Bound htstools.c pointer-destination buffer writes (batch 12)

Continues the htssafe.h pointer-destination migration: the strcpybuff/strcatbuff
macros silently fall back to a raw strcpy/strcat when the destination is a bare
char* rather than a sized array.

All four functions are internal (hidden, not HTSEXT_API), so they take explicit
destination sizes:
* lienrelatif() builds a relative link into a char* caller buffer; threads a
  size_t and bounds the "../"/path appends with strlcatbuff (the local _curr
  copy uses sizeof(_curr)).
* long_to_83() / longfile_to_83() build an 8-3 / ISO9660 name into a caller
  buffer; thread a size_t and use strl(n)catbuff.
* ident_url_relatif()'s in-place IDNA host rewrite bounds the copy by the
  remaining capacity of adrfil->adr (a pointer into that array).

Callers in htscore.c, htswizard.c, htsparse.c and htsname.c pass sizeof(dest)
(all the destinations are HTS_URLMAXSIZE*2 arrays).

Add -#7 basic_selftests for longfile_to_83 (8-3 and ISO9660), long_to_83
(per-segment path conversion) and lienrelatif (same-dir basename, parent "../").

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

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

src/htscore.c

@@ -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

@@ -431,6 +431,50 @@ static void basic_selftests(void) {
       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).

src/htsname.c

@@ -925,7 +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);
             }
           }
@@ -1307,7 +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)

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);
@@ -2720,7 +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)
@@ -2950,7 +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));

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