From b0b1c815c15e49c9172f59428810713097a65e37 Mon Sep 17 00:00:00 2001
From: hniksic <devnull@localhost>
Date: Sun, 19 Nov 2000 12:50:10 -0800
Subject: [PATCH] [svn] A bunch of new features:

- use mmap() to read whole files in core instead of allocating memory
  and read'ing it.

- use a new, more general, HTML parser (html-parse.c) and interface to
  it from Wget (html-url.c).

- respect <meta name=robots content=nofollow> (easy with the new HTML
  parser).

- use hash tables instead of linked lists in places where the lists
  were used to facilitate mappings.

- rewrite the code in host.c to be more readable and faster (hash
  tables instead of home-grown lists.)

- make convert_links properly convert partial URLs to complete ones
  for those URLs that have *not* been downloaded.

- use HTTP persistent connections where available.  very
  simple-minded, caches the last connection to the server.

Published in <sxshf533d5r.fsf@florida.arsdigita.de>.
---
 ChangeLog        |   4 +
 TODO             |  12 -
 configure        | 348 ++++++++++++++++---
 configure.in     |   1 +
 doc/ChangeLog    |   5 +
 doc/wget.texi    |   4 +-
 po/cs.gmo        | Bin 15443 -> 14838 bytes
 po/de.gmo        | Bin 15624 -> 14975 bytes
 po/hr.gmo        | Bin 21184 -> 24828 bytes
 po/it.gmo        | Bin 15483 -> 14838 bytes
 po/no.gmo        | Bin 14809 -> 14214 bytes
 po/pl.gmo        | Bin 15598 -> 14974 bytes
 po/pt_BR.gmo     | Bin 14564 -> 13920 bytes
 po/ru.gmo        | Bin 19691 -> 19040 bytes
 src/ChangeLog    | 114 +++++++
 src/Makefile.in  |  58 ++--
 src/config.h.in  |   3 +
 src/connect.c    |  31 ++
 src/ftp-ls.c     | 173 ++++++++++
 src/ftp.c        |   3 +-
 src/ftp.h        |   2 +
 src/hash.c       | 403 ++++++++++++++++++++++
 src/hash.h       |  50 +++
 src/headers.c    |   8 +
 src/headers.h    |   1 +
 src/host.c       | 340 ++++++++-----------
 src/html-parse.c | 856 +++++++++++++++++++++++++++++++++++++++++++++++
 src/html-parse.h |  44 +++
 src/html-url.c   | 569 +++++++++++++++++++++++++++++++
 src/http.c       | 203 ++++++++---
 src/main.c       |  15 +
 src/recur.c      | 148 +++++---
 src/retr.c       |  70 ++--
 src/retr.h       |   2 +-
 src/url.c        | 617 ++++++++++++----------------------
 src/url.h        |   3 +-
 src/utils.c      | 303 ++++++++++++-----
 src/utils.h      |  27 +-
 src/wget.h       |   2 +-
 39 files changed, 3518 insertions(+), 901 deletions(-)
 create mode 100644 src/hash.c
 create mode 100644 src/hash.h
 create mode 100644 src/html-parse.c
 create mode 100644 src/html-parse.h
 create mode 100644 src/html-url.c

diff --git a/ChangeLog b/ChangeLog
index 46905616..e873628f 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2000-11-10  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* configure.in: Test for MMAP.
+
 2000-11-16  Hrvoje Niksic  <hniksic@arsdigita.com>
 
 	* windows/config.h.ms: snprintf and vsnprintf exist under Windows.
diff --git a/TODO b/TODO
index d8319039..c0fecefc 100644
--- a/TODO
+++ b/TODO
@@ -49,15 +49,6 @@ changes.
 * Make `-k' check for files that were downloaded in the past and convert links 
   to them in newly-downloaded documents.
 
-* -k should convert relative references to absolute if not downloaded.
-
-* -k should convert "hostless absolute" URLs, like <A HREF="/index.html">.
-  However, Brian McMahon <bm@iucr.org> wants the old incorrect behavior to still
-  be available as an option, as he depends on it to allow mirrors of his site to
-  send CGI queries to his original site, but still get graphics off of the
-  mirror site.  Perhaps this would be better dealt with by adding an option to
-  tell -k not to convert certain URL patterns?
-
 * Add option to clobber existing file names (no `.N' suffixes).
 
 * Introduce a concept of "boolean" options.  For instance, every
@@ -85,9 +76,6 @@ changes.
 * Allow size limit to files (perhaps with an option to download oversize files 
   up through the limit or not at all, to get more functionality than [u]limit.
 
-* Recognize HTML comments correctly.  Add more options for handling
-  bogus HTML found all over the 'net.
-
 * Implement breadth-first retrieval.
 
 * Download to .in* when mirroring.
diff --git a/configure b/configure
index f7e130a2..de78c984 100755
--- a/configure
+++ b/configure
@@ -2040,15 +2040,281 @@ EOF
 
 fi
 
+for ac_hdr in unistd.h
+do
+ac_safe=`echo "$ac_hdr" | sed 'y%./+-%__p_%'`
+echo $ac_n "checking for $ac_hdr""... $ac_c" 1>&6
+echo "configure:2048: checking for $ac_hdr" >&5
+if eval "test \"`echo '$''{'ac_cv_header_$ac_safe'+set}'`\" = set"; then
+  echo $ac_n "(cached) $ac_c" 1>&6
+else
+  cat > conftest.$ac_ext <<EOF
+#line 2053 "configure"
+#include "confdefs.h"
+#include <$ac_hdr>
+EOF
+ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
+{ (eval echo configure:2058: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
+if test -z "$ac_err"; then
+  rm -rf conftest*
+  eval "ac_cv_header_$ac_safe=yes"
+else
+  echo "$ac_err" >&5
+  echo "configure: failed program was:" >&5
+  cat conftest.$ac_ext >&5
+  rm -rf conftest*
+  eval "ac_cv_header_$ac_safe=no"
+fi
+rm -f conftest*
+fi
+if eval "test \"`echo '$ac_cv_header_'$ac_safe`\" = yes"; then
+  echo "$ac_t""yes" 1>&6
+    ac_tr_hdr=HAVE_`echo $ac_hdr | sed 'y%abcdefghijklmnopqrstuvwxyz./-%ABCDEFGHIJKLMNOPQRSTUVWXYZ___%'`
+  cat >> confdefs.h <<EOF
+#define $ac_tr_hdr 1
+EOF
+ 
+else
+  echo "$ac_t""no" 1>&6
+fi
+done
+
+for ac_func in getpagesize
+do
+echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
+echo "configure:2087: checking for $ac_func" >&5
+if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
+  echo $ac_n "(cached) $ac_c" 1>&6
+else
+  cat > conftest.$ac_ext <<EOF
+#line 2092 "configure"
+#include "confdefs.h"
+/* System header to define __stub macros and hopefully few prototypes,
+    which can conflict with char $ac_func(); below.  */
+#include <assert.h>
+/* Override any gcc2 internal prototype to avoid an error.  */
+/* We use char because int might match the return type of a gcc2
+    builtin and then its argument prototype would still apply.  */
+char $ac_func();
+
+int main() {
+
+/* The GNU C library defines this for functions which it implements
+    to always fail with ENOSYS.  Some functions are actually named
+    something starting with __ and the normal name is an alias.  */
+#if defined (__stub_$ac_func) || defined (__stub___$ac_func)
+choke me
+#else
+$ac_func();
+#endif
+
+; return 0; }
+EOF
+if { (eval echo configure:2115: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+  rm -rf conftest*
+  eval "ac_cv_func_$ac_func=yes"
+else
+  echo "configure: failed program was:" >&5
+  cat conftest.$ac_ext >&5
+  rm -rf conftest*
+  eval "ac_cv_func_$ac_func=no"
+fi
+rm -f conftest*
+fi
+
+if eval "test \"`echo '$ac_cv_func_'$ac_func`\" = yes"; then
+  echo "$ac_t""yes" 1>&6
+    ac_tr_func=HAVE_`echo $ac_func | tr 'abcdefghijklmnopqrstuvwxyz' 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'`
+  cat >> confdefs.h <<EOF
+#define $ac_tr_func 1
+EOF
+ 
+else
+  echo "$ac_t""no" 1>&6
+fi
+done
+
+echo $ac_n "checking for working mmap""... $ac_c" 1>&6
+echo "configure:2140: checking for working mmap" >&5
+if eval "test \"`echo '$''{'ac_cv_func_mmap_fixed_mapped'+set}'`\" = set"; then
+  echo $ac_n "(cached) $ac_c" 1>&6
+else
+  if test "$cross_compiling" = yes; then
+  ac_cv_func_mmap_fixed_mapped=no
+else
+  cat > conftest.$ac_ext <<EOF
+#line 2148 "configure"
+#include "confdefs.h"
+
+/* Thanks to Mike Haertel and Jim Avera for this test.
+   Here is a matrix of mmap possibilities:
+	mmap private not fixed
+	mmap private fixed at somewhere currently unmapped
+	mmap private fixed at somewhere already mapped
+	mmap shared not fixed
+	mmap shared fixed at somewhere currently unmapped
+	mmap shared fixed at somewhere already mapped
+   For private mappings, we should verify that changes cannot be read()
+   back from the file, nor mmap's back from the file at a different
+   address.  (There have been systems where private was not correctly
+   implemented like the infamous i386 svr4.0, and systems where the
+   VM page cache was not coherent with the filesystem buffer cache
+   like early versions of FreeBSD and possibly contemporary NetBSD.)
+   For shared mappings, we should conversely verify that changes get
+   propogated back to all the places they're supposed to be.
+
+   Grep wants private fixed already mapped.
+   The main things grep needs to know about mmap are:
+   * does it exist and is it safe to write into the mmap'd area
+   * how to use it (BSD variants)  */
+#include <sys/types.h>
+#include <fcntl.h>
+#include <sys/mman.h>
+
+/* This mess was copied from the GNU getpagesize.h.  */
+#ifndef HAVE_GETPAGESIZE
+# ifdef HAVE_UNISTD_H
+#  include <unistd.h>
+# endif
+
+/* Assume that all systems that can run configure have sys/param.h.  */
+# ifndef HAVE_SYS_PARAM_H
+#  define HAVE_SYS_PARAM_H 1
+# endif
+
+# ifdef _SC_PAGESIZE
+#  define getpagesize() sysconf(_SC_PAGESIZE)
+# else /* no _SC_PAGESIZE */
+#  ifdef HAVE_SYS_PARAM_H
+#   include <sys/param.h>
+#   ifdef EXEC_PAGESIZE
+#    define getpagesize() EXEC_PAGESIZE
+#   else /* no EXEC_PAGESIZE */
+#    ifdef NBPG
+#     define getpagesize() NBPG * CLSIZE
+#     ifndef CLSIZE
+#      define CLSIZE 1
+#     endif /* no CLSIZE */
+#    else /* no NBPG */
+#     ifdef NBPC
+#      define getpagesize() NBPC
+#     else /* no NBPC */
+#      ifdef PAGESIZE
+#       define getpagesize() PAGESIZE
+#      endif /* PAGESIZE */
+#     endif /* no NBPC */
+#    endif /* no NBPG */
+#   endif /* no EXEC_PAGESIZE */
+#  else /* no HAVE_SYS_PARAM_H */
+#   define getpagesize() 8192	/* punt totally */
+#  endif /* no HAVE_SYS_PARAM_H */
+# endif /* no _SC_PAGESIZE */
+
+#endif /* no HAVE_GETPAGESIZE */
+
+#ifdef __cplusplus
+extern "C" { void *malloc(unsigned); }
+#else
+char *malloc();
+#endif
+
+int
+main()
+{
+	char *data, *data2, *data3;
+	int i, pagesize;
+	int fd;
+
+	pagesize = getpagesize();
+
+	/*
+	 * First, make a file with some known garbage in it.
+	 */
+	data = malloc(pagesize);
+	if (!data)
+		exit(1);
+	for (i = 0; i < pagesize; ++i)
+		*(data + i) = rand();
+	umask(0);
+	fd = creat("conftestmmap", 0600);
+	if (fd < 0)
+		exit(1);
+	if (write(fd, data, pagesize) != pagesize)
+		exit(1);
+	close(fd);
+
+	/*
+	 * Next, try to mmap the file at a fixed address which
+	 * already has something else allocated at it.  If we can,
+	 * also make sure that we see the same garbage.
+	 */
+	fd = open("conftestmmap", O_RDWR);
+	if (fd < 0)
+		exit(1);
+	data2 = malloc(2 * pagesize);
+	if (!data2)
+		exit(1);
+	data2 += (pagesize - ((int) data2 & (pagesize - 1))) & (pagesize - 1);
+	if (data2 != mmap(data2, pagesize, PROT_READ | PROT_WRITE,
+	    MAP_PRIVATE | MAP_FIXED, fd, 0L))
+		exit(1);
+	for (i = 0; i < pagesize; ++i)
+		if (*(data + i) != *(data2 + i))
+			exit(1);
+
+	/*
+	 * Finally, make sure that changes to the mapped area
+	 * do not percolate back to the file as seen by read().
+	 * (This is a bug on some variants of i386 svr4.0.)
+	 */
+	for (i = 0; i < pagesize; ++i)
+		*(data2 + i) = *(data2 + i) + 1;
+	data3 = malloc(pagesize);
+	if (!data3)
+		exit(1);
+	if (read(fd, data3, pagesize) != pagesize)
+		exit(1);
+	for (i = 0; i < pagesize; ++i)
+		if (*(data + i) != *(data3 + i))
+			exit(1);
+	close(fd);
+	unlink("conftestmmap");
+	exit(0);
+}
+
+EOF
+if { (eval echo configure:2288: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext} && (./conftest; exit) 2>/dev/null
+then
+  ac_cv_func_mmap_fixed_mapped=yes
+else
+  echo "configure: failed program was:" >&5
+  cat conftest.$ac_ext >&5
+  rm -fr conftest*
+  ac_cv_func_mmap_fixed_mapped=no
+fi
+rm -fr conftest*
+fi
+
+fi
+
+echo "$ac_t""$ac_cv_func_mmap_fixed_mapped" 1>&6
+if test $ac_cv_func_mmap_fixed_mapped = yes; then
+  cat >> confdefs.h <<\EOF
+#define HAVE_MMAP 1
+EOF
+
+fi
+
 for ac_func in strdup strstr strcasecmp strncasecmp
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:2047: checking for $ac_func" >&5
+echo "configure:2313: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2052 "configure"
+#line 2318 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -2071,7 +2337,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:2075: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2341: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -2098,12 +2364,12 @@ done
 for ac_func in gettimeofday mktime strptime
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:2102: checking for $ac_func" >&5
+echo "configure:2368: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2107 "configure"
+#line 2373 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -2126,7 +2392,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:2130: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2396: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -2153,12 +2419,12 @@ done
 for ac_func in strerror snprintf vsnprintf select signal symlink access isatty
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:2157: checking for $ac_func" >&5
+echo "configure:2423: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2162 "configure"
+#line 2428 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -2181,7 +2447,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:2185: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2451: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -2208,12 +2474,12 @@ done
 for ac_func in uname gethostname
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:2212: checking for $ac_func" >&5
+echo "configure:2478: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2217 "configure"
+#line 2483 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -2236,7 +2502,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:2240: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2506: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -2264,12 +2530,12 @@ done
 for ac_func in gethostbyname
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:2268: checking for $ac_func" >&5
+echo "configure:2534: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2273 "configure"
+#line 2539 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -2292,7 +2558,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:2296: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2562: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -2314,7 +2580,7 @@ EOF
 else
   echo "$ac_t""no" 1>&6
 echo $ac_n "checking for gethostbyname in -lnsl""... $ac_c" 1>&6
-echo "configure:2318: checking for gethostbyname in -lnsl" >&5
+echo "configure:2584: checking for gethostbyname in -lnsl" >&5
 ac_lib_var=`echo nsl'_'gethostbyname | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_lib_$ac_lib_var'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -2322,7 +2588,7 @@ else
   ac_save_LIBS="$LIBS"
 LIBS="-lnsl  $LIBS"
 cat > conftest.$ac_ext <<EOF
-#line 2326 "configure"
+#line 2592 "configure"
 #include "confdefs.h"
 /* Override any gcc2 internal prototype to avoid an error.  */
 /* We use char because int might match the return type of a gcc2
@@ -2333,7 +2599,7 @@ int main() {
 gethostbyname()
 ; return 0; }
 EOF
-if { (eval echo configure:2337: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2603: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_lib_$ac_lib_var=yes"
 else
@@ -2367,7 +2633,7 @@ done
 
 
 echo $ac_n "checking for socket in -lsocket""... $ac_c" 1>&6
-echo "configure:2371: checking for socket in -lsocket" >&5
+echo "configure:2637: checking for socket in -lsocket" >&5
 ac_lib_var=`echo socket'_'socket | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_lib_$ac_lib_var'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -2375,7 +2641,7 @@ else
   ac_save_LIBS="$LIBS"
 LIBS="-lsocket  $LIBS"
 cat > conftest.$ac_ext <<EOF
-#line 2379 "configure"
+#line 2645 "configure"
 #include "confdefs.h"
 /* Override any gcc2 internal prototype to avoid an error.  */
 /* We use char because int might match the return type of a gcc2
@@ -2386,7 +2652,7 @@ int main() {
 socket()
 ; return 0; }
 EOF
-if { (eval echo configure:2390: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2656: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_lib_$ac_lib_var=yes"
 else
@@ -2417,7 +2683,7 @@ fi
 if test "x${with_socks}" = xyes
 then
   echo $ac_n "checking for main in -lresolv""... $ac_c" 1>&6
-echo "configure:2421: checking for main in -lresolv" >&5
+echo "configure:2687: checking for main in -lresolv" >&5
 ac_lib_var=`echo resolv'_'main | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_lib_$ac_lib_var'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -2425,14 +2691,14 @@ else
   ac_save_LIBS="$LIBS"
 LIBS="-lresolv  $LIBS"
 cat > conftest.$ac_ext <<EOF
-#line 2429 "configure"
+#line 2695 "configure"
 #include "confdefs.h"
 
 int main() {
 main()
 ; return 0; }
 EOF
-if { (eval echo configure:2436: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2702: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_lib_$ac_lib_var=yes"
 else
@@ -2460,7 +2726,7 @@ else
 fi
 
   echo $ac_n "checking for Rconnect in -lsocks""... $ac_c" 1>&6
-echo "configure:2464: checking for Rconnect in -lsocks" >&5
+echo "configure:2730: checking for Rconnect in -lsocks" >&5
 ac_lib_var=`echo socks'_'Rconnect | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_lib_$ac_lib_var'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -2468,7 +2734,7 @@ else
   ac_save_LIBS="$LIBS"
 LIBS="-lsocks  $LIBS"
 cat > conftest.$ac_ext <<EOF
-#line 2472 "configure"
+#line 2738 "configure"
 #include "confdefs.h"
 /* Override any gcc2 internal prototype to avoid an error.  */
 /* We use char because int might match the return type of a gcc2
@@ -2479,7 +2745,7 @@ int main() {
 Rconnect()
 ; return 0; }
 EOF
-if { (eval echo configure:2483: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2749: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_lib_$ac_lib_var=yes"
 else
@@ -2511,7 +2777,7 @@ fi
 ALL_LINGUAS="cs de hr it no pl pt_BR ru"
 
 echo $ac_n "checking whether NLS is requested""... $ac_c" 1>&6
-echo "configure:2515: checking whether NLS is requested" >&5
+echo "configure:2781: checking whether NLS is requested" >&5
         # Check whether --enable-nls or --disable-nls was given.
 if test "${enable_nls+set}" = set; then
   enableval="$enable_nls"
@@ -2528,7 +2794,7 @@ fi
       # Extract the first word of "msgfmt", so it can be a program name with args.
 set dummy msgfmt; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:2532: checking for $ac_word" >&5
+echo "configure:2798: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_path_MSGFMT'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -2562,7 +2828,7 @@ fi
       # Extract the first word of "xgettext", so it can be a program name with args.
 set dummy xgettext; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:2566: checking for $ac_word" >&5
+echo "configure:2832: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_path_XGETTEXT'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -2597,7 +2863,7 @@ fi
       # Extract the first word of "gmsgfmt", so it can be a program name with args.
 set dummy gmsgfmt; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:2601: checking for $ac_word" >&5
+echo "configure:2867: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_path_GMSGFMT'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
@@ -2647,17 +2913,17 @@ fi
 do
 ac_safe=`echo "$ac_hdr" | sed 'y%./+-%__p_%'`
 echo $ac_n "checking for $ac_hdr""... $ac_c" 1>&6
-echo "configure:2651: checking for $ac_hdr" >&5
+echo "configure:2917: checking for $ac_hdr" >&5
 if eval "test \"`echo '$''{'ac_cv_header_$ac_safe'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2656 "configure"
+#line 2922 "configure"
 #include "confdefs.h"
 #include <$ac_hdr>
 EOF
 ac_try="$ac_cpp conftest.$ac_ext >/dev/null 2>conftest.out"
-{ (eval echo configure:2661: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
+{ (eval echo configure:2927: \"$ac_try\") 1>&5; (eval $ac_try) 2>&5; }
 ac_err=`grep -v '^ *+' conftest.out | grep -v "^conftest.${ac_ext}\$"`
 if test -z "$ac_err"; then
   rm -rf conftest*
@@ -2687,12 +2953,12 @@ done
       for ac_func in gettext
 do
 echo $ac_n "checking for $ac_func""... $ac_c" 1>&6
-echo "configure:2691: checking for $ac_func" >&5
+echo "configure:2957: checking for $ac_func" >&5
 if eval "test \"`echo '$''{'ac_cv_func_$ac_func'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
   cat > conftest.$ac_ext <<EOF
-#line 2696 "configure"
+#line 2962 "configure"
 #include "confdefs.h"
 /* System header to define __stub macros and hopefully few prototypes,
     which can conflict with char $ac_func(); below.  */
@@ -2715,7 +2981,7 @@ $ac_func();
 
 ; return 0; }
 EOF
-if { (eval echo configure:2719: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:2985: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_func_$ac_func=yes"
 else
@@ -2737,7 +3003,7 @@ EOF
 else
   echo "$ac_t""no" 1>&6
 echo $ac_n "checking for gettext in -lintl""... $ac_c" 1>&6
-echo "configure:2741: checking for gettext in -lintl" >&5
+echo "configure:3007: checking for gettext in -lintl" >&5
 ac_lib_var=`echo intl'_'gettext | sed 'y%./+-%__p_%'`
 if eval "test \"`echo '$''{'ac_cv_lib_$ac_lib_var'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
@@ -2745,7 +3011,7 @@ else
   ac_save_LIBS="$LIBS"
 LIBS="-lintl  $LIBS"
 cat > conftest.$ac_ext <<EOF
-#line 2749 "configure"
+#line 3015 "configure"
 #include "confdefs.h"
 /* Override any gcc2 internal prototype to avoid an error.  */
 /* We use char because int might match the return type of a gcc2
@@ -2756,7 +3022,7 @@ int main() {
 gettext()
 ; return 0; }
 EOF
-if { (eval echo configure:2760: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
+if { (eval echo configure:3026: \"$ac_link\") 1>&5; (eval $ac_link) 2>&5; } && test -s conftest${ac_exeext}; then
   rm -rf conftest*
   eval "ac_cv_lib_$ac_lib_var=yes"
 else
@@ -2824,7 +3090,7 @@ do
 # Extract the first word of "$ac_prog", so it can be a program name with args.
 set dummy $ac_prog; ac_word=$2
 echo $ac_n "checking for $ac_word""... $ac_c" 1>&6
-echo "configure:2828: checking for $ac_word" >&5
+echo "configure:3094: checking for $ac_word" >&5
 if eval "test \"`echo '$''{'ac_cv_prog_MAKEINFO'+set}'`\" = set"; then
   echo $ac_n "(cached) $ac_c" 1>&6
 else
diff --git a/configure.in b/configure.in
index 474e5d57..4f4440b0 100644
--- a/configure.in
+++ b/configure.in
@@ -160,6 +160,7 @@ dnl
 dnl Checks for library functions.
 dnl
 AC_FUNC_ALLOCA
+AC_FUNC_MMAP
 AC_CHECK_FUNCS(strdup strstr strcasecmp strncasecmp)
 AC_CHECK_FUNCS(gettimeofday mktime strptime)
 AC_CHECK_FUNCS(strerror snprintf vsnprintf select signal symlink access isatty)
diff --git a/doc/ChangeLog b/doc/ChangeLog
index bddc4b5d..110102a6 100644
--- a/doc/ChangeLog
+++ b/doc/ChangeLog
@@ -1,3 +1,8 @@
+2000-11-15  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* wget.texi (Robots): Document that we now support the meta tag
+	exclusion.
+
 2000-11-16  Hrvoje Niksic  <hniksic@arsdigita.com>
 
 	* wget.texi: Use --- consistently.
diff --git a/doc/wget.texi b/doc/wget.texi
index eb4d00c9..1accbb94 100644
--- a/doc/wget.texi
+++ b/doc/wget.texi
@@ -2548,8 +2548,8 @@ this:
 
 This is explained in some detail at
 @url{http://info.webcrawler.com/mak/projects/robots/meta-user.html}.
-Unfortunately, Wget does not support this method of robot exclusion yet,
-but it will be implemented in the next release.
+Wget supports this method of robot exclusion in addition to the usual
+@file{/robots.txt} exclusion.
 
 @node Security Considerations, Contributors, Robots, Appendices
 @section Security Considerations
diff --git a/po/cs.gmo b/po/cs.gmo
index 5511ce5c7c3f5e1de27487239e7c88a41b6fdce8..c276436c9b14f2f5190c5e08a9c1d84ea271c535 100644
GIT binary patch
delta 3351
zcmZA12~bs49LMp$8_MGeilQVM$itK^!4+8wRFc9%6ESl`w9G_t!6eO|nj0!gp=D}L
zqcu2MQwlW=mQFg=Oj%4(&9s;;PU2*iP2-Z)^!@P;GfZdR|9;NB_uaGn&$+j$;_5R0
zci|m(8D$UAm$=l?tO?%^=0NG<F^j@rT#o~A2p-0KJc|P`KGZA(3!Kxi2j}zL`6~2s
zz6s-TCnlTu?F5x*E?mOf&>Lo!h(j?3XJLO_gBsug?2acf0Y61O_cz9&uUo)8)W8>F
zELLJ9zJj;ndw3`P+j%PLFpgGrV*=iSWjF{cQ5_#c4YU!}@NbxgA>n2zI2MQFa*V+v
zn1g3f?S_($-k6LUNG|&5-)2%#hwHE(?nB084VZ)<x$D<asqWP?a6J<>^ZDq*YSaMh
zou8xH`4hF8og&PrS{U}iIP?#slHo4QK_+D@k+rdHn28OjhOaq$^RzND5|xQU)QlFR
zW?G4T@Q}NI8rAMOcYej)_w*+JO35G&Q`?zkqc+JNTGLwAqSp8fYKGTvC`OSk4X^;!
z(P|ulyHRW1h|1(2sP?+?CvCcZsEm%oAz0pr{Ob?3TwwoMJu*o<?aseKjl3C^slSk)
z#WL+-n1yM$6uaON<b|+$RJ$j!Wdg|lv}>rhq%-MLKRx_Z)ImI|p$vDU+_?p{hA$#N
zJIaAc*(a#AY(~8u|KLcBX4>j_3dZ1EBx$w*`HSsGZT91+cKsKqC}r1C8Hi#5m8t~P
zObSsWE<ts?2-Wds9E>lZGV(F1oo3Wt3E`mU6H)&gg&M$oBnh?=`T6Yt#~3c0LZ#{_
z)Eb9!k*bYGT`xiH{wmb<I@G|LFbL0KXZ#tp+b_HG>&T1Kk|zDXC+Ui9N1xbF@BbKv
zy^b5RaX6mGBJ9JCOT{Iq2M;=5Lp6BZoj0OV{tYT~&8W>8&JHFSmW?Sm3)Ox#(zm^j
zKKi$2DrDchOiz0u0ku0bF%ip959~$8X78X<{ypjgbrZG5aeNHaVHQrt>3BOHL%OgZ
zP)iWRPSyZ@sQqUXsp!E)sMJ<r9=?id;3}r$0D9BX6yw9V6E(A&sF^2mplU^^nfg)f
zy@4dfzQAz2ftrXnf&A;nU=~;-oP@ejflSKwB7NJtsDWN|29XvI=RVZZq$1xG%W@W?
zGPnS>S@)v`Qin8QO_+oKCX#=pHajVh>Jrp|wxc>chFY3NcmAu>L;mzO3`8wi4r*6V
zLOoxGnn(@iw=;Vay_~n_RnVRaLuD?<PbGrNQg>rB-obeTs(~A*ne^q~dawYMnYE|}
zYmv6C9?7cxh#BbR&&p69s{JBV>Ps;U{qDMd2Nj;Nmr<MQOI(0guoP#|duRL*gYh#|
z$KT>8yogG@kBt$6>8N(@LS3JYhp+;b`u==2<#_C&_kRTyrFtvs!9%Fjo<{A7vsjGn
z*-`3v4(k7lQU9w#t!*tTb045SWI-bWnTbF;vN5R5JOwq8jV;$%|7t2~=q2olC$I>c
zPy<TfyGomui)vsMDz%4^ePnN;-k$TQweP@pTn<Ft&v&lC@thw-P2hJ-rhoI0N9vY=
zWYfw~Bdo<Pcoek+4et6W)Bw()X7;tae;L_pEt7$kjXN3dB~}r;36(<P8DcAu(}}|)
zgf@}Vr(Mi<!sZjLg?6o+(3{d)G@$#5?KRn+DE|&>F~n$M3-KrsNpvS#%N{EFYG}x9
z%X&^$6C2#cB-9pFc`|@4MAB<h-T6vXik>CXwTo2Pu{MR+rj!2_Eu_lx#1bN#c#3$4
zXj{5*GS^+mLM?0oQB0^TCbT(J@&j<B|3xr?s3f#Yv_SKS*@TK-<_C!!Lc`TN*S0L+
z<N<df9;Xq}L=Evcq4xA`T}!YrZ5g4>QBB;f21|$zL=d5}E<nr9n(6A#VFa<0(3aMh
z<Q_t0c>o)S?TMl8d?ViH>K(Bcp>3@)hFD7E=5m-ploC%6DpiDD;MStglFAxl7BPp&
zBvf`0<7@VLql>3g8>UN?5aKZ+fv6x<HWL#Gy+J-gWd*T`m_$5Hw3ad|V+p-f;Y1!$
n7uoff;JP7^rJk;-$tlBQQqs~>(h^fr>yE@;2&sF0@Imi?*C0Gy

delta 3924
zcmb`Idr(|g9mg+e+J=QD;kBW>_CSE}*xdlj0t>`+Lm+G*K$oQ{*fv?%OSrJvOLrF%
z($r0BTeUCKVEV9*CbrI0r**10qp6*YI(2NEYQ{-BYE0XS&sq~rYscw0_3`=b-Z+kb
zdxr1*oO{l>=lou0`CH=L!PN29`EO~N_B1HdwACNLbDEa@T4vZ*3`;f5b|btGdSMHE
z2@b%^utw9?`d|*+V{r}^@P5F0e+cIC{sgRt&%%_8&QIw$;771l)7D{x)(Cr`f+nnl
zM<H|Tz6tZ-S=azCLcUwi=2|uMSiBWt`1-qH6?_Wf%=(vM1$=|^wXJl-V-Cz~%}nI&
zg?O8nf^|?n<lwv?Kn%@$8?tfUzhH-^<<r-iU<eYO{QIGT--Uj70dimg(l)>?5JL*a
zIA7aHC&mB=794}+@Oii%{unypuQKC>YapR6#9J9J9E9`>mmvETo`4uw_=?5fLiQ<I
z4YOepEQGAl?68&dwM}#etbqh1v_%g>Qd9H=*bT)HKC)kLk#8g-_8k!K?GcDa_9GB4
z?GmvP_#G(o)RXLM|BI$&xd}p6X1q8>*koa`ctU83_px9cieb3~<(IG;{vML+k{pPq
zC2f#AmrOtmEqN5OPsz)Wxssnla$dR?64BCT$bO~6km4)74LZqx=>v3H7+A3u$U%Hm
z`d3J}%2s9G%kWM-C=<_7nGbScSr}3`Wrtx0{01aVW$!{8lzlk3?4J+^%Ef_oP#nl1
z|K)vjIH)`bIiP$7vQhaVYg`PI`7@AQmcI=Nefbr5Bhls%$!UccRtdXdJG=vOaK+<L
z!Dk>&RJ;QDzT!>TNSG`BPKS*hwUCfG_*f#~n1F=JVL%RXEJLh!oP-?gcmZ;-<415a
zya0)a;!xQ~i9yti+6nny%|h0xgAfDMb1+45s2|dig%_d1i7eOv*+g<&xeGF1d5<-I
z5^@VFU$e&FhZt6s53hrA&8~;6k*cqvKGN^vSOM=_p{yU}{45Ts+QR@vvhiMsHlBc9
z>W)IwZh_;F0;+xy@?rI_E&ds@VYNh%@fs;Mysse&61f_Q0EgGy3^A%E1)Ja(A^X>y
zmHS`G6~2}M4y%<9Fr;<>QVX>bdWy65F4zda4rTp$h~c##K@QvGfOW7FlH*NNkOMaz
zhGXz)=!Wk>Zd09tq@b=>EW*ONNo(LJ<iolbAe+`ni2LCs$OiRTt#L@b+;>t_|0OsL
z&p|wE5CibMVGd$Y!*R$r4W}XdHGBYZw2`{eNLOQuN*7NW#bcR>K`d-MW{rOvlE=pL
zkmxvpSng!=G<zY%<rD{U;5@WL@m#nJiJ<dqkVBnP^%&xOA5vo}7goq`v}+3_w5}8+
z)UJ~dgIqs>9O#nJks9~f%)47GVcZ>n+y*yMkd(QP@B=36!~u$X^NWyuHcKRNbn_MF
zakh!Np;l5_^By|Q{6!%lY?k6Ef@iJq-@%XJ#0JO)Efl19)UuSBZ+RXPp_aE`E?HR(
zNvQ{8Bo&^munQWHNO`^j+26AQ^YPYmh7P|-o(tB*C5R!dt6&NAz&SVsF|zd~$o29{
zv1UW=v>fd9LFDa&gx-4uk`nI|kbS(*SmRO)&++~SOp$}OPtXy56%v-V*CC;8dk6Ai
z`?Zh|w<}0(w0FY2P$I{{?cal}Z<kcbx_6<2A5n<}k?Y_a`9bTLf<&ex4mqde1m{at
zcYKQhJn8t2HE|iTkuM8UG(PeuzXLv^A_n<B1*vM^mm%N#egO$>CsC2w=_I^zi#n+*
z$$jTui}yj9$NMy2m-|1MPEXo{u3b!EVb?6=W^_FX$#vJyAO?1eM{6M7ND8{8CS)9M
z#enX5h-cmH*8BhzCvYa+C`@rg8$yTCQ^h=_jiF`qFp_W^%kO0u)QcHcn~%-{s=PV?
z#otl%*=3ilwEiePg|?xG&>WKer4X++`61f3e5b7-bw52x`~%jM6H<lgMpV)*##3CG
z$h>NILEP1(qNE6<HKSY6Bl7Y;ja;SM=XpxfZ$gp-?&|V2Ie96azO*ffd}$*njHKO(
zcBgyk6j?n@GwI!o9z^6`!miDrNhGZbeH{7G5hOM9-!00^cxE6a_fxwKN&Y^E4k6jF
z4&95|P!f@I?J*QUQpMMx>yWfhW+3&V?Xmh#LMg)I$ct*wP9$wH18s*C-zr`_)__#?
zu+^7Bvm>fC-FkQ?P%pXz?MHVbX^*1nbT9MhvikSHy(ogZk+jE=6!})vxBOV{hTCta
z*NUX&paZA@EujJQ85BfZaJftAc97?%Q3xGNk7OS6U_0W<$bT2De`U=Lf67kHFU5_0
zGf7p~Q?HtRKCe@rZRuxw`q`oS;<~Oz&FSQUa9me?=0YqIP8w#+sRFSnw=FbdB-DJ|
z+!qh;S9BvrA{jR(7m|8J88MZd(bbGOuhR)9m622jjA&F%>S`gOPcNjRP8&<r?m(!2
zcr2v+L*r_<KN$26g~q#Bnw&A2)fe^5M#ld6sKLHi5)a3cOUj(K?dlEo^t0Z-Eif1e
zjVm**`U0V$-qBIjHyl)cHR2D30zG4c{-7Ee3yutr_PUiC#V1U2+jb0%sU3PukB6gb
zWMMLDOsPR*N{=OUHO*H0O?-~%$*>VkxNR#}?56F8CRd9`wRk$4TU|}<E78h7<fMAc
z|3e`4SknAYBxiMX>hg#9y{Ka0$fa4ssZQ%^av?(GP8&)6o!CE|D!h0qJZIb1o0?EN
z{R3wPh6e*`;`_m|0fHCP=gq|2QgUJTvcj2g%ur@z(LAKb{-_RJj4#YBsgdCD_THUC
z!=s^qvV~*%q^WM{AKsoBU{Z|*Uk!xbROesq9dxVh;pnN$YTk^*&B>UVNE&J;sz>zM
iq^=ycMKwIBPqFo){-Hh#38WsIG8FM8Zb8?xx&H>}UjwB8

diff --git a/po/de.gmo b/po/de.gmo
index 6dc32778c758b6edffd41764374db7d34a8e231c..63f01f82da200b1cc5f948b444ebdab9959dc249 100644
GIT binary patch
delta 3351
zcmY+`drZ}39LMpmh00Aqyg-t21T+D$1BQZVfPq?OXyGldA>I<j6tlc^@P;Co7LrDG
zvp*~?XX0wJQj2tBny$>1EwffDm!@TISc~fYIp<;1-}wGs&-0w$d9L5z(T&UhEBAdD
z6;x+j^~9aT<sh>&_+bb)uHaC!SPa2+*c;Pv501j~*c+2NnuTGZb24_OKF8HpU^w*+
zn1tIf)y!vyDa6rm8T(*(gjos>Ko3sGURZ&e;4bWhM=%*bM}6-fOvHr9fFn^8Ux4vg
zg)#Ux_QYeD!}#_S1r3-;ulisN-hpM<7pqVMA3#m?G^*pPn2BLgW?sz43|xvH+>3d5
z64h@<wj&x-Q4=Y^1je^%6g1!(?14LxIoToXhbLV74b)b5>k?=mf?D}JOu((E2_AHQ
ziR$M%>NMMSHKS+|*bNiWcNc|h*Dw=Vlr2Zj#%gc~9zu28?2P8y+L1x1ohU-BXc20q
zRoEThbe}h(`n}-lf4I*>quGCLNndW0+L+~{lBAyAbe6kOXM7U1!e$(Rv22$nScn>E
z6%NE5sIxwe+R1CE{vvruN!J6lqa!dK%e%AxdSEvV<ewcx7HN&HehxMBi>RIY8~L+%
zmVFNn#Y|j`!MGQBA?zTk-y`T>0Fs~9jCxDjvwa$;vyXxXNJ4d#?LH`XZbF^mYsjCy
z%Z)|ZDb!hBM7<p?I0)ldwg#Sn9xOpN&DJB2*ej@H{{Ypm?-vT%vKy!!h~)saRmrH8
z6rpBZj2d_$YT%7{H@=M8k<U>5Ttwwc7&m=C1@&AOY6A0+O|VMj&u6>1duaFswN*c&
z&bT9u6m1x)y%?4Kn^ElzsEM7yc6b5X<IkvUzvAjQkQc|lO?tiy+m+acF|nuK{{p7n
z#0Mp)nf52s#^N+o67ENR@hoaZKcKemva7eC1`g(7eLoVlfL!!oIVyMRupfSe{MoPE
zdNaO7uq<s|2J%~B6R|(ech;lc^Co0>?RRAI7D1yrPD35RB;+G2#{#TJ{eXRs85qLD
z9?U@{?+o;*qdE%e;C)p63(Uvsn1)&OG7RUVI^4%i1AK+_Z9PaB(#P^q6D&qAu11~p
z0o0M5Lgm;MR6i|A?7u$f!vQmv4M%;j(ABr2-t!}<iT&n`NC_li8mhzLsN@`nO0qJ%
z7i&=cpFq6@mry$q#5ok=;C}4C2HZr0R?>hRr+tIk%70NEBr*$4FdwzTQq+X1QMvGr
zYd?(Y=e+woh;w|k4Q&`s`*9LP{hmSf_m__XyJ5Z9FKzWG)Yh%yrjGWZw*D+;V+;?m
z2rEK$Sb_Q_+l-81M{y`##V}0eLuG#!>X&vjcEVY%-M5T_zEF+&MQXwvyx`j7Nagm_
zr(y`sLJeGoY>usT^|zewq55e=9mzR-3I9ap*cLKRIk5+M&S%Fduv_*mdN7m&(t|0e
zU&P@!3Vo=7PoQ3(uTjsnxOxo#08q4{*cq3gj&=h^;C^HdcGT6cW0c<iAbvM>rqQSl
z2Vx<PMdiZlsE&@Ij^qkPVOM6S4*Q~xa0)8<D$#>)pceKi4#sQ9pQUC6c5XIiFuv7N
z(1T4#f?G4{?Be-WF!o1n{UB6(9){v*)Ji9!CRB>#nt$8D|4NR-(ZmX32cfHoc%G;x
z^4fBnOK1XmWpp-L<vgPGVm?+*^lEKD<;26pv$eUQvA!*oJcLqp6Y)3^Lnzf+uX+lz
z2~GLtYaNwU#Cq4LtkQw$dM1D^K(^f`x%zU{mb^f)YVBg+{h5F@{@Sg#?ymYGK0)LX
zYl&IJ%~vFq64#*DLW!Ydm_g`TM2z>B0$CX-GX6qJV~8q3FQN`;4)GYFE1r0i$Rjje
z-sfAc`P3%5mL!}^#1XZ`(}em{7Op19Tw6jYIkpn_tHWaAcA_1jYfS)u%1v|S?bwyr
zM(AD9PtgN}uB8EN1l~praP>-j$d!Yz8<FDb_u*oqpn%&{qLf%k=-N!^MQy#ZU7-S}
z6Ele+gswW`!P=eSaWke+;+MdGDcPSSl8LAME&RU&#}aze5?tFd+(0}+Jm=a(8Iez<
s6H&xSq9LZkrI3d7n7N@H(o(${9&hGgZ)S=&tzmck`0$1i-jBoo2Tzzh@c;k-

delta 3968
zcmbW1YjB)Z8OKkcw#_Ec=Gv0mo+QnEcTG3fO&04m-E5Ymn{=~jNGvUz<W2Uao87Rx
zAv6@WV2cQfK!Ad%(9sVFqJweNQ5=~L>WGYd@HQ&QbVm80V!_)uP6dC@?ptPb#xI`9
zKfiO%`<~}Kw|C#m+_Rp2WJ~F9RDL`GYLwdYSNId93SY@@+jhfhrEJ&0)367+;kV!@
zd<!-yb+rc;!DAMW!!p)u*7`nJ%=#>BgI|DICzBsDvBOKSS*aa3p*r9Y)G!I_;RBFo
zJH84_;H$75UVxn2iE*_Dc3He0;`q+nVFP>|63osQU>*D!_p5zOBw{Ws>CQhWxe*d=
zNfx$3ImpE&=OK=kya6#>@*eD0s+76vgb_$~O3y$IzYTrxHOPf!NbQ2xLL4cZ;(pc0
zB*6w3mOTV(;g{e}_%d|BU+4GBw?k50PPDRLJ_ebWpMcmYe+uGY`AZgm3$atN1s1{z
zSPpqdRl;uWS1n9Nt&KD!wH5b5N>lL}=!N14hbp&O<cwsbG60Fbasd)i<ta#%m6EY4
z_)RF!=_j#Uc~PkX7fC3{@9)l%HhFQkL_%tIFY&?*6vv85$}iy__$H**RYj0Ut9l`I
zu9}57T6I6fPSp#LXH~y|)V%s?NJgvoLhMzKLz=Jp2y{^Y>bse^+1RvRkc&7}{TE2O
zYOctyYlu!FsFBD~jR$gJO%&2MH78*|d>&G!n%_ek6gym8^DZQSS_xnWlmLpTf9)_6
zE~*VfE~t$`4Arh%`{I~9{~@H7wQoRDU;8$^hHP_*)U-|<tA}3L2aiH7u6q<}_<2YW
zbuU5A*S!uqNORpknPAx73`v=t!;%5}EF@KS6LN`t1LD2?9OPpAbC8SeKY$;EuR$`R
zxm4_EafqJL*Fnzfn;`G$TOkhUr(u@n(BESsFJ6F}APQhR#E8_m{vhOe{hikSImlN~
z|BAK$XNY4BrEn{h&ukmy9qIZ8`XjfNz{*(fgYy0)_ZQGy4aeBvP~&Mxnj0lmc)Dj9
z9)h2Rytn5f<X{s~OGHf?lDVc1Yb~kg;-&!PeA752fTk?ea1)Z1rk_Fwyd>X0hMP&V
zMARIDMBSW$MALj0y5KVw<!hi!E%G~n!!2AQHE)q~7;agCl%VA?h`koM&=2KYD}Q#_
z14p2ScZg@*OrC@oY8B5g(E9iMx{ZcZ2WSJLP<x@&Za)-<Fxd7M<bw7}h-2+XAfH?N
zM<EWjpM#z7MM$aJ-+?&MLH4A_I^_3X40TBA<beT+q~naW|D3h{Ipp1rcOi~CNMDW;
zc{yu|9D@!yM{_zJhBTSu%kV?+M-cl?xtOoO*$&A>);Yw4Iyp~39C1>AiNq;Q)kyX+
zCaH8?3o+oDhdAi6B6K~=8i!mjLwdpWM}EM9@L#OSKxY7AuX8c~Ugsxy{v_+Kur8+l
zZfO!Oa7*ei?7j=q5AJ6mDR)arNMYAjh@md&VoK0;6b?d3F%8)DIxMD|h0qP9Brb?E
z3g^3}bVcy%uneAuyRu9qwbZ%$Px%LAC3m2Qw8{@pPZ;ilF>C*6xQ+Fjkjr}H92fWe
z15%dW59HUqI3fGJoE1C0V~}k2&cP?(NtmVC`lyFAN1rqqFZM}BsCAzKHQa#m-q&D1
zd=-u|-43~+KLYuR`lFEd`tP>Zk{P1u|2`z*{!5S&dnl9q1baN(FW;v}QY0I<!%|3%
zq;{UCAO=0(g*fH;6{IHyq<Jtj5Q3CsU=4C^;Bydr1JA=Y_&Zobm9Q;=43ZrQZ15l)
zfVau_Peg-K6CMs;gm~=bki6)X)N`5l6r^_EXCYC0zX3_T_eE=8n$N~sN=c;N3lN9A
ze}e@`npj3HI*i2eljw0ty^JZefj)txe&)tIg=K9wF}plSmRd#imp33?peE6$Hk`KV
zwg;GLM02Y9&~YU8>8<w%byrIpH`~gxXPDiAq*yW}H<BkAWF<Ee{Fip0%`eq$kg(NR
zME-JvPefghq-Os;s;rGq!xZ{3I*k?y_R6A?tTdPmJ`bt4nm|z`<7PCSn<+JBQIJH3
z(7nioNV|$5>1r7b=p)F7P9Z+s{}^#rGx?pYG>JNb?C4>%j>I6%rS3w#D1)fEdJqMX
z`~Y2vwjvqwLnNag9m~y>dJI;eM^F!HLf0V~x8$J$lIAN|(6PDkG5(KR^DVFvby({m
zm_~kd6fGn9T4dah`25}<gVyX$cq3XsUL@mDB;CIc4R1VHyz7P=m~|r=MQ9DRqZ4Qp
zeH_VWAfH4dl5s1#2g!&35V|~8m>fWEREkD6w_o*#!tBb4l(`hk=+>b&z1QREamfGP
z-2cAZ|9(B3G7LSLT+FORQ-&T+t|k_u88ex1=wM>rWsAhjv|dRimr~JX&2+&`XHw?e
zYQ|X5W<qCThK?my3?|WxHZyw7jK}qyp;yz!;%YYTu<@#%4n{`Cry|-Hn$gp~aM%}$
z%nb5sCYI!>af^{RWG=77P3+=HDw@cg(8)#HL4SB?g!g^>gJZ$Sj83NXa4-_`PfqIL
z@v!#k312u89GV*Qh4sW#cw&6g@6vjbm~hQy3xuY0z(^RWXk1UM&c)4nJ!Z}uiL{{?
zF}0i|<^>}YHREZQZS(EQ<F>0ho$fB}?i$$J?d<H^jMX15&JHCnC-RH`XBz!!y=2a1
z^r$}=@`uO#K^=%njHcnx33Hwj8D?S`Cl@cJ$g!a#zNx@iaA<@Bk*rQG7`bhS?XW*Q
z?GGK;?+eQlJv*hv;z3dy@CW?+{ULuy&rbV?M#(}jL2L<qXw`_v;heE(CJ0uixlk^#
zjhX94V%@2!xgJffm_};OSTGYytBIw-oELsGv2MgiV}_Z*K$_!HbaCcDG-H^0XnbNO
c7z)^0|976-xHml?OPR);ks!z4ah@;!H=6njb^rhX

diff --git a/po/hr.gmo b/po/hr.gmo
index 039855adb0325a946487bb936b3102d391828c3b..70c9d00192b50617fb5c68518ae2e07b0c78f018 100644
GIT binary patch
delta 7147
zcma);3vgXkxqw$fTWrLZzHFfnR{F3F?Kw@GhDX{ILL2NeElH{^V(HE~JIOwey$}15
zH08op@KqjKi3s>Ys8(cTK%)+*qd1JB&WMOQDuP$Ih#l|Tnfd@`)VcS5|Jr+}DUZsU
z{O4P*{jY!h|61$MqhGB1*Z1q<-_M+MyQ18WEI=xgmHIFEod$j=Tc#;BA8v&ofHqtU
z{|vXoh0~SV2pzZ#-Ur_bpGf&MoJIS^wEYiw7VRl#E7hb_tY%Tk(a{digJt**cr#oH
zzXmn@J#2*k2gT6F8A|b|Hu7^m+zsWsYvEFOFU-To)Aq|yHZp%^vVjZXV%AqBD)Sh)
z6)u9u;3e<}Pz;!i`O?1zo&$Ho6|e-wkWnaxKLq(xPx7PT3s4rEJX@)YU?XgXSHf0!
zGt{iFzE7nCJ`ZJ~rE`>;3pYVA>^*P^451isCtL`>262XZ60U;JrTx?1o`hl*l>Tlg
zo{6F8IR?expTJluf2Jae8d!BIoDZ=@ErSfHHn<#Kk+w%5-c(1RgzjFr0X_-EBMn$B
zde2X}4N89ribp;S`BO*d;eXljcW9jpf0s^p8Oj1P+3vbJJObszeN2*h5kIV@j=+74
z-v{MT)w7DkemZ>c48DbOHeSFQ+4*#>R10i@V)%ggUliU>#|7{q_!RsTTnxX}n7H~`
zC=2`(a%$CE7!=p8g=^u}P{wbA&(<q-H{^n->$nm`Mco3WeJ_-aeHDsFAB(Ao0sjrx
z!6^jlBDfyD1%^;=;&o6K_z098-43})>QN}y?Mb+rg`a`4;8|EH3oVACZyl8WYRVg-
zggm~Bip1uAD2MBDD6xMIvRU;qyb#V~RWV>U)bPEKD5>kAe18X&+wyhz6t4dR9K@h8
z?w~BRU`^uD928INhNzC!)l?ejI1FVccfgJCC=`R9hhosb;0ZVbf3Jr11X4V=7D@<q
zLcFft2j%b`g`)Q{DD!^~#n3w3EN5vEJdgEN8x>8*l@O!VwUD{$n3<r|(`o-dAW>4w
zT9XB~Lov7vr@&!&CcGX_g&#`WN8l{lcS4!}6}Y}msfXcw)>nV$<lO?##?NBuXW=IJ
zQz)^U!4F|o+o0^|8Ym0Ia3(yQwm*~d7?i{H5R{Po28u@}U7YA&3fYV5g|S?V!&Kzj
zJOE|lQ&1HD70OFDxh)ybLpc*YkRcVN{SQJ-`<IY#tCye{uylQLdv-w4TY{V6O;AGi
zv-RoyuiubNXoSrSbU=9*d?-8rI9vc9hN9?KDPMweIOnt{JK705Y2N_FfFD8r)Cqnh
zba@Vz=<S8_umv0AB<44z1K)shh<*vh1Ft|%scN_+S$JK_UMLDfsNtue%=<PJL;skv
z{?a7o=RpbCN+^0;;T*UnrXsH2pLX1s@-8R}AB2~|-$L=kf{uinpr&1djqqkD3qJto
z!k@s|@cFd;3Otwgbd-q!=fe$k#J`=&LOKp{;u_(dP!xR&?q|ZUq1^im@aqz|3CjH*
zfSfDUwlz6q0o*|QZYa0t87TA5+?KqQE8$DD_d*T3wi{2!%A+C++y-Tbcc=Uz<Q%F0
zNyk_0;7+q}E0i7nc2^RTzrg9VXY5Y;8{taY7em>I4GF)x0d~Mg;Z#_Eg@h(%aVlcK
zJSaI}8I%|5D%b$au)AKVAt;CD$32M){{V5G`Wu`NFX&Bn)(s`L0hIZNp&0!6wEtl!
zXXgpn!usm(RK$Q4dy*%$6-unSpt#V1E$~*@J&FGp;A9M%ePv?kxmP7FZi3>8cctxq
zC>t1sv*0}t=c#W%F}VJ{#D6xGMO6L*Ti`0VY=7dqZYcNJgA$t0!E@jb;Mwp$q0B!4
zMd8A$le2U&l(Y11DEE8>O1k|Dlr!^mIzH<F{+C#_9!MS<3ofR86O`=tK+11Id0M{*
zC3cTP+1bzG9QX{B{=Y)btZ^?e<xYNYL_U}_6;4wga*1ijiQ<pZxZQLlKOAng6FHRj
zy%QGG`bSbqOtW$@tqj3?koP042uDV3LHdye$PI{;t;i>krL1obop?c9_C+Ks;#wl2
zDu|7U>og)IL~cVqib!bQiX1?sT#w8{?m#%3>M#-@92s>fA|a7-42jR@XAd%hNNjIH
zK7(XM+$^P*Ko#I9vMX($1>c28d?hDbjY#OEY(V57&p^t^g@|le%02|$vHA}xO~?*p
z9l`-O>2v_DK<-1Nd>N^lno?hYmm_irS0i0WJ2Hq&Lq3cwL|jD5`w$CRk2r`FIX9Py
zIY+5{NO~xn6R5jkKCNF1mmtH)M&t-0<&z22y(tB6N7kh67B~y}eA>P%<)>kDTHh<~
zxC+^pcI-|$5uc-R4f0uJDzY8vM(#(lvWUtDvIg9bT!-`_(~+6Td5DxxB{2VE6mO+*
z22x5pKMucyEJkiY-h)gmZ8YAE_{jWpU^0|<<YuHJ?dydRvK$%v&D$2&ZP?{I1CDD|
zDkFN>cf!zib>E2I;tzQxTX#9-z$xergRZGuYWYEt>7kV`c-4*hu~()w)i2#-xvm%L
zg69s|ey9&D3zqB9)0V5Z?$%bZ=-WXsHfP3jb*pzpq4xT9)vkK}NWMWW3)*zIU37fA
z5IUZ#3nPV!t&4U@+b-tEPR#6_JY|($+uYnd8n@}OJ@bB8A8+=C-HK-w+ZuFIa=AXo
zE#}gd+B-IH?(XdAVNANklB{RNpcjO$RkgL}V_TDE$%<F7Dq05e=*xwx_{oL7V+ZYB
zdv;#d*_|#T2<%Y1aa8TIec6U@n|ZQAI(0>}j@*h<b;6*@sN2O?xtte;wJ6LLy+TyA
z-LU=gtvfmy5O=fNv^NkvGGab-3#<6h^>T%Z*C*?ZmzY}7bC=`rbjUvF1fk;&=zgbS
z2eP*)uyuZk4x;{k=b#-JjVMGTL1JoTkE=^6gebI#-GCh$O~mdLt)jt{TrMtpVJ-+q
zDt3EM?>@5;Ga}AOJTinsi%w9hSR>j@H0yUIo^`6SRJA7do-E0R($DR5BTm2eDn<4l
z4q}f96q-79uE(efY@hR%!}OZxvf;GUe8&w5l?-A~GC0Bhpyr@+^s7yC*-3JeII1Jz
zAQl-zd&dnWZtXptn|5_=meXX4>@pm*QEYLv*bH+daOpIoVAv=aww%{|jSA^bXU9Fd
z&kl!eZrr%zOsvdGxtvUufsxFJspjJ~`FuXJ>eTOveZRI&Dl^vXGuG66@8C$g=dSkp
z`;(n$y_ZTD`K~sDav3-}jGND{FrQr$d7<_C7h&9FZ6=+(1^I@tbLRhIT3m?yz!?%(
z$u;3n$c4@MMlTmW9nFjpp6ngA`*J~4D05=XVIFqEL0zzXyU`d|F-j`-kX?C=aaHS}
zQ;n)RPPMVEqFv+j6^_e+@^!QB=a+`#a{tUdHAdueMZ3b&le7B6i5D*EP0yq`g*=tX
zoa_L}Vdhj+jd5jTTp>=L!(7F2%fa|Avq#ePTuyGq3dKXo6i!jrWAtn{ditzFIjYIM
zmB-aCCYt(e_F!jEZg%(e-c?%W>!_x!IM93Z9*?_GRyE41j_<SgSCwGU^TUEX%{B81
z+t)-6;o_x|pyqs?>(aT3cueP9->7p5BmUwsVHZ_z$Yz_#8<agxer)>nfqXi$Jn^9U
zyk@MaSpznovsyqPG9U2-X|j=astgCKv>g`m+O_Q>d&ivYVcF8VbB7iU`GyA7ZI>j2
z5Cus-B$jTwI`+R!k4<{pnp|_MZoa6k`J$E0YsQ|)&#b>->!BgfcRlS|H7BsV390T8
z=fyWpb<Z}p){lOnbvC!f4$<I*PLk@PB+c>EL?)$;o>;ehtfl3Jy3q?;7M{DuJ8rvj
zCnd|33r@+>kuLkuKz?jStD9xFqjB3AZz8dg*E+J`l`#RAlicXtqYqgo-5K{zOk$UH
z@7k7qGR{8VE9sh7jvleF2lHf>tw~?x{60G<d#-id^_t|uzJ3bZ+|j$Mw{yE#Q*$11
z$Vg?A+iZ51j_pZK%BvFD=tDbaU)<-*H#L$j|75&K4zxlqw9EE-0<E)+<(J6vr>3n@
zt78@y`%NdSiMeX)j`96VMOL?l95R$sA*N2*!i7QEDOzsHGN!*l#*ze_7OXX2l*V^}
zI?H$PqE{qim92{F3>syp%~_(3gy{s4ge9m%-{nG7O7;+HtV7O_XfP$03u=>0{<1pe
z#IEy5Ncb`%%sQ{mUfcI{ZI*vF#!I&EEO{O0zd3d+r=X|P%r<&XNn)>2g0|y!T((Nt
zH`+N=wNFH+IM6Jaq%p2}lE+TfblesdNkWs6t(h7!x;KRD#t*5n=~b!h406@z6_?N5
z?^GPzSFV(z<D4qVbY}X*6!yB;AajK`;9IW4Jtc103CL~*+3dRZxsyikSu}qw8wy<B
zrIHSKwND=T(5YEEv=4@>2g7Pb2Yf4MtL#-rYnzswc_1&{%g2s%be$D<+vUh7q&z_B
zQudq3i?8`c5@U8zYSwovC1=QMl>2h})L4!K?5_~(V#I&P@-7+GY@Ew4Z%B6eqko?m
zIo}nFebl_<K4-U>?el=(pX9%*<T9N$D>7ZOo8#kJs}^{bDCAU2S|qjbqNI7z<*{dr
zr^{t!pOjj#WGfRBm-$YTR{Ta;i}fPeP6-nx<g9E@M#1UQRzQ~KF{wCZTk{<9z>8z}
Y|E&)XVNLSF=qZx7a@N?>J71ji{}i66mH+?%

delta 3686
zcmY+_3v}0I9mnzC#z46QY_N?n2LIVcGRD~cyD}~tFk}-LV^CxwMJX!SMWR&FR;Hpx
z-ca)nB%Pv8<>^n+MZ2ClW|U>8h*LqDDP)}zk8#M#Uf=CU2cN@d@8|Y=p6~7X^Y8EH
zeebh*1NRgZ9Z81TUYL?3!{_1ABpJ1CXnNFdvE2AD-hi)SHc7^A!)5p-tj1F~I!VS&
zOF0uK7`LWldy_H7YjFl%j03ZDZqxD6xW{oi{vLfiei&wOHdbOEPQfi$h__=6?m^#;
zKZsTM$`A`DA;SyjBZCVsz;kd1S}XjD`;)0UPnzh!7tzH<Z$a}#tFa8@!gDb`WI)l^
zkikXIpbLslU>j%0Vm&U!X1p3R_+?y(Por}uGCnEAdiRGV6T3~k88@H{C*Ft^xC_VQ
z6IhEsPv>7lAuqN!=8G905h%U_om0FM8CZNE<so!V@ktz+Bqe9#L@aWDQsQ5DOeG7_
z@kV52$#p12B@ya*7>0;I$-mGArMwtq>(G4ZS`?Aet;n&`JCMhv`*1Qwq+|V0=={=w
z*OMf#&fW-#g|bexv1}bWyzDCL@>UiGjInea<#_TxQSQrq6CRhJhmE)i8C?DdI;Z>~
zw&05>T~innH8AC66ybpiVTgk(vZyL6I<OWmM{^bTrt3dRlKejy|HfFJGyAa^f0&LB
zBPTM)P$V-EAqG^sENQ~|s75Ncpzv3I5l6{?WgNtS%I_kND#PM&coIi5_AD3A#tFtQ
ztXhmNsuGnruxfKUzc=N#P_C<ffl^iVEUJ^L{~~9rr6pcdwJ0U613D{p7*M?(Gk7=J
zR2^0Bfa)kZucG>;BpD`z)1@UYnDH0n-Hg{zL~0_mB3Kg#kHWRcftpKj0bYp=sd)?;
zQu7oJ{6puc4i9IB#X=_fgQ2gD+`6E)7pGwy;NaRj(f-<9$iUjC(cap>qJ6WXx-)n#
z)?g2+jageU)<2)lKN;`eGn)0v&_W&S!s5C)I0BcT@YgNJk+>!uuSe;qi&ql+w<*%O
zc#|<hX8!<p;1OhK{brQ5`Ui139z~8cjEMK|;0EDHiZDVL<A#(DjbdsDLuTT8kp~U;
zqk|j1iA{JMy@G6f?{#Dsp?%p)(Ye`3t!QK)NY|ri)j;-``{RQ~)<nn}9h%f(AG)|P
z);*KP&!Gbwe}rB5XUw`=c;kIHT{N`bbS>5!Ka|e@40+s?rlvWHxYduNbcEr}0}&eD
zHSb0jHot(LN{fPu)U-tCUC^=xOE7Zp(w4|!%>N42NXrT2X{#3!McB&QFu1ipWqj|v
z*1Ir+-$45YS`X_eqPAfvEk;h;;=M{mTMs(8Z8bW)Z6k_UTX-7dJt+^M1KYx5;cq*M
zJfG8^asy_JZ^la8hg0N#&QEnpO&mjY5c^}?9-d6bcuovx?^a}f$=WY8mXf*QiG$`o
zh8&ptjP>>ScjGaF7_UZ6%#ZzkC+0^g)zJL#d>%fJH6|B0BsMOHP}kxI@Kt;mGyY@T
zi9+15A06Cr2zk~K2VkdYh1H$eq4iEhI!yRGw;6L}@uMgei@%NI@OSC_OE|!yt~B&r
zQGCMQ6^1OtTTu$S4x&rC;-{AZUC*INbcLttWC<hUIW1}SOWBW|#%g707+!>5Kq*=J
z1ePa>Kn^<AEouXi!)}+x2i=il7I$yLn7<wsP4|7+h~Gg5bf3UU_&Q2akA6h3$DY{N
zvo=Wz7}E1eV+O7`g0=WJ6oFpf#JIO6-apHGeGoq=y_eu}yarj?`(soSz2Wi8coKcM
z@;I`-kJ0g5`{H6z=zBMo;Vn2GBlPz7Me3Z__bgW6e^A<2mAgOQ@v4Y`LR%HNZ^K8@
z4Tn&QR-HmQ?H7%xh5jh6v+>H5k?XUJKY>!#AD(lp{~i>{{@v;P*D;UIqx)%Kr$15Q
z(H2mKK1$Id8evGZ-E>2AG@cA5aTN+@cpgq9aalMN?V=$h7vLW1r0sONg@GI!+J3u%
zuti9d@F-e%60K+mk>BKIT1=PGJLnGj5b<=dP4o9%6j54F?G!0lM04pq6njV0^)z_@
z7Px|D#8|tFq#*K^TuUS9ZS-l17NO1`><#^iltnwAI!L`Fo?x<;6koELPPe;t-bx!N
zQa&vFzm1~Jrb=p}4^p)E(WMlHRzPE@m3GjFDcW9AY=iB>l!19?^lwc2yHfVx`)Qyn
zomhbnru`~ZM1$=de;20xsTfafI6aWgtxM_kyrDg8v^t&n_!%n~r+vj7|L=tVN%Bs*
zWw19`#mU?C8)z#nqpRpq+C+DficXf&)$}Pkvt5%8uEXVNe-0L={mCgmf|tkrEjqVS
zbGlHEH>Le}W^3pnx|eF`W7J10>2xdAX`~AJ40)RIziDzGoo*X->Zm^%mHB_mtr=UA
zm)kdPdj8J+6DH>-6+NDp8#%FKc&@kXh5X#}(|$QNH?`r{1#haKm2J#q8`~S2X4SWT
xwWTOOx2!FjpF7Y#JU@48-etpg_LgOG+ZX*YKiA$<T#$Qe#rb1$lh-|0@IQ(#dWQf2

diff --git a/po/it.gmo b/po/it.gmo
index 6e9269b1e057d94ab6ed76cddb1e3a63c2c9c169..f70efae46a9e68deace63db79a32c14ef3010905 100644
GIT binary patch
delta 3351
zcmZA3drZ}39LMpmsUYVdD1wq`90g1+)&WydFu)65x}Zs3(v-v-nu_-e*+D}Elk%Q1
zE9WvxyUbv^)M3-rKbg+yAJvAs%*%>pwI)`!_5Pgmu(_V`J+J5ao!@yb-{<-L#y_j?
zE(=_VX}8-bbwm%Mxt-Z1JQcwQrGv*T0VA*qQ!on;<5Sp#DVQ2%=EWlCRO~`~u4`9d
zB<;<Zio39%S-={pByyn{`(R|WSvvMdAI`vDSc#h80ql&YF%2)Ges>>}F{x9?(Wr?p
z#NJqg@%S-5f?r`C<J(VEG+;8l>WgvsFfPWvSc4k)bJRpHpgO*T+31Zi^Wz8{fGg04
zN3a0Tq56#?9o?`WY9b>siScbZ6%Dund*VK1PIdw_@Edpi9xBycV?)<-Q7bRQB;1CY
zV1x5As-HhmyV*X@jH*RrS4_shqf~O-g;HcuwiekN+lsk(0@ZPgvm3uwMh2lWQH)wq
zIclXf*d0G~-=9VGd)2jnci(%uk$<J6FCSCdnB}7mNgcguFAt&i_#A45E!ZCuNS7v9
zgc@ia4#d5vy}p3T<R7U1I&qT@T~Aa-M`0E&?@s=8!yztk{;UC6q@8u`?@=?qjmp&D
z$e;CQ*-zk5%*LhI0goULgf*c0J&nNyAm`IsP)|uX>C-q}0#r0WDypL#_r-GO7StZ@
zNB-<6A1unwqxSMP>go6w2Vo-1*1(g{hqI8R*+%3Rdlz-sPonw_+@hkC-9u#{felou
z(oicYM$LF8YT!kvfvfRxd<T`0Z&Cf+Mx7NeANqYd>b}9K36vp8uuaHczz*=~<H8wK
zs;;B<IEsr@Z5ZnMOw{SGMP08)P3#he;Z+RBo2b+Nt83pw9-LsBbbl=A^0#43)cLdV
zOuLpZ%20=@1;=9wCr*353bm5=P$@f%JkoX&qwzdy@2{XHdJDDU7*4Pchv8#bit2v{
zDpOx#664!-DrC*VS)MxXgW7_@s2`MK25v`9s1cQco5)xe#!EmcO+&Q{aSYB!KQ^HH
zyMZj)qUq#G%*23?N;wtXupKp#lURa3U>@f18k~$3$Xx9*PQ$-YE1AfTHGws#GqM{C
z@iTY*E^5Ltq(>RdLLI{4Y2=@CXN$Q&UsmhBIEMV$k9;Tt9`4q_Ij9wupdOzp^kN<A
zlz-qnjk@nU*S?RsKaMmG!E`LbIT_@i$=NY3=+Iq5?zLD()qwe^jLb$KR-pzuh)QK6
z>I_^%-Txc<Fpdq^!Uke>8=fMJq}{~9(0D(i78n~~gL0_kVK!DEeOLnqGk{8^Co}ZM
zDn@m%8P(xY)cuXlW*kGidsZkjv#=BGa#TjvV`n^wdW-|dsp!>t4)uJtB@I-~=bV6E
z+MCdWJ5U4fN5-;GP#L(28t4w{Ea>&90eWBoW}-5;4E4(2im`hB->0IKe}OtAH&8E>
z_5(sI^`TDdSkwe+Q1|UZO}GJh5bPr6VGBm#AXcRe7NXiqP?@VlE#NSE^!y*Eq5;pK
zQqzRY!S0|2Nanq&9~PiGT!1mS2eow{p(bzxb^5*Rqs~kUk_;Px%*j^cQ2Z2q*o@hX
zZ=S)S7e*m!pp~eV?#FOEf*r6PBk+`Ke~n#fUqEfaP2{WvCjr3&_&h#CR1kX!m11HC
z@j6k^o{xEi4w2HVz2iM$WkhSCUt3P}YQ2D(*mJ}iJM%pWfj6o7h+)JQVm=X1D4(sR
zj>;UOOX~$^6|N&Tx^@O?IV!J)u!Tr^ZHjBJMWtd7F+`_GMaOXxu~m)#6>YG}+r$zg
zpLmU!O*~k1glD-6+AM8k5m7>@loJz!wa~x6p(^7qqBf4GA$Zt>8#otVB2@IAe1Rw+
zG+jUOU|B$8vb&IqQ;9@kC$WamfzT^*J;A}WWrPmJHe!rAoJq7L!U&ZOA%bs}>8}13
z#u2*+9Z4PAv4qNs5H<=QB6M_BHWAMTYi8|GCr;0+%5Y*SF>)jyFA^^kuMjG=gx-m*
zMbC;#B{732C2|Rs-Ne&7_eCa_Orxf!PsK~DB+`i0gi19rp3sw-M5wGHHWL$w?L=!?
wOl1U-MZ^%JiTe1CzeLn$#n1C}%<ShM;PYn>@n@&|GwToau8yo<;SY=a4^6y1!~g&Q

delta 3964
zcmbuCYj9ju9l%eZQo0E=@1}X}Nl4N>cDKo9pUGlf(j;A)X47P2n+UYG$xU+G&2GB8
zp%1|Yq*#k!X)IW9Myog@j*rpKI1KcYRvnZP9d(ouVVFTjbd)N}I3nQhzkB2G$uFMC
z|NhQB_niOx>`dOuzI{A@enaKUsyLnijY@6!JN&s)o1QN&JL+JAQjV?ge&~l@_zc_)
z{|wuex-kq(;Q^aRU={Dj?Dq#@8Sjt5F8CzOyP5oii4*=Cb||$OCsYp{f*Ph_3p@+?
zw)xwz0=@vd;j55y<tSHeu;1o>h~wpV!&Z0!5={AJ*bIM(eYJy$M2x`-U-3i5K}fU}
zdDsc%AciZx3vsmK6^P=B_h3k=O6IB;CLr0VJOwp;7DnMq5Q9~S+5&HeI8rr@eRUI)
z6bl%vItQEJ*I+sP5p==d6xXY-horiiXl1>63^K1i0nt<a1&D*yS8Top(NnVlZh|$i
z8nQ>#LLc_kHYU66g)AhsHD@5Dsrej?KyidawHs}6Mlw=63W>gU9uiUQJ&-7CC1dsQ
zX(-?6C(&E`hEgRSl2B4yugjA**;pr$kea$hHq1bAtc;|*2;1QAA+@eAg+y8(fYiBu
z7UF3A!w^07mm%Nke+8*|!;O%PHuORCHN+v!*KiwjQU8Vqn0Q&ZWN#2d9BOzIlCH*U
zitmj?ClNGC<Va%}VzAMG^iAVkFa*B^DO2Mgp#zE@3^%?D37|;=*bF6rQtIC{!URK2
zlMn+<OAtj(C+&4{Oujz{sb$kEkkmI_g<HusMx>_A;#doez#zN>Vz~KHsNt6&K{Q{1
zoNxXu>><s~|6+n-X9px@P7X^3oU@QrIW340=NiO&=S7HN=hq>Ioj-)z;Y*N=XpD*;
zEe_E$`c}w!eHgM&ABQ-gzXJ0#hyFej+4w5d1W^LJAxfmiEqfr}Tkf^jFG6lX%k%d7
z-yn{)R>JF`T(gakJ<|29^he=c0;}SE2bBF&*e{_0TMw|Xg}yoqAA;f;huWk@d!ckU
z(X@RB5=q<3kc_px4l&v$8KkD|e3RO@({K_%yJVDz+tDF`wMzgV_(h2R_Mgf9C(8Es
zSfHyrBx1bpke)zshY2Y`hop=H9dghGe+qG^;~nT^Dv@%2+b|@f+oWXjehS9mQ;<Nn
zy$R9R$&Hd;=^V$t`XG}PsNpxEY!CxD(kYQ0WQlsqFJ;$BxF0?PQPNF1)om~YalZRJ
z#DVUoA!oXO0STu2FU9p9Kg8i4$w;~M-%%z#{M`phThBLOCzRUC`adA4bm6UJz%>Rj
z>=H+a(Dix9?Qne$Qd8H9unzvt=DU!6Zptg`?m@_YcLEMjfA?V~aTdM|iQFv)`zaG%
ziUS^+L=1ZJkc@bqgc^!L6nb8RIOma4&;#2!S3()KQ+GMny9zaY91>vfvwS}d#nCeA
z?>)o>!`?I`Qtt(bm)<KdRPY6_V7U0be-V<>{&Ua^MG@zHWf1X6)3V<;WK%L6<6SaE
zhJ3F>Qty*MZz3xHbxbNC#^pZyBM>EisS)?okGEpTFBua48lotG$ZZHnYB3yWhkcMe
zl8L}P#84my>50H8i2lGu7=?11@}w~EPbMe{;)HAn5{;xjNMlN~1f@jv@C@WzP#ob~
z@CAqiA%3i6UuY}D;gDnsXF?x_JK$ZAJ)z668~)gSe+&DfF#G`)h#)LALt!`qF&G|&
zI2D!>`JhCCf$&pM&iw$QFnkqO(oaMwr5o@-92htZacbZ!q<03s3hAwZ=b#_HBlrJi
z&afy=GI$hH^TBfvX9mSn3=O^waV&yT`6Y|=LVjo>?3G`*2<a@m?}oL!Cm<z=9DyZ>
zWJU%Js3a@nh-Bpg=`W0Fo@>ab5XpwM_Sa2SUE=@xkfcg4x2!KfTv2jAKE394G;~Rn
z8o32|5E0{|U&^{Zo?x=LcGOXoKgH~0$OHB#7sTblAS;ED;u%C{i?8aVkZ{#w$bh}p
z2lpcn+3)WMH7tzBd1jCgA@?H-1bc01Mc!mCgAz!6)C6K6GNcsy3Nxj~YzorIUgQiS
zb*3z836UO=(TaQoi6ZwPQmXflBrh|?g+lYm&!=>+`V4Xs5rv(|eMkV2rlZ#C5hRAp
zA=e^uZDoAC2<aPjz@C2=)*$B*KO%i|D<b30BGjlf-!;5Q^U3%G5--g7Cjin&at#YZ
zeoV5+Fmea7jNFaLco=Cb%!-dedwwrGh|D8$qhvga>_K)QBWsV8ZMpq+W<EqlDRK<y
zMou8RkxwF%2=`8Y$P42*&(nysxm=a?vBG30;zcTvU6-!E;ZK|LD<?A6;!;j`4t43i
zaM<sXXQ1#57M>wJk}*v^m0rjlGcu+gNw22njhvNExpXWw=W!&KtgK$iq!%;BvSvDO
zWpf$p&}z<{*H%jBmQ1~rUNM;%Ic??iF)NwWhfKYiH5XR%Ntc7IdS5KDD?Xjj(eW9*
zFFH9H9Z$>*vNgAq=Bs(9SyW^#uOuz>;z`Cx<xc4If@9C{<j^kmM{kLZ#S$|*ozWw)
z#Q5;klpcvs>ZqQGP9|bQ(__&|Juy8w5uY0NXgx(txaM(;j!)}RGi7Fsq@Gwkl(gpb
zm^EjnvZh`@)pDAc=gpj9C9@vKrK`0^95?j3z5UwTKhWoM_XaO5wVW=?52gSAMDATR
zv$^zoDwhrYhyUi%Vm)OhlZMW&;A!flq36i!N;;c0w3*H3O}t#qSxM97;NaBo_-s^1
z$A+fIqk1$-HfLkeL2abuP}Y>=s9H2clV#{aR_Cmofj!4ie2-)_R!nE~piZ-@qdSL3
zr=yACS<$;wPmkMKjg!@3JvBNMpNM<(l%?m=sqAVpXE`wNKT=B+df8Z9w9*+%(wj3D
MlNR}X!@aTWKft~T)Bpeg

diff --git a/po/no.gmo b/po/no.gmo
index a150d4e0a471c15387b7223f2ee97e9c5ed0e9d5..c54d1a07428ccc875caa45f6e66ec91274630608 100644
GIT binary patch
delta 3351
zcmZA3drVe!9LMpmiO5Y*Krl&p1S}r#@(7rSR}?SU1vN?yQ&UOE3#M3#+GFMo)6|qy
zv}P`&Y5y3soR+v;&MnKzvRNzJq+-)*xtgr0>-~Ao&&_ql_q=}Rch2*>e1E@l^wQD|
zRl%R)!|IHqp16;=7G`z<KZ@kU5fNpUf|0lmdt(kB#)t4n?2Z0dvuG@FmSF<d^W61H
zjNy7C`f(TbH4EBF3aM0F!}~EN&MX^q(T6j!7p_4K@E~@<X3WB`P`|s08JOPLb3AI`
z3o#9Au^YaJJ@Hc<LI2i5K^<n$s%}igd$9^Lu@=?w`>27QM>YH>=3{icSpbW0ATC26
z9>GF9hiW&LbR=V6)Ii2yI{n)W3hHnz_Q3tfnCv+A!OvWM8!FXFiC+Cs)XXa}9k-(f
z*yQ{M)z06j)$GvKjH1P15@ulV0SZH0#T;Z(R)egKZNZ^<9M$j*XEMK5Mh2lWQHq+;
zV$@7)u{*x&?w>}rd)Zz8<?cr%lYgZolhd?zW+PFXq@LEamJO&iK8Kp&4a~(9(xm~G
zpgLNO58@uwTAxQ{@;a)$&OD?|*8`Q&ahQYE-O0ZmXrO}qXHCc??X<hTgc^A(DpUU=
zf0o9y^KcmE;}VR(BghM3O{jL8F*E^Wf7%VyThfvAsh@-(1$E#@H8jNCsCI5ft>FRW
z&yI3pQuYOEEn87<$1NO$sZ3iPPeC8fMUrOgkw@%J)Mh_{YB%^B1*NPFm4Or%P^rp7
z&7>4H;@PN<7oj@dg#GaiR7O5WwbP2)E76?v`)t&6gHZ#hM3P_|kiVcE<m{v36I80M
zpw>8+N{TiLRX-cG`?sR%8&LzhfZ=!<JL0dX-TsHWZbM$2P@437BI(L(N1xbJ?|(7F
zuH{Ar_QUJ=h@NN1DHBzwAMSM?LS^cxyFP`Q={eLuzelyxfgP-Y=Ai~&ifXS0>D%6x
z^lxV<kbP@Ktw}gftAS+fgJV$*EagPowimUgAEQ!x5w$eod<@9G`LF;-Ab(cFN%!AE
zrT&Dg{|<wyxJ^ML?amfg0|i)!(@-<ohr{p;(wBwtSsaI1$e3*jY681a9ese>%;!-9
zjbPeZ`(CKcJ{&c%iGK30AFifC4efI`PN6#b-I>5B)X-q(WYn6@NA32YtKZ^0gzDfV
z>icpPmD%>$Ud9q}0@p>^<X;_cqe3=0FQGDW8@1be(wZ`ri<)5xs=gBWyxWVY0q#Ie
z=rHn;v$MFI&2|}MxSq<oGAAoXl4IL~6zIc_;b?4e*O@G=IxNM3xEQ%*?;xvX-=Z44
ziA>HS1K#_ehkAP!U=$ue4e$sm6DKhiFS_gCFBCL@Tc{cP2Y8#O*trfPsBcC!cox<1
z1=PTP#%N5+@#@o1Gw+Z3!i`0Jn5JVq?m`Wu!MhGxGX<^n*Qm|chWel+GYzFE2M1#r
zs^dD;^9NDSokgYm3Ni<4m*=f@4(c@<jhb;e>UY(scIq)s?|%aYt=)&1fL|h6wHDOt
zmdG^qWy?n&mZ4_63iaG3cmEh_m;ZnoP#f~XSQ7s-Y0ZaYAFM##ufqWSTO$SiYu1AL
zK|J}2!~zV%BJ7C8u6~lU3^m|#)Qnajdo8p<A=Js^sE^r7Vh^FClz5GJg(&R6se;e|
zh7;O!+F+H$okQubCX~)QhXyp6cy;H<sFdI~N<Lx~v6*<5P)@rLcaC}r&k!2&-N!mE
zRuk)8rE(PK$}c(>BD1q;?z#q*qSuK6{VSq_ooG{tExP#Mp`D^*FYz2Pl6aXYC+<Ev
zb1~Od3`6b65@HsiV=?hWsO0_k*DKQhB1#j9T0$?X-j#X8(}WJ~peKn!Lc`Toz5AHY
z#Z*_}$1)<7*h#D))Sh04mk9QWEhY3`ZYPS>;B2Bj5l-k>>k;}8&v514*p=8tuscKV
z!vxf^%p>$+zlU<JyWW70y0Ug-5|Qn$N8=J=%ot8j5etY_gpRF5MyTYShPd(?oJq_f
zh7vmJh=+IXk4c?1of0oz=+ImCJds5#57qFt;Uq$DTDq%y0XGtl5j$L+s3M98z9^w@
q&v<O?*6C_wV@|h<s7?d=1_t^9`2~Ue?7)D=hBSY4V@tn(qW%LDFgN}H

delta 3914
zcmbu9Yiyg<8OKkcrHvD4b8SeR<fP5biR~t~<HWtFxT%x6O_R)aTGF+&Zu2I28^;c|
zLmQx&Tgz=LErCLT4?{y@8UqCV0<_ztjR|e5VA8ZP1{-5z7i~gggH2imbic=bi?nIq
zcBH?5&wJi;&T~1>>*J}nGr32Wl>b^4#uLD<)RIr&Pn9aZP?)yWz&fRD*TTDD5A?uS
z;db~TY*Ff(04#y~EY83Ro)25k55Q8MAB7J1G|ag$-oR*tpTbt9RuF{hg#A#%6l{Vg
zA@5ea1k2!?umipi*|(CzRWtNjydDzx%A;^KJOfE)<$2fue};dx2}3f*VVSS+qHI4T
z+p-*Nhq4jJW#5AYTJ|2~;Id1wPpNWj)dgda>XaXY8lHne_%6g@1yZZvMo1tPWB6BV
zFp^B*u;L+D55EOh!XH5=e5Wv9xg1jJO0t#t${~ndxd1t*@@tR)D_^tt0py&jC9oJ)
z!Ai&)RSkXkS8W*Et%(ezv{m;*uBPfqxB*HaY^q*nkv&q8>On~M)f146s&9v6SuGW-
zg)c*SPd~}I)qhp0$W0N73iCBN$|ehIBoj(gGsS{&D1ntyl(%6s`~&2!YfB)R*7_m$
zT)P_*XzdBeIko2@?`q$I+<DzKkc!rIL(Z!kfiz#;9_Zx$>yBf1m^f=K5JzmP`xsI#
z`xS*}JK0GFcF7#s0}zMyIHYgvx57U7BIKIve}*<F=iu1>IV6F4Nniz(1WLI7`T-0a
z)kh!>>Zc(G)!%H*OJMT;hmgChe-Bdn`j6nXR2wI9rwtNV6WjoM;SCVS4X2@oUxy^o
z@ET-)!>?c`Wp22L!NHBKkdigBSt`)D8&ax96XK+C5fXjlQxM0E&p{kFz7E&IcOezg
zIF)m>1Vqp1ZIJ!?5M-S`0trBW1LkNB{R0eH_&(GmQ3N|6hsYf_?SQ;*y2F}(3i1^+
zU9jf=4hd{^IlKzWXSNKoM!J4A{gHo`#432+1Z8~`|3&Q>`<P(UntS1W@EPc)@1$vP
z-n<`DiRR}a8=HS&@dHSun&m2F{t_h97HmnNg(}H8En6T7v}7PbwJbu;Yk5V!|0ayz
zF#%hbLb|n8O2?J7h9C#D?u1Tw401s0OOUUiRRZBoTR(-AwoSfooVLk65^9sg*1_8#
z*W31r_5L@I%C}v_zr1L#E=;s<ghbjt4mqHG0S4h?kc`@Y30dP<3JJu~4+o(M`Ai+s
z{UqS{IV2FrC$NvLR7nEuV6EJJ$9_n&cjS&@kXgrrkc}PZAqRDcBbmPhany;Or+g(G
z)Op0>Daf66eiN3$?^yFcwkTJ`-p<e9Dom1)$~regDwcE71Zo=-cR?IG->~>`fi9d$
zMO+&p-R{~8DV-|~$<TGenwM^-0bSqb4~&3ZKjE3ZZV8|seg>)XTJ{zZ{8|ssrQCnl
zvlui(*UOM*?D{S2RH|E=i1Kzz*UIxPkj>rCz#jO0NN;ri3vz`X_DTgjBakdTdmwu~
z_dqZFF0?^Oj6gk~!&N!naSzhPUV@MkdHWy^y`zxKy|+S|$or(lx8QRA^f*Kg^bwRe
z_7Q{x=EIqM4L<1!nV*Mb?z<DxpuRJZz<e*j9EbS+h=F6@MaXlH9L$~fAZfy$?XVUe
zf@Ilq3<luy5C=Vfg0lV}kad0uh?4vLunWc@ckY*cd}aRAkc9nIJSQ9duQS0x{tw{_
zD1p=vB-u%G^{#<5OD|E&XV*Ig`E%-h0BR_Kka6#uP}aQzdEdvJbbFs05>THc(gvk6
zRJ-q#eE&|2bJhaMgop!ILH@M@;*bpi0~W)t!X@xgxD3jb$o$h5B@hA*NH394-~uc{
zG-`ejgxZOQkzDH;vdoV$eizYK&{AYB{;jyeAqBrYD7ccFq<~)xLQqlk;G)Y`=QxR_
z(Pku-nL%>C^w;I_1jf|jjkb!MT*htaxb=!MNbU23s^mwKUn#-v!lSwclCF~U*I9Gj
z@OmVr`R_q9<j3RurqOliZZt`<SC*9Jq^vSFA{tGNqBxRqBifa>lp3-qNTDyG`;pwM
z^vE<ifMjs*>dPpIZbw|_e~biA<As@=G=bWK_{`M9=w>7b(&g$-BpsYZ%h4$mMhDTA
zNP0xZT?NSBtlDSUk3l{)^$6-gEod8(aZ>^OZBynec<@>i_rMX$rpe@+mH&bK=;t?s
zLg)rGi;f~0Cs1?VD*UXs>^tCoG=VlC8K=<>v<VF?o+@3ncQ2L?$tXdGQ3qN;+tGa}
zg6wDwl5qsxi{!(82wfg?7+a7Bm7}d^mtXyt;@sRq+MJrs>h^wz?hXWcobv0>|Mup8
z`}9ECFmyCEnLQj&8+sr$pPYzi%~aB<!^wkgTWs3Q=(%)iDjlEIm=k6un>G*3XN?JM
zCUth&(9@|o10$Z*W>z0I6A695(DNB%az2-E+E}W0g=1Ss#$q}+Jg#>IBaz^6Y<xXS
zv(qWw8aEk*gUs2vgvq%?l8z^{3pzDv+YyTNZ)JUOb9g8m8`r6{9tg*VL(!-n7>Veh
z9t}of;r_9qU__6OMMg)WA-C31aw0UhZE$!@4;o1$9Z%@d`2z{_pdK<08p({ICpl_1
zMa~mOHf|;|Zrj<9s%LChcey-X?eVVb_PM%x&rUZvN^<?F|G$u1jBM7pT*+A@`$Z*<
zw2{%Kp8oxuPR(knNIaTI(|U3~nZ!=!7P47$%4wUWPGhmqupS=aXD}#r42L4?wK+3m
zXkJdmlT-P;qQRlT&B4glP)r+%gpsy|28Tm2d5ndJ3frTjp=dZXG^9sHM+bvDLvCBt
c(9izQdR65Mb}oOi6f1GaNK>to&VSne4L-s6F#rGn

diff --git a/po/pl.gmo b/po/pl.gmo
index 6308a0e4e7a85c8b8924b537edfd8514accaae91..96e50064dfffcf667e874c24a3eaf1f07d41f801 100644
GIT binary patch
delta 3351
zcmY+`drZ}39LMo5g&-F}K><PI<iQKKIBF*14KGX$6)Dunyx|SayJ&emYF>hbR5XpK
zIcw9UZq(r~vvfLFE6O$P4_9WJuC$!yWofQl@6S08oBhW3yq@24&Ur51=leT)bM>Pt
z|IZ2GdyT7s=tA5GH#?6fqIltI8*P?`QMd)WV;}q&i}4b6$ILjh7%Xv4!9?!ox%;&k
z%l&rD#66g6=C?0sq|<Q&dthw5SvK}X4^GFYaU*Jh2eAX5#4P+C^}T;E1HJ7-jz%qf
z5q86BOu?i06rRPQ%x_m|Xu=Fe)f*GAGcLhiSdE(aQ`ADwp$5K%`52R6=EITL4_BfG
zkKizDL5&+nI+8ILwU8q8GQUlup$Ru*S8PPqWXCZFzj6H!P^nJp80sH{+WCC+;!e~8
zo1GU><J?1?W}8lCG%X&JFa!NPX$*87vye^M8suzj2M)sHsDbZ0llitX(jS$HGSrTi
zqIO!1sd(7Ee+D(~Wp{tuy&s)S{*{tmyp)HT6{3oyfzfo9hfrtSg4*GI?2BonOA9PP
zO|%}L!~LkUK8MQW9n^U3`H+gPD=MR-un$(Hl7D^R5FON?H6xp}Gw%K(YUS5ZnR<x)
zSvR)*EDpweT#jw=2=YK!GiuzE7~B9-pLQShlthv~&6DV-p$Rfk0}XU<R5@!<XLtbl
zvrl+oQ}!L|EU%%Sj(@Q~rn7BLJQ+PW8%dgNMLuF5po;x7)VTiNXeeb5P#H+$0F|mN
z)K1D!E3QCIycjj{o0y01qcZX>YMg7RT8ZIB-_J&UZUAZl^N}RjHssH52YL0-aT=AX
zE2uM$qm!l$NA*{rs=pT1A3!baJVxMUjKp72)qd06KR_OwV4C##j-)Fqj5)E3p8sN&
z{T^>rq9zWf(k5a)>PYHPDg6@T@d7FXSKa;JsLVxCvHE-p4#HAo4y(f)Y{K4n3ALek
zc9YHgRzO2Lnu|J<ZK&$rhXr^R^_=TPN7EB|f~*vI%PNtqTb=V5j_1Aw2V)wQt7<Pt
z&9ejbv>Zjh&h82g7G;l61A6({n}lOfJ38RJ=uBW0Mf4Y=2HJ&M@K>k_{z4UF65G^S
zk4N<{KxME7naUcn$Uk*wr|IzEFYb*nN>r)KM18OrHNXmIJ?cmT7=x!!#npnE@H*=8
zY(w6anF7>PP>zMT9Oq&n#~=EQPi6Vq`3NLyHXk+NPE^%5p%(Hrs+jH|$*_oLLPwT|
zT5uuiDX4I6#nLc-2{D%bhk49_k-pGT7W-)^RVz_Na}0G>zoU+(D;1+N%*O$^3fY`B
zp$7a38QZR-iZ_KX)3mV|hZ|5Cu0<8?5$7pXru;w9P?6n659Tn6s(&1g#SbwG|HK%4
zh?+Qp{IVHK#P&D>)nA5M_)OF<+&V18-LC&U^1#^5(7oRh`h`-R>l}wV%SA{XSq%=w
zX4C}tP$>?h@^myA=)obV3{FQCStI7)QFnh4gPFo;`cp`I2R;8iXe7~*k1CQ9)P$9o
zjE9lr*eTQ*-$f13iARIAm=`@*j(UF$s&*Pt-#dZ)*)3jl_Psbq4=%(!=C@iJ8t8L%
z;9XRzy_}0C%5xT=`iq>Un8^K9RA!c87<MLf>D(t0<A}AyenM9nv72~{7}kcDN`iv6
zA%uz|_yl0<#d@rY(C%8VNE$B?b@heOY5uoqd5Gae4Y7boAvzGPR|5?_bXxM`*A{Np
z6I)%U4owHBYeNWIgrwcd-TfL=RlY;8Yh^K1_mgo)@b3RFsqXe&Tt*ZUZxC~c$1j!g
zY}cXZPQM-{#0)~$QeslD75eXQsLA|`X-y!i36+@+ZXPj{(4~@nnHWZBxjy3Y>s4-E
zavduBDMUI^Ppl?1o_=vQ5md0PAlec;i7^_mf_RdMAardG5iGrFuDuUC5qk(#oPP42
zCv>e0VWaQ~qOZH(hT~m39QCMVyL*-Ma-yh+m#M@;VjZEYme9W{t=B-;*of1KS%e-~
zU3-bK^^LLVGhU&^Pebrhv9BVsh}VNX{J#TBi7tfK^}UAMi5H1ou1_o>MiP1o6Nu47
nAf??OQGq@wbE4bz&h_>4`0@*U`PshSfseZV7#nEv4T=2^v|Bl%

delta 3943
zcmai!YjB)Z8OPt0O51H{o7*-`Hpxkwdv-VX%O=Zgx5*|=o9>3}hL8xfZ<063rkmZc
zmxhhtLb=r<P>7Ubag<R<bktEY&Nx!X(ZO-5AS&a9q1fq+@&OPK^$R25@7aBej(+gW
z{P%axdEfUu=lMV9?1#y>)0xLN7yePpwe!HHX`BBJU)Hp3FXW~zrLbJnEPLR+unV@s
zAHc)#FR)hAcDUenc+})^Sj2eM96t#289xCV;a6bBLFbors^C9iou=(%6IwGIfC|Q7
zH9P}(w)0t70AGSl@CxL+T`aEEzz&l)K{meYPPi972}x$x1y~7xOMJ~kM=~a0L1*qk
z!7)g-1sT`?<wF7&`~<Smg3FMF3qFFqnpQ|(YlR_*JB9Z_1)qancnK1)2x+_F^^lDe
zO%Pw(M<>bz0gE1k74VyI7yLQ2!&h_j#akg(7n7~b7mq^v#c9Yo#b1JKu=shCe}b%2
zvKelJC9oLsj%J0O#MkQS95yEs5Nk_LLrPQf6m&z`2p?IunB*IA#5xSg-Z}@#$a)tf
zORG3m2EPyGIp;~%wZ5Zic}`5o%gvW&uuWbpl}xavbb%M9plmE3Q(l2J@O4P7%eF%@
zE$fEVxojG;(Xxjj>y%x9JS+P>q~_&2AdZ%|LDnk|Kn`E|&CpK$%kQVt&cufKf&}8D
z^1ngsvTe$ZZDc1I*d%jgb3p>ybjZ1}-2r>ycOhl6y#Xyy)*-O%Lr4M@lE6+V32dkS
z6+?6gR51<-P_YPEsAA2WmyOBui;!AYT!vU*@h;qh+XRuCR?5bzp&RzVTOolf&q4*i
z3`wH$dC2#bufb+)u6&;k3s==aEUV&UaiD4%VpUZb5~OM!viqv1A%UyD1qodBGuQ$z
zK^##8m35SCh%=)OLB3Z@koVN>kPWD3V1~n?eo9APyaE+T<iRG$B2we(Bar9ScboH1
zLvBI!3+DWLkd5svgx5m3W?LZd$f@7U`N)nXu_DGEDDMY}pGTm*N10$p`%c62@N1C3
zH3M)0u0l#u^CBd>T7pOoYgtSjsO3Or$GjuX)gFZOYbT)_o`Ka+zPH17AP2I}CikBW
z)cKichG|Hqbzg;)q)wb+_jSL6eVjYF$CRmFGN3f|^Uw+<0NK{d#>u{30tw%OtldDi
z8h<(sQc@0k!#PNthF=h0u1mw4klHnh<y^N$UX}$LXW)(S6eOd@%Q-fcn!Ew}$U&TA
zou*$wHr$N;5}>&YVtX?#%Kt#~L+1R~Ar3ZYF4C!|^9tmi+NJgi^0Ca@_d~3+%dL>H
z1mc<fS(Cqkl*IlXBuo28kWxA%V*)xFAlK6|2yw(A4zqcOIIy4kJLJA%qeB+pW#<$m
zbLUBj)y`)i0i7?wJy2@LMp`&na+q4EqkPwb3sSO{D1TtBFvFOQw|)oWYO8Ei_z#}r
zTAMhSPyO4M=y1^5<&aU+_U9m$wf`A%csg)JYStmOqeLBIA?4|ijk)0Ku$d$k#EH&P
z=!B98hq&_;WRsm2;12jUWX;YGU@^Pw!X`PzUF=#c>yk|Rpbiy$3R1tWA3_%ImQzpm
z-Nzs{cS`~U?mhuq;5`s0x;M;u`NLu3-M@w<@GZEX95=xX4|?R9afNzrg)%+~vApLC
zCM7U6?RgF2P|y3&!+GSOO5ol(h_$`9K}yv7IjG>bArAIR>2_ik@5=Rau~v3GLVU5x
zC01FWSXBVeLMs%DsgvtRkbtf~z}=W5wWB0`;utmVn}vMecOUG6>rlaqP@ca5NvK~o
zz<2#@LT*$4QMvzQ-+zt?1usD??Ee6=kejO{W48prYPT#*Aomwde#4yqp2;6W%Hw_!
z63l%C<{@0kwhDf4M1FJ!dQvixHi2aE&!SJD@cLi36*Vr=yV_iIQmFds1l)px=#h1Y
zrM&SBJ%tXS2hnjP>*MIh?L57O_1i2(8S(1V=zjB=ctVk~4Og-)$}d+zo6ZfjPeD$e
z_5|uP=i1;+NUp_yHx5&_eV*Sq+K=u<-1qfswijd!`qHjP{5fc2NJrA-FimEA*;8-!
zG%ZH&26P&6cBCBIBAP+cIEdOOkr&;C<c$1hi!hM$M$5=))^0`|4oyz&8j=MY&`H#d
zl4vV>6meoTISbdIYmu~ja*zY19X0#sU<rC0b)i~x2uVAUgElOOZxe$KbK(IQF#DS!
zH$eVavTcCh1o9!drpxF~B<*2TlkMex`py2`@EDpyZY1q2I)XfCX#LUr-M8FAuM<h*
zh-<5;38m3t^cggcY-k^nb~`$SLg+DcwXM(@MBFj05FOmudfl7bGArqLcwsTA8U`9w
zo6FT@m*4K}Z%_8OR}IAtLj_~=$yGgWsG(RYI;SVYv8Y{*L}#6r&|)~DR^qXRxW26D
z&V>`ncz7n2H0D$|s*;O_T8yn2bo8VOC)H{=5>YdTN+pc>R3>7#@T!^|2^|bfgp}7m
zr6#@O<6eJgs-IVri!q)WCyd-e;pLS`n047nT#qKxDmHI9;u{}0$ot*{BcmgsDHV&W
zp^=c^7YwSQz_{|NG4FV2WME>{JFdni#>WCdpHrzIIk7dTW!OKVhK;BZ*CT2yH4_QX
zs?qSQ5lt9so~4#!<UD62^>8HNv~0X<U9?=+>S*s!?Hzq>osQO?4Ws(i{LENfU)7bK
zO2*=e<6+}tJC#HH{J$QKnf4{gGag&R%~gGFO^>G4l95QNH5FwTHkLeg<uTNZM+Pe8
zNhgijwP@OI$><S79q>&nk9W}TJEVs3(CZ&KG}5o4yt$^YhDmVAGixld)T)tG3s|T}
zXP3ejyc--1O!8GEHKV4z!B8L=nDqKbylOC@g3|-jzTuZg{l4LV8t{!yIMsA)MxR|%
nbFqJW*Zx1|*_A$%R;jgg;<)aiklE1TD3&hkM8!&v)06)%HgyLD

diff --git a/po/pt_BR.gmo b/po/pt_BR.gmo
index 917f90f80ab6065d6f55e85c64521cbc2ac19391..447bdf177aeae02ede9899fc377f3f9dd017e340 100644
GIT binary patch
delta 3248
zcmX}tdrVe!9LMo5i5DJ-n}{Nbhg1k3qz9D(K}?d+6mOY@kyK)lnkk}b8cz}ll9#+-
zMz^h^=GM}rWh=|98B^EjkCtnNX_?s{P1k0#Wv%z;JU{!*?|J>sc@DqJ_k7Rs+{!b{
z!l#p?+l{uHxPdqvZFUT+V>r-O#F?ezN^HihI2MxwX4A18N8t{P#Y4_7a3JSB?)+!u
zPwgV+;8h%B7PhQ-v*BErhgsNwL3{~4?8Z#IfLdT6!7K^0aWod;NUU{kMlJYVyb({}
z5d0l8uwSCtBpihq%x{b7q;a7cZ^Bnm6Lq5&b`CXQG_w|>has%O0&GJMKf}p*7B$g`
zWV69ohFZV^9D%D)^Syw>nBU%`!yMLwxp+oDV1WE7wL#SNIj9UZq6X?jE%3PWH`F*W
zDFhBh{@F+lsaS;BI2&1`t;I0Uz}o1P;D@LIFFOYfie#bym4Ugaoh(P~umy+WTkiLx
zsDXRk`9=48EVC;USy&ZiR*I^Hy@ScWo?R!u>RFyZ?d&p+!!%~o0xMAyt;bt%59--|
zh04?=)PV8iRmGK#%2+v$#d_55ok$tkL1b-qG>!c0!YM9j<>yf;x`LX}<1ikJu@G0_
z0NjTZupLAV{0aISK<dmcquz$Tq)ihgp~lZfjZ@^V*N5q-*jkY}trN+<9YIz7H+U=l
ziJB;nn>{Q>Hf`0YzhBKkZ_Q3r4Sb5qP!DQ@f1olGMY^=`AZmf(B08G595rzb-i!^%
zChSerKu1u2_>=qn5;CTF6as6riKyDCLJ!xWGSY#1l*ioff4S>v+)QzYZ3-Rz;Svs7
zQ5*KdJ=h;#K~;07JKv8Kx*bB@e-cY6s#B;1rINO7I1UT&bF9KEsM?y#ZuEC+Fj4P+
z6CDk>%lWpu;bT-SbYl=NqKD~>qVp0|rs_}w@8zIkK8&oz&Y~Vg6c_U_8TI;3M%`D(
zA%ppCD;+(XcTkVwThxG;umV%alYXy8)?k~k1oxqKd>%dQ!{|H}8;R4f3@6|gRMB;#
zhuYOd$iy&@!)oX-mo=g$cmuVNFOfIV&Y+(06;u_c@R_7*S*ZJFq841|u5Uq&_dbU3
zJ0$xSU>7nM^_I-cA^++u<btYsIcmZd<e$C9L67D*Duul`4dbXRJ%ahD??V&nH9dgL
zWxt?`H=dic!5q{ErlK-bi+p6QDM<dc(snL*_z@}t-(ymg*}tgQE|FQ7+lHZbT7g>7
zQdE(3Ah+7rs0puPDW-E%E-pYVs0B6N>&Pv(KkRPk#XGnVprQge2etDW<ZEfGFag_8
zACwN%zz0yp){FWw{)2k9<H>sr)?+MgKuz3=T5vlm1L2c&2GBW;djEe#J>#pWf#UNc
zJDP#o`BL<7ySsh>weub<!oN`y74Y3vah9QKqz0AQb*P=cj7fU`Ptc(-?FUT2KI0-E
zj)54@c`0hgm8cXhLS>{8`Dg7MG|_R?qdbocY_SxCs(u*i`h3*0Uyr)~Ma<Xx{{bEC
z-~v`)3i(k=tC2O?1|+MtAE_(*5tY&a`R<RY&J0w_vrrig;UFx+M4W|`l+_WF2(@RK
z->>*o)bn3MEFzf0PuElUD50kGM-$45cA&P4P{q$7W)Rof?R5GOJBXdcGGYtyIH5L|
zSWATO<G{lFwvwZ&mkj!?jN?o~m8_krt%~4(QYn%Dn;xCg+d|AEDv3>mT4n_QKeA%_
z&$x548un8>n+vLX6^2@C1Z#7OZN$CAG@^!3=B^{uvWaHmPMy$Z5le}B;sHV}pU~Iv
zL7mVV{G&#psX|qRYIhM26YB{Tj#@d9NAx9f2{k?I9AdgoXj=&toJTxKJV&T$V`>{C
z_-~oI1Bu<5e>Cw3kxX1`{phHO77}``77)~@-<EKE!u7>)qMBH)6Pk)m?SA5UVx3NC
zNyOd63Z2ka6aM~H%_W4Ybv_X!HWO-%#6txC$M~Pn4LFf#BBl_x5l<5p#685dHkr<1
z;xS@~yBO!(?&Nijv8kNY5?!rv8)Lc(2M-TipFbv4;Drh&gbIV9{I1%}wXt1kc_#z^
E0TWv$-T(jq

delta 3894
zcmbuBdu*Fm8HZ1|?ou~2T@sVViQ|)76DO_{$FbwCYdbGZQjc7HS(lBneo4M0PJKyj
zr(MT})9%*NvX0SifIy?0h6MaEp>ms`2u(0laobdo7?*W}iHgex65_Ir+w<CAs{YTB
zetyn5-}jyKzQ6OHD6gm9JWxEdw(2!iZf8J)Qft2pzpvD~RC&7M7_3*S;y8Q=J_mc1
zvUS1_z`d{yegUqBKe6~rSk3sdHGU6PD7E1R=!P}0xP{IRIxZ+1ZiXje7d!_wyabz+
zsss{XWiz}9?t}zUISE_g{T9zc0<L@s@^0mC;3lOuUI!Z?Qhd8{JNuU=9$=!5iO;|e
z_%+Bz8)ah`{1;@wo0wA}h*S?e00ZzDsNv6H7`_A9q-q<ig?B*$sLH_>cmlFt)${DH
zoOCWQu@Q<@9{5kld^NU;wbcXV`RXLZq3S0g3ss+o1X%q>nKf%5>(n&Cb&z9H*j#f9
zd_NSY+TjAk!J5a}UvUIAk^wK(ybM{;ZnMZUal$?dalpO^$;AE$Bt!et&;h?~J%1Il
zu>EgJt?{y_oHgIo2Fvf&Mj?*Wrg=UEiw`qFQEa*dIm1nVf}Cj`i%5ob61gAlg9KA|
z8nR*Ci;zX@ehr(Ml48O!M?YjeM*>o0j%kRaj*}2K94}bUWuFb4zvCS`c7nXIJW(&j
z!RGanIaby0glt$p5BuOz7=mAid{h4yh+Flt4jVVDg(T1*2~e>Ok^r}&VFa>I!<~?4
z4H+2X{2NxR2ND@pH^}v=gAyTGIj@JDrxP0$**UjEHgYaQ4J9Cs#Q7}bd*`<xx5xP^
zq!t>b>T#%%^OuYpw?dq0+zo9-I#QKH*ti1OxbY-p<Hpa!Zum`z1C4)#EaZ}a$k@dq
z^4uksv$0DYB1jj1Bcz60pMx5H2jYn94aiYyf|uv|R#@bL-baTj*C(Jn_%b9Q{R(6=
zEoV$2={F%&t;HD`{~huNM*j!$ep3}iwFAl#5>WGA_!zto1}G=KRa0;$q_$c<CHJ2X
zTh23qRV^1F3%2~i;@gntt=GVIh<##ht5maw)7JPwh!d^PLl$nm1S#gWYat1>bwJi{
z6U)7jAmkdi9h3VnFUp5R-S$(6t!;mWb-d$*EVx-L9)@?9pKpE^aul0?1aE;-gk;<<
zMXTZMkZ;@f!w5VMZ-y5k#oHl?XgJLNYCD|;NB|vj<^<I7J;(+fe}Dwyw!vD63v$Np
zoscSaOF+!I(@@?!0}0qI3CX-{$a?PgU=L-WVH2FRSj^L@WI{d^J_o7h&aXl??EEpr
zk<Q=2YTU<0v9OCpC9|#=<Vd<c3(3Cg0_2)@{R3jV#{(TO0ZGua1WCYi3gS@F^CdbI
zj)xD|V5R3*j9J`EHe!)?2gC{QJv=94?@`87aQA6QVBKdS8F&8x5?J@2VSr@WOakkP
zL;5`{khOZAh6L5~8gy{}zV&oi&__W?20nlJ1)l+j7(WKF-}f_!1HRuv?y^rZ#IfED
z&<<ro7Vaf`DY{<yo<i+C0y*N|Z@_g>ZVh|NHmagXhs`BozhsOP{&9FCoQ5@UKO}Si
z3S=Sw6OhgPFGDi-OU4@42~OtwBmgq+i@~jsJtV=t7a&F2_Y!2UK1rmAwSE7lL*@Yu
z*(fjuaU-w~l40Oc*a@G4WE*%9l5yY_$k_%iLmUZS14%0=iLg;{6mpcoMaa6rQ;_Np
zp0(z==0z!j;D4FmgMRwT2e~SeL4OJk!%spi?SCEime$~)p?FAOp&5v!AvvlW;HND<
z1F<~xMTn!JuRsD1ore^1=%U>JH4HwCq{%hZh()EkOKpx{qQBCb=-qi`RI*rooG}5F
z+9bapL{blOO9)8pmjKmCB#pneYFRpRHg`yewhujss0FnLNnuGlT1M&jpw-8fQlmhX
z*uL5xr^8XoAEi=-k+n)9u3V`-#P1GdmWQe+XT92{tj=NRMQ5zBoTCk`MdQ|-U=#A7
zC1jx07PUImko<#MYH^E#d2}4LNdEg}#%c+R)PTyOW7ha~Xh)wwk07a!JZeW$GeL9@
zl6D)~j;3WmOQUT_oGzf%R!!$)GD{(-0VKt9KdMD{p-Oah`v`-*D0kJQgZ>Djs!Q!Y
zes`e<%0s!%>Q<|tfo}96)Njq5mZ-$qLue57pcCjiG=ro~pd+PTshZU+{V+<Qdyxw*
zBCeqPTdhu^KE#buA4PYg4kXudH+mSYAa0UsI>1jCdICL)VyGEikFG^E=awpttSc@Z
z&L!s;3c6#!t$ka!20Zc`EdBPEenWaFXPP>eohuwNa;6^2E@x(qLNc52=tyS9Yl|-=
z^Li<lozEF*O?Nh#FXWQb%LQ{*Co{URVCseJl1axX=wv}3N~ThJ+SJQ=b8fkq^4NG)
z?}@~Bk0;_fJT|5Ggrm{$SbS<5uND@vJT(uR<%N>zrBsr2i6m!a3Ws%e&Nez29oWtL
z;oBl3k@%F(=JZe`J~kMO>7ntc4(o|<G#(j9jD(|lA`zV!j}3aYj*%0gd2PdE2|a9P
z%$$+Z6U)=7<cuCk&X}3JspnWKoh9d4vtT4sd9Ur<+xBML`+BzY`gO1WW?$cyp8j)A
zeee3>K=%JPaxh;oSB*6E@2?p7)@)YKWYapUv(t-aA$c%MK>EP4sq^Ndq2DbeQ$>$0
zZRB-qa8yqw!XuN3!KfY%i;I!)Hk~okI&J3DS#9WfLySur`Yo|?CRs38%r-DCE=Ecg
z?h5Z3)Y)a(BN4~a(ijiJIx(szqT>@6W0CmyRjzt%ckGZrjIwc;4L$L$_-Np(xGD~J
clx&_DaqA@l&}l;o1y@rGrr3JA>xZ`g0hdnmumAu6

diff --git a/po/ru.gmo b/po/ru.gmo
index 6019b4af6213f8139ff01898cc4429452f4c45d0..df7cd87d720b7afaecef05978aa369e788f25e61 100644
GIT binary patch
delta 3472
zcmZwJdra0<9LMo5-cT-zAXhIuB#MX%h#298FwDd&3gP{JtJKI#W}=TJsTG=vB&aD{
z%xuj+8oEqpb4yFHnX}wXUD#!=X{Fugl<NKAcWD06ncwGi&hL4C=XcKcob$t5)dAOc
z2Kc`YZ&hiO8lsD$wONA<;)7C!ov|9r@Cf$D?Dl4(a0&Ln4>1@UoVPH7_Ft|Y+`%k_
zb~GkqBK9@&+gK{`Tqwn!xEE9KWAx!|Ou(=(Z-D)=Bkgh68)u{LE5}5vb6!Rbyb&M5
zSZ<2J!PpIpFo);c3MzWwVfV##jHUe#_QKe3@4>@R1D%QL*pFGb2h;FN%*57A+lPZN
z7pI{fxDTW81Zp6cushGUpQ-49;hoI7VjpDCmXCdKs=K}!mFffT`e&$_-^K3ei}VJV
z=bVS?rwrR+CGu0-k5PCS{Yg~LxeLD`leQqfWo@hr4#s@c-q?i7K&|tXyZ${YBLPfL
zGwX_)@j#5jBGmUKsQ$}cdtWs9*AI?yfq!<1kI4aMKcP0u2-2c8Ekv#LYSc_?F$2$_
z26`XWPj5a3;&9ZOm!dMdAJyLp)aI*4W%l<t^536I0>je}3Xu(Kvyo(2iED2}jl2?-
zv3HO@>~kD|H&J^bo;0<^NytI5*{FUOW77nX{cN?U({k8PCAEdwNz?=HkT*T34S80F
zF{tYW&V{HoU5ngeTan4zd#JTNjXEt?@L{}<8o&Ubw<&W`OEd>{zkdxC?fxpf*wXA2
z>H)jRyHa=%m63C()LlW%tQ9*^GYH3_*bOz1>8OFM!aG=wskoM%s*D{%W&8}%zu)dr
zVUX61ysG1&s2fJ3Mp%ku*Q!y!JAyvEgi7h3sJ)QDS=IMNsOu|{XIqWC{t0TpH?R%<
zih(-+cd6)j1Teio3`S0*MWAj-!d`634Acy!(fa{hg&Jr#(wC28QJb$CgYW_>b5~JI
zdegP<yYGY9fhj!SqNu2&VW<ZzL;AKY?)p*GuKxkazJ;*mHPdwDonr-<hVxJ}u0-8`
z9QFGfs8bN!&wHUIquLYDuOF098IF~hi&s%I>B^{-`a)##R)V^5H);S!QTJcQQP_w>
za3rHqMmC`ac9D<K*fQOWu5CQ3KYu#;R|oHMLA&)lGFWTF4)Ngt)C`M|zHJeb1gk=I
zaN1qJjhb;h8$l1uM-6a|a}TzseH??Z9>ei+2KndZWw*GXj$3e;HN#ZY55}N6oQ0av
zcGQEep#~6`=^eLp)cq4t8C&nJ??o-$N2sOw7Jb;3^hRQ`pNfu64k{CqQF~w=&Sle8
zV+id=(!)NrAYKYg)-q57Sb$3THq-#lqh7uDa4^O_;vMf|)M?s)>aPY#w)qcHVNlk9
zTKmXsFGa<ul+8zFWCcdx4%B<$0BT9jqB_2dBQR=^_uv_*)3gWM;oGPOA42A2$Go=R
zexafc{>EC~gq<0-W^j_XiXPm6Be4<rvtT1~yiCo+IND|C!~NJ9Pht#SLuK?2<e$Yo
z>J4BDYDpJjqRxL6l>{!FKyAVXJj<Ifnsg`w-{g5SzJuC45gZj|ECJOn#weVR%)v@g
z13ik`Ll^KKeur!`Yk1t7SOn?H<@uIQMQgYmlkk1ijb9<lXg{LXy2lgVW*Ub*XfH?2
zq#Cs+>QPJBh&pCDq)ltQ6!m=#YL9$@8n||}za6%xq6;yoHR|q6L#_D`R3-{A0GWtY
z5^oUdP(>4nCtg&8GMm`lRO7#NxZc&ZWX#9L5gmAb(@ZqV?Zg^FWr~LtJ9C`PY@*WJ
zHD1S9LTkUoT~k(7ltmS7*fL@mv4L1gloQWs{e!5eXwyw65{an<2g2SWl+tWsCZXR3
z6YB_-rNknlml~9<g!T}xczc1+ADTj<f>3#$V2}S-7B`QSC29II+6kW`iU_5AA<<m6
zQJF~0AUGt(9<|p9m41Ye&<1;!P<fh|OK5YpBDAzRdYg%5gv!VoKEhO^O!8>@9P~i_
zC0F0$WPjN+#AB{K4|i4Mg{1i3bXO)jwR;{c)ies+g|!cSG0WAHFwfPe;VZ5lh))vQ
z?j=M%v7FdXEFiWJky>ZYmaQOOCH}9ht_TZ_9$D(XT7~n8OkyL!TfM2I^7*o>kH!f^
z6|sXzCRD}}V=78QqZ9X0YfJPbb`s5H4i)`18&A|73~d`!*AR2AV`zHcv`k-GR(4ue
SN?Ll|2faTHsjJUQ==3iib5>~p

delta 4110
zcmbuBdvMgn9l$q&3I~`F@`5CEqb1~VIr0vWhrOgEx#W15M`=`$5H932<l^NbQpLm8
zqEuSYK!u75wc2*-XgfHajymm}=A=iHJ4k|9v{*1H1TrKdGaa!?Ki}Msj-9E0b!Wc$
z?Cx*(``&}U?}@m$Kf=3rYV>Ip9*01ZQa64Rex#Ji9Nsql3MR7tHGB;EVXjiQ)WBM}
z6)u1QI0KFw{2b0vD)Q#=IuadfChLVT9p*!C7L&D1z5#c_RQL+ago9ASYcNHrDCXi|
zR5_dt?|~Wce#pJ30}$t<-Z6L{Vsy3%-U>0Q5}+N<hYswk+nErR7@XZ}e3*c9m5Ppp
zX;3a=IJyqvXtWEWIC>Xc41WW&;V>*9=vb&>EwsXRh`~9p!g$yZab(U2_SJ1nzGMS~
zF?gwxVF`?acS19~&)DAwNp;Ncjr|WHk;hJl=!wM%aWJ;Qpa-HSb`RuC>`O2Pz6Rr9
zFZLDni2WeE5jO>r+PHa;(!^n0;Y^%3LI&cVff$H;-QbY1e+80}c<xGM@!XY&<0~P$
z;z{-7_id0f@duQOut2Y*go_D(gB(ow43eU`cp1?I#etc8+77AhydFr+=luZ^@w|^=
z9x9~CFq&8l(Ua(eg-}XMRukWZWH#|ML|@{+AZ7I?N$N0?gc9lMq;-%3Nn+?0co0(0
zBuSa9{g4QfByuvB^bd%^WWK3<h*If^<h5`bl#J1U$p;~NlciS(AXx&4r2feg!3;J&
zRqBRC@EU6jYUW}{FM}x5tBw8L249BMR=)))jh3dOhx7$V_EM%oUQG&nst8JEa3G}%
zYPbz1Q2&%eOi+;W2F!#fm6|d@)}$9Ou;4vN3KyJ%M43u7lDbq;%sjOM5<u!5unew&
zd`mq5aU@lmbprkc64XL*n2aqPgI-deCbh#r+EPe!rHKL*r)`B?NP80EV4Bo05xx)Q
z+!d(t1tXHube^8{Li#$$_w+r+{_Ai7>yN_w88<>4&S1~0Zeo(d22Ny<W_czV-!j%U
za1LvE_i~|?rsVe^!_ArqGM|N%AyXp6QL{MC9rKSN&1V)zP4s0XBy(BZk&<RfkI1@M
z-oG3WLz%D<qA2T8$l<Kt!3@}E?Ee$qrj&)olh<IWg+yxUgwb#}%!V&QqO%M^?pv-x
z&M#t5Ucn-<M<9#Zus^x+lyTrSxQq{S(TXndm`JkkhNM2b2jYFUoI_FeUm*@;i(?qb
ziHA6qQvyrj6A)uL??N2QC3<Ct<<J7XyO^Lb_m2<-x#wXKWi&zFa~{q}jq{{yi7-!U
zN`~@&3C-{%<bM9t@P0l@CE|Q}ei+Q(0dX+@4TGm3sn3^;V8dH5oe7Ey#Pc}FWl>x}
z^b%n~hw*(UL}9@Tum&E3oGm0P;y_^~Oo5${`-M+Kz7_t-*#7`hy239YWhmk-fq7L?
z6%*cjQ5(%g>WUs=&GT7&6q1p}!;ms8!O@5o$ez68rIOk-xEJO^$smp_{REQwVsQ*d
zikskkxE)fK;+G)3R(uvFdD$Z!qOgQbl>_A!@RUmKg=C@RAT&X_z}b=sNM=f|L0(5G
zb9ug{+?CRlHb9<h>0^-KN)JH{myW`l`H4t{1XE^#UJRFsL7J>gYAx%BAqvWN@Qe5h
z_%drEu#z1yY!%0BP%=Q-tOJm&ST8^_RUQd7%z$K}T<#^njgWKY+u-exOnAj$`3M`-
zw)}GVKm~P@T2@p;ny~_dQv&ektjR!CJ0#+&hoA{Q56M{7E5=$fMK4uJrizuSCTelC
zx&lVSyWr=rL*9QL!C*ilvxy>7ZhH%s!Y|=M`i=C-#hML}Y}IUr>F~D@XKSRZ7eF~r
zAhkp(JyF{T(N`;(BHgt|A!VxlEd1S`iG67j`&vjO_9x+VxZha62+5HBRfBIrYHt5C
zBop=zVFV(j`zf*;LEq#c0C{@qq3}|zhtC@G=isA=Q`Xc>8bz%_q;^CiUaF^&hY=Zf
zg`vJ{kow7gM*lM;!VF`z6V62vkq%>TTaV8)$NNK8<%k1$0FfVRH$v!>V?KW+6_UTZ
zktN9U2su<sk#&fioq<T|WVjFsBTW{JT?kDhZ%l1LZbGCE{Bnh_Nbzn!u8&Q|#0JZd
z8022$I|w;eZOHZU6cc%0_aM>~QxK_{3=5(W8cKB{GVVj9Y3N2ZRc=vCd24D1(uK%4
zghWlwlxi|4DE`X)`^Nl$!D`r!R2XYJ+}D#cGt>Kwv2&+E+U38-&sfzNJCA(jOS3Vj
z-_^H``5L(0n9CztjbtPDBQ=NzIf!gTeuBixyR2l=jr<7tU*mf{U(AfRZ8bhU2wRZ?
z<Z&b)k+BH*!Q@P-<***vkMtnvh>VrUik>yI;!|E=HVsKd_9E9u3lshsQ+FKRH*1>7
z+quQ<+}P&PX%*=@XUUQRv-~Zb{982ncd@Q?I~=;+)#};Y>~`o%S9iyTW{=a=Vb->e
z7K^E&&Do_p-L8%9=1rRE24|PY?OfmOact1e4((}k=r&iUgGsYTJ3V@{v%Oufcj)de
zM{BpY-E88hUTtfrUfI~7tt*=JYHMAcbwxu{F-JXZF1|V*aD){(H+8l<(TgYU<_^ym
z?P@hGUshL9&3S9N&2DRG(k{2Iv^A_)R$s3xSJr8(US+LouvIkLt#x`;W8JEi^~)?;
z*Ao-2Sxi+c8g-SU!{Kgj*Q>hMw>w+3-Pz*k=yK>*RBdt*^9F~f+1cJ@F&%y`J~(?;
zc2;hl&dpnzlb@AcbojyaBQw1fuKz!g!@gk9cU&JE2pt*l_4oVqfNzk+;Lw0Cc-%iU
zcwC><L0_-`y<WdrpB@V7P|(NeK+r$v^$+P|p}zM|oiLk%A$@k_s=j#U;^_ITvC*^R
z`rMhzXD(eCy-<vTp-{g*9T@EMbG~=Lcj`pH|D?~~A3Ca!_^>k+2$&|u$40JR8vpdl
zs1Ec6^u*=SbK}F|_3+pydVJ#Y$neP6$fZl;7xc#?W1|*5yu1&;LwzSs`-6ew`t4Ak
p-)|EAp|^u*4g}u~4fOlo2?X?De`shRFyz-?pLVibR(Na7zX8!Wh0_24

diff --git a/src/ChangeLog b/src/ChangeLog
index 29f99623..acbc0331 100644
--- a/src/ChangeLog
+++ b/src/ChangeLog
@@ -1,3 +1,117 @@
+2000-11-19  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* retr.c (get_contents): If use_expected, make sure that the
+	appropriate amount of data is being read.
+
+	* http.c (gethttp): Check for both `Keep-Alive: ...' and
+	`Connection: Keep-Alive'.
+
+	* wget.h (DEBUGP): Call debug_logprintf only if opt.debug is
+	turned on.
+
+2000-11-19  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* http.c (connection_available_p): Use it.
+
+	* connect.c (test_socket_open): New function.
+
+	* http.c (gethttp): Support persistent connections.  Based on the
+	ideas, and partly on code, by Sam Horrocks <sam@daemoninc.com>.
+	(register_persistent): New function.
+	(connection_available_p): Ditto.
+	(invalidate_connection): Ditto.
+
+2000-11-19  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* url.c (convert_links): Handle UREL2ABS case.
+
+	* recur.c (recursive_retrieve): Instead of the list
+	urls_downloaded, use hash tables dl_file_url_map and
+	dl_url_file_map.
+	(convert_all_links): Use them to retrieve data.
+
+	* host.c (clean_hosts): Free the hash tables.
+
+	* main.c (private_initialize): Call host_init().
+
+	* host.c (store_hostaddress): Use a saner, hash table-based data
+	model.
+	(realhost): Ditto.
+	(host_init): Initialize the hash tables.
+
+2000-11-18  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* utils.c (slist_append): Eviscerate NOSORT.  Hash tables are now
+	used for what the sorted slists used to be used for.
+	(slist_contains): Don't rely on the list being sorted.
+	(slist_append): Simplify the code.
+
+	* recur.c (recursive_cleanup): Use free_string_set.
+
+	* utils.c (string_set_add, string_set_exists, string_set_free):
+	New functions for easier freeing of hash tables whose keys are
+	strdup'ed strings.
+
+	* recur.c (recursive_retrieve): Use the hash table functions for
+	storing undesirable URLs.
+
+	* hash.c: New file.
+
+2000-11-17  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* main.c (private_initialize): Call url_init.
+	(main): Call private_initialize.
+
+	* url.c (unsafe_char_table): New table.
+	(UNSAFE_CHAR): Use it.
+	(init_unsafe_char_table): New function.
+	(url_init): New function; call init_unsafe_char_table.
+
+2000-11-15  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* html-url.c (handle_link): Handle HTML fragment identifiers.
+
+	* recur.c (recursive_retrieve): If norobot info is respected and
+	the file is specified not to be followed by robots, respect that.
+
+	* html-url.c (collect_tags_mapper): Handle <meta name=robots
+	content=X>.  For us the important cases are where X is NONE or
+	where X contains NOFOLLOW.
+	(get_urls_html): Propagate that information to the caller.
+
+2000-11-13  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* url.c (convert_links): Unlink the file we might be reading from
+	before writing to it.
+	(convert_links): Use alloca instead of malloc for
+	filename_plus_orig_suffix.
+
+2000-11-10  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* url.c (get_urls_file): Ditto.
+	(convert_links): Ditto.
+
+	* html-url.c (get_urls_html): Use read_file() instead of
+	load_file().
+
+	* utils.c (read_file): New function, instead of the old
+	load_file().
+	(read_file_free): Ditto.
+
+	* url.c (findurl): Search only for the supported protocols.
+	(convert_links): Use fwrite() when writing out a region of
+	characters.
+
+2000-11-10  Hrvoje Niksic  <hniksic@arsdigita.com>
+
+	* ftp-ls.c: Move html_quote_string and ftp_index here.
+
+	* url.c: Remove get_urls_html, since that's now in html-url.c.
+
+	* html-url.c: New file.
+
+	* html-parse.c: New file.
+
 2000-11-16  Hrvoje Niksic  <hniksic@arsdigita.com>
 
 	* mswindows.h: Define snprintf and vsnprintf to _snprintf and
diff --git a/src/Makefile.in b/src/Makefile.in
index e3b433b0..bfe9868d 100644
--- a/src/Makefile.in
+++ b/src/Makefile.in
@@ -57,9 +57,10 @@ MD5_OBJ = @MD5_OBJ@
 OPIE_OBJ = @OPIE_OBJ@
 
 OBJ = $(ALLOCA) cmpt$o connect$o fnmatch$o ftp$o ftp-basic$o  \
-      ftp-ls$o $(OPIE_OBJ) getopt$o headers$o host$o html$o   \
-      http$o init$o log$o main$o $(MD5_OBJ) netrc$o rbuf$o    \
-      recur$o retr$o snprintf$o url$o utils$o version$o
+      ftp-ls$o $(OPIE_OBJ) getopt$o hash$o headers$o host$o   \
+      html-parse$o html-url$o http$o init$o log$o main$o      \
+      $(MD5_OBJ) netrc$o rbuf$o recur$o retr$o snprintf$o     \
+      url$o utils$o version$o
 
 .SUFFIXES:
 .SUFFIXES: .c .o ._c ._o
@@ -133,26 +134,31 @@ TAGS: *.c *.h
 
 # DO NOT DELETE THIS LINE -- make depend depends on it.
 
-cmpt$o: config.h wget.h sysdep.h options.h
-connect$o: config.h wget.h sysdep.h options.h connect.h host.h
-fnmatch$o: config.h wget.h sysdep.h options.h fnmatch.h
-ftp-basic$o: config.h wget.h sysdep.h options.h utils.h rbuf.h connect.h host.h
-ftp-ls$o: config.h wget.h sysdep.h options.h utils.h ftp.h rbuf.h
-ftp-opie$o: config.h wget.h sysdep.h options.h md5.h
-ftp$o: config.h wget.h sysdep.h options.h utils.h url.h rbuf.h retr.h ftp.h html.h connect.h host.h fnmatch.h netrc.h
-getopt$o: wget.h sysdep.h options.h
-headers$o: config.h wget.h sysdep.h options.h connect.h rbuf.h headers.h
-host$o: config.h wget.h sysdep.h options.h utils.h host.h url.h
-html$o: config.h wget.h sysdep.h options.h url.h utils.h ftp.h rbuf.h html.h
-http$o: config.h wget.h sysdep.h options.h utils.h url.h host.h rbuf.h retr.h headers.h connect.h fnmatch.h netrc.h
-init$o: config.h wget.h sysdep.h options.h utils.h init.h host.h recur.h netrc.h
-log$o: config.h wget.h sysdep.h options.h utils.h
-main$o: config.h wget.h sysdep.h options.h utils.h getopt.h init.h retr.h rbuf.h recur.h host.h
-md5$o: wget.h sysdep.h options.h md5.h
-mswindows$o: config.h winsock.h wget.h sysdep.h options.h url.h
-netrc$o: wget.h sysdep.h options.h utils.h netrc.h init.h
-rbuf$o: config.h wget.h sysdep.h options.h rbuf.h connect.h
-recur$o: config.h wget.h sysdep.h options.h url.h recur.h utils.h retr.h rbuf.h ftp.h fnmatch.h host.h
-retr$o: config.h wget.h sysdep.h options.h utils.h retr.h rbuf.h url.h recur.h ftp.h host.h connect.h
-url$o: config.h wget.h sysdep.h options.h utils.h url.h host.h html.h
-utils$o: config.h wget.h sysdep.h options.h utils.h fnmatch.h
+cmpt$o: wget.h
+connect$o: wget.h connect.h host.h
+fnmatch$o: wget.h fnmatch.h
+ftp-basic$o: wget.h utils.h rbuf.h connect.h host.h
+ftp-ls$o: wget.h utils.h ftp.h url.h
+ftp-opie$o: wget.h md5.h
+ftp$o: wget.h utils.h url.h rbuf.h retr.h ftp.h connect.h host.h fnmatch.h netrc.h
+getopt$o: wget.h getopt.h
+hash$o: wget.h utils.h hash.h
+headers$o: wget.h connect.h rbuf.h headers.h
+host$o: wget.h utils.h host.h url.h hash.h
+html-parse$o: wget.h html-parse.h
+html-url$o: wget.h html-parse.h url.h utils.h
+html$o: wget.h url.h utils.h ftp.h
+http$o: wget.h utils.h url.h host.h rbuf.h retr.h headers.h connect.h fnmatch.h netrc.h md5.h
+init$o: wget.h utils.h init.h host.h recur.h netrc.h
+log$o: wget.h utils.h
+main$o: wget.h utils.h getopt.h init.h retr.h recur.h host.h
+md5$o: wget.h md5.h
+mswindows$o: wget.h url.h
+netrc$o: wget.h utils.h netrc.h init.h
+rbuf$o: wget.h rbuf.h connect.h
+recur$o: wget.h url.h recur.h utils.h retr.h ftp.h fnmatch.h host.h hash.h
+retr$o: wget.h utils.h retr.h url.h recur.h ftp.h host.h connect.h hash.h
+snprintf$o:
+url$o: wget.h utils.h url.h host.h
+utils$o: wget.h utils.h fnmatch.h hash.h
+version$o:
diff --git a/src/config.h.in b/src/config.h.in
index 2038acde..ed200e32 100644
--- a/src/config.h.in
+++ b/src/config.h.in
@@ -101,6 +101,9 @@ char *alloca ();
 /* Define if you have the uname function.  */
 #undef HAVE_UNAME
 
+/* Define if you have a working version of mmap.  */
+#undef HAVE_MMAP
+
 /* Define if you have the gethostname function.  */
 #undef HAVE_GETHOSTNAME
 
diff --git a/src/connect.c b/src/connect.c
index 28ce2043..feb2bb52 100644
--- a/src/connect.c
+++ b/src/connect.c
@@ -107,6 +107,37 @@ make_connection (int *sock, char *hostname, unsigned short port)
   return NOCONERROR;
 }
 
+int
+test_socket_open (int sock)
+{
+#ifdef HAVE_SELECT
+  fd_set check_set;
+  struct timeval to;
+
+  /* Check if we still have a valid (non-EOF) connection.  From Andrew
+   * Maholski's code in the Unix Socket FAQ.  */
+
+  FD_ZERO (&check_set);
+  FD_SET (sock, &check_set);
+
+  /* Wait one microsecond */
+  to.tv_sec = 0;
+  to.tv_usec = 1;
+
+  /* If we get a timeout, then that means still connected */
+  if (select (sock + 1, &check_set, NULL, NULL, &to) == 0)
+    {
+      /* Connection is valid (not EOF), so continue */
+      return 1;
+    }
+  else
+    return 0;
+#else
+  /* Without select, it's hard to know for sure. */
+  return 1;
+#endif
+}
+
 /* Bind the local port PORT.  This does all the necessary work, which
    is creating a socket, setting SO_REUSEADDR option on it, then
    calling bind() and listen().  If *PORT is 0, a random port is
diff --git a/src/ftp-ls.c b/src/ftp-ls.c
index 16a7f7d6..884cf3d8 100644
--- a/src/ftp-ls.c
+++ b/src/ftp-ls.c
@@ -36,6 +36,7 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #include "wget.h"
 #include "utils.h"
 #include "ftp.h"
+#include "url.h"
 
 /* Converts symbolic permissions to number-style ones, e.g. string
    rwxr-xr-x to 755.  For now, it knows nothing of
@@ -388,3 +389,175 @@ ftp_parse_ls (const char *file)
 {
   return ftp_parse_unix_ls (file);
 }
+
+/* Stuff for creating FTP index. */
+
+/* The function returns the pointer to the malloc-ed quoted version of
+   string s.  It will recognize and quote numeric and special graphic
+   entities, as per RFC1866:
+
+   `&' -> `&amp;'
+   `<' -> `&lt;'
+   `>' -> `&gt;'
+   `"' -> `&quot;'
+
+   No other entities are recognized or replaced.  */
+static char *
+html_quote_string (const char *s)
+{
+  const char *b = s;
+  char *p, *res;
+  int i;
+
+  /* Pass through the string, and count the new size.  */
+  for (i = 0; *s; s++, i++)
+    {
+      if (*s == '&')
+	i += 4;                /* `amp;' */
+      else if (*s == '<' || *s == '>')
+	i += 3;                /* `lt;' and `gt;' */
+      else if (*s == '\"')
+	i += 5;                /* `quot;' */
+    }
+  res = (char *)xmalloc (i + 1);
+  s = b;
+  for (p = res; *s; s++)
+    {
+      switch (*s)
+	{
+	case '&':
+	  *p++ = '&';
+	  *p++ = 'a';
+	  *p++ = 'm';
+	  *p++ = 'p';
+	  *p++ = ';';
+	  break;
+	case '<': case '>':
+	  *p++ = '&';
+	  *p++ = (*s == '<' ? 'l' : 'g');
+	  *p++ = 't';
+	  *p++ = ';';
+	  break;
+	case '\"':
+	  *p++ = '&';
+	  *p++ = 'q';
+	  *p++ = 'u';
+	  *p++ = 'o';
+	  *p++ = 't';
+	  *p++ = ';';
+	  break;
+	default:
+	  *p++ = *s;
+	}
+    }
+  *p = '\0';
+  return res;
+}
+
+/* The function creates an HTML index containing references to given
+   directories and files on the appropriate host.  The references are
+   FTP.  */
+uerr_t
+ftp_index (const char *file, struct urlinfo *u, struct fileinfo *f)
+{
+  FILE *fp;
+  char *upwd;
+  char *htclfile;		/* HTML-clean file name */
+
+  if (!opt.dfp)
+    {
+      fp = fopen (file, "wb");
+      if (!fp)
+	{
+	  logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
+	  return FOPENERR;
+	}
+    }
+  else
+    fp = opt.dfp;
+  if (u->user)
+    {
+      char *tmpu, *tmpp;        /* temporary, clean user and passwd */
+
+      tmpu = CLEANDUP (u->user);
+      tmpp = u->passwd ? CLEANDUP (u->passwd) : NULL;
+      upwd = (char *)xmalloc (strlen (tmpu)
+			     + (tmpp ? (1 + strlen (tmpp)) : 0) + 2);
+      sprintf (upwd, "%s%s%s@", tmpu, tmpp ? ":" : "", tmpp ? tmpp : "");
+      free (tmpu);
+      FREE_MAYBE (tmpp);
+    }
+  else
+    upwd = xstrdup ("");
+  fprintf (fp, "<!DOCTYPE HTML PUBLIC \"-//IETF//DTD HTML 2.0//EN\">\n");
+  fprintf (fp, "<html>\n<head>\n<title>");
+  fprintf (fp, _("Index of /%s on %s:%d"), u->dir, u->host, u->port);
+  fprintf (fp, "</title>\n</head>\n<body>\n<h1>");
+  fprintf (fp, _("Index of /%s on %s:%d"), u->dir, u->host, u->port);
+  fprintf (fp, "</h1>\n<hr>\n<pre>\n");
+  while (f)
+    {
+      fprintf (fp, "  ");
+      if (f->tstamp != -1)
+	{
+	  /* #### Should we translate the months? */
+	  static char *months[] = {
+	    "Jan", "Feb", "Mar", "Apr", "May", "Jun",
+	    "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"
+	  };
+	  struct tm *ptm = localtime ((time_t *)&f->tstamp);
+
+	  fprintf (fp, "%d %s %02d ", ptm->tm_year + 1900, months[ptm->tm_mon],
+		  ptm->tm_mday);
+	  if (ptm->tm_hour)
+	    fprintf (fp, "%02d:%02d  ", ptm->tm_hour, ptm->tm_min);
+	  else
+	    fprintf (fp, "       ");
+	}
+      else
+	fprintf (fp, _("time unknown       "));
+      switch (f->type)
+	{
+	case FT_PLAINFILE:
+	  fprintf (fp, _("File        "));
+	  break;
+	case FT_DIRECTORY:
+	  fprintf (fp, _("Directory   "));
+	  break;
+	case FT_SYMLINK:
+	  fprintf (fp, _("Link        "));
+	  break;
+	default:
+	  fprintf (fp, _("Not sure    "));
+	  break;
+	}
+      htclfile = html_quote_string (f->name);
+      fprintf (fp, "<a href=\"ftp://%s%s:%hu", upwd, u->host, u->port);
+      if (*u->dir != '/')
+	putc ('/', fp);
+      fprintf (fp, "%s", u->dir);
+      if (*u->dir)
+	putc ('/', fp);
+      fprintf (fp, "%s", htclfile);
+      if (f->type == FT_DIRECTORY)
+	putc ('/', fp);
+      fprintf (fp, "\">%s", htclfile);
+      if (f->type == FT_DIRECTORY)
+	putc ('/', fp);
+      fprintf (fp, "</a> ");
+      if (f->type == FT_PLAINFILE)
+	fprintf (fp, _(" (%s bytes)"), legible (f->size));
+      else if (f->type == FT_SYMLINK)
+	fprintf (fp, "-> %s", f->linkto ? f->linkto : "(nil)");
+      putc ('\n', fp);
+      free (htclfile);
+      f = f->next;
+    }
+  fprintf (fp, "</pre>\n</body>\n</html>\n");
+  free (upwd);
+  if (!opt.dfp)
+    fclose (fp);
+  else
+    fflush (fp);
+  return FTPOK;
+}
diff --git a/src/ftp.c b/src/ftp.c
index 4c26cf7d..aa283cf8 100644
--- a/src/ftp.c
+++ b/src/ftp.c
@@ -40,7 +40,6 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #include "rbuf.h"
 #include "retr.h"
 #include "ftp.h"
-#include "html.h"
 #include "connect.h"
 #include "host.h"
 #include "fnmatch.h"
@@ -722,7 +721,7 @@ Error in server response, closing control connection.\n"));
     }
   reset_timer ();
   /* Get the contents of the document.  */
-  res = get_contents (dtsock, fp, len, restval, expected_bytes, &con->rbuf);
+  res = get_contents (dtsock, fp, len, restval, expected_bytes, &con->rbuf, 0);
   con->dltime = elapsed_time ();
   tms = time_str (NULL);
   tmrate = rate (*len - restval, con->dltime);
diff --git a/src/ftp.h b/src/ftp.h
index c2e6d44c..064e6354 100644
--- a/src/ftp.h
+++ b/src/ftp.h
@@ -92,4 +92,6 @@ typedef struct
 struct fileinfo *ftp_parse_ls PARAMS ((const char *));
 uerr_t ftp_loop PARAMS ((struct urlinfo *, int *));
 
+uerr_t ftp_index (const char *, struct urlinfo *, struct fileinfo *);
+
 #endif /* FTP_H */
diff --git a/src/hash.c b/src/hash.c
new file mode 100644
index 00000000..e54fb33a
--- /dev/null
+++ b/src/hash.c
@@ -0,0 +1,403 @@
+/* Hash tables.
+   Copyright (C) 2000 Free Software Foundation, Inc.
+
+This file is part of Wget.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
+
+#ifdef HAVE_CONFIG_H
+# include <config.h>
+#endif
+
+#include <stdlib.h>
+#include <assert.h>
+
+#include "wget.h"
+#include "utils.h"
+
+#include "hash.h"
+
+#ifdef STANDALONE
+# define xmalloc malloc
+# define xrealloc realloc
+#endif
+
+/* This file implements simple hash tables based on linear probing.
+   The hash table stores key-value pairs in a contiguous array.  Both
+   key and value are void pointers that the hash and test functions
+   know how to handle.
+
+   Although Knuth & co. recommend double hashing over linear probing,
+   we use the latter because it accesses array elements sequentially
+   in case of a collision, yielding in better cache behaviour and
+   ultimately in better speed.  To avoid collision problems with
+   linear probing, we make sure that the table grows as soon as the
+   fullness/size ratio exceeds 75%.  */
+
+struct ht_pair {
+  void *key;
+  void *value;
+};
+
+struct hash_table {
+  unsigned long (*hash_function) (const void *);
+  int (*test_function) (const void *, const void *);
+
+  int size;			/* size of the array */
+  int fullness;			/* number of non-empty fields */
+  int count;			/* number of non-empty, non-deleted
+                                   fields. */
+
+  struct ht_pair *pairs;
+};
+
+#define ENTRY_DELETED ((void *)0xdeadbeef)
+
+#define DELETED_ENTRY_P(ptr) ((ptr) == ENTRY_DELETED)
+#define EMPTY_ENTRY_P(ptr)   ((ptr) == NULL)
+
+/* Find a prime near, but greather than or equal to SIZE. */
+
+int
+prime_size (int size)
+{
+  static const unsigned long primes [] = {
+    19, 29, 41, 59, 79, 107, 149, 197, 263, 347, 457, 599, 787, 1031,
+    1361, 1777, 2333, 3037, 3967, 5167, 6719, 8737, 11369, 14783,
+    19219, 24989, 32491, 42257, 54941, 71429, 92861, 120721, 156941,
+    204047, 265271, 344857, 448321, 582821, 757693, 985003, 1280519,
+    1664681, 2164111, 2813353, 3657361, 4754591, 6180989, 8035301,
+    10445899, 13579681, 17653589, 22949669, 29834603, 38784989,
+    50420551, 65546729, 85210757, 110774011, 144006217, 187208107,
+    243370577, 316381771, 411296309, 534685237, 695090819, 903618083,
+    1174703521, 1527114613, 1985248999, 2580823717UL, 3355070839UL
+  };
+  int i;
+  for (i = 0; i < ARRAY_SIZE (primes); i++)
+    if (primes[i] >= size)
+      return primes[i];
+  /* huh? */
+  return size;
+}
+
+/* Create a hash table of INITIAL_SIZE with hash function
+   HASH_FUNCTION and test function TEST_FUNCTION.  If you wish to
+   start out with a "small" table which will be regrown as needed,
+   specify 0 as INITIAL_SIZE.  */
+
+struct hash_table *
+hash_table_new (int initial_size,
+		unsigned long (*hash_function) (const void *),
+		int (*test_function) (const void *, const void *))
+{
+  struct hash_table *ht
+    = (struct hash_table *)xmalloc (sizeof (struct hash_table));
+  ht->hash_function = hash_function;
+  ht->test_function = test_function;
+  ht->size = prime_size (initial_size);
+  ht->fullness = 0;
+  ht->count    = 0;
+  ht->pairs = xmalloc (ht->size * sizeof (struct ht_pair));
+  memset (ht->pairs, '\0', ht->size * sizeof (struct ht_pair));
+  return ht;
+}
+
+/* Free the data associated with hash table HT. */
+
+void
+hash_table_destroy (struct hash_table *ht)
+{
+  free (ht->pairs);
+  free (ht);
+}
+
+/* Get the value that corresponds to the key KEY in the hash table HT.
+   If no value is found, return NULL.  Note that NULL is a legal value
+   for value; if you are storing NULLs in your hash table, you can use
+   hash_table_exists to be sure that a (possibly NULL) value exists in
+   the table.  */
+
+void *
+hash_table_get (struct hash_table *ht, const void *key)
+{
+  int location = ht->hash_function (key) % ht->size;
+  while (1)
+    {
+      struct ht_pair *the_pair = ht->pairs + location;
+      if (EMPTY_ENTRY_P (the_pair->key))
+	return NULL;
+      else if (DELETED_ENTRY_P (the_pair->key)
+	       || !ht->test_function (key, the_pair->key))
+	{
+	  ++location;
+	  if (location == ht->size)
+	    location = 0;
+	}
+      else
+	return the_pair->value;
+    }
+}
+
+/* Return 1 if KEY exists in HT, 0 otherwise. */
+
+int
+hash_table_exists (struct hash_table *ht, const void *key)
+{
+  int location = ht->hash_function (key) % ht->size;
+  while (1)
+    {
+      struct ht_pair *the_pair = ht->pairs + location;
+      if (EMPTY_ENTRY_P (the_pair->key))
+	return 0;
+      else if (DELETED_ENTRY_P (the_pair->key)
+	       || !ht->test_function (key, the_pair->key))
+	{
+	  ++location;
+	  if (location == ht->size)
+	    location = 0;
+	}
+      else
+	return 1;
+    }
+}
+
+#define MAX(i, j) (((i) >= (j)) ? (i) : (j))
+
+/* Grow hash table HT as necessary, and rehash all the key-value
+   pairs.  */
+
+static void
+grow_hash_table (struct hash_table *ht)
+{
+  int i;
+  struct ht_pair *old_pairs = ht->pairs;
+  int old_count = ht->count;	/* for assert() below */
+  int old_size = ht->size;
+
+  /* Normally, the idea is to double ht->size (and round it to next
+     prime) on each regrow:
+
+         ht->size = prime_size (ht->size * 2);
+
+     But it is possible that the table has large fullness because of
+     the many deleted entries.  If that is the case, we don't want to
+     blindly grow the table; we just want to rehash it.  For that
+     reason, we use ht->count as the relevant parameter.  MAX is used
+     only because we don't want to actually shrink the table.  (But
+     maybe that's wrong.)  */
+
+  int needed_size = prime_size (ht->count * 2);
+  ht->size = MAX (old_size, needed_size);
+
+  ht->pairs = xmalloc (ht->size * sizeof (struct ht_pair));
+  memset (ht->pairs, '\0', ht->size * sizeof (struct ht_pair));
+
+  /* Need to reset these two; hash_table_put will reinitialize them.  */
+  ht->fullness = 0;
+  ht->count    = 0;
+  for (i = 0; i < old_size; i++)
+    {
+      struct ht_pair *the_pair = old_pairs + i;
+      if (!EMPTY_ENTRY_P (the_pair->key)
+	  && !DELETED_ENTRY_P (the_pair->key))
+	hash_table_put (ht, the_pair->key, the_pair->value);
+    }
+  assert (ht->count == old_count);
+  free (old_pairs);
+}
+
+/* Put VALUE in the hash table HT under the key KEY.  This regrows the
+   table if necessary.  */
+
+void
+hash_table_put (struct hash_table *ht, const void *key, void *value)
+{
+  int location = ht->hash_function (key) % ht->size;
+  while (1)
+    {
+      struct ht_pair *the_pair = ht->pairs + location;
+      if (EMPTY_ENTRY_P (the_pair->key))
+	{
+	  ++ht->fullness;
+	  ++ht->count;
+	just_insert:
+	  the_pair->key = (void *)key; /* const? */
+	  the_pair->value = value;
+	  break;
+	}
+      else if (DELETED_ENTRY_P (the_pair->key))
+	{
+	  /* We're replacing a deleteed entry, so ht->count gets
+             increased, but ht->fullness remains unchanged.  */
+	  ++ht->count;
+	  goto just_insert;
+	}
+      else if (ht->test_function (key, the_pair->key))
+	{
+	  /* We're replacing an existing entry, so ht->count and
+             ht->fullness remain unchanged.  */
+	  goto just_insert;
+	}
+      else
+	{
+	  ++location;
+	  if (location == ht->size)
+	    location = 0;
+	}
+    }
+  if (ht->fullness * 4 > ht->size * 3)
+    /* When fullness exceeds 75% of size, regrow the table. */
+    grow_hash_table (ht);
+}
+
+/* Remove KEY from HT. */
+
+int
+hash_table_remove (struct hash_table *ht, const void *key)
+{
+  int location = ht->hash_function (key) % ht->size;
+  while (1)
+    {
+      struct ht_pair *the_pair = ht->pairs + location;
+      if (EMPTY_ENTRY_P (the_pair->key))
+	return 0;
+      else if (DELETED_ENTRY_P (the_pair->key)
+	       || !ht->test_function (key, the_pair->key))
+	{
+	  ++location;
+	  if (location == ht->size)
+	    location = 0;
+	}
+      else
+	{
+	  /* We don't really remove an entry from the hash table: we
+	     just mark it as deleted.  This is because there may be
+	     other entries located after this entry whose hash number
+	     points to a location before this entry.  (Example: keys
+	     A, B and C have the same hash.  If you were to really
+	     *delete* B from the table, C could no longer be found.)
+
+	     As an optimization, it might be worthwhile to check
+	     whether the immediately preceding entry is empty and, if
+	     so, really delete the pair (set it to empty and decrease
+	     the fullness along with the count).  I *think* it should
+	     be safe.  */
+	  the_pair->key = ENTRY_DELETED;
+	  --ht->count;
+	  return 1;
+	}
+    }
+}
+
+void
+hash_table_clear (struct hash_table *ht)
+{
+  memset (ht->pairs, '\0', ht->size * sizeof (struct ht_pair));
+  ht->fullness = 0;
+  ht->count    = 0;
+}
+
+void
+hash_table_map (struct hash_table *ht,
+		int (*mapfun) (void *, void *, void *),
+		void *closure)
+{
+  int i;
+  for (i = 0; i < ht->size; i++)
+    {
+      struct ht_pair *the_pair = ht->pairs + i;
+      if (!EMPTY_ENTRY_P (the_pair->key)
+	  && !DELETED_ENTRY_P (the_pair->key))
+	if (mapfun (the_pair->key, the_pair->value, closure))
+	  return;
+    }
+}
+
+/* Support for hash tables whose keys are strings.  */
+
+/* supposedly from the Dragon Book P436. */
+unsigned long
+string_hash (const void *sv)
+{
+  unsigned int h = 0;
+  unsigned const char *x = (unsigned const char *) sv;
+
+  while (*x)
+    {
+      unsigned int g;
+      h = (h << 4) + *x++;
+      if ((g = h & 0xf0000000) != 0)
+	h = (h ^ (g >> 24)) ^ g;
+    }
+
+  return h;
+}
+
+int
+string_cmp (const void *s1, const void *s2)
+{
+  return !strcmp ((const char *)s1, (const char *)s2);
+}
+
+struct hash_table *
+make_string_hash_table (int initial_size)
+{
+  return hash_table_new (initial_size, string_hash, string_cmp);
+}
+
+
+#ifdef STANDALONE
+
+#include <stdio.h>
+#include <string.h>
+
+int
+print_hash_table_mapper (const void *key, void *value, void *count)
+{
+  ++*(int *)count;
+  printf ("%s: %s\n", (const char *)key, (char *)value);
+  return 0;
+}
+
+void
+print_hash (struct hash_table *sht)
+{
+  int debug_count = 0;
+  hash_table_map (sht, print_hash_table_mapper, &debug_count);
+  assert (debug_count == sht->count);
+}
+
+int
+main (void)
+{
+  struct hash_table *ht = make_string_hash_table (0);
+  char line[80];
+  while ((fgets (line, sizeof (line), stdin)))
+    {
+      int len = strlen (line);
+      if (len <= 1)
+	continue;
+      line[--len] = '\0';
+      hash_table_put (ht, strdup (line), "here I am!");
+      if (len % 2)
+	hash_table_remove (ht, line);
+    }
+  print_hash (ht);
+#if 0
+  printf ("%d %d %d\n", ht->count, ht->fullness, ht->size);
+#endif
+  return 0;
+}
+#endif
diff --git a/src/hash.h b/src/hash.h
new file mode 100644
index 00000000..ab3136aa
--- /dev/null
+++ b/src/hash.h
@@ -0,0 +1,50 @@
+/* Hash table declarations.
+   Copyright (C) 2000 Free Software Foundation, Inc.
+
+This file is part of Wget.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
+
+/* From XEmacs, and hence from Dragon book. */
+
+#define GOOD_HASH 65599 /* prime number just over 2^16; Dragon book, p. 435 */
+#define HASH2(a,b)               (GOOD_HASH * (a)                     + (b))
+#define HASH3(a,b,c)             (GOOD_HASH * HASH2 (a,b)             + (c))
+#define HASH4(a,b,c,d)           (GOOD_HASH * HASH3 (a,b,c)           + (d))
+#define HASH5(a,b,c,d,e)         (GOOD_HASH * HASH4 (a,b,c,d)         + (e))
+#define HASH6(a,b,c,d,e,f)       (GOOD_HASH * HASH5 (a,b,c,d,e)       + (f))
+#define HASH7(a,b,c,d,e,f,g)     (GOOD_HASH * HASH6 (a,b,c,d,e,f)     + (g))
+#define HASH8(a,b,c,d,e,f,g,h)   (GOOD_HASH * HASH7 (a,b,c,d,e,f,g)   + (h))
+#define HASH9(a,b,c,d,e,f,g,h,i) (GOOD_HASH * HASH8 (a,b,c,d,e,f,g,h) + (i))
+
+struct hash_table;
+
+struct hash_table *hash_table_new PARAMS ((int,
+					   unsigned long (*) (const void *),
+					   int (*) (const void *,
+						    const void *)));
+void hash_table_destroy PARAMS ((struct hash_table *));
+void *hash_table_get PARAMS ((struct hash_table *, const void *));
+int hash_table_exists PARAMS ((struct hash_table *, const void *));
+void hash_table_put PARAMS ((struct hash_table *, const void *, void *));
+int hash_table_remove PARAMS ((struct hash_table *, const void *));
+void hash_table_clear PARAMS ((struct hash_table *));
+void hash_table_map PARAMS ((struct hash_table *,
+			     int (*) (void *, void *, void *),
+			     void *));
+
+unsigned long string_hash PARAMS ((const void *));
+int string_cmp PARAMS ((const void *, const void *));
+struct hash_table *make_string_hash_table PARAMS ((int));
diff --git a/src/headers.c b/src/headers.c
index 6b1a670f..521073df 100644
--- a/src/headers.c
+++ b/src/headers.c
@@ -165,6 +165,14 @@ header_strdup (const char *header, void *closure)
   return 1;
 }
 
+/* Write the value 1 into the integer pointed to by CLOSURE.  */
+int
+header_exists (const char *header, void *closure)
+{
+  *(int *)closure = 1;
+  return 1;
+}
+
 /* Skip LWS (linear white space), if present.  Returns number of
    characters to skip.  */
 int
diff --git a/src/headers.h b/src/headers.h
index cc66e49b..5f85c6eb 100644
--- a/src/headers.h
+++ b/src/headers.h
@@ -31,5 +31,6 @@ int header_process PARAMS ((const char *, const char *,
 
 int header_extract_number PARAMS ((const char *, void *));
 int header_strdup PARAMS ((const char *, void *));
+int header_exists PARAMS ((const char *, void *));
 
 int skip_lws PARAMS ((const char *));
diff --git a/src/host.c b/src/host.c
index 3fa1bb84..eeb4940d 100644
--- a/src/host.c
+++ b/src/host.c
@@ -1,5 +1,5 @@
 /* Dealing with host names.
-   Copyright (C) 1995, 1996, 1997 Free Software Foundation, Inc.
+   Copyright (C) 1995, 1996, 1997, 2000 Free Software Foundation, Inc.
 
 This file is part of Wget.
 
@@ -48,35 +48,38 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #include "utils.h"
 #include "host.h"
 #include "url.h"
+#include "hash.h"
 
 #ifndef errno
 extern int errno;
 #endif
 
-/* Host list entry */
-struct host
+/* Mapping between all known hosts to their addresses (n.n.n.n). */
+struct hash_table *host_name_address_map;
+
+/* Mapping between all known addresses (n.n.n.n) to their hosts.  This
+   is the inverse of host_name_address_map.  These two tables share
+   the strdup'ed strings. */
+struct hash_table *host_address_name_map;
+
+/* Mapping between auxilliary (slave) and master host names. */
+struct hash_table *host_slave_master_map;
+
+/* Utility function: like xstrdup(), but also lowercases S.  */
+
+static char *
+xstrdup_lower (const char *s)
 {
-  /* Host's symbolical name, as encountered at the time of first
-     inclusion, e.g. "fly.cc.fer.hr".  */
-  char *hostname;
-  /* Host's "real" name, i.e. its IP address, written out in ASCII
-     form of N.N.N.N, e.g. "161.53.70.130".  */
-  char *realname;
-  /* More than one HOSTNAME can correspond to the same REALNAME.  For
-     our purposes, the canonical name of the host is its HOSTNAME when
-     it was first encountered.  This entry is said to have QUALITY.  */
-  int quality;
-  /* Next entry in the list.  */
-  struct host *next;
-};
-
-static struct host *hlist;
-
-static struct host *add_hlist PARAMS ((struct host *, const char *,
-				       const char *, int));
+  char *copy = xstrdup (s);
+  char *p = copy;
+  for (; *p; p++)
+    *p = TOLOWER (*p);
+  return copy;
+}
 
 /* The same as gethostbyname, but supports internet addresses of the
-   form `N.N.N.N'.  */
+   form `N.N.N.N'.  On some systems gethostbyname() knows how to do
+   this automatically.  */
 struct hostent *
 ngethostbyname (const char *name)
 {
@@ -91,42 +94,51 @@ ngethostbyname (const char *name)
   return hp;
 }
 
-/* Search for HOST in the linked list L, by hostname.  Return the
-   entry, if found, or NULL.  The search is case-insensitive.  */
-static struct host *
-search_host (struct host *l, const char *host)
-{
-  for (; l; l = l->next)
-    if (strcasecmp (l->hostname, host) == 0)
-      return l;
-  return NULL;
-}
+/* Add host name HOST with the address ADDR_TEXT to the cache.
+   Normally this means that the (HOST, ADDR_TEXT) pair will be to
+   host_name_address_map and to host_address_name_map.  (It is the
+   caller's responsibility to make sure that HOST is not already in
+   host_name_address_map.)
 
-/* Like search_host, but searches by address.  */
-static struct host *
-search_address (struct host *l, const char *address)
+   If the ADDR_TEXT has already been seen and belongs to another host,
+   HOST will be added to host_slave_master_map instead.  */
+
+static void
+add_host_to_cache (const char *host, const char *addr_text)
 {
-  for (; l; l = l->next)
+  char *canonical_name = hash_table_get (host_address_name_map, addr_text);
+  if (canonical_name)
+    {
+      DEBUGP (("Mapping %s to %s in host_slave_master_map.\n",
+	       host, canonical_name));
+      /* We've already dealt with that host under another name. */
+      hash_table_put (host_slave_master_map,
+		      xstrdup_lower (host),
+		      xstrdup_lower (canonical_name));
+    }
+  else
     {
-      int cmp = strcmp (l->realname, address);
-      if (cmp == 0)
-	return l;
-      else if (cmp > 0)
-	return NULL;
+      /* This is really the first time we're dealing with that host.  */
+      char *h_copy = xstrdup_lower (host);
+      char *a_copy = xstrdup (addr_text);
+      DEBUGP (("Caching %s <-> %s\n", h_copy, a_copy));
+      hash_table_put (host_name_address_map, h_copy, a_copy);
+      hash_table_put (host_address_name_map, a_copy, h_copy);
     }
-  return NULL;
 }
 
-/* Store the address of HOSTNAME, internet-style, to WHERE.  First
-   check for it in the host list, and (if not found), use
-   ngethostbyname to get it.
+/* Store the address of HOSTNAME, internet-style (four octets in
+   network order), to WHERE.  First try to get the address from the
+   cache; if it is not available, call the DNS functions and update
+   the cache.
 
    Return 1 on successful finding of the hostname, 0 otherwise.  */
 int
 store_hostaddress (unsigned char *where, const char *hostname)
 {
-  struct host *t;
   unsigned long addr;
+  char *addr_text;
+  char *canonical_name;
   struct hostent *hptr;
   struct in_addr in;
   char *inet_s;
@@ -134,178 +146,119 @@ store_hostaddress (unsigned char *where, const char *hostname)
   /* If the address is of the form d.d.d.d, there will be no trouble
      with it.  */
   addr = (unsigned long)inet_addr (hostname);
-  if ((int)addr == -1)
-    {
-      /* If it is not of that form, try to find it in the cache.  */
-      t = search_host (hlist, hostname);
-      if (t)
-	addr = (unsigned long)inet_addr (t->realname);
-    }
   /* If we have the numeric address, just store it.  */
   if ((int)addr != -1)
     {
-      /* ADDR is in network byte order, meaning the code works on
-         little and big endian 32-bit architectures without change.
-         On big endian 64-bit architectures we need to be careful to
-         copy the correct four bytes.  */
-      int offset = 0;
+      /* ADDR is defined to be in network byte order, meaning the code
+         works on little and big endian 32-bit architectures without
+         change.  On big endian 64-bit architectures we need to be
+         careful to copy the correct four bytes.  */
+      int offset;
+    have_addr:
 #ifdef WORDS_BIGENDIAN
       offset = sizeof (unsigned long) - 4;
+#else
+      offset = 0;
 #endif
       memcpy (where, (char *)&addr + offset, 4);
       return 1;
     }
+
+  /* By now we know that the address is not of the form d.d.d.d.  Try
+     to find it in our cache of host addresses.  */
+  addr_text = hash_table_get (host_name_address_map, hostname);
+  if (addr_text)
+    {
+      DEBUGP (("Found %s in host_name_address_map: %s\n",
+	       hostname, addr_text));
+      addr = (unsigned long)inet_addr (addr_text);
+      goto have_addr;
+    }
+
+  /* Maybe this host is known to us under another name.  If so, we'll
+     find it in host_slave_master_map, and use the master name to find
+     its address in host_name_address_map. */
+  canonical_name = hash_table_get (host_slave_master_map, hostname);
+  if (canonical_name)
+    {
+      addr_text = hash_table_get (host_name_address_map, canonical_name);
+      assert (addr_text != NULL);
+      DEBUGP (("Found %s as slave of %s -> %s\n",
+	       hostname, canonical_name, addr_text));
+      addr = (unsigned long)inet_addr (addr_text);
+      goto have_addr;
+    }
+
   /* Since all else has failed, let's try gethostbyname().  Note that
      we use gethostbyname() rather than ngethostbyname(), because we
-     *know* the address is not numerical.  */
+     already know that the address is not numerical.  */
   hptr = gethostbyname (hostname);
   if (!hptr)
     return 0;
   /* Copy the address of the host to socket description.  */
   memcpy (where, hptr->h_addr_list[0], hptr->h_length);
-  /* Now that we're here, we could as well cache the hostname for
-     future use, as in realhost().  First, we have to look for it by
-     address to know if it's already in the cache by another name.  */
+  assert (hptr->h_length == 4);
 
+  /* Now that we've gone through the truoble of calling
+     gethostbyname(), we can store this valuable information to the
+     cache.  First, we have to look for it by address to know if it's
+     already in the cache by another name.  */
   /* Originally, we copied to in.s_addr, but it appears to be missing
      on some systems.  */
   memcpy (&in, *hptr->h_addr_list, sizeof (in));
-  STRDUP_ALLOCA (inet_s, inet_ntoa (in));
-  t = search_address (hlist, inet_s);
-  if (t) /* Found in the list, as realname.  */
-    {
-      /* Set the default, 0 quality.  */
-      hlist = add_hlist (hlist, hostname, inet_s, 0);
-      return 1;
-    }
-  /* Since this is really the first time this host is encountered,
-     set quality to 1.  */
-  hlist = add_hlist (hlist, hostname, inet_s, 1);
+  inet_s = inet_ntoa (in);
+  add_host_to_cache (hostname, inet_s);
   return 1;
 }
 
-/* Add a host to the host list.  The list is sorted by addresses.  For
-   equal addresses, the entries with quality should bubble towards the
-   beginning of the list.  */
-static struct host *
-add_hlist (struct host *l, const char *nhost, const char *nreal, int quality)
-{
-  struct host *t, *old, *beg;
-
-  /* The entry goes to the beginning of the list if the list is empty
-     or the order requires it.  */
-  if (!l || (strcmp (nreal, l->realname) < 0))
-    {
-      t = (struct host *)xmalloc (sizeof (struct host));
-      t->hostname = xstrdup (nhost);
-      t->realname = xstrdup (nreal);
-      t->quality = quality;
-      t->next = l;
-      return t;
-    }
-
-  beg = l;
-  /* Second two one-before-the-last element.  */
-  while (l->next)
-    {
-      int cmp;
-      old = l;
-      l = l->next;
-      cmp = strcmp (nreal, l->realname);
-      if (cmp >= 0)
-	continue;
-      /* If the next list element is greater than s, put s between the
-	 current and the next list element.  */
-      t = (struct host *)xmalloc (sizeof (struct host));
-      old->next = t;
-      t->next = l;
-      t->hostname = xstrdup (nhost);
-      t->realname = xstrdup (nreal);
-      t->quality = quality;
-      return beg;
-    }
-  t = (struct host *)xmalloc (sizeof (struct host));
-  t->hostname = xstrdup (nhost);
-  t->realname = xstrdup (nreal);
-  t->quality = quality;
-  /* Insert the new element after the last element.  */
-  l->next = t;
-  t->next = NULL;
-  return beg;
-}
-
 /* Determine the "real" name of HOST, as perceived by Wget.  If HOST
    is referenced by more than one name, "real" name is considered to
-   be the first one encountered in the past.
-
-   If the host cannot be found in the list of already dealt-with
-   hosts, try with its INET address.  If this fails too, add it to the
-   list.  The routine does not call gethostbyname twice for the same
-   host if it can possibly avoid it.  */
+   be the first one encountered in the past.  */
 char *
 realhost (const char *host)
 {
-  struct host *l, *l_real;
   struct in_addr in;
   struct hostent *hptr;
-  char *inet_s;
+  char *master_name;
 
-  DEBUGP (("Checking for %s.\n", host));
-  /* Look for the host, looking by the host name.  */
-  l = search_host (hlist, host);
-  if (l && l->quality)		/* Found it with quality */
+  DEBUGP (("Checking for %s in host_name_address_map.\n", host));
+  if (hash_table_exists (host_name_address_map, host))
     {
-      DEBUGP (("%s was already used, by that name.\n", host));
-      /* Here we return l->hostname, not host, because of the possible
-         case differences (e.g. jaGOR.srce.hr and jagor.srce.hr are
-         the same, but we want the one that was first.  */
-      return xstrdup (l->hostname);
+      DEBUGP (("Found; %s was already used, by that name.\n", host));
+      return xstrdup_lower (host);
     }
-  else if (!l)			/* Not found, with or without quality */
-    {
-      /* The fact that gethostbyname will get called makes it
-	 necessary to store it to the list, to ensure that
-	 gethostbyname will not be called twice for the same string.
-	 However, the quality argument must be set appropriately.
-
-	 Note that add_hlist must be called *after* the realname
-	 search, or the quality would be always set to 0 */
-      DEBUGP (("This is the first time I hear about host %s by that name.\n",
-	       host));
-      hptr = ngethostbyname (host);
-      if (!hptr)
-	return xstrdup (host);
-      /* Originally, we copied to in.s_addr, but it appears to be
-         missing on some systems.  */
-      memcpy (&in, *hptr->h_addr_list, sizeof (in));
-      STRDUP_ALLOCA (inet_s, inet_ntoa (in));
-    }
-  else				/* Found, without quality */
+
+  DEBUGP (("Checking for %s in host_slave_master_map.\n", host));
+  master_name = hash_table_get (host_slave_master_map, host);
+  if (master_name)
     {
-      /* This case happens when host is on the list,
-	 but not as first entry (the one with quality).
-	 Then we just get its INET address and pick
-	 up the first entry with quality.  */
-      DEBUGP (("We've dealt with host %s, but under the name %s.\n",
-	       host, l->realname));
-      STRDUP_ALLOCA (inet_s, l->realname);
+    has_master:
+      DEBUGP (("Found; %s was already used, by the name %s.\n",
+	       host, master_name));
+      return xstrdup (master_name);
     }
 
-  /* Now we certainly have the INET address.  The following loop is
-     guaranteed to pick either an entry with quality (because it is
-     the first one), or none at all.  */
-  l_real = search_address (hlist, inet_s);
-  if (l_real)			/* Found in the list, as realname.  */
+  DEBUGP (("First time I hear about %s by that name; looking it up.\n",
+	   host));
+  hptr = ngethostbyname (host);
+  if (hptr)
     {
-      if (!l)
-	/* Set the default, 0 quality.  */
-	hlist = add_hlist (hlist, host, inet_s, 0);
-      return xstrdup (l_real->hostname);
+      char *inet_s;
+      /* Originally, we copied to in.s_addr, but it appears to be
+	 missing on some systems.  */
+      memcpy (&in, *hptr->h_addr_list, sizeof (in));
+      inet_s = inet_ntoa (in);
+
+      add_host_to_cache (host, inet_s);
+
+      /* add_host_to_cache() can establish a slave-master mapping. */
+      DEBUGP (("Checking again for %s in host_slave_master_map.\n", host));
+      master_name = hash_table_get (host_slave_master_map, host);
+      if (master_name)
+	goto has_master;
     }
-  /* Since this is really the first time this host is encountered,
-     set quality to 1.  */
-  hlist = add_hlist (hlist, host, inet_s, 1);
-  return xstrdup (host);
+
+  return xstrdup_lower (host);
 }
 
 /* Compare two hostnames (out of URL-s if the arguments are URL-s),
@@ -547,20 +500,23 @@ herrmsg (int error)
     return _("Unknown error");
 }
 
-/* Clean the host list.  This is a separate function, so we needn't
-   export HLIST and its implementation.  Ha!  */
 void
 clean_hosts (void)
 {
-  struct host *l = hlist;
+  /* host_name_address_map and host_address_name_map share the
+     strings.  Because of that, calling free_keys_and_values once
+     suffices for both.  */
+  free_keys_and_values (host_name_address_map);
+  hash_table_destroy (host_name_address_map);
+  hash_table_destroy (host_address_name_map);
+  free_keys_and_values (host_slave_master_map);
+  hash_table_destroy (host_slave_master_map);
+}
 
-  while (l)
-    {
-      struct host *p = l->next;
-      free (l->hostname);
-      free (l->realname);
-      free (l);
-      l = p;
-    }
-  hlist = NULL;
+void
+host_init (void)
+{
+  host_name_address_map = make_string_hash_table (0);
+  host_address_name_map = make_string_hash_table (0);
+  host_slave_master_map = make_string_hash_table (0);
 }
diff --git a/src/html-parse.c b/src/html-parse.c
new file mode 100644
index 00000000..b5efa7f2
--- /dev/null
+++ b/src/html-parse.c
@@ -0,0 +1,856 @@
+/* HTML parser for Wget.
+   Copyright (C) 1998, 2000 Free Software Foundation, Inc.
+
+This file is part of Wget.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or (at
+your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
+
+/* The only entry point to this module is map_html_tags(), which see.  */
+
+/* TODO:
+
+   - Allow hooks for callers to process contents outside tags.  This
+     is needed to implement handling <style> and <script>.  The
+     taginfo structure already carries the information about where the
+     tags are, but this is not enough, because one would also want to
+     skip the comments.  (The funny thing is that for <style> and
+     <script> you *don't* want to skip comments!)
+
+   - Create a test suite for regression testing. */
+
+/* HISTORY:
+
+   This is the third HTML parser written for Wget.  The first one was
+   written some time during the Geturl 1.0 beta cycle, and was very
+   inefficient and buggy.  It also contained some very complex code to
+   remember a list of parser states, because it was supposed to be
+   reentrant.  The idea was that several parsers would be running
+   concurrently, and you'd have pass the function a unique ID string
+   (for example, the URL) by which it found the relevant parser state
+   and returned the next URL.  Over-engineering at its best.
+
+   The second HTML parser was written for Wget 1.4 (the first version
+   by the name `Wget'), and was a complete rewrite.  Although the new
+   parser behaved much better and made no claims of reentrancy, it
+   still shared many of the fundamental flaws of the old version -- it
+   only regarded HTML in terms tag-attribute pairs, where the
+   attribute's value was a URL to be returned.  Any other property of
+   HTML, such as <base href=...>, or strange way to specify a URL,
+   such as <meta http-equiv=Refresh content="0; URL=..."> had to be
+   crudely hacked in -- and the caller had to be aware of these hacks.
+   Like its predecessor, this parser did not support HTML comments.
+
+   After Wget 1.5.1 was released, I set out to write a third HTML
+   parser.  The objectives of the new parser were to: (1) provide a
+   clean way to analyze HTML lexically, (2) separate interpretation of
+   the markup from the parsing process, (3) be as correct as possible,
+   e.g. correctly skipping comments and other SGML declarations, (4)
+   understand the most common errors in markup and skip them or be
+   relaxed towrds them, and (5) be reasonably efficient (no regexps,
+   minimum copying and minimum or no heap allocation).
+
+   I believe this parser meets all of the above goals.  It is
+   reasonably well structured, and could be relatively easily
+   separated from Wget and used elsewhere.  While some of its
+   intrinsic properties limit its value as a general-purpose HTML
+   parser, I believe that, with minimum modifications, it could serve
+   as a backend for one.
+
+   Due to time and other constraints, this parser was not integrated
+   into Wget until the version ???. */
+
+/* DESCRIPTION:
+
+   The single entry point of this parser is map_html_tags(), which
+   works by calling a function you specify for each tag.  The function
+   gets called with the pointer to a structure describing the tag and
+   its attributes.  */
+
+/* To test as standalone, compile with `-DSTANDALONE -I.'.  You'll
+   still need Wget headers to compile.  */
+
+#include <config.h>
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <ctype.h>
+#ifdef HAVE_STRING_H
+# include <string.h>
+#else
+# include <strings.h>
+#endif
+#include <assert.h>
+
+#include "wget.h"
+#include "html-parse.h"
+
+#ifdef STANDALONE
+# define xmalloc malloc
+# define xrealloc realloc
+#endif /* STANDALONE */
+
+/* Pool support.  For efficiency, map_html_tags() stores temporary
+   string data to a single stack-allocated pool.  If the pool proves
+   too small, additional memory is allocated/resized with
+   malloc()/realloc().  */
+
+struct pool {
+  char *contents;		/* pointer to the contents. */
+  int size;			/* size of the pool. */
+  int index;			/* next unoccupied position in
+                                   contents. */
+
+  int alloca_p;			/* whether contents was allocated
+                                   using alloca(). */
+  char *orig_contents;		/* orig_contents, allocated by
+                                   alloca().  this is used by
+                                   POOL_FREE to restore the pool to
+                                   the "initial" state. */
+  int orig_size;
+};
+
+/* Initialize the pool to hold INITIAL_SIZE bytes of storage. */
+
+#define POOL_INIT(pool, initial_size) do {		\
+  (pool).size = (initial_size);				\
+  (pool).contents = ALLOCA_ARRAY (char, (pool).size);	\
+  (pool).index = 0;					\
+  (pool).alloca_p = 1;					\
+  (pool).orig_contents = (pool).contents;		\
+  (pool).orig_size = (pool).size;			\
+} while (0)
+
+/* Grow the pool to accomodate at least SIZE new bytes.  If the pool
+   already has room to accomodate SIZE bytes of data, this is a no-op.  */
+
+#define POOL_GROW(pool, increase) do {					\
+  int PG_newsize = (pool).index + increase;				\
+  DO_REALLOC_FROM_ALLOCA ((pool).contents, (pool).size, PG_newsize,	\
+			  (pool).alloca_p, char);			\
+} while (0)
+
+/* Append text in the range [beg, end) to POOL.  No zero-termination
+   is done.  */
+
+#define POOL_APPEND(pool, beg, end) do {			\
+  const char *PA_beg = beg;					\
+  int PA_size = end - PA_beg;					\
+  POOL_GROW (pool, PA_size);					\
+  memcpy ((pool).contents + (pool).index, PA_beg, PA_size);	\
+  (pool).index += PA_size;					\
+} while (0)
+
+/* The same as the above, but with zero termination. */
+
+#define POOL_APPEND_ZT(pool, beg, end) do {			\
+  const char *PA_beg = beg;					\
+  int PA_size = end - PA_beg;					\
+  POOL_GROW (pool, PA_size + 1);				\
+  memcpy ((pool).contents + (pool).index, PA_beg, PA_size);	\
+  (pool).contents[(pool).index + PA_size] = '\0';		\
+  (pool).index += PA_size + 1;					\
+} while (0)
+
+/* Forget old pool contents.  The allocated memory is not freed. */
+#define POOL_REWIND(pool) pool.index = 0
+
+/* Free heap-allocated memory for contents of POOL.  This calls free()
+   if the memory was allocated through malloc.  It also restores
+   `contents' and `size' to their original, pre-malloc values.  That
+   way after POOL_FREE, the pool is fully usable, just as if it were
+   freshly initialized with POOL_INIT.  */
+
+#define POOL_FREE(pool) do {			\
+  if (!(pool).alloca_p)				\
+    free ((pool).contents);			\
+  (pool).contents = (pool).orig_contents;	\
+  (pool).size = (pool).orig_size;		\
+  (pool).index = 0;				\
+  (pool).alloca_p = 1;				\
+} while (0)
+
+
+#define AP_DOWNCASE		1
+#define AP_PROCESS_ENTITIES	2
+#define AP_SKIP_BLANKS		4
+
+/* Copy the text in the range [BEG, END) to POOL, optionally
+   performing operations specified by FLAGS.  FLAGS may be any
+   combination of AP_DOWNCASE, AP_PROCESS_ENTITIES and AP_SKIP_BLANKS
+   with the following meaning:
+
+   * AP_DOWNCASE -- downcase all the letters;
+
+   * AP_PROCESS_ENTITIES -- process the SGML entities and write out
+   the decoded string.  Recognized entities are &lt, &gt, &amp, &quot,
+   &nbsp and the numerical entities.
+
+   * AP_SKIP_BLANKS -- ignore blanks at the beginning and at the end
+   of text.  */
+static void
+convert_and_copy (struct pool *pool, const char *beg, const char *end, int flags)
+{
+  int old_index = pool->index;
+  int size;
+
+  /* First, skip blanks if required.  We must do this before entities
+     are processed, so that blanks can still be inserted as, for
+     instance, `&#32;'.  */
+  if (flags & AP_SKIP_BLANKS)
+    {
+      while (beg < end && ISSPACE (*beg))
+	++beg;
+      while (end > beg && ISSPACE (end[-1]))
+	--end;
+    }
+  size = end - beg;
+
+  if (flags & AP_PROCESS_ENTITIES)
+    {
+      /* Stack-allocate a copy of text, process entities and copy it
+         to the pool.  */
+      char *local_copy = (char *)alloca (size + 1);
+      const char *from = beg;
+      char *to = local_copy;
+
+      while (from < end)
+	{
+	  if (*from != '&')
+	    *to++ = *from++;
+	  else
+	    {
+	      const char *save = from;
+	      int remain;
+
+	      if (++from == end) goto lose;
+	      remain = end - from;
+
+	      if (*from == '#')
+		{
+		  int numeric;
+		  ++from;
+		  if (from == end || !ISDIGIT (*from)) goto lose;
+		  for (numeric = 0; from < end && ISDIGIT (*from); from++)
+		    numeric = 10 * numeric + (*from) - '0';
+		  if (from < end && ISALPHA (*from)) goto lose;
+		  numeric &= 0xff;
+		  *to++ = numeric;
+		}
+#define FROB(x) (remain >= (sizeof (x) - 1)			\
+		 && !memcmp (from, x, sizeof (x) - 1)		\
+		 && (*(from + sizeof (x) - 1) == ';'		\
+		     || remain == sizeof (x) - 1		\
+		     || !ISALNUM (*(from + sizeof (x) - 1))))
+	      else if (FROB ("lt"))
+		*to++ = '<', from += 2;
+	      else if (FROB ("gt"))
+		*to++ = '>', from += 2;
+	      else if (FROB ("amp"))
+		*to++ = '&', from += 3;
+	      else if (FROB ("quot"))
+		*to++ = '\"', from += 4;
+	      /* We don't implement the proposed "Added Latin 1"
+		 entities (except for nbsp), because it is unnecessary
+		 in the context of Wget, and would require hashing to
+		 work efficiently.  */
+	      else if (FROB ("nbsp"))
+		*to++ = 160, from += 4;
+	      else
+		goto lose;
+#undef FROB
+	      /* If the entity was followed by `;', we step over the
+		 `;'.  Otherwise, it was followed by either a
+		 non-alphanumeric or EOB, in which case we do nothing.	*/
+	      if (from < end && *from == ';')
+		++from;
+	      continue;
+
+	    lose:
+	      /* This was not an entity after all.  Back out.  */
+	      from = save;
+	      *to++ = *from++;
+	    }
+	}
+      *to++ = '\0';
+      POOL_APPEND (*pool, local_copy, to);
+    }
+  else
+    {
+      /* Just copy the text to the pool.  */
+      POOL_APPEND_ZT (*pool, beg, end);
+    }
+
+  if (flags & AP_DOWNCASE)
+    {
+      char *p = pool->contents + old_index;
+      for (; *p; p++)
+	*p = TOLOWER (*p);
+    }
+}
+
+/* Check whether the contents of [POS, POS+LENGTH) match any of the
+   strings in the ARRAY.  */
+static int
+array_allowed (const char **array, const char *beg, const char *end)
+{
+  int length = end - beg;
+  if (array)
+    {
+      for (; *array; array++)
+	if (length >= strlen (*array)
+	    && !strncasecmp (*array, beg, length))
+	  break;
+      if (!*array)
+	return 0;
+    }
+  return 1;
+}
+
+/* RFC1866: name [of attribute or tag] consists of letters, digits,
+   periods, or hyphens.  We also allow _, for compatibility with
+   brain-damaged generators.  */
+#define NAME_CHAR_P(x) (ISALNUM (x) || (x) == '.' || (x) == '-' || (x) == '_')
+
+/* States while advancing through comments. */
+#define AC_S_DONE	0
+#define AC_S_BACKOUT	1
+#define AC_S_BANG	2
+#define AC_S_DEFAULT	3
+#define AC_S_DCLNAME	4
+#define AC_S_DASH1	5
+#define AC_S_DASH2	6
+#define AC_S_COMMENT	7
+#define AC_S_DASH3	8
+#define AC_S_DASH4	9
+#define AC_S_QUOTE1	10
+#define AC_S_IN_QUOTE	11
+#define AC_S_QUOTE2	12
+
+#ifdef STANDALONE
+static int comment_backout_count;
+#endif
+
+/* Advance over an SGML declaration (the <!...> forms you find in HTML
+   documents).  The function returns the location after the
+   declaration.  The reason we need this is that HTML comments are
+   expressed as comments in so-called "empty declarations".
+
+   To recap: any SGML declaration may have comments associated with
+   it, e.g.
+       <!MY-DECL -- isn't this fun? -- foo bar>
+
+   An HTML comment is merely an empty declaration (<!>) with a comment
+   attached, like this:
+       <!-- some stuff here -->
+
+   Several comments may be embedded in one comment declaration:
+       <!-- have -- -- fun -->
+
+   Whitespace is allowed between and after the comments, but not
+   before the first comment.
+
+   Additionally, this function attempts to handle double quotes in
+   SGML declarations correctly.  */
+static const char *
+advance_declaration (const char *beg, const char *end)
+{
+  const char *p = beg;
+  char quote_char = '\0';	/* shut up, gcc! */
+  char ch;
+  int state = AC_S_BANG;
+
+  if (beg == end)
+    return beg;
+  ch = *p++;
+
+  /* It looked like a good idea to write this as a state machine, but
+     now I wonder...  */
+
+  while (state != AC_S_DONE && state != AC_S_BACKOUT)
+    {
+      if (p == end)
+	state = AC_S_BACKOUT;
+      switch (state)
+	{
+	case AC_S_DONE:
+	case AC_S_BACKOUT:
+	  break;
+	case AC_S_BANG:
+	  if (ch == '!')
+	    {
+	      ch = *p++;
+	      state = AC_S_DEFAULT;
+	    }
+	  else
+	    state = AC_S_BACKOUT;
+	  break;
+	case AC_S_DEFAULT:
+	  switch (ch)
+	    {
+	    case '-':
+	      state = AC_S_DASH1;
+	      break;
+	    case ' ':
+	    case '\t':
+	    case '\r':
+	    case '\n':
+	      ch = *p++;
+	      break;
+	    case '>':
+	      state = AC_S_DONE;
+	      break;
+	    case '\'':
+	    case '\"':
+	      state = AC_S_QUOTE1;
+	      break;
+	    default:
+	      if (NAME_CHAR_P (ch))
+		state = AC_S_DCLNAME;
+	      else
+		state = AC_S_BACKOUT;
+	      break;
+	    }
+	  break;
+	case AC_S_DCLNAME:
+	  if (NAME_CHAR_P (ch))
+	    ch = *p++;
+	  else if (ch == '-')
+	    state = AC_S_DASH1;
+	  else
+	    state = AC_S_DEFAULT;
+	  break;
+	case AC_S_QUOTE1:
+	  assert (ch == '\'' || ch == '\"');
+	  quote_char = ch;	/* cheating -- I really don't feel like
+				   introducing more different states for
+				   different quote characters. */
+	  ch = *p++;
+	  state = AC_S_IN_QUOTE;
+	  break;
+	case AC_S_IN_QUOTE:
+	  if (ch == quote_char)
+	    state = AC_S_QUOTE2;
+	  else
+	    ch = *p++;
+	  break;
+	case AC_S_QUOTE2:
+	  assert (ch == quote_char);
+	  ch = *p++;
+	  state = AC_S_DEFAULT;
+	  break;
+	case AC_S_DASH1:
+	  assert (ch == '-');
+	  ch = *p++;
+	  state = AC_S_DASH2;
+	  break;
+	case AC_S_DASH2:
+	  switch (ch)
+	    {
+	    case '-':
+	      ch = *p++;
+	      state = AC_S_COMMENT;
+	      break;
+	    default:
+	      state = AC_S_BACKOUT;
+	    }
+	  break;
+	case AC_S_COMMENT:
+	  switch (ch)
+	    {
+	    case '-':
+	      state = AC_S_DASH3;
+	      break;
+	    default:
+	      ch = *p++;
+	      break;
+	    }
+	  break;
+	case AC_S_DASH3:
+	  assert (ch == '-');
+	  ch = *p++;
+	  state = AC_S_DASH4;
+	  break;
+	case AC_S_DASH4:
+	  switch (ch)
+	    {
+	    case '-':
+	      ch = *p++;
+	      state = AC_S_DEFAULT;
+	      break;
+	    default:
+	      state = AC_S_COMMENT;
+	      break;
+	    }
+	  break;
+	}
+    }
+
+  if (state == AC_S_BACKOUT)
+    {
+#ifdef STANDALONE
+      ++comment_backout_count;
+#endif
+      return beg + 1;
+    }
+  return p;
+}
+
+/* Advance P (a char pointer), with the explicit intent of being able
+   to read the next character.  If this is not possible, go to finish.  */
+
+#define ADVANCE(p) do {				\
+  ++p;						\
+  if (p >= end)					\
+    goto finish;				\
+} while (0)
+
+/* Skip whitespace, if any. */
+
+#define SKIP_WS(p) do {				\
+  while (ISSPACE (*p)) {			\
+    ADVANCE (p);				\
+  }						\
+} while (0)
+
+/* Skip non-whitespace, if any. */
+
+#define SKIP_NON_WS(p) do {			\
+  while (!ISSPACE (*p)) {			\
+    ADVANCE (p);				\
+  }						\
+} while (0)
+
+#ifdef STANDALONE
+static int tag_backout_count;
+#endif
+
+/* Map MAPFUN over HTML tags in TEXT, which is SIZE characters long.
+   MAPFUN will be called with two arguments: pointer to an initialized
+   struct taginfo, and CLOSURE.
+
+   ALLOWED_TAG_NAMES should be a NULL-terminated array of tag names to
+   be processed by this function.  If it is NULL, all the tags are
+   allowed.  The same goes for attributes and ALLOWED_ATTRIBUTE_NAMES.
+
+   (Obviously, the caller can filter out unwanted tags and attributes
+   just as well, but this is just an optimization designed to avoid
+   unnecessary copying for tags/attributes which the caller doesn't
+   want to know about.  These lists are searched linearly; therefore,
+   if you're interested in a large number of tags or attributes, you'd
+   better set these to NULL and filter them out yourself with a
+   hashing process most appropriate for your application.)  */
+
+void
+map_html_tags (const char *text, int size,
+	       const char **allowed_tag_names,
+	       const char **allowed_attribute_names,
+	       void (*mapfun) (struct taginfo *, void *),
+	       void *closure)
+{
+  const char *p = text;
+  const char *end = text + size;
+
+  int attr_pair_count = 8;
+  int attr_pair_alloca_p = 1;
+  struct attr_pair *pairs = ALLOCA_ARRAY (struct attr_pair, attr_pair_count);
+  struct pool pool;
+
+  if (!size)
+    return;
+
+  POOL_INIT (pool, 256);
+
+  {
+    int nattrs, end_tag;
+    const char *tag_name_begin, *tag_name_end;
+    const char *tag_start_position;
+    int uninteresting_tag;
+
+  look_for_tag:
+    POOL_REWIND (pool);
+
+    nattrs = 0;
+    end_tag = 0;
+
+    /* Find beginning of tag.  We use memchr() instead of the usual
+       looping with ADVANCE() for speed. */
+    p = memchr (p, '<', end - p);
+    if (!p)
+      goto finish;
+
+    tag_start_position = p;
+    ADVANCE (p);
+
+    /* Establish the type of the tag (start-tag, end-tag or
+       declaration).  */
+    if (*p == '!')
+      {
+	/* This is an SGML declaration -- just skip it.  */
+	p = advance_declaration (p, end);
+	if (p == end)
+	  goto finish;
+	goto look_for_tag;
+      }
+    else if (*p == '/')
+      {
+	end_tag = 1;
+	ADVANCE (p);
+      }
+    tag_name_begin = p;
+    while (NAME_CHAR_P (*p))
+      ADVANCE (p);
+    if (p == tag_name_begin)
+      goto look_for_tag;
+    tag_name_end = p;
+    SKIP_WS (p);
+    if (end_tag && *p != '>')
+      goto backout_tag;
+
+    if (!array_allowed (allowed_tag_names, tag_name_begin, tag_name_end))
+      /* We can't just say "goto look_for_tag" here because we need
+         the loop below to properly advance over the tag's attributes.  */
+      uninteresting_tag = 1;
+    else
+      {
+	uninteresting_tag = 0;
+	convert_and_copy (&pool, tag_name_begin, tag_name_end, AP_DOWNCASE);
+      }
+
+    /* Find the attributes. */
+    while (1)
+      {
+	const char *attr_name_begin, *attr_name_end;
+	const char *attr_value_begin, *attr_value_end;
+	const char *attr_raw_value_begin, *attr_raw_value_end;
+	int operation = AP_DOWNCASE; /* stupid compiler. */
+
+	SKIP_WS (p);
+
+	/* Check for end of tag definition. */
+	if (*p == '>')
+	  break;
+
+	/* Establish bounds of attribute name. */
+	attr_name_begin = p;	/* <foo bar ...> */
+				/*      ^        */
+	while (NAME_CHAR_P (*p))
+	  ADVANCE (p);
+	attr_name_end = p;	/* <foo bar ...> */
+				/*         ^     */
+	if (attr_name_begin == attr_name_end)
+	  goto backout_tag;
+
+	/* Establish bounds of attribute value. */
+	SKIP_WS (p);
+	if (NAME_CHAR_P (*p) || *p == '>')
+	  {
+	    /* Minimized attribute syntax allows `=' to be omitted.
+               For example, <UL COMPACT> is a valid shorthand for <UL
+               COMPACT="compact">.  Even if such attributes are not
+               useful to Wget, we need to support them, so that the
+               tags containing them can be parsed correctly. */
+	    attr_raw_value_begin = attr_value_begin = attr_name_begin;
+	    attr_raw_value_end = attr_value_end = attr_name_end;
+	  }
+	else if (*p == '=')
+	  {
+	    ADVANCE (p);
+	    SKIP_WS (p);
+	    if (*p == '\"' || *p == '\'')
+	      {
+		int newline_seen = 0;
+		char quote_char = *p;
+		attr_raw_value_begin = p;
+		ADVANCE (p);
+		attr_value_begin = p; /* <foo bar="baz"> */
+				      /*           ^     */
+		while (*p != quote_char)
+		  {
+		    if (!newline_seen && *p == '\n')
+		      {
+			/* If a newline is seen within the quotes, it
+			   is most likely that someone forgot to close
+			   the quote.  In that case, we back out to
+			   the value beginning, and terminate the tag
+			   at either `>' or the delimiter, whichever
+			   comes first.  Such a tag terminated at `>'
+			   is discarded.  */
+			p = attr_value_begin;
+			newline_seen = 1;
+			continue;
+		      }
+		    else if (newline_seen && *p == '>')
+		      break;
+		    ADVANCE (p);
+		  }
+		attr_value_end = p; /* <foo bar="baz"> */
+				    /*              ^  */
+		if (*p == quote_char)
+		  ADVANCE (p);
+		else
+		  goto look_for_tag;
+		attr_raw_value_end = p;	/* <foo bar="baz"> */
+					/*               ^ */
+		/* The AP_SKIP_BLANKS part is not entirely correct,
+		   because we don't want to skip blanks for all the
+		   attribute values.  */
+		operation = AP_PROCESS_ENTITIES | AP_SKIP_BLANKS;
+	      }
+	    else
+	      {
+		attr_value_begin = p; /* <foo bar=baz> */
+				      /*          ^    */
+		/* According to SGML, a name token should consist only
+		   of alphanumerics, . and -.  However, this is often
+		   violated by, for instance, `%' in `width=75%'.
+		   We'll be liberal and allow just about anything as
+		   an attribute value.  */
+		while (!ISSPACE (*p) && *p != '>')
+		  ADVANCE (p);
+		attr_value_end = p; /* <foo bar=baz qux=quix> */
+				    /*             ^          */
+		if (attr_value_begin == attr_value_end)
+		  /* <foo bar=> */
+		  /*          ^ */
+		  goto backout_tag;
+		attr_raw_value_begin = attr_value_begin;
+		attr_raw_value_end = attr_value_end;
+		operation = AP_PROCESS_ENTITIES;
+	      }
+	  }
+	else
+	  {
+	    /* We skipped the whitespace and found something that is
+	       neither `=' nor the beginning of the next attribute's
+	       name.  Back out.  */
+	    goto backout_tag;	/* <foo bar /... */
+				/*          ^    */
+	  }
+
+	/* If we're not interested in the tag, don't bother with any
+           of the attributes.  */
+	if (uninteresting_tag)
+	  continue;
+
+	/* If we aren't interested in the attribute, skip it.  We
+           cannot do this test any sooner, because our text pointer
+           needs to correctly advance over the attribute.  */
+	if (allowed_attribute_names
+	    && !array_allowed (allowed_attribute_names, attr_name_begin,
+			       attr_name_end))
+	  continue;
+
+	DO_REALLOC_FROM_ALLOCA (pairs, attr_pair_count, nattrs + 1,
+				attr_pair_alloca_p, struct attr_pair);
+
+	pairs[nattrs].name_pool_index = pool.index;
+	convert_and_copy (&pool, attr_name_begin, attr_name_end, AP_DOWNCASE);
+
+	pairs[nattrs].value_pool_index = pool.index;
+	convert_and_copy (&pool, attr_value_begin, attr_value_end, operation);
+	pairs[nattrs].value_raw_beginning = attr_raw_value_begin;
+	pairs[nattrs].value_raw_size = (attr_raw_value_end
+					- attr_raw_value_begin);
+	++nattrs;
+      }
+
+    if (uninteresting_tag)
+      {
+	ADVANCE (p);
+	goto look_for_tag;
+      }
+
+    /* By now, we have a valid tag with a name and zero or more
+       attributes.  Fill in the data and call the mapper function.  */
+    {
+      int i;
+      struct taginfo taginfo;
+
+      taginfo.name      = pool.contents;
+      taginfo.end_tag_p = end_tag;
+      taginfo.nattrs    = nattrs;
+      /* We fill in the char pointers only now, when pool can no
+	 longer get realloc'ed.  If we did that above, we could get
+	 hosed by reallocation.  Obviously, after this point, the pool
+	 may no longer be grown.  */
+      for (i = 0; i < nattrs; i++)
+	{
+	  pairs[i].name = pool.contents + pairs[i].name_pool_index;
+	  pairs[i].value = pool.contents + pairs[i].value_pool_index;
+	}
+      taginfo.attrs = pairs;
+      taginfo.start_position = tag_start_position;
+      taginfo.end_position   = p + 1;
+      /* Ta-dam! */
+      (*mapfun) (&taginfo, closure);
+      ADVANCE (p);
+    }
+    goto look_for_tag;
+
+  backout_tag:
+#ifdef STANDALONE
+    ++tag_backout_count;
+#endif
+    /* The tag wasn't really a tag.  Treat its contents as ordinary
+       data characters. */
+    p = tag_start_position + 1;
+    goto look_for_tag;
+  }
+
+ finish:
+  POOL_FREE (pool);
+  if (!attr_pair_alloca_p)
+    free (pairs);
+}
+
+#undef ADVANCE
+#undef SKIP_WS
+#undef SKIP_NON_WS
+
+#ifdef STANDALONE
+static void
+test_mapper (struct taginfo *taginfo, void *arg)
+{
+  int i;
+
+  printf ("%s%s", taginfo->end_tag_p ? "/" : "", taginfo->name);
+  for (i = 0; i < taginfo->nattrs; i++)
+    printf (" %s=%s", taginfo->attrs[i].name, taginfo->attrs[i].value);
+  putchar ('\n');
+  ++*(int *)arg;
+}
+
+int main ()
+{
+  int size = 256;
+  char *x = (char *)xmalloc (size);
+  int length = 0;
+  int read_count;
+  int tag_counter = 0;
+
+  while ((read_count = fread (x + length, 1, size - length, stdin)))
+    {
+      length += read_count;
+      size <<= 1;
+      x = (char *)xrealloc (x, size);
+    }
+
+  map_html_tags (x, length, NULL, NULL, test_mapper, &tag_counter);
+  printf ("TAGS: %d\n", tag_counter);
+  printf ("Tag backouts:     %d\n", tag_backout_count);
+  printf ("Comment backouts: %d\n", comment_backout_count);
+  return 0;
+}
+#endif /* STANDALONE */
diff --git a/src/html-parse.h b/src/html-parse.h
new file mode 100644
index 00000000..5810a2af
--- /dev/null
+++ b/src/html-parse.h
@@ -0,0 +1,44 @@
+/* Declarations for html-parse.c.
+   Copyright (C) 1998 Free Software Foundation, Inc.
+
+This file is part of Wget.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
+
+struct attr_pair {
+  char *name;			/* attribute name */
+  char *value;			/* attribute value */
+
+  /* Needed for URL conversion; the places where the value begins and
+     ends, including the quotes and everything. */
+  const char *value_raw_beginning;
+  int value_raw_size;
+
+  /* Used internally by map_html_tags. */
+  int name_pool_index, value_pool_index;
+};
+
+struct taginfo {
+  char *name;			/* tag name */
+  int end_tag_p;		/* whether this is an end-tag */
+  int nattrs;			/* number of attributes */
+  struct attr_pair *attrs;	/* attributes */
+
+  const char *start_position;	/* start position of tag */
+  const char *end_position;	/* end position of tag */
+};
+
+void map_html_tags PARAMS ((const char *, int, const char **, const char **,
+			    void (*) (struct taginfo *, void *), void *));
diff --git a/src/html-url.c b/src/html-url.c
new file mode 100644
index 00000000..0441b470
--- /dev/null
+++ b/src/html-url.c
@@ -0,0 +1,569 @@
+/* Collect URLs from HTML source.
+   Copyright (C) 1998, 2000 Free Software Foundation, Inc.
+
+This file is part of Wget.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2 of the License, or
+(at your option) any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
+
+#include <config.h>
+
+#include <stdio.h>
+#ifdef HAVE_STRING_H
+# include <string.h>
+#else
+# include <strings.h>
+#endif
+#include <stdlib.h>
+#include <ctype.h>
+#include <errno.h>
+#include <assert.h>
+
+#include "wget.h"
+#include "html-parse.h"
+#include "url.h"
+#include "utils.h"
+
+#ifndef errno
+extern int errno;
+#endif
+
+enum tag_category { TC_LINK, TC_SPEC };
+
+/* Here we try to categorize the known tags.  Each tag has its ID and
+   cetegory.  Category TC_LINK means that one or more of its
+   attributes contain links that should be retrieved.  TC_SPEC means
+   that the tag is specific in some way, and has to be handled
+   specially. */
+static struct {
+  const char *name;
+  enum tag_category category;
+} known_tags[] = {
+#define TAG_A		0
+  { "a",	TC_LINK },
+#define TAG_APPLET	1
+  { "applet",	TC_LINK },
+#define TAG_AREA	2
+  { "area",	TC_LINK },
+#define TAG_BASE	3
+  { "base",	TC_SPEC },
+#define TAG_BGSOUND	4
+  { "bgsound",	TC_LINK },
+#define TAG_BODY	5
+  { "body",	TC_LINK },
+#define TAG_EMBED	6
+  { "embed",	TC_LINK },
+#define TAG_FIG		7
+  { "fig",	TC_LINK },
+#define TAG_FRAME	8
+  { "frame",	TC_LINK },
+#define TAG_IFRAME	9
+  { "iframe",	TC_LINK },
+#define TAG_IMG		10
+  { "img",	TC_LINK },
+#define TAG_INPUT	11
+  { "input",	TC_LINK },
+#define TAG_LAYER	12
+  { "layer",	TC_LINK },
+#define TAG_LINK	13
+  { "link",	TC_SPEC },
+#define TAG_META	14
+  { "meta",	TC_SPEC },
+#define TAG_OVERLAY	15
+  { "overlay",	TC_LINK },
+#define TAG_SCRIPT	16
+  { "script",	TC_LINK },
+#define TAG_TABLE	17
+  { "table",	TC_LINK },
+#define TAG_TD		18
+  { "td",	TC_LINK },
+#define TAG_TH		19
+  { "th",	TC_LINK }
+};
+
+/* Flags for specific url-attr pairs handled through TC_LINK: */
+#define AF_EXTERNAL 1
+
+/* For tags handled by TC_LINK: attributes that contain URLs to
+   download. */
+static struct {
+  int tagid;
+  const char *attr_name;
+  int flags;
+} url_tag_attr_map[] = {
+  { TAG_A,		"href",		AF_EXTERNAL },
+  { TAG_APPLET,		"code",		0 },
+  { TAG_AREA,		"href",		AF_EXTERNAL },
+  { TAG_BGSOUND,	"src",		0 },
+  { TAG_BODY,		"background",	0 },
+  { TAG_EMBED,		"src",		0 },
+  { TAG_FIG,		"src",		0 },
+  { TAG_FRAME,		"src",		0 },
+  { TAG_IFRAME,		"src",		0 },
+  { TAG_IMG,		"href",		0 },
+  { TAG_IMG,		"lowsrc",	0 },
+  { TAG_IMG,		"src",		0 },
+  { TAG_INPUT,		"src",		0 },
+  { TAG_LAYER,		"src",		0 },
+  { TAG_OVERLAY,	"src",		0 },
+  { TAG_SCRIPT,		"src",		0 },
+  { TAG_TABLE,		"background",	0 },
+  { TAG_TD,		"background",	0 },
+  { TAG_TH,		"background",	0 }
+};
+
+/* The lists of interesting tags and attributes are built dynamically,
+   from the information above.  However, some places in the code refer
+   to the attributes not mentioned here.  We add them manually.  */
+static const char *additional_attributes[] = {
+  "rel",			/* for TAG_LINK */
+  "http-equiv",			/* for TAG_META */
+  "name",			/* for TAG_META */
+  "content"			/* for TAG_META */
+};
+
+static const char **interesting_tags;
+static const char **interesting_attributes;
+
+void
+init_interesting (void)
+{
+  /* Init the variables interesting_tags and interesting_attributes
+     that are used by the HTML parser to know which tags and
+     attributes we're interested in.  We initialize this only once,
+     for performance reasons.
+
+     Here we also make sure that what we put in interesting_tags
+     matches the user's preferences as specified through --ignore-tags
+     and --follow-tags.  */
+
+  {
+    int i, ind = 0;
+    int size = ARRAY_SIZE (known_tags);
+    interesting_tags = (const char **)xmalloc ((size + 1) * sizeof (char *));
+
+    for (i = 0; i < size; i++)
+      {
+	const char *name = known_tags[i].name;
+
+	/* Normally here we could say:
+	   interesting_tags[i] = name;
+	   But we need to respect the settings of --ignore-tags and
+	   --follow-tags, so the code gets a bit harier.  */
+
+	if (opt.ignore_tags)
+	  {
+	    /* --ignore-tags was specified.  Do not match these
+	       specific tags.  --ignore-tags takes precedence over
+	       --follow-tags, so we process --ignore first and fall
+	       through if there's no match. */
+	    int j, lose = 0;
+	    for (j = 0; opt.ignore_tags[j] != NULL; j++)
+	      /* Loop through all the tags this user doesn't care
+                 about. */
+	      if (strcasecmp(opt.ignore_tags[j], name) == EQ)
+		{
+		  lose = 1;
+		  break;
+		}
+	    if (lose)
+	      continue;
+	  }
+
+	if (opt.follow_tags)
+	  {
+	    /* --follow-tags was specified.  Only match these specific
+	       tags, so return FALSE if we don't match one of them. */
+	    int j, win = 0;
+	    for (j = 0; opt.follow_tags[j] != NULL; j++)
+	      /* Loop through all the tags this user cares about. */
+	      if (strcasecmp(opt.follow_tags[j], name) == EQ)
+		{
+		  win = 1;
+		  break;
+		}
+	    if (!win)
+	      continue;		/* wasn't one of the explicitly
+                                   desired tags */
+	  }
+
+	/* If we get to here, --follow-tags isn't being used or the
+	   tag is among the ones that are follwed, and --ignore-tags,
+	   if specified, didn't include this tag, so it's an
+	   "interesting" one. */
+	interesting_tags[ind++] = name;
+      }
+    interesting_tags[ind] = NULL;
+  }
+
+  /* The same for attributes, except we loop through url_tag_attr_map.
+     Here we also need to make sure that the list of attributes is
+     unique, and to include the attributes from additional_attributes.  */
+  {
+    int i, ind;
+    const char **att = xmalloc ((ARRAY_SIZE (additional_attributes) + 1)
+				* sizeof (char *));
+    /* First copy the "additional" attributes. */
+    for (i = 0; i < ARRAY_SIZE (additional_attributes); i++)
+      att[i] = additional_attributes[i];
+    ind = i;
+    att[ind] = NULL;
+    for (i = 0; i < ARRAY_SIZE (url_tag_attr_map); i++)
+      {
+	int j, seen = 0;
+	const char *look_for = url_tag_attr_map[i].attr_name;
+	for (j = 0; j < ind - 1; j++)
+	  if (!strcmp (att[j], look_for))
+	    {
+	      seen = 1;
+	      break;
+	    }
+	if (!seen)
+	  {
+	    att = xrealloc (att, (ind + 2) * sizeof (*att));
+	    att[ind++] = look_for;
+	    att[ind] = NULL;
+	  }
+      }
+    interesting_attributes = att;
+  }
+}
+
+static int
+find_tag (const char *tag_name)
+{
+  int i;
+
+  /* This is linear search; if the number of tags grow, we can switch
+     to binary search.  */
+
+  for (i = 0; i < ARRAY_SIZE (known_tags); i++)
+    {
+      int cmp = strcasecmp (known_tags[i].name, tag_name);
+      /* known_tags are sorted alphabetically, so we can
+         micro-optimize.  */
+      if (cmp > 0)
+	break;
+      else if (cmp == 0)
+	return i;
+    }
+  return -1;
+}
+
+/* Find the value of attribute named NAME in the taginfo TAG.  If the
+   attribute is not present, return NULL.  If ATTRID is non-NULL, the
+   exact identity of the attribute will be returned.  */
+static char *
+find_attr (struct taginfo *tag, const char *name, int *attrid)
+{
+  int i;
+  for (i = 0; i < tag->nattrs; i++)
+    if (!strcasecmp (tag->attrs[i].name, name))
+      {
+	if (attrid)
+	  *attrid = i;
+	return tag->attrs[i].value;
+      }
+  return NULL;
+}
+
+struct collect_urls_closure {
+  char *text;			/* HTML text. */
+  char *base;			/* Base URI of the document, possibly
+				   changed through <base href=...>. */
+  urlpos *head, *tail;		/* List of URLs */
+  const char *parent_base;	/* Base of the current document. */
+  const char *document_file;	/* File name of this document. */
+  int dash_p_leaf_HTML;		/* Whether -p is specified, and this
+                                   document is the "leaf" node of the
+                                   HTML tree. */
+  int nofollow;			/* whether NOFOLLOW was specified in a
+                                   <meta name=robots> tag. */
+};
+
+/* Resolve LINK_URI and append it to closure->tail.  TAG and ATTRID
+   are the necessary context to store the position and size.  */
+
+static void
+handle_link (struct collect_urls_closure *closure, const char *link_uri,
+	     struct taginfo *tag, int attrid)
+{
+  int no_proto = !has_proto (link_uri);
+  urlpos *newel;
+
+  const char *base = closure->base ? closure->base : closure->parent_base;
+  char *complete_uri;
+
+  char *fragment = strrchr (link_uri, '#');
+
+  if (fragment)
+    {
+      /* Nullify the fragment identifier, i.e. everything after the
+         last occurrence of `#', inclusive.  This copying is
+         relatively inefficient, but it doesn't matter because
+         fragment identifiers don't come up all that often.  */
+      int hashlen = fragment - link_uri;
+      char *p = alloca (hashlen + 1);
+      memcpy (p, link_uri, hashlen);
+      p[hashlen] = '\0';
+      link_uri = p;
+    }
+
+  if (!base)
+    {
+      if (no_proto)
+	{
+	  /* We have no base, and the link does not have a protocol or
+             a host attached to it.  Nothing we can do.  */
+	  /* #### Should we print a warning here?  Wget 1.5.x used to.  */
+	  return;
+	}
+      else
+	complete_uri = xstrdup (link_uri);
+    }
+  else
+    complete_uri = url_concat (base, link_uri);
+
+  DEBUGP (("%s: merge(\"%s\", \"%s\") -> %s\n",
+	   closure->document_file, base ? base : "(null)",
+	   link_uri, complete_uri));
+
+  newel = (urlpos *)xmalloc (sizeof (urlpos));
+
+  memset (newel, 0, sizeof (*newel));
+  newel->next = NULL;
+  newel->url = complete_uri;
+  newel->pos = tag->attrs[attrid].value_raw_beginning - closure->text;
+  newel->size = tag->attrs[attrid].value_raw_size;
+
+  /* A URL is relative if the host and protocol are not named, and the
+     name does not start with `/'.
+     #### This logic might need some rethinking.  */
+  if (no_proto && *link_uri != '/')
+    newel->flags |= (URELATIVE | UNOPROTO);
+  else if (no_proto)
+    newel->flags |= UNOPROTO;
+
+  if (closure->tail)
+    {
+      closure->tail->next = newel;
+      closure->tail = newel;
+    }
+  else
+    closure->tail = closure->head = newel;
+}
+
+/* #### Document what this does.
+   #### It would be nice to split this into several functions.  */
+
+static void
+collect_tags_mapper (struct taginfo *tag, void *arg)
+{
+  struct collect_urls_closure *closure = (struct collect_urls_closure *)arg;
+  int tagid = find_tag (tag->name);
+  assert (tagid != -1);
+
+  switch (known_tags[tagid].category)
+    {
+    case TC_LINK:
+      {
+	int i;
+	int size = ARRAY_SIZE (url_tag_attr_map);
+	for (i = 0; i < size; i++)
+	  if (url_tag_attr_map[i].tagid == tagid)
+	    break;
+	/* We've found the index of url_tag_attr_map where the
+           attributes of our tags begin.  Now, look for every one of
+           them, and handle it.  */
+	for (; (i < size && url_tag_attr_map[i].tagid == tagid); i++)
+	  {
+	    char *attr_value;
+	    int id;
+	    if (closure->dash_p_leaf_HTML
+		&& (url_tag_attr_map[i].flags & AF_EXTERNAL))
+	      /* If we're at a -p leaf node, we don't want to retrieve
+                 links to references we know are external, such as <a
+                 href=...>.  */
+	      continue;
+
+	    /* This find_attr() buried in a loop may seem inefficient
+               (O(n^2)), but it's not, since the number of attributes
+               (n) we loop over is extremely small.  In the worst case
+               of IMG with all its possible attributes, n^2 will be
+               only 9.  */
+	    attr_value = find_attr (tag, url_tag_attr_map[i].attr_name, &id);
+	    if (attr_value)
+	      handle_link (closure, attr_value, tag, id);
+	  }
+      }
+      break;
+    case TC_SPEC:
+      switch (tagid)
+	{
+	case TAG_BASE:
+	  {
+	    char *newbase = find_attr (tag, "href", NULL);
+	    if (!newbase)
+	      break;
+	    if (closure->base)
+	      free (closure->base);
+	    if (closure->parent_base)
+	      closure->base = url_concat (closure->parent_base, newbase);
+	    else
+	      closure->base = xstrdup (newbase);
+	  }
+	  break;
+	case TAG_LINK:
+	  {
+	    int id;
+	    char *rel  = find_attr (tag, "rel", NULL);
+	    char *href = find_attr (tag, "href", &id);
+	    if (href)
+	      {
+		/* In the normal case, all <link href=...> tags are
+		   fair game.
+
+		   In the special case of when -p is active, however,
+		   and we're at a leaf node (relative to the -l
+		   max. depth) in the HTML document tree, the only
+		   <LINK> tag we'll follow is a <LINK REL=
+		   "stylesheet">, as it's necessary for displaying
+		   this document properly.  We won't follow other
+		   <LINK> tags, like <LINK REL="home">, for instance,
+		   as they refer to external documents.  */
+		if (!closure->dash_p_leaf_HTML
+		    || (rel && !strcasecmp (rel, "stylesheet")))
+		  handle_link (closure, href, tag, id);
+	      }
+	  }
+	  break;
+	case TAG_META:
+	  /* Some pages use a META tag to specify that the page be
+	     refreshed by a new page after a given number of seconds.
+	     The general format for this is:
+
+	     <meta http-equiv=Refresh content="NUMBER; URL=index2.html">
+
+	     So we just need to skip past the "NUMBER; URL=" garbage
+	     to get to the URL.  */
+	  {
+	    int id;
+	    char *name = find_attr (tag, "name", NULL);
+	    char *http_equiv = find_attr (tag, "http-equiv", &id);
+	    if (http_equiv && !strcasecmp (http_equiv, "refresh"))
+	      {
+		char *refresh = find_attr (tag, "content", NULL);
+		char *p = refresh;
+		int offset;
+		while (ISDIGIT (*p))
+		  ++p;
+		if (*p++ != ';')
+		  return;
+		while (ISSPACE (*p))
+		  ++p;
+		if (!(TOUPPER (*p) == 'U'
+		      && TOUPPER (*(p + 1)) == 'R'
+		      && TOUPPER (*(p + 2)) == 'L'
+		      && *(p + 3) == '='))
+		  return;
+		p += 4;
+		while (ISSPACE (*p))
+		  ++p;
+		offset = p - refresh;
+		tag->attrs[id].value_raw_beginning += offset;
+		tag->attrs[id].value_raw_size -= offset;
+		handle_link (closure, p, tag, id);
+	      }
+	    else if (name && !strcasecmp (name, "robots"))
+	      {
+		/* Handle stuff like:
+		   <meta name="robots" content="index,nofollow"> */
+		char *content = find_attr (tag, "content", NULL);
+		if (!content)
+		  return;
+		if (!strcasecmp (content, "none"))
+		  closure->nofollow = 1;
+		else
+		  {
+		    while (*content)
+		      {
+			/* Find the next occurrence of ',' or the end of
+			   the string.  */
+			char *end = strchr (content, ',');
+			if (end)
+			  ++end;
+			else
+			  end = content + strlen (content);
+			if (!strncasecmp (content, "nofollow", end - content))
+			  closure->nofollow = 1;
+			content = end;
+		      }
+		  }
+	      }
+	  }
+	  break;
+	default:
+	  /* Category is TC_SPEC, but tag name is unhandled.  This
+             must not be.  */
+	  abort ();
+	}
+      break;
+    }
+}
+
+/* Scan FILE, retrieving links to HTML documents from it.  Each link is 
+
+  Similar to get_urls_file, but for HTML files.  FILE is scanned as
+   an HTML document.  get_urls_html() constructs the URLs from the
+   relative href-s.
+
+   If SILENT is non-zero, do not barf on baseless relative links.  */
+urlpos *
+get_urls_html (const char *file, const char *this_url, int dash_p_leaf_HTML,
+	       int *meta_disallow_follow)
+{
+  struct file_memory *fm;
+  struct collect_urls_closure closure;
+
+  /* Load the file. */
+  fm = read_file (file);
+  if (!fm)
+    {
+      logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
+      return NULL;
+    }
+  DEBUGP (("Loaded %s (size %ld).\n", file, fm->length));
+
+  closure.text = fm->content;
+  closure.head = closure.tail = NULL;
+  closure.base = NULL;
+  closure.parent_base = this_url ? this_url : opt.base_href;
+  closure.document_file = file;
+  closure.dash_p_leaf_HTML = dash_p_leaf_HTML;
+  closure.nofollow = 0;
+
+  if (!interesting_tags)
+    init_interesting ();
+
+  map_html_tags (fm->content, fm->length, interesting_tags,
+		 interesting_attributes, collect_tags_mapper, &closure);
+
+  DEBUGP (("no-follow in %s: %d\n", file, closure.nofollow));
+  if (meta_disallow_follow)
+    *meta_disallow_follow = closure.nofollow;
+
+  FREE_MAYBE (closure.base);
+  read_file_free (fm);
+  return closure.head;
+}
diff --git a/src/http.c b/src/http.c
index 48c1bc5e..912f4923 100644
--- a/src/http.c
+++ b/src/http.c
@@ -254,6 +254,85 @@ http_process_type (const char *hdr, void *arg)
   return 1;
 }
 
+/* Check whether the `Connection' header is set to "keep-alive". */
+static int
+http_process_connection (const char *hdr, void *arg)
+{
+  int *flag = (int *)arg;
+  if (!strcasecmp (hdr, "Keep-Alive"))
+    *flag = 1;
+  return 1;
+}
+
+/* Persistent connections (pc). */
+
+static unsigned char pc_last_host[4];
+static unsigned short pc_last_port;
+static int pc_last_fd;
+
+static void
+register_persistent (const char *host, unsigned short port, int fd)
+{
+  if (!store_hostaddress (pc_last_host, host))
+    return;
+  pc_last_port = port;
+  pc_last_fd = fd;
+}
+
+static void
+invalidate_persistent (void)
+{
+  pc_last_port = 0;
+}
+
+static int
+persistent_available_p (const char *host, unsigned short port)
+{
+  unsigned char this_host[4];
+  if (port != pc_last_port)
+    return 0;
+  if (!store_hostaddress (this_host, host))
+    return 0;
+  if (memcmp (pc_last_host, this_host, 4))
+    return 0;
+  if (!test_socket_open (pc_last_fd))
+    {
+      invalidate_persistent ();
+      return 0;
+    }
+  return 1;
+}
+
+/* The idea behind these two CLOSE macros is to distinguish between
+   two cases: one when the job we've been doing is finished, and we
+   want to close the connection and leave, and two when something is
+   seriously wrong and we're closing the connection as part of
+   cleanup.
+
+   In case of keep_alive, CLOSE_FINISH should leave the connection
+   open, while CLOSE_INVALIDATE should still close it.
+
+   The semantic difference between the flags `keep_alive' and
+   `reused_connection' is that keep_alive defines the state of HTTP:
+   whether the connection *will* be preservable.  reused_connection,
+   on the other hand, reflects the present: whether the *current*
+   connection is the result of preserving.  */
+
+#define CLOSE_FINISH(fd) do {			\
+  if (!keep_alive)				\
+    {						\
+      CLOSE (fd);				\
+      if (reused_connection)			\
+	invalidate_persistent ();		\
+    }						\
+} while (0)
+
+#define CLOSE_INVALIDATE(fd) do {		\
+  CLOSE (fd);					\
+  if (reused_connection)			\
+    invalidate_persistent ();			\
+} while (0)
+
 
 struct http_stat
 {
@@ -317,6 +396,8 @@ gethttp (struct urlinfo *u, struct http_stat *hs, int *dt)
   FILE *fp;
   int auth_tried_already;
   struct rbuf rbuf;
+  int keep_alive, http_keep_alive_1, http_keep_alive_2;
+  int reused_connection;
 
   if (!(*dt & HEAD_ONLY))
     /* If we're doing a GET on the URL, as opposed to just a HEAD, we need to
@@ -329,6 +410,9 @@ gethttp (struct urlinfo *u, struct http_stat *hs, int *dt)
  again:
   /* We need to come back here when the initial attempt to retrieve
      without authorization header fails.  */
+  keep_alive = 0;
+  http_keep_alive_1 = http_keep_alive_2 = 0;
+  reused_connection = 0;
 
   /* Initialize certain elements of struct http_stat.  */
   hs->len = 0L;
@@ -345,40 +429,49 @@ gethttp (struct urlinfo *u, struct http_stat *hs, int *dt)
     ou = u;
 
   /* First: establish the connection.  */
-  logprintf (LOG_VERBOSE, _("Connecting to %s:%hu... "), u->host, u->port);
-  err = make_connection (&sock, u->host, u->port);
-  switch (err)
+  if (u->proxy || !persistent_available_p (u->host, u->port))
     {
-    case HOSTERR:
-      logputs (LOG_VERBOSE, "\n");
-      logprintf (LOG_NOTQUIET, "%s: %s.\n", u->host, herrmsg (h_errno));
-      return HOSTERR;
-      break;
-    case CONSOCKERR:
-      logputs (LOG_VERBOSE, "\n");
-      logprintf (LOG_NOTQUIET, "socket: %s\n", strerror (errno));
-      return CONSOCKERR;
-      break;
-    case CONREFUSED:
-      logputs (LOG_VERBOSE, "\n");
-      logprintf (LOG_NOTQUIET,
-		 _("Connection to %s:%hu refused.\n"), u->host, u->port);
-      CLOSE (sock);
-      return CONREFUSED;
-    case CONERROR:
-      logputs (LOG_VERBOSE, "\n");
-      logprintf (LOG_NOTQUIET, "connect: %s\n", strerror (errno));
-      CLOSE (sock);
-      return CONERROR;
-      break;
-    case NOCONERROR:
-      /* Everything is fine!  */
-      logputs (LOG_VERBOSE, _("connected!\n"));
-      break;
-    default:
-      abort ();
-      break;
-    } /* switch */
+      logprintf (LOG_VERBOSE, _("Connecting to %s:%hu... "), u->host, u->port);
+      err = make_connection (&sock, u->host, u->port);
+      switch (err)
+	{
+	case HOSTERR:
+	  logputs (LOG_VERBOSE, "\n");
+	  logprintf (LOG_NOTQUIET, "%s: %s.\n", u->host, herrmsg (h_errno));
+	  return HOSTERR;
+	  break;
+	case CONSOCKERR:
+	  logputs (LOG_VERBOSE, "\n");
+	  logprintf (LOG_NOTQUIET, "socket: %s\n", strerror (errno));
+	  return CONSOCKERR;
+	  break;
+	case CONREFUSED:
+	  logputs (LOG_VERBOSE, "\n");
+	  logprintf (LOG_NOTQUIET,
+		     _("Connection to %s:%hu refused.\n"), u->host, u->port);
+	  CLOSE (sock);
+	  return CONREFUSED;
+	case CONERROR:
+	  logputs (LOG_VERBOSE, "\n");
+	  logprintf (LOG_NOTQUIET, "connect: %s\n", strerror (errno));
+	  CLOSE (sock);
+	  return CONERROR;
+	  break;
+	case NOCONERROR:
+	  /* Everything is fine!  */
+	  logputs (LOG_VERBOSE, _("connected!\n"));
+	  break;
+	default:
+	  abort ();
+	  break;
+	}
+    }
+  else
+    {
+      logprintf (LOG_VERBOSE, _("Reusing connection to %s:%hu.\n"), u->host, u->port);
+      sock = pc_last_fd;
+      reused_connection = 1;
+    }
 
   if (u->proxy)
     path = u->proxy->url;
@@ -487,6 +580,7 @@ gethttp (struct urlinfo *u, struct http_stat *hs, int *dt)
 User-Agent: %s\r\n\
 Host: %s%s\r\n\
 Accept: %s\r\n\
+Connection: Keep-Alive\r\n\
 %s%s%s%s%s%s\r\n",
 	   command, path, useragent, remhost,
 	   host_port ? host_port : "",
@@ -505,8 +599,9 @@ Accept: %s\r\n\
   num_written = iwrite (sock, request, strlen (request));
   if (num_written < 0)
     {
-      logputs (LOG_VERBOSE, _("Failed writing HTTP request.\n"));
-      CLOSE (sock);
+      logprintf (LOG_VERBOSE, _("Failed writing HTTP request: %s.\n"),
+		 strerror (errno));
+      CLOSE_INVALIDATE (sock);
       return WRITEFAILED;
     }
   logprintf (LOG_VERBOSE, _("%s request sent, awaiting response... "),
@@ -553,7 +648,7 @@ Accept: %s\r\n\
 	  FREE_MAYBE (type);
 	  FREE_MAYBE (hs->newloc);
 	  FREE_MAYBE (all_headers);
-	  CLOSE (sock);
+	  CLOSE_INVALIDATE (sock);
 	  return HEOF;
 	}
       else if (status == HG_ERROR)
@@ -565,7 +660,7 @@ Accept: %s\r\n\
 	  FREE_MAYBE (type);
 	  FREE_MAYBE (hs->newloc);
 	  FREE_MAYBE (all_headers);
-	  CLOSE (sock);
+	  CLOSE_INVALIDATE (sock);
 	  return HERR;
 	}
 
@@ -672,12 +767,32 @@ Accept: %s\r\n\
 	      goto done_header;
 	    }
 	}
+      /* Check for the `Keep-Alive' header. */
+      if (!http_keep_alive_1)
+	{
+	  if (header_process (hdr, "Keep-Alive", header_exists,
+			      &http_keep_alive_1))
+	    goto done_header;
+	}
+      /* Check for `Connection: Keep-Alive'. */
+      if (!http_keep_alive_2)
+	{
+	  if (header_process (hdr, "Connection", http_process_connection,
+			      &http_keep_alive_2))
+	    goto done_header;
+	}
     done_header:
       free (hdr);
     }
 
   logputs (LOG_VERBOSE, "\n");
 
+  if (contlen != -1
+      && (http_keep_alive_1 || http_keep_alive_2))
+    keep_alive = 1;
+  if (keep_alive && !reused_connection)
+    register_persistent (u->host, u->port, sock);
+
   if ((statcode == HTTP_STATUS_UNAUTHORIZED)
       && authenticate_h)
     {
@@ -685,7 +800,7 @@ Accept: %s\r\n\
       FREE_MAYBE (type);
       type = NULL;
       FREEHSTAT (*hs);
-      CLOSE (sock);
+      CLOSE_FINISH (sock);
       if (auth_tried_already)
 	{
 	  /* If we have tried it already, then there is not point
@@ -753,7 +868,7 @@ Accept: %s\r\n\
       FREE_MAYBE (type);
       FREE_MAYBE (hs->newloc);
       FREE_MAYBE (all_headers);
-      CLOSE (sock);
+      CLOSE_INVALIDATE (sock);
       return RANGEERR;
     }
 
@@ -783,7 +898,7 @@ Accept: %s\r\n\
 		     _("Location: %s%s\n"),
 		     hs->newloc ? hs->newloc : _("unspecified"),
 		     hs->newloc ? _(" [following]") : "");
-	  CLOSE (sock);
+	  CLOSE_FINISH (sock);
 	  FREE_MAYBE (type);
 	  FREE_MAYBE (all_headers);
 	  return NEWLOCATION;
@@ -824,7 +939,7 @@ Accept: %s\r\n\
       hs->res = 0;
       FREE_MAYBE (type);
       FREE_MAYBE (all_headers);
-      CLOSE (sock);
+      CLOSE_FINISH (sock);
       return RETRFINISHED;
     }
 
@@ -838,7 +953,7 @@ Accept: %s\r\n\
       if (!fp)
 	{
 	  logprintf (LOG_NOTQUIET, "%s: %s\n", u->local, strerror (errno));
-	  CLOSE (sock);
+	  CLOSE_FINISH (sock);
 	  FREE_MAYBE (all_headers);
 	  return FOPENERR;
 	}
@@ -863,7 +978,7 @@ Accept: %s\r\n\
   /* Get the contents of the document.  */
   hs->res = get_contents (sock, fp, &hs->len, hs->restval,
 			  (contlen != -1 ? contlen : 0),
-			  &rbuf);
+			  &rbuf, keep_alive);
   hs->dltime = elapsed_time ();
   {
     /* Close or flush the file.  We have to be careful to check for
@@ -878,7 +993,7 @@ Accept: %s\r\n\
       hs->res = -2;
   }
   FREE_MAYBE (all_headers);
-  CLOSE (sock);
+  CLOSE_FINISH (sock);
   if (hs->res == -2)
     return FWRITEERR;
   return RETRFINISHED;
diff --git a/src/main.c b/src/main.c
index 300a82e4..ca4f7f23 100644
--- a/src/main.c
+++ b/src/main.c
@@ -97,6 +97,20 @@ i18n_initialize (void)
   textdomain ("wget");
 #endif /* HAVE_NLS */
 }
+
+/* It's kosher to declare these here because their interface _has_ to
+   be void foo(void).  */
+void url_init PARAMS ((void));
+void host_init PARAMS ((void));
+
+/* This just calls the various initialization functions from the
+   modules that need one-time initialization. */
+static void
+private_initialize (void)
+{
+  url_init ();
+  host_init ();
+}
 
 /* Print the usage message.  */
 static void
@@ -293,6 +307,7 @@ main (int argc, char *const *argv)
   };
 
   i18n_initialize ();
+  private_initialize ();
 
   append_to_log = 0;
 
diff --git a/src/recur.c b/src/recur.c
index 52cc8e12..695ce59d 100644
--- a/src/recur.c
+++ b/src/recur.c
@@ -42,21 +42,20 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #include "ftp.h"
 #include "fnmatch.h"
 #include "host.h"
+#include "hash.h"
 
 extern char *version_string;
 
 #define ROBOTS_FILENAME "robots.txt"
 
-/* #### Many of these lists should really be hashtables!  */
-
-/* List of downloaded URLs.  */
-static urlpos *urls_downloaded;
+static struct hash_table *dl_file_url_map;
+static struct hash_table *dl_url_file_map;
 
 /* List of HTML URLs.  */
 static slist *urls_html;
 
 /* List of undesirable-to-load URLs.  */
-static slist *ulist;
+static struct hash_table *undesirable_urls;
 
 /* List of forbidden locations.  */
 static char **forbidden = NULL;
@@ -84,14 +83,28 @@ static int robots_match PARAMS ((struct urlinfo *, char **));
 void
 recursive_cleanup (void)
 {
-  free_slist (ulist);
-  ulist = NULL;
+  if (undesirable_urls)
+    {
+      string_set_free (undesirable_urls);
+      undesirable_urls = NULL;
+    }
+  if (dl_file_url_map)
+    {
+      free_keys_and_values (dl_file_url_map);
+      hash_table_destroy (dl_file_url_map);
+      dl_file_url_map = NULL;
+    }
+  if (dl_url_file_map)
+    {
+      free_keys_and_values (dl_url_file_map);
+      hash_table_destroy (dl_url_file_map);
+      dl_url_file_map = NULL;
+    }
+  undesirable_urls = NULL;
   free_vec (forbidden);
   forbidden = NULL;
-  free_slist (urls_html);
+  slist_free (urls_html);
   urls_html = NULL;
-  free_urlpos (urls_downloaded);
-  urls_downloaded = NULL;
   FREE_MAYBE (base_dir);
   FREE_MAYBE (robots_host);
   first_time = 1;
@@ -117,6 +130,7 @@ recursive_retrieve (const char *file, const char *this_url)
   char *constr, *filename, *newloc;
   char *canon_this_url = NULL;
   int dt, inl, dash_p_leaf_HTML = FALSE;
+  int meta_disallow_follow;
   int this_url_ftp;            /* See below the explanation */
   uerr_t err;
   struct urlinfo *rurl;
@@ -132,17 +146,29 @@ recursive_retrieve (const char *file, const char *this_url)
   /* Cache the current URL in the list.  */
   if (first_time)
     {
-      ulist = add_slist (ulist, this_url, 0);
-      urls_downloaded = NULL;
+      /* These three operations need to be done only once per Wget
+         run.  They should probably be at a different location.  */
+      if (!undesirable_urls)
+	undesirable_urls = make_string_hash_table (0);
+      if (!dl_file_url_map)
+	dl_file_url_map = make_string_hash_table (0);
+      if (!dl_url_file_map)
+	dl_url_file_map = make_string_hash_table (0);
+
+      hash_table_clear (undesirable_urls);
+      string_set_add (undesirable_urls, this_url);
+      hash_table_clear (dl_file_url_map);
+      hash_table_clear (dl_url_file_map);
       urls_html = NULL;
-      /* Enter this_url to the slist, in original and "enhanced" form.  */
+      /* Enter this_url to the hash table, in original and "enhanced" form.  */
       u = newurl ();
       err = parseurl (this_url, u, 0);
       if (err == URLOK)
 	{
-	  ulist = add_slist (ulist, u->url, 0);
-	  urls_downloaded = add_url (urls_downloaded, u->url, file);
-	  urls_html = add_slist (urls_html, file, NOSORT);
+	  string_set_add (undesirable_urls, u->url);
+	  hash_table_put (dl_file_url_map, xstrdup (file), xstrdup (u->url));
+	  hash_table_put (dl_url_file_map, xstrdup (u->url), xstrdup (file));
+	  urls_html = slist_append (urls_html, file);
 	  if (opt.no_parent)
 	    base_dir = xstrdup (u->dir); /* Set the base dir.  */
 	  /* Set the canonical this_url to be sent as referer.  This
@@ -191,7 +217,15 @@ recursive_retrieve (const char *file, const char *this_url)
 
   /* Get the URL-s from an HTML file: */
   url_list = get_urls_html (file, canon_this_url ? canon_this_url : this_url,
-			    0, dash_p_leaf_HTML);
+			    dash_p_leaf_HTML, &meta_disallow_follow);
+
+  if (opt.use_robots && meta_disallow_follow)
+    {
+      /* The META tag says we are not to follow this file.  Respect
+         that.  */
+      free_urlpos (url_list);
+      url_list = NULL;
+    }
 
   /* Decide what to do with each of the URLs.  A URL will be loaded if
      it meets several requirements, discussed later.  */
@@ -240,16 +274,16 @@ recursive_retrieve (const char *file, const char *this_url)
 	 the list.  */
 
       /* inl is set if the URL we are working on (constr) is stored in
-	 ulist.  Using it is crucial to avoid the incessant calls to
-	 in_slist, which is quite slow.  */
-      inl = in_slist (ulist, constr);
+	 undesirable_urls.  Using it is crucial to avoid unnecessary
+	 repeated continuous hits to the hash table.  */
+      inl = string_set_exists (undesirable_urls, constr);
 
       /* If it is FTP, and FTP is not followed, chuck it out.  */
       if (!inl)
 	if (u->proto == URLFTP && !opt.follow_ftp && !this_url_ftp)
 	  {
 	    DEBUGP (("Uh, it is FTP but i'm not in the mood to follow FTP.\n"));
-	    ulist = add_slist (ulist, constr, 0);
+	    string_set_add (undesirable_urls, constr);
 	    inl = 1;
 	  }
       /* If it is absolute link and they are not followed, chuck it
@@ -258,7 +292,7 @@ recursive_retrieve (const char *file, const char *this_url)
 	if (opt.relative_only && !(cur_url->flags & URELATIVE))
 	  {
 	    DEBUGP (("It doesn't really look like a relative link.\n"));
-	    ulist = add_slist (ulist, constr, 0);
+	    string_set_add (undesirable_urls, constr);
 	    inl = 1;
 	  }
       /* If its domain is not to be accepted/looked-up, chuck it out.  */
@@ -266,7 +300,7 @@ recursive_retrieve (const char *file, const char *this_url)
 	if (!accept_domain (u))
 	  {
 	    DEBUGP (("I don't like the smell of that domain.\n"));
-	    ulist = add_slist (ulist, constr, 0);
+	    string_set_add (undesirable_urls, constr);
 	    inl = 1;
 	  }
       /* Check for parent directory.  */
@@ -286,7 +320,7 @@ recursive_retrieve (const char *file, const char *this_url)
 		{
 		  /* Failing that too, kill the URL.  */
 		  DEBUGP (("Trying to escape parental guidance with no_parent on.\n"));
-		  ulist = add_slist (ulist, constr, 0);
+		  string_set_add (undesirable_urls, constr);
 		  inl = 1;
 		}
 	      freeurl (ut, 1);
@@ -300,7 +334,7 @@ recursive_retrieve (const char *file, const char *this_url)
 	  if (!accdir (u->dir, ALLABS))
 	    {
 	      DEBUGP (("%s (%s) is excluded/not-included.\n", constr, u->dir));
-	      ulist = add_slist (ulist, constr, 0);
+	      string_set_add (undesirable_urls, constr);
 	      inl = 1;
 	    }
 	}
@@ -330,7 +364,7 @@ recursive_retrieve (const char *file, const char *this_url)
 		{
 		  DEBUGP (("%s (%s) does not match acc/rej rules.\n",
 			  constr, u->file));
-		  ulist = add_slist (ulist, constr, 0);
+		  string_set_add (undesirable_urls, constr);
 		  inl = 1;
 		}
 	    }
@@ -353,12 +387,12 @@ recursive_retrieve (const char *file, const char *this_url)
 	    }
 	  free (constr);
 	  constr = xstrdup (u->url);
-	  inl = in_slist (ulist, constr);
+	  string_set_add (undesirable_urls, constr);
 	  if (!inl && !((u->proto == URLFTP) && !this_url_ftp))
 	    if (!opt.spanhost && this_url && !same_host (this_url, constr))
 	      {
 		DEBUGP (("This is not the same hostname as the parent's.\n"));
-		ulist = add_slist (ulist, constr, 0);
+		string_set_add (undesirable_urls, constr);
 		inl = 1;
 	      }
 	}
@@ -398,7 +432,7 @@ recursive_retrieve (const char *file, const char *this_url)
 	    {
 	      DEBUGP (("Stuffing %s because %s forbids it.\n", this_url,
 		       ROBOTS_FILENAME));
-	      ulist = add_slist (ulist, constr, 0);
+	      string_set_add (undesirable_urls, constr);
 	      inl = 1;
 	    }
 	}
@@ -409,7 +443,7 @@ recursive_retrieve (const char *file, const char *this_url)
 	{
 	  DEBUGP (("I've decided to load it -> "));
 	  /* Add it to the list of already-loaded URL-s.  */
-	  ulist = add_slist (ulist, constr, 0);
+	  string_set_add (undesirable_urls, constr);
 	  /* Automatically followed FTPs will *not* be downloaded
 	     recursively.  */
 	  if (u->proto == URLFTP)
@@ -439,10 +473,13 @@ recursive_retrieve (const char *file, const char *this_url)
 	    {
 	      if (dt & RETROKF)
 		{
-		  urls_downloaded = add_url (urls_downloaded, constr, filename);
+		  hash_table_put (dl_file_url_map,
+				  xstrdup (filename), xstrdup (constr));
+		  hash_table_put (dl_url_file_map,
+				  xstrdup (constr), xstrdup (filename));
 		  /* If the URL is HTML, note it.  */
 		  if (dt & TEXTHTML)
-		    urls_html = add_slist (urls_html, filename, NOSORT);
+		    urls_html = slist_append (urls_html, filename);
 		}
 	    }
 	  /* If there was no error, and the type is text/html, parse
@@ -489,6 +526,10 @@ recursive_retrieve (const char *file, const char *this_url)
       /* Increment the pbuf for the appropriate size.  */
     }
   if (opt.convert_links && !opt.delete_after)
+    /* This is merely the first pass: the links that have been
+       successfully downloaded are converted.  In the second pass,
+       convert_all_links() will also convert those links that have NOT
+       been downloaded to their canonical form.  */
     convert_links (file, url_list);
   /* Free the linked list of URL-s.  */
   free_urlpos (url_list);
@@ -531,30 +572,37 @@ void
 convert_all_links (void)
 {
   uerr_t res;
-  urlpos *l1, *l2, *urls;
+  urlpos *l1, *urls;
   struct urlinfo *u;
   slist *html;
-  urlpos *urlhtml;
 
   for (html = urls_html; html; html = html->next)
     {
+      int meta_disallow_follow;
+      char *url;
+
       DEBUGP (("Rescanning %s\n", html->string));
       /* Determine the URL of the HTML file.  get_urls_html will need
 	 it.  */
-      for (urlhtml = urls_downloaded; urlhtml; urlhtml = urlhtml->next)
-	if (!strcmp (urlhtml->local_name, html->string))
-	  break;
-      if (urlhtml)
-	DEBUGP (("It should correspond to %s.\n", urlhtml->url));
+      url = hash_table_get (dl_file_url_map, html->string);
+      if (url)
+	DEBUGP (("It should correspond to %s.\n", url));
       else
 	DEBUGP (("I cannot find the corresponding URL.\n"));
       /* Parse the HTML file...  */
-      urls = get_urls_html (html->string, urlhtml ? urlhtml->url : NULL, 1,
-			    FALSE);
+      urls = get_urls_html (html->string, url, FALSE, &meta_disallow_follow);
+      if (opt.use_robots && meta_disallow_follow)
+	{
+	  /* The META tag says we are not to follow this file.
+	     Respect that.  */
+	  free_urlpos (urls);
+	  urls = NULL;
+	}
       if (!urls)
 	continue;
       for (l1 = urls; l1; l1 = l1->next)
 	{
+	  char *local_name;
 	  /* The URL must be in canonical form to be compared.  */
 	  u = newurl ();
 	  res = parseurl (l1->url, u, 0);
@@ -565,22 +613,18 @@ convert_all_links (void)
 	    }
 	  /* We decide the direction of conversion according to whether
 	     a URL was downloaded.  Downloaded URLs will be converted
-	     ABS2REL, whereas non-downloaded will be converted REL2ABS.
-	     Note: not yet implemented; only ABS2REL works.  */
-	  for (l2 = urls_downloaded; l2; l2 = l2->next)
-	    if (!strcmp (l2->url, u->url))
-	      {
-		DEBUGP (("%s flagged for conversion, local %s\n",
-			 l2->url, l2->local_name));
-		break;
-	      }
+	     ABS2REL, whereas non-downloaded will be converted REL2ABS.  */
+	  local_name = hash_table_get (dl_url_file_map, u->url);
+	  if (local_name)
+	    DEBUGP (("%s flagged for conversion, local %s\n",
+		     u->url, local_name));
 	  /* Clear the flags.  */
 	  l1->flags &= ~ (UABS2REL | UREL2ABS);
 	  /* Decide on the conversion direction.  */
-	  if (l2)
+	  if (local_name)
 	    {
 	      l1->flags |= UABS2REL;
-	      l1->local_name = xstrdup (l2->local_name);
+	      l1->local_name = xstrdup (local_name);
 	    }
 	  else
 	    {
diff --git a/src/retr.c b/src/retr.c
index f60976d8..30f4556f 100644
--- a/src/retr.c
+++ b/src/retr.c
@@ -42,6 +42,7 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #include "ftp.h"
 #include "host.h"
 #include "connect.h"
+#include "hash.h"
 
 #ifdef WINDOWS
 LARGE_INTEGER internal_time;
@@ -60,6 +61,8 @@ enum spflags { SP_NONE, SP_INIT, SP_FINISH };
 
 static int show_progress PARAMS ((long, long, enum spflags));
 
+#define MIN(i, j) ((i) <= (j) ? (i) : (j))
+
 /* Reads the contents of file descriptor FD, until it is closed, or a
    read error occurs.  The data is read in 8K chunks, and stored to
    stream fp, which should have been open for writing.  If BUF is
@@ -83,9 +86,9 @@ static int show_progress PARAMS ((long, long, enum spflags));
    from fd immediately, flush or discard the buffer.  */
 int
 get_contents (int fd, FILE *fp, long *len, long restval, long expected,
-	      struct rbuf *rbuf)
+	      struct rbuf *rbuf, int use_expected)
 {
-  int res;
+  int res = 0;
   static char c[8192];
 
   *len = restval;
@@ -105,10 +108,17 @@ get_contents (int fd, FILE *fp, long *len, long restval, long expected,
 	  *len += res;
 	}
     }
-  /* Read from fd while there is available data.  */
-  do
+  /* Read from fd while there is available data.
+
+     Normally, if expected is 0, it means that it is not known how
+     much data is expected.  However, if use_expected is specified,
+     then expected being zero means exactly that.  */
+  while (!use_expected || (*len < expected))
     {
-      res = iread (fd, c, sizeof (c));
+      int amount_to_read = (use_expected
+			    ? MIN (expected - *len, sizeof (c))
+			    : sizeof (c));
+      res = iread (fd, c, amount_to_read);
       if (res > 0)
 	{
 	  if (fwrite (c, sizeof (char), res, fp) < res)
@@ -120,7 +130,9 @@ get_contents (int fd, FILE *fp, long *len, long restval, long expected,
 	    }
 	  *len += res;
 	}
-    } while (res > 0);
+      else
+	break;
+    }
   if (res < -1)
     res = -1;
   if (opt.verbose)
@@ -323,7 +335,7 @@ retrieve_url (const char *origurl, char **file, char **newloc,
   int local_use_proxy;
   char *mynewloc, *proxy;
   struct urlinfo *u;
-  slist *redirections;
+  struct hash_table *redirections = NULL;
 
   /* If dt is NULL, just ignore it.  */
   if (!dt)
@@ -334,8 +346,6 @@ retrieve_url (const char *origurl, char **file, char **newloc,
   if (file)
     *file = NULL;
 
-  redirections = NULL;
-
   u = newurl ();
   /* Parse the URL. */
   result = parseurl (url, u, 0);
@@ -343,7 +353,8 @@ retrieve_url (const char *origurl, char **file, char **newloc,
     {
       logprintf (LOG_NOTQUIET, "%s: %s.\n", url, uerrmsg (result));
       freeurl (u, 1);
-      free_slist (redirections);
+      if (redirections)
+	string_set_free (redirections);
       free (url);
       return result;
     }
@@ -379,7 +390,8 @@ retrieve_url (const char *origurl, char **file, char **newloc,
 	{
 	  logputs (LOG_NOTQUIET, _("Could not find proxy host.\n"));
 	  freeurl (u, 1);
-	  free_slist (redirections);
+	  if (redirections)
+	    string_set_free (redirections);
 	  free (url);
 	  return PROXERR;
 	}
@@ -392,7 +404,8 @@ retrieve_url (const char *origurl, char **file, char **newloc,
 	  else
 	    logprintf (LOG_NOTQUIET, _("Proxy %s: Must be HTTP.\n"), proxy);
 	  freeurl (u, 1);
-	  free_slist (redirections);
+	  if (redirections)
+	    string_set_free (redirections);
 	  free (url);
 	  return PROXERR;
 	}
@@ -454,7 +467,8 @@ retrieve_url (const char *origurl, char **file, char **newloc,
 	  logprintf (LOG_NOTQUIET, "%s: %s.\n", mynewloc, uerrmsg (newloc_result));
 	  freeurl (newloc_struct, 1);
 	  freeurl (u, 1);
-	  free_slist (redirections);
+	  if (redirections)
+	    string_set_free (redirections);
 	  free (url);
 	  free (mynewloc);
 	  return result;
@@ -466,34 +480,29 @@ retrieve_url (const char *origurl, char **file, char **newloc,
       free (mynewloc);
       mynewloc = xstrdup (newloc_struct->url);
 
-      /* Check for redirection to back to itself.  */
-      if (!strcmp (u->url, newloc_struct->url))
+      if (!redirections)
 	{
-	  logprintf (LOG_NOTQUIET, _("%s: Redirection to itself.\n"),
-		     mynewloc);
-	  freeurl (newloc_struct, 1);
-	  freeurl (u, 1);
-	  free_slist (redirections);
-	  free (url);
-	  free (mynewloc);
-	  return WRONGCODE;
+	  redirections = make_string_hash_table (0);
+	  /* Add current URL immediately so we can detect it as soon
+             as possible in case of a cycle. */
+	  string_set_add (redirections, u->url);
 	}
 
       /* The new location is OK.  Let's check for redirection cycle by
          peeking through the history of redirections. */
-      if (in_slist (redirections, newloc_struct->url))
+      if (string_set_exists (redirections, newloc_struct->url))
 	{
 	  logprintf (LOG_NOTQUIET, _("%s: Redirection cycle detected.\n"),
 		     mynewloc);
 	  freeurl (newloc_struct, 1);
 	  freeurl (u, 1);
-	  free_slist (redirections);
+	  if (redirections)
+	    string_set_free (redirections);
 	  free (url);
 	  free (mynewloc);
 	  return WRONGCODE;
 	}
-
-      redirections = add_slist (redirections, newloc_struct->url, NOSORT);
+      string_set_add (redirections, newloc_struct->url);
 
       free (url);
       url = mynewloc;
@@ -510,7 +519,8 @@ retrieve_url (const char *origurl, char **file, char **newloc,
 	*file = NULL;
     }
   freeurl (u, 1);
-  free_slist (redirections);
+  if (redirections)
+    string_set_free (redirections);
 
   if (newloc)
     *newloc = url;
@@ -531,9 +541,7 @@ retrieve_from_file (const char *file, int html, int *count)
   uerr_t status;
   urlpos *url_list, *cur_url;
 
-  /* If spider-mode is on, we do not want get_urls_html barfing
-     errors on baseless links.  */
-  url_list = (html ? get_urls_html (file, NULL, opt.spider, FALSE)
+  url_list = (html ? get_urls_html (file, NULL, FALSE, NULL)
 	      : get_urls_file (file));
   status = RETROK;             /* Suppose everything is OK.  */
   *count = 0;                  /* Reset the URL count.  */
diff --git a/src/retr.h b/src/retr.h
index 308eeedf..9866f246 100644
--- a/src/retr.h
+++ b/src/retr.h
@@ -22,7 +22,7 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 
 #include "rbuf.h"
 
-int get_contents PARAMS ((int, FILE *, long *, long, long, struct rbuf *));
+int get_contents PARAMS ((int, FILE *, long *, long, long, struct rbuf *, int));
 
 uerr_t retrieve_url PARAMS ((const char *, char **, char **,
 			     const char *, int *));
diff --git a/src/url.c b/src/url.c
index 44e7280d..c077f000 100644
--- a/src/url.c
+++ b/src/url.c
@@ -38,7 +38,6 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #include "utils.h"
 #include "url.h"
 #include "host.h"
-#include "html.h"
 
 #ifndef errno
 extern int errno;
@@ -48,22 +47,12 @@ extern int errno;
 #define DEFAULT_HTTP_PORT 80
 #define DEFAULT_FTP_PORT 21
 
-/* URL separator (for findurl) */
-#define URL_SEPARATOR "!\"#'(),>`{}|<>"
+/* Table of Unsafe chars.  This is intialized in
+   init_unsafe_char_table.  */
 
-/* A list of unsafe characters for encoding, as per RFC1738.  '@' and
-   ':' (not listed in RFC) were added because of user/password
-   encoding.  */
+static char unsafe_char_table[256];
 
-#ifndef WINDOWS
-# define URL_UNSAFE_CHARS "<>\"#%{}|\\^~[]`@:"
-#else  /* WINDOWS */
-# define URL_UNSAFE_CHARS "<>\"%{}|\\^[]`"
-#endif /* WINDOWS */
-
-#define UNSAFE_CHAR(c) (   ((unsigned char)(c) <= ' ')  /* ASCII 32  */  \
-   			|| ((unsigned char)(c) >  '~')  /* ASCII 127 */  \
-			|| strchr (URL_UNSAFE_CHARS, c))
+#define UNSAFE_CHAR(c) (unsafe_char_table[(unsigned char)(c)])
 
 /* If S contains unsafe characters, free it and replace it with a
    version that doesn't.  */
@@ -176,6 +165,34 @@ skip_url (const char *url)
     return 0;
 }
 
+/* Unsafe chars:
+   - anything <= 32;
+   - stuff from rfc1738 ("<>\"#%{}|\\^~[]`");
+   - @ and :, for user/password encoding.
+   - everything over 127 (but we don't bother with recording those.  */
+void
+init_unsafe_char_table (void)
+{
+  int i;
+  for (i = 0; i < 256; i++)
+    if (i < 32 || i >= 127
+	|| i == '<'
+	|| i == '>'
+	|| i == '\"'
+	|| i == '#'
+	|| i == '%'
+	|| i == '{'
+	|| i == '}'
+	|| i == '|'
+	|| i == '\\'
+	|| i == '^'
+	|| i == '~'
+	|| i == '['
+	|| i == ']'
+	|| i == '`')
+      unsafe_char_table[i] = 1;
+}
+
 /* Returns 1 if the string contains unsafe characters, 0 otherwise.  */
 int
 contains_unsafe (const char *s)
@@ -296,7 +313,7 @@ skip_proto (const char *url)
 
 /* Returns 1 if the URL begins with a protocol (supported or
    unsupported), 0 otherwise.  */
-static int
+int
 has_proto (const char *url)
 {
   char **s;
@@ -765,297 +782,54 @@ url_equal (const char *url1, const char *url2)
   return res;
 }
 
-/* Find URL of format scheme:hostname[:port]/dir in a buffer.  The
-   buffer may contain pretty much anything; no errors are signaled.  */
-static const char *
-findurl (const char *buf, int howmuch, int *count)
-{
-  char **prot;
-  const char *s1, *s2;
-
-  for (s1 = buf; howmuch; s1++, howmuch--)
-    for (prot = protostrings; *prot; prot++)
-      if (howmuch <= strlen (*prot))
-	continue;
-      else if (!strncasecmp (*prot, s1, strlen (*prot)))
-	{
-	  for (s2 = s1, *count = 0;
-	       howmuch && *s2 && *s2 >= 32 && *s2 < 127 && !ISSPACE (*s2) &&
-		 !strchr (URL_SEPARATOR, *s2);
-	       s2++, (*count)++, howmuch--);
-	  return s1;
-	}
-  return NULL;
-}
-
-/* Scans the file for signs of URL-s.  Returns a vector of pointers,
-   each pointer representing a URL string.  The file is *not* assumed
-   to be HTML.  */
 urlpos *
 get_urls_file (const char *file)
 {
-  long nread;
-  FILE *fp;
-  char *buf;
-  const char *pbuf;
-  int size;
-  urlpos *first, *current, *old;
-
-  if (file && !HYPHENP (file))
-    {
-      fp = fopen (file, "rb");
-      if (!fp)
-	{
-	  logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
-	  return NULL;
-	}
-    }
-  else
-    fp = stdin;
-  /* Load the file.  */
-  load_file (fp, &buf, &nread);
-  if (file && !HYPHENP (file))
-    fclose (fp);
-  DEBUGP (("Loaded %s (size %ld).\n", file, nread));
-  first = current = NULL;
-  /* Fill the linked list with URLs.  */
-  for (pbuf = buf; (pbuf = findurl (pbuf, nread - (pbuf - buf), &size));
-       pbuf += size)
-    {
-      /* Allocate the space.  */
-      old = current;
-      current = (urlpos *)xmalloc (sizeof (urlpos));
-      if (old)
-	old->next = current;
-      memset (current, 0, sizeof (*current));
-      current->next = NULL;
-      current->url = (char *)xmalloc (size + 1);
-      memcpy (current->url, pbuf, size);
-      current->url[size] = '\0';
-      if (!first)
-	first = current;
-    }
-  /* Free the buffer.  */
-  free (buf);
-
-  return first;
-}
-
-/* Similar to get_urls_file, but for HTML files.  FILE is scanned as
-   an HTML document using htmlfindurl(), which see.  get_urls_html()
-   constructs the HTML-s from the relative href-s.
+  struct file_memory *fm;
+  urlpos *head, *tail;
+  const char *text, *text_end;
 
-   If SILENT is non-zero, do not barf on baseless relative links.  */
-urlpos *
-get_urls_html (const char *file, const char *this_url, int silent,
-	       int dash_p_leaf_HTML)
-{
-  long nread;
-  FILE *fp;
-  char *orig_buf;
-  const char *buf;
-  int step, first_time;
-  urlpos *first, *current, *old;
-
-  if (file && !HYPHENP (file))
+  /* Load the file.  */
+  fm = read_file (file);
+  if (!fm)
     {
-      fp = fopen (file, "rb");
-      if (!fp)
-	{
-	  logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
-	  return NULL;
-	}
+      logprintf (LOG_NOTQUIET, "%s: %s\n", file, strerror (errno));
+      return NULL;
     }
-  else
-    fp = stdin;
-  /* Load the file.  */
-  load_file (fp, &orig_buf, &nread);
-  if (file && !HYPHENP (file))
-    fclose (fp);
-  DEBUGP (("Loaded HTML file %s (size %ld).\n", file, nread));
-  first = current = NULL;
-  first_time = 1;
-  /* Iterate over the URLs in BUF, picked by htmlfindurl().  */
-  for (buf = orig_buf;
-       (buf = htmlfindurl (buf, nread - (buf - orig_buf), &step, first_time,
-			   dash_p_leaf_HTML));
-       buf += step)
+  DEBUGP (("Loaded %s (size %ld).\n", file, fm->length));
+  head = tail = NULL;
+  text = fm->content;
+  text_end = fm->content + fm->length;
+  while (text < text_end)
     {
-      int i, no_proto;
-      int size = step;
-      const char *pbuf = buf;
-      char *constr, *base;
-      const char *cbase;
-      char *needs_freeing, *url_data;
-
-      first_time = 0;
-
-      /* A frequent phenomenon that needs to be handled are pages
-         generated by brain-damaged HTML generators, which refer to to
-         URI-s as <a href="<spaces>URI<spaces>">.  We simply ignore
-         any spaces at the beginning or at the end of the string.
-         This is probably not strictly correct, but that's what the
-         browsers do, so we may follow.  May the authors of "WYSIWYG"
-         HTML tools burn in hell for the damage they've inflicted!  */
-      while ((pbuf < buf + step) && ISSPACE (*pbuf))
-        {
-          ++pbuf;
-          --size;
-        }
-      while (size && ISSPACE (pbuf[size - 1]))
-	--size;
-      if (!size)
-	break;
-
-      /* It would be nice if we could avoid allocating memory in this
-         loop, but I don't see an easy way.  To process the entities,
-         we need to either copy the data, or change it destructively.
-         I choose the former.
-
-	 We have two pointers: needs_freeing and url_data, because the
-	 code below does thing like url_data += <something>, and we
-	 want to pass the original string to free(). */
-      needs_freeing = url_data = html_decode_entities (pbuf, pbuf + size);
-      size = strlen (url_data);
-
-      for (i = 0; protostrings[i]; i++)
-	{
-	  if (!strncasecmp (protostrings[i], url_data,
-			    MINVAL (strlen (protostrings[i]), size)))
-	    break;
-	}
-      /* Check for http:RELATIVE_URI.  See below for details.  */
-      if (protostrings[i]
-	  && !(strncasecmp (url_data, "http:", 5) == 0
-	       && strncasecmp (url_data, "http://", 7) != 0))
-	{
-	  no_proto = 0;
-	}
+      const char *line_beg = text;
+      const char *line_end = memchr (text, '\n', text_end - text);
+      if (!line_end)
+	line_end = text_end;
       else
+	++line_end;
+      text = line_end;
+      while (line_beg < line_end
+	     && ISSPACE (*line_beg))
+	++line_beg;
+      while (line_end > line_beg + 1
+	     && ISSPACE (*(line_end - 1)))
+	--line_end;
+      if (line_end > line_beg)
 	{
-	  no_proto = 1;
-	  /* This is for extremely brain-damaged pages that refer to
-	     relative URI-s as <a href="http:URL">.  Just strip off the
-	     silly leading "http:" (as well as any leading blanks
-	     before it).  */
-	  if ((size > 5) && !strncasecmp ("http:", url_data, 5))
-	    url_data += 5, size -= 5;
-	}
-      if (!no_proto)
-	{
-	  for (i = 0; i < ARRAY_SIZE (sup_protos); i++)
-	    {
-	      if (!strncasecmp (sup_protos[i].name, url_data,
-			       MINVAL (strlen (sup_protos[i].name), size)))
-		break;
-	    }
-	  /* Do *not* accept a non-supported protocol.  */
-	  if (i == ARRAY_SIZE (sup_protos))
-	    {
-	      free (needs_freeing);
-	      continue;
-	    }
-	}
-      if (no_proto)
-	{
-	  /* First, construct the base, which can be relative itself.
-
-	     Criteria for creating the base are:
-	     1) html_base created by <base href="...">
-	     2) current URL
-	     3) base provided from the command line */
-	  cbase = html_base ();
-	  if (!cbase)
-	    cbase = this_url;
-	  if (!cbase)
-	    cbase = opt.base_href;
-	  if (!cbase)             /* Error condition -- a baseless
-				     relative link.  */
-	    {
-	      if (!opt.quiet && !silent)
-		{
-		  /* Use malloc, not alloca because this is called in
-                     a loop. */
-		  char *temp = (char *)malloc (size + 1);
-		  strncpy (temp, url_data, size);
-		  temp[size] = '\0';
-		  logprintf (LOG_NOTQUIET,
-			     _("Error (%s): Link %s without a base provided.\n"),
-			     file, temp);
-		  free (temp);
-		}
-	      free (needs_freeing);
-	      continue;
-	    }
-	  if (this_url)
-	    base = construct (this_url, cbase, strlen (cbase),
-			      !has_proto (cbase));
+	  urlpos *entry = (urlpos *)xmalloc (sizeof (urlpos));
+	  memset (entry, 0, sizeof (*entry));
+	  entry->next = NULL;
+	  entry->url = strdupdelim (line_beg, line_end);
+	  if (!head)
+	    head = entry;
 	  else
-	    {
-	      /* Base must now be absolute, with host name and
-		 protocol.  */
-	      if (!has_proto (cbase))
-		{
-		  logprintf (LOG_NOTQUIET, _("\
-Error (%s): Base %s relative, without referer URL.\n"),
-			     file, cbase);
-		  free (needs_freeing);
-		  continue;
-		}
-	      base = xstrdup (cbase);
-	    }
-	  constr = construct (base, url_data, size, no_proto);
-	  free (base);
-	}
-      else /* has proto */
-	{
-	  constr = (char *)xmalloc (size + 1);
-	  strncpy (constr, url_data, size);
-	  constr[size] = '\0';
+	    tail->next = entry;
+	  tail = entry;
 	}
-#ifdef DEBUG
-      if (opt.debug)
-	{
-	  char *tmp;
-	  const char *tmp2;
-
-	  tmp2 = html_base ();
-	  /* Use malloc, not alloca because this is called in a loop. */
-	  tmp = (char *)xmalloc (size + 1);
-	  strncpy (tmp, url_data, size);
-	  tmp[size] = '\0';
-	  logprintf (LOG_ALWAYS,
-		     "file %s; this_url %s; base %s\nlink: %s; constr: %s\n",
-		     file, this_url ? this_url : "(null)",
-		     tmp2 ? tmp2 : "(null)", tmp, constr);
-	  free (tmp);
-	}
-#endif
-
-      /* Allocate the space.  */
-      old = current;
-      current = (urlpos *)xmalloc (sizeof (urlpos));
-      if (old)
-	old->next = current;
-      if (!first)
-	first = current;
-      /* Fill the values.  */
-      memset (current, 0, sizeof (*current));
-      current->next = NULL;
-      current->url = constr;
-      current->size = step;
-      current->pos = buf - orig_buf;
-      /* A URL is relative if the host and protocol are not named,
-	 and the name does not start with `/'.  */
-      if (no_proto && *url_data != '/')
-	current->flags |= (URELATIVE | UNOPROTO);
-      else if (no_proto)
-	current->flags |= UNOPROTO;
-      free (needs_freeing);
     }
-  free (orig_buf);
-
-  return first;
+  read_file_free (fm);
+  return head;
 }
 
 /* Free the linked list of urlpos.  */
@@ -1527,103 +1301,59 @@ no_proxy_match (const char *host, const char **no_proxy)
     return !sufmatch (no_proxy, host);
 }
 
+static void write_backup_file PARAMS ((const char *, downloaded_file_t));
+
 /* Change the links in an HTML document.  Accepts a structure that
    defines the positions of all the links.  */
 void
 convert_links (const char *file, urlpos *l)
 {
+  struct file_memory *fm;
   FILE               *fp;
-  char               *buf, *p, *p2;
+  char               *p;
   downloaded_file_t  downloaded_file_return;
-  long               size;
+
+  {
+    /* First we do a "dry run": go through the list L and see whether
+       any URL needs to be converted in the first place.  If not, just
+       leave the file alone.  */
+    int count = 0;
+    urlpos *dry = l;
+    for (dry = l; dry; dry = dry->next)
+      if (dry->flags & (UABS2REL | UREL2ABS))
+	++count;
+    if (!count)
+      {
+	logprintf (LOG_VERBOSE, _("Nothing to do while converting %s.\n"),
+		   file);
+	return;
+      }
+  }
 
   logprintf (LOG_VERBOSE, _("Converting %s... "), file);
-  /* Read from the file....  */
-  fp = fopen (file, "rb");
-  if (!fp)
+
+  fm = read_file (file);
+  if (!fm)
     {
       logprintf (LOG_NOTQUIET, _("Cannot convert links in %s: %s\n"),
 		 file, strerror (errno));
       return;
     }
-  /* ...to a buffer.  */
-  load_file (fp, &buf, &size);
-  fclose (fp);
-
-  downloaded_file_return = downloaded_file(CHECK_FOR_FILE, file);
 
+  downloaded_file_return = downloaded_file (CHECK_FOR_FILE, file);
   if (opt.backup_converted && downloaded_file_return)
-    /* Rather than just writing over the original .html file with the converted
-       version, save the former to *.orig.  Note we only do this for files we've
-       _successfully_ downloaded, so we don't clobber .orig files sitting around
-       from previous invocations. */
-    {
-      /* Construct the backup filename as the original name plus ".orig". */
-      size_t         filename_len = strlen(file);
-      char*          filename_plus_orig_suffix;
-      boolean        already_wrote_backup_file = FALSE;
-      slist*         converted_file_ptr;
-      static slist*  converted_files = NULL;
-
-      if (downloaded_file_return == FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED)
-	{
-	  /* Just write "orig" over "html".  We need to do it this way because
-	     when we're checking to see if we've downloaded the file before (to
-	     see if we can skip downloading it), we don't know if it's a
-	     text/html file.  Therefore we don't know yet at that stage that -E
-	     is going to cause us to tack on ".html", so we need to compare
-	     vs. the original URL plus ".orig", not the original URL plus
-	     ".html.orig". */
-	  filename_plus_orig_suffix = xmalloc(filename_len + 1);
-	  strcpy(filename_plus_orig_suffix, file);
-	  strcpy((filename_plus_orig_suffix + filename_len) - 4, "orig");
-	}
-      else /* downloaded_file_return == FILE_DOWNLOADED_NORMALLY */
-	{
-	  /* Append ".orig" to the name. */
-	  filename_plus_orig_suffix = xmalloc(filename_len + sizeof(".orig"));
-	  strcpy(filename_plus_orig_suffix, file);
-	  strcpy(filename_plus_orig_suffix + filename_len, ".orig");
-	}
-
-      /* We can get called twice on the same URL thanks to the
-	 convert_all_links() call in main().  If we write the .orig file each
-	 time in such a case, it'll end up containing the first-pass conversion,
-	 not the original file.  So, see if we've already been called on this
-	 file. */
-      converted_file_ptr = converted_files;
-      while (converted_file_ptr != NULL)
-	if (strcmp(converted_file_ptr->string, file) == 0)
-	  {
-	    already_wrote_backup_file = TRUE;
-	    break;
-	  }
-	else
-	  converted_file_ptr = converted_file_ptr->next;
-
-      if (!already_wrote_backup_file)
-	{
-	  /* Rename <file> to <file>.orig before former gets written over. */
-	  if (rename(file, filename_plus_orig_suffix) != 0)
-	    logprintf (LOG_NOTQUIET, _("Cannot back up %s as %s: %s\n"),
-		       file, filename_plus_orig_suffix, strerror (errno));
-
-	  /* Remember that we've already written a .orig backup for this file.
-	     Note that we never free this memory since we need it till the
-	     convert_all_links() call, which is one of the last things the
-	     program does before terminating.  BTW, I'm not sure if it would be
-	     safe to just set 'converted_file_ptr->string' to 'file' below,
-	     rather than making a copy of the string...  Another note is that I
-	     thought I could just add a field to the urlpos structure saying
-	     that we'd written a .orig file for this URL, but that didn't work,
-	     so I had to make this separate list. */
-	  converted_file_ptr = xmalloc(sizeof(*converted_file_ptr));
-	  converted_file_ptr->string = xstrdup(file);  /* die on out-of-mem. */
-	  converted_file_ptr->next = converted_files;
-	  converted_files = converted_file_ptr;
-	}
+    write_backup_file (file, downloaded_file_return);
 
-      free(filename_plus_orig_suffix);
+  /* Before opening the file for writing, unlink the file.  This is
+     important if the data in FM is mmaped.  In such case, nulling the
+     file, which is what fopen() below does, would make us read all
+     zeroes from the mmaped region.  */
+  if (unlink (file) < 0 && errno != ENOENT)
+    {
+      logprintf (LOG_NOTQUIET, _("Unable to delete `%s': %s\n"),
+		 file, strerror (errno));
+      read_file_free (fm);
+      return;
     }
   /* Now open the file for writing.  */
   fp = fopen (file, "wb");
@@ -1631,50 +1361,63 @@ convert_links (const char *file, urlpos *l)
     {
       logprintf (LOG_NOTQUIET, _("Cannot convert links in %s: %s\n"),
 		 file, strerror (errno));
-      free (buf);
+      read_file_free (fm);
       return;
     }
-  /* Presumably we have to loop through multiple URLs here (even though we're
-     only talking about a single local file) because of the -O option. */
-  for (p = buf; l; l = l->next)
+  /* Here we loop through all the URLs in file, replacing those of
+     them that are downloaded with relative references.  */
+  p = fm->content;
+  for (; l; l = l->next)
     {
-      if (l->pos >= size)
+      char *url_start = fm->content + l->pos;
+      if (l->pos >= fm->length)
 	{
 	  DEBUGP (("Something strange is going on.  Please investigate."));
 	  break;
 	}
-      /* If the URL already is relative or it is not to be converted
-	 for some other reason (e.g. because of not having been
-	 downloaded in the first place), skip it.  */
-      if ((l->flags & URELATIVE) || !(l->flags & UABS2REL))
+      /* If the URL is not to be converted, skip it.  */
+      if (!(l->flags & (UABS2REL | UREL2ABS)))
 	{
 	  DEBUGP (("Skipping %s at position %d (flags %d).\n", l->url,
 		   l->pos, l->flags));
 	  continue;
 	}
-      /* Else, reach the position of the offending URL, echoing
-	 everything up to it to the outfile.  */
-      for (p2 = buf + l->pos; p < p2; p++)
-	putc (*p, fp);
+
+      /* Echo the file contents, up to the offending URL's opening
+         quote, to the outfile.  */
+      fwrite (p, 1, url_start - p, fp);
+      p = url_start;
       if (l->flags & UABS2REL)
-	/* Convert absolute URL to relative. */
 	{
+	  /* Convert absolute URL to relative. */
 	  char *newname = construct_relative (file, l->local_name);
-	  fprintf (fp, "%s", newname);
+	  putc (*p, fp);	/* quoting char */
+	  fputs (newname, fp);
+	  p += l->size - 1;
+	  putc (*p, fp);	/* close quote */
+	  ++p;
 	  DEBUGP (("ABS2REL: %s to %s at position %d in %s.\n",
 		   l->url, newname, l->pos, file));
 	  free (newname);
 	}
-      p += l->size;
+      else if (l->flags & UREL2ABS)
+	{
+	  /* Convert the link to absolute URL. */
+	  char *newlink = l->url;
+	  putc (*p, fp);	/* quoting char */
+	  fputs (newlink, fp);
+	  p += l->size - 1;
+	  putc (*p, fp);	/* close quote */
+	  ++p;
+	  DEBUGP (("REL2ABS: <something> to %s at position %d in %s.\n",
+		   newlink, l->pos, file));
+	}
     }
   /* Output the rest of the file. */
-  if (p - buf < size)
-    {
-      for (p2 = buf + size; p < p2; p++)
-	putc (*p, fp);
-    }
+  if (p - fm->content < fm->length)
+    fwrite (p, 1, fm->length - (p - fm->content), fp);
   fclose (fp);
-  free (buf);
+  read_file_free (fm);
   logputs (LOG_VERBOSE, _("done.\n"));
 }
 
@@ -1746,6 +1489,79 @@ add_url (urlpos *l, const char *url, const char *file)
   return t;
 }
 
+static void
+write_backup_file (const char *file, downloaded_file_t downloaded_file_return)
+{
+  /* Rather than just writing over the original .html file with the
+     converted version, save the former to *.orig.  Note we only do
+     this for files we've _successfully_ downloaded, so we don't
+     clobber .orig files sitting around from previous invocations. */
+
+  /* Construct the backup filename as the original name plus ".orig". */
+  size_t         filename_len = strlen(file);
+  char*          filename_plus_orig_suffix;
+  boolean        already_wrote_backup_file = FALSE;
+  slist*         converted_file_ptr;
+  static slist*  converted_files = NULL;
+
+  if (downloaded_file_return == FILE_DOWNLOADED_AND_HTML_EXTENSION_ADDED)
+    {
+      /* Just write "orig" over "html".  We need to do it this way
+	 because when we're checking to see if we've downloaded the
+	 file before (to see if we can skip downloading it), we don't
+	 know if it's a text/html file.  Therefore we don't know yet
+	 at that stage that -E is going to cause us to tack on
+	 ".html", so we need to compare vs. the original URL plus
+	 ".orig", not the original URL plus ".html.orig". */
+      filename_plus_orig_suffix = alloca (filename_len + 1);
+      strcpy(filename_plus_orig_suffix, file);
+      strcpy((filename_plus_orig_suffix + filename_len) - 4, "orig");
+    }
+  else /* downloaded_file_return == FILE_DOWNLOADED_NORMALLY */
+    {
+      /* Append ".orig" to the name. */
+      filename_plus_orig_suffix = alloca (filename_len + sizeof(".orig"));
+      strcpy(filename_plus_orig_suffix, file);
+      strcpy(filename_plus_orig_suffix + filename_len, ".orig");
+    }
+
+  /* We can get called twice on the same URL thanks to the
+     convert_all_links() call in main().  If we write the .orig file
+     each time in such a case, it'll end up containing the first-pass
+     conversion, not the original file.  So, see if we've already been
+     called on this file. */
+  converted_file_ptr = converted_files;
+  while (converted_file_ptr != NULL)
+    if (strcmp(converted_file_ptr->string, file) == 0)
+      {
+	already_wrote_backup_file = TRUE;
+	break;
+      }
+    else
+      converted_file_ptr = converted_file_ptr->next;
+
+  if (!already_wrote_backup_file)
+    {
+      /* Rename <file> to <file>.orig before former gets written over. */
+      if (rename(file, filename_plus_orig_suffix) != 0)
+	logprintf (LOG_NOTQUIET, _("Cannot back up %s as %s: %s\n"),
+		   file, filename_plus_orig_suffix, strerror (errno));
+
+      /* Remember that we've already written a .orig backup for this file.
+	 Note that we never free this memory since we need it till the
+	 convert_all_links() call, which is one of the last things the
+	 program does before terminating.  BTW, I'm not sure if it would be
+	 safe to just set 'converted_file_ptr->string' to 'file' below,
+	 rather than making a copy of the string...  Another note is that I
+	 thought I could just add a field to the urlpos structure saying
+	 that we'd written a .orig file for this URL, but that didn't work,
+	 so I had to make this separate list. */
+      converted_file_ptr = xmalloc(sizeof(*converted_file_ptr));
+      converted_file_ptr->string = xstrdup(file);  /* die on out-of-mem. */
+      converted_file_ptr->next = converted_files;
+      converted_files = converted_file_ptr;
+    }
+}
 
 /* Remembers which files have been downloaded.  In the standard case, should be
    called with mode == FILE_DOWNLOADED_NORMALLY for each file we actually
@@ -1798,3 +1614,10 @@ downloaded_file (downloaded_file_t  mode, const char*  file)
       return FILE_NOT_ALREADY_DOWNLOADED;
     }
 }
+
+/* Initialization of static stuff. */
+void
+url_init (void)
+{
+  init_unsafe_char_table ();
+}
diff --git a/src/url.h b/src/url.h
index 0f55ec35..648193fa 100644
--- a/src/url.h
+++ b/src/url.h
@@ -88,6 +88,7 @@ struct urlinfo *newurl PARAMS ((void));
 void freeurl PARAMS ((struct urlinfo *, int));
 uerr_t urlproto PARAMS ((const char *));
 int skip_proto PARAMS ((const char *));
+int has_proto PARAMS ((const char *));
 int skip_uname PARAMS ((const char *));
 
 uerr_t parseurl PARAMS ((const char *, struct urlinfo *, int));
@@ -95,7 +96,7 @@ char *str_url PARAMS ((const struct urlinfo *, int));
 int url_equal PARAMS ((const char *, const char *));
 
 urlpos *get_urls_file PARAMS ((const char *));
-urlpos *get_urls_html PARAMS ((const char *, const char *, int, int));
+urlpos *get_urls_html PARAMS ((const char *, const char *, int, int *));
 void free_urlpos PARAMS ((urlpos *));
 
 char *url_concat PARAMS ((const char *, const char *));
diff --git a/src/utils.c b/src/utils.c
index 795ecb75..a6a08e2a 100644
--- a/src/utils.c
+++ b/src/utils.c
@@ -31,6 +31,9 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #ifdef HAVE_UNISTD_H
 # include <unistd.h>
 #endif
+#ifdef HAVE_MMAP
+# include <sys/mman.h>
+#endif
 #ifdef HAVE_PWD_H
 # include <pwd.h>
 #endif
@@ -45,11 +48,13 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #ifdef NeXT
 # include <libc.h>		/* for access() */
 #endif
+#include <fcntl.h>
 #include <assert.h>
 
 #include "wget.h"
 #include "utils.h"
 #include "fnmatch.h"
+#include "hash.h"
 
 #ifndef errno
 extern int errno;
@@ -736,28 +741,149 @@ read_whole_line (FILE *fp)
     line = xrealloc (line, length + 1);
   return line;
 }
+
+/* Read FILE into memory.  A pointer to `struct file_memory' are
+   returned; use struct element `content' to access file contents, and
+   the element `length' to know the file length.  `content' is *not*
+   zero-terminated, and you should *not* read or write beyond the [0,
+   length) range of characters.
 
-/* Load file pointed to by FP to memory and return the malloc-ed
-   buffer with the contents.  *NREAD will contain the number of read
-   bytes.  The file is loaded in chunks, allocated exponentially,
-   starting with FILE_BUFFER_SIZE bytes.  */
-void
-load_file (FILE *fp, char **buf, long *nread)
-{
-  long bufsize;
+   After you are done with the file contents, call read_file_free to
+   release the memory.
+
+   Depending on the operating system and the type of file that is
+   being read, read_file() either mmap's the file into memory, or
+   reads the file into the core using read().
 
-  bufsize = 512;
-  *nread = 0;
-  *buf = NULL;
-  while (!feof (fp) && !ferror (fp))
+   If file is named "-", fileno(stdin) is used for reading instead.
+   If you want to read from a real file named "-", use "./-" instead.  */
+
+struct file_memory *
+read_file (const char *file)
+{
+  int fd;
+  struct file_memory *fm;
+  long size;
+  int inhibit_close = 0;
+
+  /* Some magic in the finest tradition of Perl and its kin: if FILE
+     is "-", just use stdin.  */
+  if (HYPHENP (file))
     {
-      *buf = (char *)xrealloc (*buf, bufsize + *nread);
-      *nread += fread (*buf + *nread, sizeof (char), bufsize, fp);
-      bufsize <<= 1;
+      fd = fileno (stdin);
+      inhibit_close = 1;
+      /* Note that we don't inhibit mmap() in this case.  If stdin is
+         redirected from a regular file, mmap() will still work.  */
+    }
+  else
+    fd = open (file, O_RDONLY);
+  if (fd < 0)
+    return NULL;
+  fm = xmalloc (sizeof (struct file_memory));
+
+#ifdef HAVE_MMAP
+  {
+    struct stat buf;
+    if (fstat (fd, &buf) < 0)
+      goto mmap_lose;
+    fm->length = buf.st_size;
+    /* NOTE: As far as I know, the callers of this function never
+       modify the file text.  Relying on this would enable us to
+       specify PROT_READ and MAP_SHARED for a marginal gain in
+       efficiency, but at some cost to generality.  */
+    fm->content = mmap (NULL, fm->length, PROT_READ | PROT_WRITE,
+			MAP_PRIVATE, fd, 0);
+    if (fm->content == MAP_FAILED)
+      goto mmap_lose;
+    if (!inhibit_close)
+      close (fd);
+
+    fm->mmap_p = 1;
+    return fm;
+  }
+
+ mmap_lose:
+  /* The most common reason why mmap() fails is that FD does not point
+     to a plain file.  However, it's also possible that mmap() doesn't
+     work for a particular type of file.  Therefore, whenever mmap()
+     fails, we just fall back to the regular method.  */
+#endif /* HAVE_MMAP */
+
+  fm->length = 0;
+  size = 512;			/* number of bytes fm->contents can
+                                   hold at any given time. */
+  fm->content = xmalloc (size);
+  while (1)
+    {
+      long nread;
+      if (fm->length > size / 2)
+	{
+	  /* #### I'm not sure whether the whole exponential-growth
+             thing makes sense with kernel read.  On Linux at least,
+             read() refuses to read more than 4K from a file at a
+             single chunk anyway.  But other Unixes might optimize it
+             better, and it doesn't *hurt* anything, so I'm leaving
+             it.  */
+
+	  /* Normally, we grow SIZE exponentially to make the number
+             of calls to read() and realloc() logarithmic in relation
+             to file size.  However, read() can read an amount of data
+             smaller than requested, and it would be unreasonably to
+             double SIZE every time *something* was read.  Therefore,
+             we double SIZE only when the length exceeds half of the
+             entire allocated size.  */
+	  size <<= 1;
+	  fm->content = xrealloc (fm->content, size);
+	}
+      nread = read (fd, fm->content + fm->length, size - fm->length);
+      if (nread > 0)
+	/* Successful read. */
+	fm->length += nread;
+      else if (nread < 0)
+	/* Error. */
+	goto lose;
+      else
+	/* EOF */
+	break;
     }
-  /* #### No indication of encountered error??  */
+  if (!inhibit_close)
+    close (fd);
+  if (size > fm->length && fm->length != 0)
+    /* Due to exponential growth of fm->content, the allocated region
+       might be much larger than what is actually needed.  */
+    fm->content = xrealloc (fm->content, fm->length);
+  fm->mmap_p = 0;
+  return fm;
+
+ lose:
+  if (!inhibit_close)
+    close (fd);
+  free (fm->content);
+  free (fm);
+  return NULL;
 }
 
+/* Release the resources held by FM.  Specifically, this calls
+   munmap() or free() on fm->content, depending whether mmap or
+   malloc/read were used to read in the file.  It also frees the
+   memory needed to hold the FM structure itself.  */
+
+void
+read_file_free (struct file_memory *fm)
+{
+#ifdef HAVE_MMAP
+  if (fm->mmap_p)
+    {
+      munmap (fm->content, fm->length);
+    }
+  else
+#endif
+    {
+      free (fm->content);
+    }
+  free (fm);
+}
+
 /* Free the pointers in a NULL-terminated vector of pointers, then
    free the pointer itself.  */
 void
@@ -801,97 +927,42 @@ merge_vecs (char **v1, char **v2)
   return v1;
 }
 
-/* A set of simple-minded routines to store and search for strings in
-   a linked list.  You may add a string to the slist, and peek whether
-   it's still in there at any time later.  */
+/* A set of simple-minded routines to store strings in a linked list.
+   This used to also be used for searching, but now we have hash
+   tables for that.  */
 
-/* Add an element to the list.  If flags is NOSORT, the list will not
-   be sorted.  */
+/* Append an element to the list.  */
 slist *
-add_slist (slist *l, const char *s, int flags)
+slist_append (slist *l, const char *s)
 {
-  slist *t, *old, *beg;
-  int cmp;
+  slist *newel = (slist *)xmalloc (sizeof (slist));
+  slist *beg = l;
 
-  if (flags & NOSORT)
-    {
-      if (!l)
-	{
-	  t = (slist *)xmalloc (sizeof (slist));
-	  t->string = xstrdup (s);
-	  t->next = NULL;
-	  return t;
-	}
-      beg = l;
-      /* Find the last element.  */
-      while (l->next)
-	l = l->next;
-      t = (slist *)xmalloc (sizeof (slist));
-      l->next = t;
-      t->string = xstrdup (s);
-      t->next = NULL;
-      return beg;
-    }
-  /* Empty list or changing the first element.  */
-  if (!l || (cmp = strcmp (l->string, s)) > 0)
-    {
-      t = (slist *)xmalloc (sizeof (slist));
-      t->string = xstrdup (s);
-      t->next = l;
-      return t;
-    }
+  newel->string = xstrdup (s);
+  newel->next = NULL;
 
-  beg = l;
-  if (cmp == 0)
-    return beg;
-
-  /* Second two one-before-the-last element.  */
+  if (!l)
+    return newel;
+  /* Find the last element.  */
   while (l->next)
-    {
-      old = l;
-      l = l->next;
-      cmp = strcmp (s, l->string);
-      if (cmp == 0)             /* no repeating in the list */
-	return beg;
-      else if (cmp > 0)
-	continue;
-      /* If the next list element is greater than s, put s between the
-	 current and the next list element.  */
-      t = (slist *)xmalloc (sizeof (slist));
-      old->next = t;
-      t->next = l;
-      t->string = xstrdup (s);
-      return beg;
-    }
-  t = (slist *)xmalloc (sizeof (slist));
-  t->string = xstrdup (s);
-  /* Insert the new element after the last element.  */
-  l->next = t;
-  t->next = NULL;
+    l = l->next;
+  l->next = newel;
   return beg;
 }
 
 /* Is there a specific entry in the list?  */
 int
-in_slist (slist *l, const char *s)
+slist_contains (slist *l, const char *s)
 {
-  int cmp;
-
-  while (l)
-    {
-      cmp = strcmp (l->string, s);
-      if (cmp == 0)
-	return 1;
-      else if (cmp > 0)         /* the list is ordered!  */
-	return 0;
-      l = l->next;
-    }
+  for (; l; l = l->next)
+    if (!strcmp (l->string, s))
+      return 1;
   return 0;
 }
 
 /* Free the whole slist.  */
 void
-free_slist (slist *l)
+slist_free (slist *l)
 {
   slist *n;
 
@@ -903,6 +974,58 @@ free_slist (slist *l)
       l = n;
     }
 }
+
+/* Sometimes it's useful to create "sets" of strings, i.e. special
+   hash tables where you want to store strings as keys and merely
+   query for their existence.  Here is a set of utility routines that
+   makes that transparent.  */
+
+void
+string_set_add (struct hash_table *ht, const char *s)
+{
+  /* We use "1" as value.  It provides us a useful and clear arbitrary
+     value, and it consumes no memory -- the pointers to the same
+     string "1" will be shared by all the key-value pairs in the hash
+     table.  */
+  hash_table_put (ht, xstrdup (s), "1");
+}
+
+int
+string_set_exists (struct hash_table *ht, const char *s)
+{
+  return hash_table_exists (ht, s);
+}
+
+static int
+string_set_free_mapper (void *key, void *value_ignored, void *arg_ignored)
+{
+  free (key);
+  return 0;
+}
+
+void
+string_set_free (struct hash_table *ht)
+{
+  hash_table_map (ht, string_set_free_mapper, NULL);
+  hash_table_destroy (ht);
+}
+
+static int
+free_keys_and_values_mapper (void *key, void *value, void *arg_ignored)
+{
+  free (key);
+  free (value);
+  return 0;
+}
+
+/* Another utility function: call free() on all keys and values of HT.  */
+
+void
+free_keys_and_values (struct hash_table *ht)
+{
+  hash_table_map (ht, free_keys_and_values_mapper, NULL);
+}
+
 
 /* Engine for legible and legible_long_long; this function works on
    strings.  */
diff --git a/src/utils.h b/src/utils.h
index eb3c99cc..c9de38bb 100644
--- a/src/utils.h
+++ b/src/utils.h
@@ -20,11 +20,6 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 #ifndef UTILS_H
 #define UTILS_H
 
-/* Flags for slist.  */
-enum {
-   NOSORT     = 1
-};
-
 enum accd {
    ALLABS = 1
 };
@@ -36,6 +31,14 @@ typedef struct _slist
   struct _slist *next;
 } slist;
 
+struct hash_table;
+
+struct file_memory {
+  char *content;
+  long length;
+  int mmap_p;
+};
+
 char *time_str PARAMS ((time_t *));
 const char *uerrmsg PARAMS ((uerr_t));
 
@@ -58,13 +61,19 @@ int accdir PARAMS ((const char *s, enum accd));
 char *suffix PARAMS ((const char *s));
 
 char *read_whole_line PARAMS ((FILE *));
-void load_file PARAMS ((FILE *, char **, long *));
+struct file_memory *read_file PARAMS ((const char *));
+void read_file_free PARAMS ((struct file_memory *));
 
 void free_vec PARAMS ((char **));
 char **merge_vecs PARAMS ((char **, char **));
-slist *add_slist PARAMS ((slist *, const char *, int));
-int in_slist PARAMS ((slist *, const char *));
-void free_slist PARAMS ((slist *));
+slist *slist_append PARAMS ((slist *, const char *));
+int slist_contains PARAMS ((slist *, const char *));
+void slist_free PARAMS ((slist *));
+
+void string_set_add PARAMS ((struct hash_table *, const char *));
+int string_set_exists PARAMS ((struct hash_table *, const char *));
+void string_set_free PARAMS ((struct hash_table *));
+void free_keys_and_values PARAMS ((struct hash_table *));
 
 char *legible PARAMS ((long));
 char *legible_very_long PARAMS ((VERY_LONG_TYPE));
diff --git a/src/wget.h b/src/wget.h
index 196f913c..718ee0b1 100644
--- a/src/wget.h
+++ b/src/wget.h
@@ -71,7 +71,7 @@ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 
 /* Print X if debugging is enabled; a no-op otherwise.  */
 #ifdef DEBUG
-# define DEBUGP(x) do { debug_logprintf x; } while (0)
+# define DEBUGP(x) do { if (opt.debug) { debug_logprintf x; } } while (0)
 #else  /* not DEBUG */
 # define DEBUGP(x) DO_NOTHING
 #endif /* not DEBUG */
-- 
2.39.2