htsparse: extract HT_ADD_END into hts_finish_html_file, drop dead macros

The final HTML-flush step lived in the ~60-line HT_ADD_END macro, captured by
lexical scope in htsparse(). Move it to a real function in htscore.c (next to
its sibling hts_finish_makeindex), passing the output buffer as a plain
pointer+length since the TypedArray is an anonymous struct type that can't cross
a function boundary; the caller keeps the size guard and the trailing
TypedArrayFree. Behavior-preserving: a two-crawl mirror produces byte-identical
output against master.

Also delete the _FILTERS/_FILTERS_PTR macros, dead in htsparse.c (htscore.c and
htswizard.c keep their own local copies); _ROBOTS stays.

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

src/htscore.c

@@ -441,6 +441,72 @@ 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

@@ -370,6 +370,12 @@ 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/htsparse.c

@@ -106,73 +106,6 @@ Please visit our Website: http://www.httrack.com
 // does nothing
 #define XH_uninit do {} while(0)
 
-/* clang-format off: an edit realigns all backslashes, churning the macro. */
-/* clang-format off */
-#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() \
@@ -232,8 +165,6 @@ Please visit our Website: http://www.httrack.com
   /* */ \
   *stre->stat_fragment_ = stat_fragment
 
-#define _FILTERS     (*opt->filters.filters)
-#define _FILTERS_PTR (opt->filters.filptr)
 #define _ROBOTS      ((robots_wizard*)opt->robotsptr)
 
 /* Apply current *adr character for the script automate */
@@ -3528,7 +3459,12 @@ int htsparse(htsmoduleStruct * str, htsmoduleStructExtended * stre) {
         }
 
         /* Flush and save to disk */
-        HT_ADD_END;             // achever
+        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);
       }
       //
       //