htsparse: keep makestat_time out of ENGINE_SET_CONTEXT

makestat_time throttles the makestat/maketrack stats to once per minute:
the wait loop compares time_local() against it and, when it fires, writes
it back to the local. But the field is by-value in the extended context,
so it can't round-trip through ENGINE_SAVE_CONTEXT, while ENGINE_SET_CONTEXT
re-read it from the load-once baseline on every loop iteration. That reset
the local before the next compare, so under -%v / maketrack the throttle
never held and the stats line plus the full back-stack dump were emitted
every iteration.

Drop makestat_time (and the never-changing makestat_fp) from SET_CONTEXT;
they belong to the load-once set. Wrapped the macro in clang-format off/on
for the same backslash-realignment reason as HT_ADD_END.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Signed-off-by: Xavier Roche <roche@httrack.com>

debian/control

@@ -1,8 +1,9 @@
 Source: httrack
 Section: web
+Priority: optional
 Maintainer: Xavier Roche <roche@httrack.com>
 Standards-Version: 4.7.4
-Build-Depends: debhelper-compat (= 14), autoconf, autoconf-archive, automake, libtool, zlib1g-dev, libssl-dev
+Build-Depends: debhelper-compat (= 13), autoconf, autoconf-archive, automake, libtool, zlib1g-dev, libssl-dev
 Rules-Requires-Root: no
 Homepage: http://www.httrack.com
 Vcs-Git: https://github.com/xroche/httrack.git

debian/source/lintian-overrides

@@ -1,6 +1,4 @@
-# Maintainer uploads sign the changelog as xavier@debian.org while the control
-# Maintainer is roche@httrack.com; lintian reads the address mismatch as an NMU.
-httrack source: no-nmu-in-changelog
+httrack source: changelog-should-mention-nmu
 httrack source: source-nmu-has-incorrect-version-number
 
 # The bundled HTML pages are the genuine upstream documentation taken from

debian/upstream/metadata

@@ -1,6 +0,0 @@
----
-Repository: https://github.com/xroche/httrack.git
-Repository-Browse: https://github.com/xroche/httrack
-Bug-Database: https://github.com/xroche/httrack/issues
-Bug-Submit: https://github.com/xroche/httrack/issues/new
-Contact: Xavier Roche <roche@httrack.com>

src/htscore.c

@@ -441,72 +441,6 @@ void hts_finish_makeindex(httrackp *opt, int *makeindex_done,
   *makeindex_done = 1;
 }
 
-/* Flush the parsed HTML output buffer to disk, skipping the rewrite when the
- * on-disk MD5 is unchanged. */
-void hts_finish_html_file(httrackp *opt, cache_back *cache, htsblk *r,
-                          FILE **fp, const char *ht_buff, size_t ht_len,
-                          const char *adr, const char *fil, const char *save) {
-  char digest[32 + 2];
-  off_t fsize_old =
-      fsize(fconv(OPT_GET_BUFF(opt), OPT_GET_BUFF_SIZE(opt), save));
-  int ok = 0;
-
-  digest[0] = '\0';
-  domd5mem(ht_buff, ht_len, digest, 1);
-  if (fsize_old == (off_t) ht_len) {
-    int mlen = 0;
-    char *mbuff;
-
-    cache_readdata(cache, "//[HTML-MD5]//", save, &mbuff, &mlen);
-    if (mlen)
-      mbuff[mlen] = '\0';
-    if ((mlen == 32) && (strcmp(((mbuff != NULL) ? mbuff : ""), digest) == 0)) {
-      ok = 1;
-      hts_log_print(opt, LOG_DEBUG, "File not re-written (md5): %s", save);
-    }
-    freet(mbuff);
-  }
-  if (!ok) {
-    file_notify(opt, adr, fil, save, 1, 1, r->notmodified);
-    *fp = filecreate(&opt->state.strc, save);
-    if (*fp) {
-      if (ht_len > 0 && fwrite(ht_buff, 1, ht_len, *fp) != ht_len) {
-        int fcheck = check_fatal_io_errno();
-
-        if (fcheck)
-          opt->state.exit_xh = -1;
-        if (opt->log) {
-          hts_log_print(opt, LOG_ERROR | LOG_ERRNO,
-                        "Unable to write HTML file %s", save);
-          if (fcheck)
-            hts_log_print(opt, LOG_ERROR, "* * Fatal write error, giving up");
-        }
-      }
-      fclose(*fp);
-      *fp = NULL;
-      if (strnotempty(r->lastmodified))
-        set_filetime_rfc822(save, r->lastmodified);
-    } else {
-      int fcheck = check_fatal_io_errno();
-
-      if (fcheck) {
-        hts_log_print(opt, LOG_ERROR,
-                      "Mirror aborted: disk full or filesystem problems");
-        opt->state.exit_xh = -1;
-      }
-      hts_log_print(opt, LOG_ERROR | LOG_ERRNO, "Unable to save file %s", save);
-      if (fcheck)
-        hts_log_print(opt, LOG_ERROR, "* * Fatal write error, giving up");
-    }
-  } else {
-    file_notify(opt, adr, fil, save, 0, 0, r->notmodified);
-    filenote(&opt->state.strc, save, NULL);
-  }
-  if (cache->ndx)
-    cache_writedata(cache->ndx, cache->dat, "//[HTML-MD5]//", save, digest,
-                    (int) strlen(digest));
-}
-
 /* does it look like XML ? (SVG et al.) */
 static int look_like_xml(const char *s) {
   return strncmp(s, "<?xml", 5) == 0

src/htscore.h

@@ -370,12 +370,6 @@ void hts_finish_makeindex(httrackp *opt, int *makeindex_done,
                           const char *template_footer, const char *adr,
                           const char *fil);
 
-// Flush ht_buff[0..ht_len] to save on disk (skip if MD5 unchanged); *fp
-// closed+NULLed on write. Precondition: ht_len>0.
-void hts_finish_html_file(httrackp *opt, cache_back *cache, htsblk *r,
-                          FILE **fp, const char *ht_buff, size_t ht_len,
-                          const char *adr, const char *fil, const char *save);
-
 int filters_init(char ***ptrfilters, int maxfilter, int filterinc);
 
 int fspc(httrackp * opt, FILE * fp, const char *type);

src/htsftp.c

@@ -128,31 +128,6 @@ void launch_ftp(FTPDownloadStruct * params) {
   return 0; \
   }
 
-/* Bounded split of a hostile-URL "user[:pass]@" prefix (see htsftp.h). */
-void ftp_split_userpass(const char *src, const char *end, char *user,
-                        size_t user_size, char *pass, size_t pass_size) {
-  size_t n = 0;
-
-  while (src[n] != '\0' && src[n] != ':') {
-    if (n < user_size - 1)
-      user[n] = src[n];
-    n++;
-  }
-  user[n < user_size ? n : user_size - 1] = '\0';
-  pass[0] = '\0';
-  if (src[n] == ':') { // password follows the colon
-    const size_t base = n + 1;
-    size_t k = 0;
-
-    while (&src[base + k + 1] < end && src[base + k] != '\0') {
-      if (k < pass_size - 1)
-        pass[k] = src[base + k];
-      k++;
-    }
-    pass[k < pass_size ? k : pass_size - 1] = '\0';
-  }
-}
-
 // la véritable fonction une fois lancées les routines thread/fork
 int run_launch_ftp(FTPDownloadStruct * pStruct) {
   lien_back *back = pStruct->pBack;
@@ -198,7 +173,24 @@ int run_launch_ftp(FTPDownloadStruct * pStruct) {
   while(*real_adr == '/')
     real_adr++;                 // sauter /
   if ((adr = jump_identification(real_adr)) != real_adr) {      // user
-    ftp_split_userpass(real_adr, adr, user, sizeof(user), pass, sizeof(pass));
+    int i = -1;
+
+    pass[0] = '\0';
+    do {
+      i++;
+      user[i] = real_adr[i];
+    } while((real_adr[i] != ':') && (real_adr[i]));
+    user[i] = '\0';
+    if (real_adr[i] == ':') {   // pass
+      int j = -1;
+
+      i++;                      // oui on saute aussi le :
+      do {
+        j++;
+        pass[j] = real_adr[i + j];
+      } while(((&real_adr[i + j + 1]) < adr) && (real_adr[i + j]));
+      pass[j] = '\0';
+    }
   }
   // Calculer RETR <nom>
   {
@@ -992,8 +984,8 @@ int get_ftp_line(T_SOC soc, char *ptrline, size_t line_size, int timeout) {
       //case 0: break;    // pas encore --> erreur (on attend)!
     case 1:
       HTS_STAT.HTS_TOTAL_RECV += 1;     // compter flux entrant
-      if ((b != 10) && (b != 13) && (i < (int) sizeof(data) - 1))
-        data[i++] = b; // truncate hostile over-long reply lines
+      if ((b != 10) && (b != 13))
+        data[i++] = b;
       break;
     default:
       if (ptrline)

src/htsftp.h

@@ -70,10 +70,6 @@ int back_launch_ftp(FTPDownloadStruct * params);
 int run_launch_ftp(FTPDownloadStruct * params);
 int send_line(T_SOC soc, const char *data);
 int get_ftp_line(T_SOC soc, char *line, size_t line_size, int timeout);
-/* Split a "user[:pass]@" prefix (end = jump_identification result) into
-   bounded, NUL-terminated user/pass buffers, truncating to fit. */
-void ftp_split_userpass(const char *src, const char *end, char *user,
-                        size_t user_size, char *pass, size_t pass_size);
 T_SOC get_datasocket(char *to_send, size_t to_send_size);
 int stop_ftp(lien_back * back);
 char *linejmp(char *line);

src/htsjava.c

@@ -63,9 +63,6 @@ Please visit our Website: http://www.httrack.com
 /* This file */
 #include "htsjava.h"
 
-/* calloct/freet wrappers */
-#include "htssafe.h"
-
 static int reverse_endian(void) {
   int endian = 1;
 
@@ -207,16 +204,7 @@ static int hts_parse_java(t_hts_callbackarg * carg, httrackp * opt,
         return 0;
       }
 
-      /* A constant-pool entry is >= 1 byte on disk; reject a count exceeding
-         the file size (hostile .class ~68 MB alloc DoS). */
-      if (!hts_count_fits(header.count, (LLint) fsize(file))) {
-        fclose(fpout);
-        sprintf(str->err_msg,
-                "Invalid constant pool count %u (file len " LLintP ")",
-                (unsigned) header.count, (LLint) fsize(file));
-        return 0;
-      }
-      tab = (RESP_STRUCT *) calloct(header.count, sizeof(RESP_STRUCT));
+      tab = (RESP_STRUCT *) calloc(header.count, sizeof(RESP_STRUCT));
       if (!tab) {
         sprintf(str->err_msg, "Unable to alloc %d bytes",
                 (int) sizeof(RESP_STRUCT));
@@ -242,7 +230,7 @@ static int hts_parse_java(t_hts_callbackarg * carg, httrackp * opt,
           } else {              // ++ une erreur est survenue!
             if (strnotempty(str->err_msg) == 0)
               strcpy(str->err_msg, "Internal readtable error");
-            freet(tab);
+            free(tab);
             if (fpout) {
               fclose(fpout);
               fpout = NULL;
@@ -300,7 +288,7 @@ static int hts_parse_java(t_hts_callbackarg * carg, httrackp * opt,
 #if JAVADEBUG
       printf("end\n");
 #endif
-      freet(tab);
+      free(tab);
       if (fpout) {
         fclose(fpout);
         fpout = NULL;

src/htsparse.c

@@ -106,99 +106,159 @@ Please visit our Website: http://www.httrack.com
 // does nothing
 #define XH_uninit do {} while(0)
 
-#define HT_ADD_FOP
-
-/* Mutable extended-context fields: one source of truth so the DEFINE/SET/SAVE
-   load and store lists can't drift apart. */
+/* clang-format off: an edit realigns all backslashes, churning the macro. */
 /* clang-format off */
-#define ENGINE_MUTABLE_FIELDS(X) \
-  X(int, error, stre->error_) \
-  X(int, store_errpage, stre->store_errpage_) \
-  X(int, makeindex_done, stre->makeindex_done_) \
-  X(FILE *, makeindex_fp, stre->makeindex_fp_) \
-  X(int, makeindex_links, stre->makeindex_links_) \
-  X(LLint, stat_fragment, stre->stat_fragment_)
-
-#define ENGINE_FIELD_DECLARE(type, name, src) type name = *(src);
-#define ENGINE_FIELD_LOAD(type, name, src) name = *(src);
-#define ENGINE_FIELD_STORE(type, name, src) *(src) = name;
+#define HT_ADD_END { \
+  int ok=0;\
+  if (TypedArraySize(output_buffer) != 0) { \
+    const size_t ht_len = TypedArraySize(output_buffer); \
+    const char *const ht_buff = TypedArrayElts(output_buffer); \
+    char digest[32+2];\
+    off_t fsize_old = fsize(fconv(OPT_GET_BUFF(opt),OPT_GET_BUFF_SIZE(opt),savename()));\
+    digest[0] = '\0';\
+    domd5mem(TypedArrayElts(output_buffer), ht_len, digest, 1);\
+    if (fsize_old == (off_t) ht_len) { \
+      int mlen = 0;\
+      char* mbuff;\
+      cache_readdata(cache,"//[HTML-MD5]//",savename(),&mbuff,&mlen);\
+      if (mlen) \
+        mbuff[mlen]='\0';\
+      if ((mlen == 32) && (strcmp(((mbuff!=NULL)?mbuff:""),digest)==0)) {\
+        ok=1;\
+        hts_log_print(opt, LOG_DEBUG, "File not re-written (md5): %s",savename());\
+      } else {\
+        ok=0;\
+      } \
+      freet(mbuff);\
+    }\
+    if (!ok) { \
+      file_notify(opt,urladr(), urlfil(), savename(), 1, 1, r->notmodified); \
+      fp=filecreate(&opt->state.strc, savename()); \
+      if (fp) { \
+        if (ht_len>0) {\
+        if (fwrite(ht_buff,1,ht_len,fp) != ht_len) { \
+          int fcheck;\
+          if ((fcheck=check_fatal_io_errno())) {\
+            opt->state.exit_xh=-1;\
+          }\
+          if (opt->log) {   \
+            hts_log_print(opt, LOG_ERROR | LOG_ERRNO, "Unable to write HTML file %s", savename());\
+            if (fcheck) {\
+              hts_log_print(opt, LOG_ERROR, "* * Fatal write error, giving up");\
+            }\
+          }\
+        }\
+        }\
+        fclose(fp); fp=NULL; \
+        if (strnotempty(r->lastmodified)) \
+        set_filetime_rfc822(savename(),r->lastmodified); \
+      } else {\
+        int fcheck;\
+        if ((fcheck=check_fatal_io_errno())) {\
+  				hts_log_print(opt, LOG_ERROR, "Mirror aborted: disk full or filesystem problems"); \
+          opt->state.exit_xh=-1;\
+        }\
+        hts_log_print(opt, LOG_ERROR | LOG_ERRNO, "Unable to save file %s", savename());\
+        if (fcheck) {\
+          hts_log_print(opt, LOG_ERROR, "* * Fatal write error, giving up");\
+        }\
+      }\
+    } else {\
+      file_notify(opt,urladr(), urlfil(), savename(), 0, 0, r->notmodified); \
+      filenote(&opt->state.strc, savename(),NULL); \
+    }\
+    if (cache->ndx)\
+      cache_writedata(cache->ndx,cache->dat,"//[HTML-MD5]//",savename(),digest,(int)strlen(digest));\
+  } \
+  TypedArrayFree(output_buffer); \
+}
+/* clang-format on */
+#define HT_ADD_FOP
 
 #define ENGINE_DEFINE_CONTEXT() \
   ENGINE_DEFINE_CONTEXT_BASE(); \
+  /* */ \
   htsblk* const r HTS_UNUSED = stre->r_; \
   hash_struct* const hash HTS_UNUSED = stre->hash_; \
   char* const codebase HTS_UNUSED = stre->codebase; \
   char* const base HTS_UNUSED = stre->base; \
+  /* */ \
   const char * const template_header HTS_UNUSED = stre->template_header_; \
   const char * const template_body HTS_UNUSED = stre->template_body_; \
   const char * const template_footer HTS_UNUSED = stre->template_footer_; \
+  /* */ \
   HTS_UNUSED char* const makeindex_firstlink = stre->makeindex_firstlink_; \
-  ENGINE_MUTABLE_FIELDS(ENGINE_FIELD_DECLARE) \
-  /* load-once (kept out of SET/SAVE): re-reading would reset the throttle */ \
+  /* */ \
+  /* */ \
+  int error = * stre->error_; \
+  int store_errpage = * stre->store_errpage_; \
+  /* */ \
+  int makeindex_done = *stre->makeindex_done_; \
+  FILE* makeindex_fp = *stre->makeindex_fp_; \
+  int makeindex_links = *stre->makeindex_links_; \
+  /* */ \
+  LLint stat_fragment = *stre->stat_fragment_; \
   HTS_UNUSED TStamp makestat_time = stre->makestat_time; \
   HTS_UNUSED FILE* makestat_fp = stre->makestat_fp
 
+/* clang-format off: an edit realigns all backslashes, churning the macro. */
+/* clang-format off */
+/* Load-once: re-reading resets makestat_time (mutated locally, never SAVEd). */
 #define ENGINE_SET_CONTEXT() \
   ENGINE_SET_CONTEXT_BASE(); \
-  ENGINE_MUTABLE_FIELDS(ENGINE_FIELD_LOAD)
+  /* */ \
+  error = * stre->error_; \
+  store_errpage = * stre->store_errpage_; \
+  /* */ \
+  makeindex_done = *stre->makeindex_done_; \
+  makeindex_fp = *stre->makeindex_fp_; \
+  makeindex_links = *stre->makeindex_links_; \
+  /* */ \
+  stat_fragment = *stre->stat_fragment_
+/* clang-format on */
 
 #define ENGINE_LOAD_CONTEXT() \
   ENGINE_DEFINE_CONTEXT()
 
 #define ENGINE_SAVE_CONTEXT() \
   ENGINE_SAVE_CONTEXT_BASE(); \
-  ENGINE_MUTABLE_FIELDS(ENGINE_FIELD_STORE)
-/* clang-format on */
+  /* */ \
+  * stre->error_ = error; \
+  * stre->store_errpage_ = store_errpage; \
+  /* */ \
+  *stre->makeindex_done_ = makeindex_done; \
+  *stre->makeindex_fp_ = makeindex_fp; \
+  *stre->makeindex_links_ = makeindex_links; \
+  /* */ \
+  *stre->stat_fragment_ = stat_fragment
 
+#define _FILTERS     (*opt->filters.filters)
+#define _FILTERS_PTR (opt->filters.filptr)
 #define _ROBOTS      ((robots_wizard*)opt->robotsptr)
 
-/* JS-detection automaton states; INSCRIPT_DEFAULT is the synthetic "any other
-   char" column of the transition table. */
-typedef enum {
-  INSCRIPT_START = 0,
-  INSCRIPT_ANTISLASH,
-  INSCRIPT_INQUOTE,
-  INSCRIPT_INQUOTE2,
-  INSCRIPT_SLASH,
-  INSCRIPT_SLASHSLASH,
-  INSCRIPT_COMMENT,
-  INSCRIPT_COMMENT2,
-  INSCRIPT_ANTISLASH_IN_QUOTE,
-  INSCRIPT_ANTISLASH_IN_QUOTE2,
-  INSCRIPT_DEFAULT = 256
-} INSCRIPT;
+/* Apply current *adr character for the script automate */
+#define AUTOMATE_LOOKUP_CURRENT_ADR() do { \
+  if (inscript) { \
+  int new_state_pos; \
+  new_state_pos=inscript_state[inscript_state_pos][(unsigned char)*html]; \
+  if (new_state_pos < 0) { \
+  new_state_pos=inscript_state[inscript_state_pos][INSCRIPT_DEFAULT]; \
+  } \
+  assertf(new_state_pos >= 0); \
+  assertf(new_state_pos*sizeof(inscript_state[0]) < sizeof(inscript_state)); \
+  inscript_state_pos=new_state_pos; \
+  } \
+} while(0)
 
-#define INSCRIPT_NSTATES 10 /* rows in the transition table */
-
-/* Live view of the parser's automaton locals, set up once so the helpers below
-   can drive it without capturing them by lexical scope. */
-typedef struct {
-  const int *inscript;             /* nonzero while inside a script body */
-  const signed char (*table)[257]; /* [INSCRIPT_NSTATES][257] transitions */
-  INSCRIPT *pos;                   /* current state */
-  const char **html;               /* parse cursor */
-} script_automate;
-
-/* Feed the current *html byte to the automaton. No-op outside a script body. */
-static void hts_automate_lookup(const script_automate *aut) {
-  if (*aut->inscript) {
-    int next = aut->table[*aut->pos][(unsigned char) **aut->html];
-    if (next < 0) {
-      next = aut->table[*aut->pos][INSCRIPT_DEFAULT];
-    }
-    assertf(next >= 0 && next < INSCRIPT_NSTATES);
-    *aut->pos = (INSCRIPT) next;
-  }
-}
-
-/* Advance the cursor by 'steps' bytes, feeding each to the automaton. */
-static void hts_automate_increment(const script_automate *aut, int steps) {
-  while (steps > 0) {
-    (*aut->html)++;
-    hts_automate_lookup(aut);
-    steps--;
-  }
-}
+/* Increment current pointer to 'steps' characters, modifying automate if necessary */
+#define INCREMENT_CURRENT_ADR(steps) do { \
+  int steps__ = (int) ( steps ); \
+  while(steps__ > 0) { \
+  html++; \
+  AUTOMATE_LOOKUP_CURRENT_ADR(); \
+  steps__ --; \
+  } \
+} while(0)
 
 /* Percent-encode the angle brackets of a string so it is safe to embed inside
    an HTML comment (the default footer) or any other HTML context. A URL holding
@@ -343,7 +403,20 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
       int incomment = 0;        // dans un <!--
       int inscript = 0;         // dans un scipt pour applets javascript)
       int inscript_locked = 0;  // in locked script (ie. js file)
-      signed char inscript_state[INSCRIPT_NSTATES][257];
+      signed char inscript_state[10][257];
+      typedef enum {
+        INSCRIPT_START = 0,
+        INSCRIPT_ANTISLASH,
+        INSCRIPT_INQUOTE,
+        INSCRIPT_INQUOTE2,
+        INSCRIPT_SLASH,
+        INSCRIPT_SLASHSLASH,
+        INSCRIPT_COMMENT,
+        INSCRIPT_COMMENT2,
+        INSCRIPT_ANTISLASH_IN_QUOTE,
+        INSCRIPT_ANTISLASH_IN_QUOTE2,
+        INSCRIPT_DEFAULT = 256
+      } INSCRIPT;
       INSCRIPT inscript_state_pos = INSCRIPT_START;
       const char *inscript_name = NULL;       // script tag name
       int inscript_tag = 0;     // on est dans un <body onLoad="... terminé par >
@@ -404,8 +477,6 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
       inscript_state[INSCRIPT_COMMENT2]['*'] = INSCRIPT_COMMENT2;
       inscript_state[INSCRIPT_ANTISLASH_IN_QUOTE][INSCRIPT_DEFAULT] = INSCRIPT_INQUOTE; /* #8: escape in '' */
       inscript_state[INSCRIPT_ANTISLASH_IN_QUOTE2][INSCRIPT_DEFAULT] = INSCRIPT_INQUOTE2;       /* #9: escape in "" */
-      const script_automate saut = {&inscript, inscript_state,
-                                    &inscript_state_pos, &html};
 
       /* Primary list or URLs */
       if (ptr == 0) {
@@ -844,7 +915,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
             }
 
             /* automate */
-            hts_automate_lookup(&saut);
+            AUTOMATE_LOOKUP_CURRENT_ADR();
 
             // Note:
             // Certaines pages ne respectent pas le html
@@ -1760,7 +1831,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
             }
             // sauter espaces
             // adr+=p;
-            hts_automate_increment(&saut, p);
+            INCREMENT_CURRENT_ADR(p);
             while((is_space(*html)
                    || (inscriptgen && html[0] == '\\' && is_space(html[1])
                    )
@@ -1775,7 +1846,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
                 }
               // puis quitter
               // html++;    // sauter les espaces, "" et cie
-              hts_automate_increment(&saut, 1);
+              INCREMENT_CURRENT_ADR(1);
             }
 
             /* Stop at \n (LF) if primary links or link lists */
@@ -1790,7 +1861,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
               if (*html == '\\') {
                 if ((*(html + 1) == '\'') || (*(html + 1) == '"')) {      // \" ou \'
                   // html+=2;    // sauter
-                  hts_automate_increment(&saut, 2);
+                  INCREMENT_CURRENT_ADR(2);
                 }
               }
             }
@@ -1838,7 +1909,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
             if (srcset_p) {
               while(html < r->adr + r->size
                     && (is_realspace(*html) || *html == ','))
-                hts_automate_increment(&saut, 1);
+                INCREMENT_CURRENT_ADR(1);
             }
             eadr = html;
 
@@ -3298,7 +3369,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
 
             assertf(eadr - html >= 0);   // Should not go back
             if (eadr > html) {
-              hts_automate_increment(&saut, (int) (eadr - 1 - html));
+              INCREMENT_CURRENT_ADR(eadr - 1 - html);
             }
             // adr=eadr-1;  // ** sauter
 
@@ -3317,8 +3388,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
                   q++;          // skip whitespace and empty candidates
                 if (q < endp && *q != '\0' && *q != ',' && *q != quote
                     && *q != '<' && *q != '>' && (unsigned char) *q >= 32) {
-                  hts_automate_increment(
-                      &saut, (int) (q - html)); // keep the automate in sync
+                  INCREMENT_CURRENT_ADR(q - html);   // keep the automate in sync
                   ok = 1;
                   goto srcset_next;
                 }
@@ -3458,12 +3528,7 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
         }
 
         /* Flush and save to disk */
-        if (TypedArraySize(output_buffer) != 0) {
-          hts_finish_html_file(
-              opt, cache, r, &fp, TypedArrayElts(output_buffer),
-              TypedArraySize(output_buffer), urladr(), urlfil(), savename());
-        }
-        TypedArrayFree(output_buffer);
+        HT_ADD_END;             // achever
       }
       //
       //
@@ -3488,24 +3553,6 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
   return 0;
 }
 
-/* Mirror the savename to tell whether a redirect saves to the same file (#159);
- * contract in htsparse.h. */
-hts_boolean hts_redirect_same_savefile(httrackp *opt, const char *cur_adr,
-                                       const char *cur_fil,
-                                       const char *moved_adr,
-                                       const char *moved_fil) {
-  const int norm_slash = opt->urlhack && !opt->no_slash_dedup;
-  const int norm_query = opt->urlhack && !opt->no_query_dedup;
-  char BIGSTK n_fil[HTS_URLMAXSIZE * 2], pn_fil[HTS_URLMAXSIZE * 2];
-
-  if (strcasecmp(jump_identification_const(moved_adr),
-                 jump_identification_const(cur_adr)) != 0)
-    return HTS_FALSE;
-  fil_normalized_filtered_ex(moved_fil, n_fil, NULL, norm_slash, norm_query);
-  fil_normalized_filtered_ex(cur_fil, pn_fil, NULL, norm_slash, norm_query);
-  return strcasecmp(n_fil, pn_fil) == 0;
-}
-
 /*
 Check 301, 302, .. statuscodes (moved)
 */
@@ -3551,9 +3598,36 @@ int hts_mirror_check_moved(htsmoduleStruct * str,
         if ((reponse =
              ident_url_relatif(mov_url, urladr(), urlfil(), moved)) >= 0) {
           int set_prio_to = 0;  // pas de priotité fixéd par wizard
-          // A same-file alias redirect must be followed, not stubbed (#159).
-          const hts_boolean same_savefile = hts_redirect_same_savefile(
-              opt, urladr(), urlfil(), moved->adr, moved->fil);
+
+          // check whether URLHack is harmless or not (per the effective
+          // sub-flags)
+          if (opt->urlhack && (!opt->no_www_dedup || !opt->no_slash_dedup ||
+                               !opt->no_query_dedup)) {
+            const int norm_host = !opt->no_www_dedup;
+            const int norm_slash = !opt->no_slash_dedup;
+            const int norm_query = !opt->no_query_dedup;
+            char BIGSTK n_adr[HTS_URLMAXSIZE * 2], n_fil[HTS_URLMAXSIZE * 2];
+            char BIGSTK pn_adr[HTS_URLMAXSIZE * 2], pn_fil[HTS_URLMAXSIZE * 2];
+
+            strlcpybuff(n_adr,
+                        norm_host ? jump_normalized_const(moved->adr)
+                                  : jump_identification_const(moved->adr),
+                        sizeof(n_adr));
+            strlcpybuff(pn_adr,
+                        norm_host ? jump_normalized_const(urladr())
+                                  : jump_identification_const(urladr()),
+                        sizeof(pn_adr));
+            fil_normalized_filtered_ex(moved->fil, n_fil, NULL, norm_slash,
+                                       norm_query);
+            fil_normalized_filtered_ex(urlfil(), pn_fil, NULL, norm_slash,
+                                       norm_query);
+            if (strcasecmp(n_adr, pn_adr) == 0
+                && strcasecmp(n_fil, pn_fil) == 0) {
+              hts_log_print(opt, LOG_WARNING,
+                            "Redirected link is identical because of 'URL Hack' option: %s%s and %s%s",
+                            urladr(), urlfil(), moved->adr, moved->fil);
+            }
+          }
           //if (ident_url_absolute(mov_url,moved->adr,moved->fil)!=-1) {    // ok URL reconnue
           // c'est (en gros) la même URL..
           // si c'est un problème de casse dans le host c'est que le serveur est buggé
@@ -3581,17 +3655,7 @@ int hts_mirror_check_moved(htsmoduleStruct * str,
                 hts_log_print(opt, LOG_DEBUG, "moved link accepted: %s%s",
                               moved->adr, moved->fil);
               }
-            } else if (same_savefile) {
-              // A stub would point at itself; follow the redirect instead.
-              if (hts_acceptlink(opt, ptr, moved->adr, moved->fil, NULL, NULL,
-                                 &set_prio_to, NULL) != 1) {
-                get_it = 1;
-                hts_log_print(opt, LOG_WARNING,
-                              "Redirect to a same-file alias, fetching real "
-                              "content: %s%s -> %s%s",
-                              urladr(), urlfil(), moved->adr, moved->fil);
-              }
-            } /* sinon traité normalement */
+            }                   /* sinon traité normalement */
           }
 
           //if ((strfield2(moved->adr,urladr())!=0) && (strfield2(moved->fil,urlfil())!=0)) {  // identique à casse près
@@ -3614,11 +3678,7 @@ int hts_mirror_check_moved(htsmoduleStruct * str,
                    heap(heap(ptr)->precedent)->adr,
                    heap(heap(ptr)->precedent)->fil, opt,
                    sback, cache, hash, ptr, numero_passe, NULL) != -1) {
-                // Same-file alias: the reserved name is the invalidated source,
-                // so record anyway.
-                if (same_savefile ||
-                    hash_read(hash, savedmoved.save, NULL,
-                              HASH_STRUCT_FILENAME) < 0) { // n'existe pas déja
+                if (hash_read(hash, savedmoved.save, NULL, HASH_STRUCT_FILENAME) < 0) {   // n'existe pas déja
                   // enregistrer lien avec SAV IDENTIQUE
                   if (hts_record_link(opt, moved->adr, moved->fil, heap(ptr)->sav, "", "", NULL)) {
                     // mode test?
@@ -3642,6 +3702,7 @@ int hts_mirror_check_moved(htsmoduleStruct * str,
                                 "moving %s to an existing file %s",
                                 heap(ptr)->fil, urlfil());
                 }
+
               }
             }
 

src/htsparse.h

@@ -116,19 +116,6 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre);
 int hts_mirror_check_moved(htsmoduleStruct * str,
                            htsmoduleStructExtended * stre);
 
-/*
-  Non-zero if a redirect (cur_adr,cur_fil)->(moved_adr,moved_fil) saves to the
-  same local file, so it must be followed rather than turned into a
-  self-pointing "moved" stub (#159). Mirrors the savename: scheme+userinfo
-  stripped, www kept (www dedup is the crawl layer's job), path
-  slash/query-normalized per the URL-hack flags. Not hash_url_equals: that keys
-  on the dedup hash, which folds www and never collapses http<->https.
-*/
-hts_boolean hts_redirect_same_savefile(httrackp *opt, const char *cur_adr,
-                                       const char *cur_fil,
-                                       const char *moved_adr,
-                                       const char *moved_fil);
-
 /*
   Process user intercations: pause, add link, delete link..
 */

src/htssafe.h

@@ -456,13 +456,6 @@ static HTS_INLINE HTS_UNUSED const char *htsbuff_str(const htsbuff *b) {
   return b->buf;
 }
 
-/** True if 'count' records of >= 1 byte each fit in 'available' bytes; guards
-    an attacker-controlled count driving a large allocation. */
-static HTS_INLINE HTS_UNUSED hts_boolean hts_count_fits(size_t count,
-                                                        LLint available) {
-  return (available >= 0 && (LLint) count <= available) ? HTS_TRUE : HTS_FALSE;
-}
-
 /* Thin aliases over the libc allocator/memcpy (historical "t" suffix); no
    added bounds checking. freet() also NULLs the freed pointer and tolerates
    NULL. memcpybuff() despite the name is a raw memcpy: the caller owns the

src/htsselftest.c

@@ -45,12 +45,10 @@ Please visit our Website: http://www.httrack.com
 #include "htscore.h"
 #include "htsdefines.h"
 #include "htslib.h"
-#include "htsparse.h"
 #include "htscache_selftest.h"
 #include "htsdns_selftest.h"
 #include "htscharset.h"
 #include "htsencoding.h"
-#include "htsftp.h"
 #include "htsmd5.h"
 #if HTS_USEZLIB
 #include "htszlib.h"
@@ -62,10 +60,6 @@ Please visit our Website: http://www.httrack.com
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
-#ifndef _WIN32
-#include <sys/socket.h>
-#include <unistd.h>
-#endif
 
 /* very minimalistic internal tests */
 static void basic_selftests(void) {
@@ -1346,37 +1340,6 @@ static int st_urlhack(httrackp *opt, int argc, char **argv) {
   return 0;
 }
 
-/* #159: hts_redirect_same_savefile decides whether a redirect is a same-file
- * alias. */
-static int st_redirect_samefile(httrackp *opt, int argc, char **argv) {
-  (void) argc;
-  (void) argv;
-#define SAME(aa, fa, ab, fb) hts_redirect_same_savefile(opt, aa, fa, ab, fb)
-  /* scheme and userinfo collapse (the #159 case); a different path does not */
-  assertf(SAME("http://foo.com", "/a/b", "https://foo.com", "/a/b"));
-  assertf(SAME("http://user@foo.com", "/a", "http://foo.com", "/a"));
-  assertf(!SAME("http://foo.com", "/a", "http://foo.com", "/b"));
-  /* www stays distinct here; the crawl's dedup layer folds www, not this helper
-   */
-  opt->urlhack = HTS_TRUE;
-  opt->no_www_dedup = opt->no_slash_dedup = opt->no_query_dedup = HTS_FALSE;
-  assertf(!SAME("http://www.foo.com", "/a", "http://foo.com", "/a"));
-  /* slash/query fold only when the dedup flag is on */
-  assertf(SAME("https://foo.com", "/a//b", "http://foo.com", "/a/b"));
-  assertf(
-      SAME("https://foo.com", "/p?b=2&a=1", "http://foo.com", "/p?a=1&b=2"));
-  opt->no_slash_dedup = opt->no_query_dedup = HTS_TRUE;
-  assertf(!SAME("https://foo.com", "/a//b", "http://foo.com", "/a/b"));
-  assertf(
-      !SAME("https://foo.com", "/p?b=2&a=1", "http://foo.com", "/p?a=1&b=2"));
-  /* but a pure scheme alias still collapses regardless of dedup opt-outs */
-  assertf(SAME("http://foo.com", "/a/b", "https://foo.com", "/a/b"));
-  opt->no_slash_dedup = opt->no_query_dedup = HTS_FALSE;
-#undef SAME
-  printf("redirect-samefile self-test OK\n");
-  return 0;
-}
-
 // hts_finish_makeindex writes the footer, emits the refresh meta only when
 // makeindex_links==1, and clears *fp / sets *done. argv[0] is a writable dir.
 static int st_makeindex(httrackp *opt, int argc, char **argv) {
@@ -1774,75 +1737,6 @@ static int st_robots(httrackp *opt, int argc, char **argv) {
   return 0;
 }
 
-/* get_ftp_line must bound a hostile, CRLF-less reply into its internal
-   1024-byte buffer; ASan turns the pre-fix overflow into an abort here. */
-#ifndef _WIN32
-static int st_ftpline(httrackp *opt, int argc, char **argv) {
-  int sv[2];
-  char line[2048];
-  char flood[4096];
-
-  (void) opt;
-  (void) argc;
-  (void) argv;
-  memset(flood, 'x', sizeof(flood));
-  assertf(socketpair(AF_UNIX, SOCK_STREAM, 0, sv) == 0);
-  assertf(write(sv[1], "220 ", 4) == 4); // valid 3-digit code
-  assertf(write(sv[1], flood, sizeof(flood)) == (ssize_t) sizeof(flood));
-  assertf(write(sv[1], "\r\n", 2) == 2); // end the line so we return
-  close(sv[1]);
-  line[0] = '\0';
-  get_ftp_line(sv[0], line, sizeof(line), 5);
-  close(sv[0]);
-  printf("ftp-line self-test OK (bounded %d-byte reply)\n",
-         (int) sizeof(flood));
-  return 0;
-}
-#endif
-
-/* ftp_split_userpass: well-formed split, plus a hostile over-long userinfo
-   that pre-fix overran user[256]/pass[256]. */
-static int st_ftpuser(httrackp *opt, int argc, char **argv) {
-  char user[256], pass[256];
-  char in[1200];
-
-  (void) opt;
-  (void) argc;
-  (void) argv;
-  {
-    const char ok[] = "bob:secret@host/f"; // '@' at index 10
-
-    ftp_split_userpass(ok, ok + 11, user, sizeof(user), pass, sizeof(pass));
-    assertf(strcmp(user, "bob") == 0);
-    assertf(strcmp(pass, "secret") == 0);
-  }
-  memset(in, 'u', 400);
-  in[400] = ':';
-  memset(in + 401, 'p', 400);
-  in[801] = '@';
-  in[802] = '\0';
-  ftp_split_userpass(in, in + 802, user, sizeof(user), pass, sizeof(pass));
-  assertf(strlen(user) == sizeof(user) - 1);
-  assertf(strlen(pass) == sizeof(pass) - 1);
-  printf("ftp-userpass self-test OK\n");
-  return 0;
-}
-
-/* hts_count_fits caps the .class constant-pool entry count to the file size,
-   rejecting the ~68 MB-per-file calloc DoS. */
-static int st_java(httrackp *opt, int argc, char **argv) {
-  (void) opt;
-  (void) argc;
-  (void) argv;
-  assertf(hts_count_fits(10, 1000) == HTS_TRUE);
-  assertf(hts_count_fits(0, 10) == HTS_TRUE);
-  assertf(hts_count_fits(65535, 10) == HTS_FALSE);
-  assertf(hts_count_fits(1, 0) == HTS_FALSE);
-  assertf(hts_count_fits(1, -1) == HTS_FALSE);
-  printf("java constant-pool cap self-test OK\n");
-  return 0;
-}
-
 /* ------------------------------------------------------------ */
 /* Registry: name -> handler, with a usage hint and a one-line description. */
 /* ------------------------------------------------------------ */
@@ -1863,8 +1757,6 @@ static const struct selftest_entry {
      st_stripquery},
     {"urlhack", "", "-%u url-hack sub-flag (www/slash/query) self-test",
      st_urlhack},
-    {"redirect-samefile", "", "same-file redirect detection self-test (#159)",
-     st_redirect_samefile},
     {"mime", "<filename>", "MIME type for a filename", st_mime},
     {"charset", "<charset> <string>",
      "convert a string to UTF-8 from a charset", st_charset},
@@ -1903,12 +1795,6 @@ static const struct selftest_entry {
      "Accept-Encoding advertises gzip+deflate, both decode", st_acceptencoding},
     {"robots", "", "robots.txt RFC 9309 Allow/Disallow precedence self-test",
      st_robots},
-#ifndef _WIN32
-    {"ftp-line", "", "get_ftp_line bounds a hostile FTP reply line",
-     st_ftpline},
-#endif
-    {"ftp-userpass", "", "ftp_split_userpass bounds URL userinfo", st_ftpuser},
-    {"java", "", "java .class constant-pool count cap self-test", st_java},
 };
 
 static void list_selftests(void) {

tests/01_engine-ftp-line.test

@@ -1,7 +0,0 @@
-#!/bin/bash
-#
-
-set -euo pipefail
-
-# get_ftp_line bounds a hostile CRLF-less FTP reply into its 1024-byte buffer.
-httrack -O /dev/null -#test=ftp-line run | grep -q "ftp-line self-test OK"

tests/01_engine-ftp-userpass.test

@@ -1,7 +0,0 @@
-#!/bin/bash
-#
-
-set -euo pipefail
-
-# ftp_split_userpass bounds an over-long user:pass@ from a hostile ftp:// URL.
-httrack -O /dev/null -#test=ftp-userpass run | grep -q "ftp-userpass self-test OK"

tests/01_engine-java.test

@@ -1,7 +0,0 @@
-#!/bin/bash
-#
-
-set -euo pipefail
-
-# .class constant-pool count is capped to the file size (calloc DoS).
-httrack -O /dev/null -#test=java run | grep -q "java constant-pool cap self-test OK"

tests/01_engine-redirect.test

@@ -1,9 +0,0 @@
-#!/bin/bash
-#
-
-set -euo pipefail
-
-# #159: a redirect to a same-file alias (http<->https, user@host, ..) must be
-# followed through, not turned into a self-pointing "moved" stub. The decision
-# helper is exercised by the engine self-test.
-httrack -O /dev/null -#test=redirect-samefile run | grep -q "redirect-samefile self-test OK"

tests/30_local-fragment-link.test

@@ -1,13 +0,0 @@
-#!/bin/bash
-# Issue #279: an anchored link (target.html#sec, quoted or bare) fetches the
-# target with the fragment dropped (strict server 400s on a '#' in the request)
-# but keeps it in the rewritten local link so the anchor still works.
-set -e
-
-: "${top_srcdir:=..}"
-
-bash "$top_srcdir/tests/local-crawl.sh" --errors 0 \
-    --found 'fraglink/target.html' \
-    --file-matches 'fraglink/index.html' 'href=target\.html#sec' \
-    --file-matches 'fraglink/index.html' 'href="target\.html#sec2"' \
-    httrack 'BASEURL/fraglink/index.html'

tests/Makefile.am

@@ -6,7 +6,6 @@ EXTRA_DIST = $(TESTS) crawl-test.sh run-all-tests.sh check-network.sh \
 	local-crawl.sh local-server.py server.crt server.key \
 	server-root/simple/basic.html server-root/simple/link.html \
 	server-root/stripquery/index.html server-root/stripquery/a.html \
-	server-root/fraglink/index.html server-root/fraglink/target.html \
 	fixtures/cache-golden/hts-cache/new.zip
 
 TESTS_ENVIRONMENT =
@@ -35,19 +34,15 @@ TESTS = \
 	01_engine-entities.test \
 	01_engine-filelist.test \
 	01_engine-filter.test \
-	01_engine-ftp-line.test \
-	01_engine-ftp-userpass.test \
 	01_engine-hashtable.test \
 	01_engine-idna.test \
 	01_engine-escape-room.test \
 	01_engine-inplace-escape.test \
-	01_engine-java.test \
 	01_engine-makeindex.test \
 	01_engine-mime.test \
 	01_engine-parse.test \
 	01_engine-pause.test \
 	01_engine-rcfile.test \
-	01_engine-redirect.test \
 	01_engine-relative.test \
 	01_engine-robots.test \
 	01_engine-savename.test \
@@ -88,7 +83,6 @@ TESTS = \
 	26_local-strip-query.test \
 	27_local-cookies-file.test \
 	28_local-pause.test \
-	29_local-redirect-fragment.test \
-	30_local-fragment-link.test
+	29_local-redirect-fragment.test
 
 CLEANFILES = check-network_sh.cache

tests/local-crawl.sh

@@ -15,11 +15,8 @@
 #   bash local-crawl.sh [--tls] [--root DIR] [--cookie NAME=VALUE ...] \
 #       --errors N --files N --found PATH ... --directory PATH ... \
 #       --log-found REGEX ... --log-not-found REGEX ... \
-#       --file-matches PATH REGEX ... --file-not-matches PATH REGEX ... \
 #       httrack BASEURL/some/path [httrack-args...]
 # --log-found/--log-not-found grep (ERE) the crawl's hts-log.txt.
-# --file-matches/--file-not-matches grep (ERE) a mirrored file (PATH under the
-# host root), to assert rewritten link/content survived the crawl.
 # --cookie writes a Netscape cookies.txt (scoped to the discovered host:port,
 # which the ephemeral port forces into the cookie domain) and passes it to
 # httrack via --cookies-file, to exercise preloaded cookies.
@@ -124,10 +121,6 @@ while test "$pos" -lt "$nargs"; do
         audit+=("${args[$pos]}" "${args[$((pos + 1))]}")
         pos=$((pos + 1))
         ;;
-    --file-matches | --file-not-matches)
-        audit+=("${args[$pos]}" "${args[$((pos + 1))]}" "${args[$((pos + 2))]}")
-        pos=$((pos + 2))
-        ;;
     httrack)
         pos=$((pos + 1))
         break
@@ -301,24 +294,6 @@ while test "$i" -lt "${#audit[@]}"; do
             exit 1
         else result "OK"; fi
         ;;
-    --file-matches)
-        path="${audit[$((i + 1))]}"
-        i=$((i + 2))
-        info "checking ${path} matches ${audit[$i]}"
-        if grep -aqE "${audit[$i]}" "${hostroot}/${path}"; then result "OK"; else
-            result "no match"
-            exit 1
-        fi
-        ;;
-    --file-not-matches)
-        path="${audit[$((i + 1))]}"
-        i=$((i + 2))
-        info "checking ${path} lacks ${audit[$i]}"
-        if grep -aqE "${audit[$i]}" "${hostroot}/${path}"; then
-            result "matched"
-            exit 1
-        else result "OK"; fi
-        ;;
     esac
     i=$((i + 1))
 done

tests/server-root/fraglink/index.html

@@ -1,4 +0,0 @@
-<html><body>
-<a href=target.html#sec>unquoted fragment link</a>
-<a href="target.html#sec2">quoted fragment link</a>
-</body></html>

tests/server-root/fraglink/target.html

@@ -1 +0,0 @@
-<html><body><a name="sec"></a><a name="sec2"></a>target</body></html>