Move Wget from IDN2003 (libidn) to IDN2008 (libidn2)

* .travis.yml: Install libidn2-dev instead libidn11-dev.
* bootstrap.conf: Add modules libunistring-optional, unistr/base,
  unicase/tolower.
* configure.ac: Check for libidn2.
* src/Makefile.am: Add $(LTLIBUNISTRING) to LDADD.
* tests/Makefile.am: Set LDADD similar to LDADD in src/Makefile.am
* src/connect.c: Use libidn2 code instead of libidn.
* src/host.c: Likewise.
* src/iri.c: Likewise.
* src/iri.h: Likewise.
* src/options.h: Likewise.
* src/url.c: Likewise.
* src/url.h: Likewise.
* src/log.c: Fix C99 comment.

IDN2003 should not be used any more due to security concerns.
We use libunistring (resp. the unicode code from gnulib) for
lowercasing UTF-8 before we give data to libidn2.
TR#46 is missing, no support in libidn2 nor in libunistring.

.travis.yml

@@ -36,7 +36,7 @@ addons:
             - make
             - libhttp-daemon-perl
             - libio-socket-ssl-perl
-            - libidn11-dev
+            - libidn2-dev
             - gettext
             - texlive
             - python3

bootstrap.conf

@@ -101,6 +101,9 @@ tmpdir
 unlink
 unlocked-io
 update-copyright
+libunistring-optional
+unistr/base
+unicase/tolower
 vasprintf
 vsnprintf
 write

configure.ac

@@ -640,9 +640,6 @@ AC_ARG_ENABLE(iri,
   ]
 )
 
-AC_ARG_WITH(libidn, AC_HELP_STRING([--with-libidn=[DIR]],
-                                   [Support IDN/IRIs (needs GNU Libidn)]),
-                                   libidn=$withval, libidn="")
 AS_IF([test "X$iri" != "Xno"],[
   AM_ICONV
 
@@ -660,33 +657,26 @@ AS_IF([test "X$iri" != "Xno"],[
   LIBICONV=
 ])
 
+AC_ARG_WITH(libidn, AC_HELP_STRING([--with-libidn=[DIR]],
+                                   [Support IDN2008/IRIs (needs GNU libidn2 + libunicode)]),
+                                   libidn=$withval, libidn="")
 if test "X$iri" != "Xno"; then
-  if test "$libidn" != ""; then
-    LDFLAGS="${LDFLAGS} -L$libidn/lib"
-    CPPFLAGS="${CPPFLAGS} -I$libidn/include"
-  fi
+  AS_IF([test "x$with_libidn2" != xno], [
+    AC_SEARCH_LIBS(idn2_lookup_u8, idn2,
+      [with_libidn2=yes; AC_DEFINE([ENABLE_IRI], 1, [Define if IRI support is enabled.])],
+      [with_libidn2=no; iri=no; AC_MSG_WARN(*** LIBIDN2 was not found. You will not be able to use IDN2008 support)])
 
-  # If idna.h can't be found, check to see if it was installed under
-  # /usr/include/idn (OpenSolaris, at least, places it there).
-  # Check for idn-int.h in that case, because idna.h won't find
-  # idn-int.h until we've decided to add -I/usr/include/idn.
-  AC_CHECK_HEADER(idna.h, ,
-    [AC_CHECK_HEADER(idn/idn-int.h,
-                    [CPPFLAGS="${CPPFLAGS} -I/usr/include/idn"],
-                    [iri=no])]
-  )
-
-  if test "X$iri" != "Xno"; then
-    AC_CHECK_LIB(idn, stringprep_check_version,
-      [iri=yes LIBS="${LIBS} -lidn"], iri=no)
-  fi
-
-  if test "X$iri" != "Xno" ; then
-    AC_DEFINE([ENABLE_IRI], 1, [Define if IRI support is enabled.])
-    AC_MSG_NOTICE([Enabling support for IRI.])
-  else
-    AC_MSG_WARN([Libidn not found])
-  fi
+#    AS_IF([test "x$with_libidn2" = xyes], [
+#      AC_SEARCH_LIBS(u8_tolower, unistring,
+#        [AC_DEFINE([ENABLE_IRI], 1, [Define if IRI support is enabled.])],
+#        [iri=no; AC_MSG_WARN(*** LIBUNISTRING was not found. You will not be able to use IDN2008 support)])
+#    ])
+  ])
+fi
+if test "X$iri" = "Xno"; then
+  # we don't need libunistring - clear settings from gnulib module
+  LIBUNISTRING=""
+  LTLIBUNISTRING=""
 fi
 
 dnl
@@ -755,7 +745,7 @@ dnl
 dnl Check for libcares (resolver library)
 dnl
 
-AS_IF([test "X$with_cares" == "Xyes"],[
+AS_IF([test "X$with_cares" = "Xyes"],[
   PKG_CHECK_MODULES([CARES], libcares, [
     CFLAGS="$CARES_CFLAGS $CFLAGS"
     AC_CHECK_HEADER(ares.h, [
@@ -839,4 +829,5 @@ AC_MSG_NOTICE([Summary of build options:
   Metalink:          $with_metalink
   Resolver:          $RESOLVER_INFO
   GPGME:             $have_gpg
+  IRI:               $iri
 ])

src/Makefile.am

@@ -65,7 +65,7 @@ nodist_wget_SOURCES = version.c
 EXTRA_wget_SOURCES = iri.c
 LDADD = $(LIBOBJS) ../lib/libgnu.a $(GETADDRINFO_LIB) $(HOSTENT_LIB) $(INET_NTOP_LIB) $(LIBSOCKET)\
  $(LIB_CLOCK_GETTIME) $(LIB_CRYPTO) $(LIB_SELECT) $(LTLIBICONV) $(LTLIBINTL) $(LTLIBTHREAD)\
- $(SERVENT_LIB)
+ $(LTLIBUNISTRING) $(SERVENT_LIB)
 AM_CPPFLAGS = -I$(top_builddir)/lib -I$(top_srcdir)/lib
 
 

src/connect.c

@@ -56,7 +56,7 @@ as that of the covered work.  */
 #include <sys/time.h>
 
 #ifdef ENABLE_IRI
-#include <idn-free.h>
+#include <idn2.h>
 #endif
 
 #include "utils.h"
@@ -283,7 +283,7 @@ connect_to_ip (const ip_address *ip, int port, const char *print)
               str = xmalloc (len);
               snprintf (str, len, "%s (%s)", name, print);
               str[len-1] = '\0';
-              idn_free (name);
+              idn2_free (name);
             }
 
           logprintf (LOG_VERBOSE, _("Connecting to %s|%s|:%d... "),

src/host.c

@@ -58,7 +58,7 @@ as that of the covered work.  */
 #include <errno.h>
 
 #ifdef ENABLE_IRI
-#include <idn-free.h>
+#include <idn2.h>
 #endif
 
 #include "utils.h"
@@ -852,7 +852,7 @@ lookup_host (const char *host, int flags)
           str = xmalloc (len);
           snprintf (str, len, "%s (%s)", name, host);
           str[len-1] = '\0';
-          idn_free (name);
+          idn2_free (name);
         }
 
       logprintf (LOG_VERBOSE, _("Resolving %s... "),

src/iri.c

@@ -33,13 +33,14 @@ as that of the covered work.  */
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
+#include <langinfo.h>
+#include <errno.h>
 #ifdef HAVE_ICONV
 # include <iconv.h>
 #endif
-#include <stringprep.h>
-#include <idna.h>
-#include <idn-free.h>
-#include <errno.h>
+#include <idn2.h>
+#include <unicase.h>
+#include <unistr.h>
 
 #include "utils.h"
 #include "url.h"
@@ -90,10 +91,15 @@ parse_charset (const char *str)
 }
 
 /* Find the locale used, or fall back on a default value */
-char *
+const char *
 find_locale (void)
 {
-  return (char *) stringprep_locale_charset ();
+	const char *encoding = nl_langinfo(CODESET);
+
+	if (!encoding || !*encoding)
+		return "ASCII";
+
+   return encoding;
 }
 
 /* Basic check of an encoding name. */
@@ -284,33 +290,47 @@ idn_encode (const struct iri *i, const char *host)
   int ret;
   char *ascii_encoded;
   char *utf8_encoded = NULL;
+  const char *src;
+  uint8_t *lower;
+  size_t len = 0;
 
   /* Encode to UTF-8 if not done */
   if (!i->utf8_encode)
     {
       if (!remote_to_utf8 (i, host, &utf8_encoded))
           return NULL;  /* Nothing to encode or an error occured */
+      src = utf8_encoded;
     }
+  else
+    src = host;
 
-  if (!_utf8_is_valid(utf8_encoded ? utf8_encoded : host))
+  if (!_utf8_is_valid (src))
     {
       logprintf (LOG_VERBOSE, _("Invalid UTF-8 sequence: %s\n"),
-                 quote(utf8_encoded ? utf8_encoded : host));
+                 quote (src));
       xfree (utf8_encoded);
       return NULL;
     }
 
-  /* Store in ascii_encoded the ASCII UTF-8 NULL terminated string */
-  ret = idna_to_ascii_8z (utf8_encoded ? utf8_encoded : host, &ascii_encoded, IDNA_FLAGS);
+  /* we need a conversion to lowercase */
+  lower = u8_tolower ((uint8_t *) src, u8_strlen ((uint8_t *) src) + 1, 0, UNINORM_NFKC, NULL, &len);
+  if (!lower)
+    {
+      logprintf (LOG_VERBOSE, _("Failed to convert to lower: %d: %s\n"),
+                 errno, quote (src));
       xfree (utf8_encoded);
+      return NULL;
+    }
 
-  if (ret != IDNA_SUCCESS)
+  if ((ret = idn2_lookup_u8 (lower, (uint8_t **) &ascii_encoded, IDN2_NFC_INPUT)) != IDN2_OK)
     {
       logprintf (LOG_VERBOSE, _("idn_encode failed (%d): %s\n"), ret,
-                 quote (idna_strerror (ret)));
-      return NULL;
+                 quote (idn2_strerror (ret)));
+      ascii_encoded = NULL;
     }
 
+  xfree (lower);
+
   return ascii_encoded;
 }
 
@@ -319,18 +339,24 @@ idn_encode (const struct iri *i, const char *host)
 char *
 idn_decode (const char *host)
 {
+/*
   char *new;
   int ret;
 
-  ret = idna_to_unicode_8zlz (host, &new, IDNA_FLAGS);
-  if (ret != IDNA_SUCCESS)
+  ret = idn2_register_u8 (NULL, host, (uint8_t **) &new, 0);
+  if (ret != IDN2_OK)
     {
-      logprintf (LOG_VERBOSE, _("idn_decode failed (%d): %s\n"), ret,
-                 quote (idna_strerror (ret)));
+      logprintf (LOG_VERBOSE, _("idn2_register_u8 failed (%d): %s: %s\n"), ret,
+                 quote (idn2_strerror (ret)), host);
       return NULL;
     }
 
   return new;
+*/
+  /* idn2_register_u8() just works label by label.
+   * That is pretty much overhead for just displaying the original ulabels.
+   * To keep at least the debug output format, return a cloned host. */
+  return xstrdup(host);
 }
 
 /* Try to transcode string str from remote encoding to UTF-8. On success, *new

src/iri.h

@@ -40,11 +40,10 @@ struct iri {
 
 #ifdef ENABLE_IRI
 
-# include <idna.h>
-# include <idn-free.h>
+# include <idn2.h>
 
 char *parse_charset (const char *str);
-char *find_locale (void);
+const char *find_locale (void);
 bool check_encoding_name (const char *encoding);
 const char *locale_to_utf8 (const char *str);
 char *idn_encode (const struct iri *i, const char *host);
@@ -66,7 +65,7 @@ extern struct iri dummy_iri;
 #define locale_to_utf8(str)         (str)
 #define idn_encode(a,b)             NULL
 #define idn_decode(str)             NULL
-#define idn_free(str)               ((void)0)
+#define idn2_free(str)              ((void)0)
 #define remote_to_utf8(a,b,c)       false
 #define iri_new()                   (&dummy_iri)
 #define iri_dup(a)                  (&dummy_iri)

src/log.c

@@ -964,12 +964,12 @@ check_redirect_output (void)
     {
       if (tcgetpgrp (STDIN_FILENO) != getpgrp ())
         {
-          // Process backgrounded
+          /* Process backgrounded */
           redirect_output (true,NULL);
         }
       else
         {
-          // Process foregrounded
+          /* Process foregrounded */
           redirect_output (false,NULL);
         }
     }

src/options.h

@@ -311,7 +311,7 @@ struct options
 
   bool enable_iri;
   char *encoding_remote;
-  char *locale;
+  const char *locale;
 
   bool trustservernames;
 #ifdef __VMS

src/url.c

@@ -1213,7 +1213,7 @@ url_free (struct url *url)
   if (url)
     {
       if (url->idn_allocated) {
-        idn_free (url->host);      /* A dummy if !defined(ENABLE_IRI) */
+        idn2_free (url->host);      /* A dummy if !defined(ENABLE_IRI) */
         url->host = NULL;
       }
       else

src/url.h

@@ -100,8 +100,8 @@ struct url
   char *user;
   char *passwd;
 
-  /* 'host' is allocated by idna_to_ascii_8z() via idn_encode().
-   * Call 'idn_free()' to free this memory. */
+  /* 'host' is allocated by idn2_lookup_u8() via idn_encode().
+   * Call 'idn2_free()' to free this memory. */
   bool idn_allocated;
 };
 

tests/Makefile.am

@@ -136,7 +136,10 @@ EXTRA_DIST = FTPServer.pm FTPTest.pm HTTPServer.pm HTTPTest.pm \
 
 check_PROGRAMS = unit-tests
 unit_tests_SOURCES =
-LDADD = ../src/libunittest.a ../lib/libgnu.a @LIBICONV@ @LIBINTL@ $(LIBS) $(LIB_CLOCK_GETTIME)
+LDADD = ../src/libunittest.a ../lib/libgnu.a $(GETADDRINFO_LIB) $(HOSTENT_LIB) $(INET_NTOP_LIB) $(LIBSOCKET)\
+ $(LIB_CLOCK_GETTIME) $(LIB_CRYPTO) $(LIB_SELECT) $(LTLIBICONV) $(LTLIBINTL) $(LTLIBTHREAD)\
+ $(LTLIBUNISTRING) $(SERVENT_LIB)
+
 
 CLEANFILES = *~ *.bak core core.[0-9]*