizeof(guessed), file))
+      return 0;
     return strfield2(guessed, reference);
   }
   return 0;

src/htsname.c

@@ -51,12 +51,13 @@ Please visit our Website: http://www.httrack.com
       url_savename_addstr(afs->save, buff);\
     }
 
-#define ADD_STANDARD_NAME(shortname) \
-    {  /* ajout nom */\
-      char BIGSTK buff[HTS_URLMAXSIZE*2];\
-      standard_name(buff,dot_pos,nom_pos,fil_complete,(shortname));\
-      url_savename_addstr(afs->save, buff);\
-    }
+#define ADD_STANDARD_NAME(shortname)                                           \
+  { /* add name */                                                             \
+    char BIGSTK buff[HTS_URLMAXSIZE * 2];                                      \
+    standard_name(buff, sizeof(buff), dot_pos, nom_pos, fil_complete,          \
+                  (shortname));                                                \
+    url_savename_addstr(afs->save, buff);                                      \
+  }
 
 /* Avoid stupid DOS system folders/file such as 'nul' */
 /* Based on linux/fs/umsdos/mangle.c */
@@ -200,7 +201,7 @@ int url_savename(lien_adrfilsave *const afs,
   // foo.com/bar//foobar -> foo.com/bar/foobar
   if (opt->urlhack) {
     // copy of adr (without protocol), used for lookups (see urlhack)
-    normadr = adr_normalized(adr, normadr_);
+    normadr = adr_normalized_sized(adr, normadr_, sizeof(normadr_));
     normfil = fil_normalized(fil_complete, normfil_);
   } else {
     if (link_has_authority(adr_complete)) {     // https or other protocols : in "http/" subfolder
@@ -344,8 +345,7 @@ int url_savename(lien_adrfilsave *const afs,
         mime[0] = ext[0] = '\0';
         get_userhttptype(opt, mime, fil);
         if (strnotempty(mime)) {
-          give_mimext(ext, mime);
-          if (strnotempty(ext)) {
+          if (give_mimext(ext, sizeof(ext), mime)) {
             ext_chg = 1;
           }
         }
@@ -378,8 +378,8 @@ int url_savename(lien_adrfilsave *const afs,
                 ext_chg = 2;      /* change filename */
                 strcpybuff(ext, r.cdispo);
               } else if (!may_unknown2(opt, r.contenttype, fil)) {        // on peut patcher à priori?
-                give_mimext(s, r.contenttype);    // obtenir extension
-                if (strnotempty(s) > 0) { // on a reconnu l'extension
+                if (give_mimext(s, sizeof(s),
+                                r.contenttype)) { // recognized extension
                   ext_chg = 1;
                   strcpybuff(ext, s);
                 }
@@ -403,8 +403,7 @@ int url_savename(lien_adrfilsave *const afs,
             mime[0] = ext[0] = '\0';
             get_userhttptype(opt, mime, fil);
             if (strnotempty(mime)) {
-              give_mimext(ext, mime);
-              if (strnotempty(ext)) {
+              if (give_mimext(ext, sizeof(ext), mime)) {
                 ext_chg = 1;
               }
             }
@@ -420,9 +419,9 @@ int url_savename(lien_adrfilsave *const afs,
                 strcpybuff(ext, headers->r.cdispo);
               } else if (!may_unknown2(opt, headers->r.contenttype, headers->url_fil)) {    // on peut patcher à priori? (pas interdit ou pas de type)
                 char s[16];
-                s[0] = '\0';
-                give_mimext(s, headers->r.contenttype);    // obtenir extension
-                if (strnotempty(s) > 0) { // on a reconnu l'extension
+                if (give_mimext(
+                        s, sizeof(s),
+                        headers->r.contenttype)) { // recognized extension
                   ext_chg = 1;
                   strcpybuff(ext, s);
                 }
@@ -431,13 +430,14 @@ int url_savename(lien_adrfilsave *const afs,
             else if (mime_type != NULL) {
               ext[0] = '\0';
               if (*mime_type) {
-                give_mimext(ext, mime_type);
+                give_mimext(ext, sizeof(ext), mime_type);
               }
               if (strnotempty(ext)) {
                 char mime_from_file[128];
 
                 mime_from_file[0] = 0;
-                get_httptype(opt, mime_from_file, fil, 1);
+                get_httptype_sized(opt, mime_from_file, sizeof(mime_from_file),
+                                   fil, 1);
                 if (!strnotempty(mime_from_file) || strcasecmp(mime_type, mime_from_file) != 0) {       /* different mime for this type */
                   /* type change not forbidden (or no extension at all) */
                   if (!may_unknown2(opt, mime_type, fil)) {
@@ -646,8 +646,9 @@ int url_savename(lien_adrfilsave *const afs,
                       ext_chg = 2;      /* change filename */
                       strcpybuff(ext, back[b].r.cdispo);
                     } else if (!may_unknown2(opt, back[b].r.contenttype, back[b].url_fil)) {    // on peut patcher à priori? (pas interdit ou pas de type)
-                      give_mimext(s, back[b].r.contenttype);    // obtenir extension
-                      if (strnotempty(s) > 0) { // on a reconnu l'extension
+                      if (give_mimext(
+                              s, sizeof(s),
+                              back[b].r.contenttype)) { // recognized extension
                         ext_chg = 1;
                         strcpybuff(ext, s);
                       }
@@ -924,7 +925,7 @@ int url_savename(lien_adrfilsave *const afs,
 
               pth[0] = n83[0] = '\0';
               strncatbuff(pth, fil, (int) (nom_pos - fil) - 1);
-              long_to_83(opt->savename_83, n83, pth);
+              long_to_83(opt->savename_83, n83, sizeof(n83), pth);
               htsbuff_cat(&sb, n83);
             }
           }
@@ -1306,7 +1307,7 @@ int url_savename(lien_adrfilsave *const afs,
   if (opt->savename_83) {
     char BIGSTK n83[HTS_URLMAXSIZE * 2];
 
-    long_to_83(opt->savename_83, n83, afs->save);
+    long_to_83(opt->savename_83, n83, sizeof(n83), afs->save);
     strcpybuff(afs->save, n83);
   }
   // enforce stricter ISO9660 compliance (bug reported by Steffo Carlsson)
@@ -1377,7 +1378,9 @@ int url_savename(lien_adrfilsave *const afs,
     if (lastDot == NULL) {
       strcatbuff(afs->save, "." DELAYED_EXT);
     } else if (!IS_DELAYED_EXT(afs->save)) {
-      strcatbuff(lastDot, "." DELAYED_EXT);
+      /* lastDot points within afs->save; bound by the remaining capacity */
+      strlcatbuff(lastDot, "." DELAYED_EXT,
+                  sizeof(afs->save) - (size_t) (lastDot - afs->save));
     }
   }
   // enforce 260-character path limit before inserting destination path
@@ -1582,41 +1585,41 @@ int url_savename(lien_adrfilsave *const afs,
   return 0;
 }
 
-/* nom avec md5 urilisé partout */
-void standard_name(char *b, const char *dot_pos, const char *nom_pos, const char *fil,
-                   int short_ver) {
+/* md5-based name used everywhere; builds into b (capacity bsize) */
+void standard_name(char *b, size_t bsize, const char *dot_pos,
+                   const char *nom_pos, const char *fil, int short_ver) {
   char md5[32 + 2];
+  htsbuff bb = htsbuff_ptr(b, bsize);
 
-  b[0] = '\0';
-  /* Nom */
+  /* Name */
   if (dot_pos) {
-    if (!short_ver)             // Noms longs
-      strncatbuff(b, nom_pos, (dot_pos - nom_pos));
+    if (!short_ver) // long names
+      htsbuff_catn(&bb, nom_pos, (size_t) (dot_pos - nom_pos));
     else
-      strncatbuff(b, nom_pos, min(dot_pos - nom_pos, 8));
+      htsbuff_catn(&bb, nom_pos, (size_t) min(dot_pos - nom_pos, 8));
   } else {
-    if (!short_ver)             // Noms longs
-      strcatbuff(b, nom_pos);
+    if (!short_ver) // long names
+      htsbuff_cat(&bb, nom_pos);
     else
-      strncatbuff(b, nom_pos, 8);
+      htsbuff_catn(&bb, nom_pos, 8);
   }
   /* MD5 - 16 bits */
-  strncatbuff(b, url_md5(md5, fil), 4);
+  htsbuff_catn(&bb, url_md5(md5, fil), 4);
   /* Ext */
   if (dot_pos) {
-    strcatbuff(b, ".");
-    if (!short_ver)             // Noms longs
-      strcatbuff(b, dot_pos + 1);
+    htsbuff_catc(&bb, '.');
+    if (!short_ver) // long names
+      htsbuff_cat(&bb, dot_pos + 1);
     else
-      strncatbuff(b, dot_pos + 1, 3);
+      htsbuff_catn(&bb, dot_pos + 1, 3);
   }
   // Allow extensionless
 #ifdef DO_NOT_ALLOW_EXTENSIONLESS
   else {
-    if (!short_ver)             // Noms longs
-      strcatbuff(b, DEFAULT_EXT);
+    if (!short_ver) // long names
+      htsbuff_cat(&bb, DEFAULT_EXT);
     else
-      strcatbuff(b, DEFAULT_EXT_SHORT);
+      htsbuff_cat(&bb, DEFAULT_EXT_SHORT);
   }
 #endif
 }

src/htsname.h

@@ -96,8 +96,8 @@ int url_savename(lien_adrfilsave *const afs,
                  httrackp * opt, struct_back * sback, cache_back * cache,
                  hash_struct * hash, int ptr, int numero_passe,
                  const lien_back * headers);
-void standard_name(char *b, const char *dot_pos, const char *nom_pos,
-                   const char *fil_complete,
+void standard_name(char *b, size_t bsize, const char *dot_pos,
+                   const char *nom_pos, const char *fil_complete,
                    int short_ver);
 void url_savename_addstr(char *d, const char *s);
 char *url_md5(char *digest_buffer, const char *fil_complete);

src/htsopt.h

@@ -499,9 +499,9 @@ struct htsblk {
   FILE *out;                    // écriture directe sur disque (si is_write=1)
   LLint size;                   // taille fichier
   char msg[80];                 // message éventuel si échec ("\0"=non précisé)
-  char contenttype[64];         // content-type ("text/html" par exemple)
-  char charset[64];             // charset ("iso-8859-1" par exemple)
-  char contentencoding[64];     // content-encoding ("gzip" par exemple)
+  char contenttype[HTS_MIMETYPE_SIZE];     // content-type (e.g. "text/html")
+  char charset[HTS_MIMETYPE_SIZE];         // charset (e.g. "iso-8859-1")
+  char contentencoding[HTS_MIMETYPE_SIZE]; // content-encoding (e.g. "gzip")
   char *location;               // on copie dedans éventuellement la véritable 'location'
   LLint totalsize;              // taille totale à télécharger (-1=inconnue)
   short int is_file;            // ce n'est pas une socket mais un descripteur de fichier si 1

src/htsparse.c

@@ -610,11 +610,11 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
                           b = strchr(a, '<');   // prochain tag
                         }
                       }
-                      if (lienrelatif
-                          (tempo, heap(ptr)->sav,
-                           concat(OPT_GET_BUFF(opt), OPT_GET_BUFF_SIZE(opt),
-                                  StringBuff(opt->path_html_utf8),
-                                  "index.html")) == 0) {
+                      if (lienrelatif(tempo, sizeof(tempo), heap(ptr)->sav,
+                                      concat(OPT_GET_BUFF(opt),
+                                             OPT_GET_BUFF_SIZE(opt),
+                                             StringBuff(opt->path_html_utf8),
+                                             "index.html")) == 0) {
                         detect_title = 1;       // ok détecté pour cette page!
                         makeindex_links++;      // un de plus
                         strcpybuff(makeindex_firstlink, tempo);
@@ -1649,8 +1649,9 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
                                 }
                                 // Prendre si extension reconnue
                                 if (!url_ok) {
-                                  get_httptype(opt, type, tempo, 0);
-                                  if (strnotempty(type))        // type reconnu!
+                                  if (get_httptype_sized(opt, type,
+                                                         sizeof(type), tempo,
+                                                         0)) // recognized type
                                     url_ok = 1;
                                   else if (is_dyntype(get_ext(OPT_GET_BUFF(opt), OPT_GET_BUFF_SIZE(opt), tempo)))       // reconnu php,cgi,asp..
                                     url_ok = 1;
@@ -2719,7 +2720,8 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
 
                               strcpybuff(save, StringBuff(opt->path_html_utf8));
                               strcatbuff(save, cat_name);
-                              if (lienrelatif(tempo, save, relativesavename()) == 0) {
+                              if (lienrelatif(tempo, sizeof(tempo), save,
+                                              relativesavename()) == 0) {
                                 /* Never escape high-chars (we don't know the encoding!!) */
                                 inplace_escape_uri_utf(tempo, sizeof(tempo));  // escape with %xx
                                 //if (!no_esc_utf)
@@ -2949,7 +2951,8 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
                       tempo[0] = '\0';
                       // calculer le lien relatif
 
-                      if (lienrelatif(tempo, afs.save, relativesavename()) == 0) {
+                      if (lienrelatif(tempo, sizeof(tempo), afs.save,
+                                      relativesavename()) == 0) {
                         if (!in_media) {        // In media (such as real audio): don't patch
                           /* Never escape high-chars (we don't know the encoding!!) */
                           inplace_escape_uri_utf(tempo, sizeof(tempo));
@@ -3507,9 +3510,9 @@ int hts_mirror_check_moved(htsmoduleStruct * str,
             char BIGSTK pn_adr[HTS_URLMAXSIZE * 2], pn_fil[HTS_URLMAXSIZE * 2];
 
             n_adr[0] = n_fil[0] = '\0';
-            (void) adr_normalized(moved->adr, n_adr);
+            (void) adr_normalized_sized(moved->adr, n_adr, sizeof(n_adr));
             (void) fil_normalized(moved->fil, n_fil);
-            (void) adr_normalized(urladr(), pn_adr);
+            (void) adr_normalized_sized(urladr(), pn_adr, sizeof(pn_adr));
             (void) fil_normalized(urlfil(), pn_fil);
             if (strcasecmp(n_adr, pn_adr) == 0
                 && strcasecmp(n_fil, pn_fil) == 0) {

src/htssafe.h

@@ -237,6 +237,15 @@ static char *strncatbuff_ptr_(char *dest, const char *src, size_t n) {
   HTS_IS_NOT_CHAR_BUFFER(B) ? (size_t) -1 : sizeof(B), (size_t) -1, \
   "overflow while appending '" #B "' to '"#A"'", __FILE__, __LINE__)
 
+/**
+ * Append at most "N" characters of "B" to "A", "A" having a maximum capacity
+ * of "S".
+ */
+#define strlncatbuff(A, B, S, N)                                               \
+  strncat_safe_(A, S, B, HTS_IS_NOT_CHAR_BUFFER(B) ? (size_t) -1 : sizeof(B),  \
+                N, "overflow while appending '" #B "' to '" #A "'", __FILE__,  \
+                __LINE__)
+
 /**
  * Copy characters of "B" to "A", "A" having a maximum capacity of "S".
  */

src/htstools.c

@@ -274,7 +274,9 @@ int ident_url_relatif(const char *lien, const char *origin_adr,
       char *const idna = hts_convertStringUTF8ToIDNA(a, strlen(a));
       if (idna != NULL) {
         if (strlen(idna) < HTS_URLMAXSIZE) {
-          strcpybuff(a, idna);
+          /* a points within adrfil->adr; bound by the remaining capacity */
+          strlcpybuff(a, idna,
+                      sizeof(adrfil->adr) - (size_t) (a - adrfil->adr));
         }
         free(idna);
       }
@@ -286,7 +288,7 @@ int ident_url_relatif(const char *lien, const char *origin_adr,
 
 // créer dans s, à partir du chemin courant curr_fil, le lien vers link (absolu)
 // un ident_url_relatif a déja été fait avant, pour que link ne soit pas un chemin relatif
-int lienrelatif(char *s, const char *link, const char *curr_fil) {
+int lienrelatif(char *s, size_t ssize, const char *link, const char *curr_fil) {
   char BIGSTK _curr[HTS_URLMAXSIZE * 2];
   char BIGSTK newcurr_fil[HTS_URLMAXSIZE * 2], newlink[HTS_URLMAXSIZE * 2];
   char *curr;
@@ -314,9 +316,9 @@ int lienrelatif(char *s, const char *link, const char *curr_fil) {
     }
   }
 
-  // recopier uniquement le chemin courant
+  // copy only the current path
   curr = _curr;
-  strcpybuff(curr, curr_fil);
+  strlcpybuff(curr, curr_fil, sizeof(_curr));
   if ((a = strchr(curr, '?')) == NULL)  // couper au ? (params)
     a = curr + strlen(curr) - 1;        // pas de params: aller à la fin
   while((*a != '/') && (a > curr))
@@ -359,14 +361,14 @@ int lienrelatif(char *s, const char *link, const char *curr_fil) {
     a++;
   while(*a)
     if (*(a++) == '/')
-      strcatbuff(s, "../");
+      strlcatbuff(s, "../", ssize);
   //if (strlen(s)==0) strcatbuff(s,"/");
 
   if (slash)
-    strcatbuff(s, "/");         // garder absolu!!
+    strlcatbuff(s, "/", ssize); // keep it absolute!
 
-  // on est dans le répertoire de départ, copier
-  strcatbuff(s, link + ((*link == '/') ? 1 : 0));
+  // we are in the starting directory, copy
+  strlcatbuff(s, link + ((*link == '/') ? 1 : 0), ssize);
 
   /* Security check */
   if (strlen(s) >= HTS_URLMAXSIZE)
@@ -410,7 +412,7 @@ int link_has_authorization(const char *lien) {
 }
 
 // conversion chemin de fichier/dossier vers 8-3 ou ISO9660
-void long_to_83(int mode, char *n83, char *save) {
+void long_to_83(int mode, char *n83, size_t n83size, char *save) {
   n83[0] = '\0';
 
   while(*save) {
@@ -425,19 +427,19 @@ void long_to_83(int mode, char *n83, char *save) {
     }
     fnl[j] = '\0';
     // conversion
-    longfile_to_83(mode, fn83, fnl);
-    strcatbuff(n83, fn83);
+    longfile_to_83(mode, fn83, sizeof(fn83), fnl);
+    strlcatbuff(n83, fn83, n83size);
 
     save += i;
     if (*save == '/') {
-      strcatbuff(n83, "/");
+      strlcatbuff(n83, "/", n83size);
       save++;
     }
   }
 }
 
 // conversion nom de fichier/dossier isolé vers 8-3 ou ISO9660
-void longfile_to_83(int mode, char *n83, char *save) {
+void longfile_to_83(int mode, char *n83, size_t n83size, char *save) {
   int j = 0, max = 0;
   int i = 0;
   char nom[256];
@@ -526,10 +528,10 @@ void longfile_to_83(int mode, char *n83, char *save) {
   }
   // corriger vers 8-3
   n83[0] = '\0';
-  strncatbuff(n83, nom, max);
+  strlncatbuff(n83, nom, n83size, max);
   if (strnotempty(ext)) {
-    strcatbuff(n83, ".");
-    strncatbuff(n83, ext, 3);
+    strlcatbuff(n83, ".", n83size);
+    strlncatbuff(n83, ext, n83size, 3);
   }
 }
 

src/htstools.h

@@ -61,11 +61,11 @@ typedef struct lien_adrfilsave lien_adrfilsave;
 int ident_url_relatif(const char *lien, const char *origin_adr,
                       const char *origin_fil,
                       lien_adrfil* const adrfil);
-int lienrelatif(char *s, const char *link, const char *curr);
+int lienrelatif(char *s, size_t ssize, const char *link, const char *curr);
 int link_has_authority(const char *lien);
 int link_has_authorization(const char *lien);
-void long_to_83(int mode, char *n83, char *save);
-void longfile_to_83(int mode, char *n83, char *save);
+void long_to_83(int mode, char *n83, size_t n83size, char *save);
+void longfile_to_83(int mode, char *n83, size_t n83size, char *save);
 HTS_INLINE int __rech_tageq(const char *adr, const char *s);
 HTS_INLINE int __rech_tageqbegdigits(const char *adr, const char *s);
 HTS_INLINE int rech_tageq_all(const char *adr, const char *s);

src/htswizard.c

@@ -223,8 +223,9 @@ static int hts_acceptlink_(httrackp * opt, int ptr,
       // note (up/down): on calcule à partir du lien primaire, ET du lien précédent.
       // ex: si on descend 2 fois on peut remonter 1 fois
 
-      if (lienrelatif(tempo, fil, heap(heap(ptr)->premier)->fil) == 0) {
-        if (lienrelatif(tempo2, fil, heap(ptr)->fil) == 0) {
+      if (lienrelatif(tempo, sizeof(tempo), fil,
+                      heap(heap(ptr)->premier)->fil) == 0) {
+        if (lienrelatif(tempo2, sizeof(tempo2), fil, heap(ptr)->fil) == 0) {
           hts_log_print(opt, LOG_DEBUG,
                         "build relative links to test: %s %s (with %s and %s)",
                         tempo, tempo2, heap(heap(ptr)->premier)->fil,
@@ -326,8 +327,9 @@ static int hts_acceptlink_(httrackp * opt, int ptr,
       char BIGSTK tempo[HTS_URLMAXSIZE * 2];
       char BIGSTK tempo2[HTS_URLMAXSIZE * 2];
 
-      if (lienrelatif(tempo, fil, heap(heap(ptr)->premier)->fil) == 0) {
-        if (lienrelatif(tempo2, fil, heap(ptr)->fil) == 0) {
+      if (lienrelatif(tempo, sizeof(tempo), fil,
+                      heap(heap(ptr)->premier)->fil) == 0) {
+        if (lienrelatif(tempo2, sizeof(tempo2), fil, heap(ptr)->fil) == 0) {
         } else {
           hts_log_print(opt, LOG_ERROR,
                         "Error building relative link %s and %s", fil,
@@ -336,7 +338,6 @@ static int hts_acceptlink_(httrackp * opt, int ptr,
       } else {
         hts_log_print(opt, LOG_ERROR, "Error building relative link %s and %s",
                       fil, heap(heap(ptr)->premier)->fil);
-
       }
     }                           // fin tester interdiction de monter
 

src/httrack-library.h

@@ -207,6 +207,9 @@ HTSEXT_API const char *jump_normalized_const(const char *);
 HTSEXT_API char *jump_toport(char *);
 HTSEXT_API const char *jump_toport_const(const char *);
 HTSEXT_API char *fil_normalized(const char *source, char *dest);
+HTSEXT_API char *adr_normalized_sized(const char *source, char *dest,
+                                      size_t destsize);
+HTS_DEPRECATED("use adr_normalized_sized(source, dest, destsize)")
 HTSEXT_API char *adr_normalized(const char *source, char *dest);
 HTSEXT_API const char *hts_rootdir(char *file);
 
@@ -244,6 +247,9 @@ HTSEXT_API char *unescape_http_unharm(char *const catbuff, const size_t size, co
 HTSEXT_API char *antislash_unescaped(char *catbuff, const char *s);
 
 HTSEXT_API void escape_remove_control(char *s);
+HTSEXT_API int get_httptype_sized(httrackp *opt, char *s, size_t ssize,
+                                  const char *fil, int flag);
+HTS_DEPRECATED("use get_httptype_sized(opt, s, ssize, fil, flag)")
 HTSEXT_API void get_httptype(httrackp * opt, char *s, const char *fil,
                              int flag);
 HTSEXT_API int is_knowntype(httrackp * opt, const char *fil);
@@ -251,6 +257,9 @@ HTSEXT_API int is_userknowntype(httrackp * opt, const char *fil);
 HTSEXT_API int is_dyntype(const char *fil);
 HTSEXT_API const char *get_ext(char *catbuff, size_t size, const char *fil);
 HTSEXT_API int may_unknown(httrackp * opt, const char *st);
+HTSEXT_API int guess_httptype_sized(httrackp *opt, char *s, size_t ssize,
+                                    const char *fil);
+HTS_DEPRECATED("use guess_httptype_sized(opt, s, ssize, fil)")
 HTSEXT_API void guess_httptype(httrackp * opt, char *s, const char *fil);
 
 /* Ugly string tools */

tests/01_engine-cmdline.test

@@ -30,6 +30,17 @@ run() {
     RC=$?
 }
 
+# crawl using exactly the given args as the only URL(s), no implicit primary URL;
+# leaves the exit status in RC
+run_only() {
+    local out="$1"
+    shift
+    rm -rf "$out"
+    mkdir -p "$out"
+    httrack -O "$out" --quiet -n "$@" >"$out/.log" 2>&1
+    RC=$?
+}
+
 # assert the value was accepted: clean exit and the fixture was mirrored
 accepted() {
     { test "$RC" -eq 0 && test -n "$(find "$1" -type f -path '*/index.html' -print -quit)"; } ||
@@ -68,4 +79,15 @@ refused "#152: over-cap -F not refused cleanly"
 run "$tmp/ov-l" --user-agent "$over"
 refused "#152: over-cap --user-agent not refused cleanly"
 
+# Quote handling on the sole URL (run_only, so the quoted arg is the only URL and
+# can't be masked by an implicit one). A fully "-quoted URL has its surrounding
+# quotes stripped in place and is mirrored; a dangling opening quote, and a lone
+# quote (empty after the opening "), are refused cleanly and never crash.
+run_only "$tmp/q-ok" "\"file://$tmp/index.html\""
+accepted "$tmp/q-ok" "quoted URL not stripped/mirrored"
+run_only "$tmp/q-bad" '"foo'
+refused "dangling-quote argument not refused cleanly"
+run_only "$tmp/q-lone" '"'
+refused "lone-quote argument not refused cleanly"
+
 exit 0