updated for version 7.2-201

runtime/doc/options.txt

@@ -1443,6 +1443,14 @@ A jump table for the options with a short description can be found at |Q_op|.
 	autoselectml	Like "autoselect", but for the modeless selection
 			only.  Compare to the 'A' flag in 'guioptions'.
 
+	html		When the clipboard contains HTML, use this when
+			pasting.  When putting text on the clipboard, mark it
+			as HTML.  This works to copy rendered HTML from
+			Firefox, paste it as raw HTML in Vim, select the HTML
+			in Vim and paste it in a rich edit box in Firefox.
+			Only supported for GTK version 2 and later.
+			Only available with the |+multi_byte| feature.
+
 	exclude:{pattern}
 			Defines a pattern that is matched against the name of
 			the terminal 'term'.  If there is a match, no

src/globals.h

@@ -509,6 +509,7 @@ EXTERN VimClipboard clip_plus;	/* CLIPBOARD selection in X11 */
 EXTERN int	clip_unnamed INIT(= FALSE);
 EXTERN int	clip_autoselect INIT(= FALSE);
 EXTERN int	clip_autoselectml INIT(= FALSE);
+EXTERN int	clip_html INIT(= FALSE);
 EXTERN regprog_T *clip_exclude_prog INIT(= NULL);
 #endif
 

src/gui_gtk_x11.c

@@ -107,6 +107,7 @@ enum
     TARGET_UTF8_STRING,
     TARGET_STRING,
     TARGET_COMPOUND_TEXT,
+    TARGET_HTML,
     TARGET_TEXT,
     TARGET_TEXT_URI_LIST,
     TARGET_TEXT_PLAIN,
@@ -123,6 +124,7 @@ static const GtkTargetEntry selection_targets[] =
     {VIMENC_ATOM_NAME,	0, TARGET_VIMENC},
     {VIM_ATOM_NAME,	0, TARGET_VIM},
 #ifdef FEAT_MBYTE
+    {"text/html",	0, TARGET_HTML},
     {"UTF8_STRING",	0, TARGET_UTF8_STRING},
 #endif
     {"COMPOUND_TEXT",	0, TARGET_COMPOUND_TEXT},
@@ -140,6 +142,7 @@ static const GtkTargetEntry dnd_targets[] =
 {
     {"text/uri-list",	0, TARGET_TEXT_URI_LIST},
 # ifdef FEAT_MBYTE
+    {"text/html",	0, TARGET_HTML},
     {"UTF8_STRING",	0, TARGET_UTF8_STRING},
 # endif
     {"STRING",		0, TARGET_STRING},
@@ -178,6 +181,7 @@ static GdkAtom save_yourself_atom = GDK_NONE;
  * Atoms used to control/reference X11 selections.
  */
 #ifdef FEAT_MBYTE
+static GdkAtom html_atom = GDK_NONE;
 static GdkAtom utf8_string_atom = GDK_NONE;
 #endif
 #ifndef HAVE_GTK2
@@ -1364,6 +1368,24 @@ selection_received_cb(GtkWidget		*widget UNUSED,
 	    else
 		text = tmpbuf_utf8;
 	}
+	else if (len >= 2 && text[0] == 0xff && text[1] == 0xfe)
+	{
+	    vimconv_T conv;
+
+	    /* UTF-16, we get this for HTML */
+	    conv.vc_type = CONV_NONE;
+	    convert_setup_ext(&conv, (char_u *)"utf-16le", FALSE, p_enc, TRUE);
+
+	    if (conv.vc_type != CONV_NONE)
+	    {
+		text += 2;
+		len -= 2;
+		tmpbuf = string_convert(&conv, text, &len);
+		convert_setup(&conv, NULL, NULL);
+	    }
+	    if (tmpbuf != NULL)
+		text = tmpbuf;
+	}
     }
 #else /* !HAVE_GTK2 */
 # ifdef FEAT_MBYTE
@@ -1451,6 +1473,7 @@ selection_get_cb(GtkWidget	    *widget UNUSED,
 
     if (info != (guint)TARGET_STRING
 #ifdef FEAT_MBYTE
+	    && (!clip_html || info != (guint)TARGET_HTML)
 	    && info != (guint)TARGET_UTF8_STRING
 	    && info != (guint)TARGET_VIMENC
 #endif
@@ -1486,6 +1509,40 @@ selection_get_cb(GtkWidget	    *widget UNUSED,
     }
 
 #ifdef FEAT_MBYTE
+    else if (info == (guint)TARGET_HTML)
+    {
+	vimconv_T conv;
+
+	/* Since we get utf-16, we probably should set it as well. */
+	conv.vc_type = CONV_NONE;
+	convert_setup_ext(&conv, p_enc, TRUE, (char_u *)"utf-16le", FALSE);
+	if (conv.vc_type != CONV_NONE)
+	{
+	    tmpbuf = string_convert(&conv, string, &length);
+	    convert_setup(&conv, NULL, NULL);
+	    vim_free(string);
+	    string = tmpbuf;
+	}
+
+	/* Prepend the BOM: "fffe" */
+	if (string != NULL)
+	{
+	    tmpbuf = alloc(length + 2);
+	    tmpbuf[0] = 0xff;
+	    tmpbuf[1] = 0xfe;
+	    mch_memmove(tmpbuf + 2, string, (size_t)length);
+	    vim_free(string);
+	    string = tmpbuf;
+	    length += 2;
+
+	    selection_data->type = selection_data->target;
+	    selection_data->format = 16;	/* 16 bits per char */
+	    gtk_selection_data_set(selection_data, html_atom, 16,
+							      string, length);
+	    vim_free(string);
+	}
+	return;
+    }
     else if (info == (guint)TARGET_VIMENC)
     {
 	int l = STRLEN(p_enc);
@@ -3464,6 +3521,7 @@ gui_mch_init(void)
 
     /* Initialise atoms */
 #ifdef FEAT_MBYTE
+    html_atom = gdk_atom_intern("text/html", FALSE);
     utf8_string_atom = gdk_atom_intern("UTF8_STRING", FALSE);
 #endif
 #ifndef HAVE_GTK2
@@ -6665,6 +6723,10 @@ clip_mch_request_selection(VimClipboard *cbd)
 
     for (i = 0; i < N_SELECTION_TARGETS; ++i)
     {
+#ifdef FEAT_MBYTE
+	if (!clip_html && selection_targets[i].info == TARGET_HTML)
+	    continue;
+#endif
 	received_selection = RS_NONE;
 	target = gdk_atom_intern(selection_targets[i].target, FALSE);
 

src/mbyte.c

@@ -3265,7 +3265,7 @@ encname2codepage(name)
 
 # if defined(USE_ICONV) || defined(PROTO)
 
-static char_u *iconv_string __ARGS((vimconv_T *vcp, char_u *str, int slen, int *unconvlenp));
+static char_u *iconv_string __ARGS((vimconv_T *vcp, char_u *str, int slen, int *unconvlenp, int *resultlenp));
 
 /*
  * Call iconv_open() with a check if iconv() works properly (there are broken
@@ -3326,13 +3326,15 @@ my_iconv_open(to, from)
  * If "unconvlenp" is not NULL handle the string ending in an incomplete
  * sequence and set "*unconvlenp" to the length of it.
  * Returns the converted string in allocated memory.  NULL for an error.
+ * If resultlenp is not NULL, sets it to the result length in bytes.
  */
     static char_u *
-iconv_string(vcp, str, slen, unconvlenp)
+iconv_string(vcp, str, slen, unconvlenp, resultlenp)
     vimconv_T	*vcp;
     char_u	*str;
     int		slen;
     int		*unconvlenp;
+    int		*resultlenp;
 {
     const char	*from;
     size_t	fromlen;
@@ -3418,6 +3420,9 @@ iconv_string(vcp, str, slen, unconvlenp)
 	/* Not enough room or skipping illegal sequence. */
 	done = to - (char *)result;
     }
+
+    if (resultlenp != NULL)
+	*resultlenp = (int)(to - (char *)result);
     return result;
 }
 
@@ -5836,9 +5841,26 @@ convert_setup(vcp, from, to)
     vimconv_T	*vcp;
     char_u	*from;
     char_u	*to;
+{
+    return convert_setup_ext(vcp, from, TRUE, to, TRUE);
+}
+
+/*
+ * As convert_setup(), but only when from_unicode_is_utf8 is TRUE will all
+ * "from" unicode charsets be considered utf-8.  Same for "to".
+ */
+    int
+convert_setup_ext(vcp, from, from_unicode_is_utf8, to, to_unicode_is_utf8)
+    vimconv_T	*vcp;
+    char_u	*from;
+    int		from_unicode_is_utf8;
+    char_u	*to;
+    int		to_unicode_is_utf8;
 {
     int		from_prop;
     int		to_prop;
+    int		from_is_utf8;
+    int		to_is_utf8;
 
     /* Reset to no conversion. */
 # ifdef USE_ICONV
@@ -5856,37 +5878,46 @@ convert_setup(vcp, from, to)
 
     from_prop = enc_canon_props(from);
     to_prop = enc_canon_props(to);
-    if ((from_prop & ENC_LATIN1) && (to_prop & ENC_UNICODE))
+    if (from_unicode_is_utf8)
+	from_is_utf8 = from_prop & ENC_UNICODE;
+    else
+	from_is_utf8 = from_prop == ENC_UNICODE;
+    if (to_unicode_is_utf8)
+	to_is_utf8 = to_prop & ENC_UNICODE;
+    else
+	to_is_utf8 = to_prop == ENC_UNICODE;
+
+    if ((from_prop & ENC_LATIN1) && to_is_utf8)
     {
 	/* Internal latin1 -> utf-8 conversion. */
 	vcp->vc_type = CONV_TO_UTF8;
 	vcp->vc_factor = 2;	/* up to twice as long */
     }
-    else if ((from_prop & ENC_LATIN9) && (to_prop & ENC_UNICODE))
+    else if ((from_prop & ENC_LATIN9) && to_is_utf8)
     {
 	/* Internal latin9 -> utf-8 conversion. */
 	vcp->vc_type = CONV_9_TO_UTF8;
 	vcp->vc_factor = 3;	/* up to three as long (euro sign) */
     }
-    else if ((from_prop & ENC_UNICODE) && (to_prop & ENC_LATIN1))
+    else if (from_is_utf8 && (to_prop & ENC_LATIN1))
     {
 	/* Internal utf-8 -> latin1 conversion. */
 	vcp->vc_type = CONV_TO_LATIN1;
     }
-    else if ((from_prop & ENC_UNICODE) && (to_prop & ENC_LATIN9))
+    else if (from_is_utf8 && (to_prop & ENC_LATIN9))
     {
 	/* Internal utf-8 -> latin9 conversion. */
 	vcp->vc_type = CONV_TO_LATIN9;
     }
 #ifdef WIN3264
     /* Win32-specific codepage <-> codepage conversion without iconv. */
-    else if (((from_prop & ENC_UNICODE) || encname2codepage(from) > 0)
+    else if ((from_is_utf8 || encname2codepage(from) > 0)
-	    && ((to_prop & ENC_UNICODE) || encname2codepage(to) > 0))
+	    && (to_is_utf8 || encname2codepage(to) > 0))
     {
 	vcp->vc_type = CONV_CODEPAGE;
 	vcp->vc_factor = 2;	/* up to twice as long */
-	vcp->vc_cpfrom = (from_prop & ENC_UNICODE) ? 0 : encname2codepage(from);
+	vcp->vc_cpfrom = from_is_utf8 ? 0 : encname2codepage(from);
-	vcp->vc_cpto = (to_prop & ENC_UNICODE) ? 0 : encname2codepage(to);
+	vcp->vc_cpto = to_is_utf8 ? 0 : encname2codepage(to);
     }
 #endif
 #ifdef MACOS_X
@@ -5894,7 +5925,7 @@ convert_setup(vcp, from, to)
     {
 	vcp->vc_type = CONV_MAC_LATIN1;
     }
-    else if ((from_prop & ENC_MACROMAN) && (to_prop & ENC_UNICODE))
+    else if ((from_prop & ENC_MACROMAN) && to_is_utf8)
     {
 	vcp->vc_type = CONV_MAC_UTF8;
 	vcp->vc_factor = 2;	/* up to twice as long */
@@ -5903,7 +5934,7 @@ convert_setup(vcp, from, to)
     {
 	vcp->vc_type = CONV_LATIN1_MAC;
     }
-    else if ((from_prop & ENC_UNICODE) && (to_prop & ENC_MACROMAN))
+    else if (from_is_utf8 && (to_prop & ENC_MACROMAN))
     {
 	vcp->vc_type = CONV_UTF8_MAC;
     }
@@ -5913,8 +5944,8 @@ convert_setup(vcp, from, to)
     {
 	/* Use iconv() for conversion. */
 	vcp->vc_fd = (iconv_t)my_iconv_open(
-		(to_prop & ENC_UNICODE) ? (char_u *)"utf-8" : to,
+		to_is_utf8 ? (char_u *)"utf-8" : to,
-		(from_prop & ENC_UNICODE) ? (char_u *)"utf-8" : from);
+		from_is_utf8 ? (char_u *)"utf-8" : from);
 	if (vcp->vc_fd != (iconv_t)-1)
 	{
 	    vcp->vc_type = CONV_ICONV;
@@ -6170,9 +6201,7 @@ string_convert_ext(vcp, ptr, lenp, unconvlenp)
 
 # ifdef USE_ICONV
 	case CONV_ICONV:	/* conversion with output_conv.vc_fd */
-	    retval = iconv_string(vcp, ptr, len, unconvlenp);
+	    retval = iconv_string(vcp, ptr, len, unconvlenp, lenp);
-	    if (retval != NULL && lenp != NULL)
-		*lenp = (int)STRLEN(retval);
 	    break;
 # endif
 # ifdef WIN3264

src/option.c

@@ -7024,6 +7024,7 @@ check_clipboard_option()
     int		new_unnamed = FALSE;
     int		new_autoselect = FALSE;
     int		new_autoselectml = FALSE;
+    int		new_html = FALSE;
     regprog_T	*new_exclude_prog = NULL;
     char_u	*errmsg = NULL;
     char_u	*p;
@@ -7047,6 +7048,11 @@ check_clipboard_option()
 	    new_autoselectml = TRUE;
 	    p += 12;
 	}
+	else if (STRNCMP(p, "html", 4) == 0 && (p[4] == ',' || p[4] == NUL))
+	{
+	    new_html = TRUE;
+	    p += 4;
+	}
 	else if (STRNCMP(p, "exclude:", 8) == 0 && new_exclude_prog == NULL)
 	{
 	    p += 8;
@@ -7068,6 +7074,7 @@ check_clipboard_option()
 	clip_unnamed = new_unnamed;
 	clip_autoselect = new_autoselect;
 	clip_autoselectml = new_autoselectml;
+	clip_html = new_html;
 	vim_free(clip_exclude_prog);
 	clip_exclude_prog = new_exclude_prog;
     }

src/version.c

@@ -676,6 +676,8 @@ static char *(features[]) =
 
 static int included_patches[] =
 {   /* Add new patch number below this line */
+/**/
+    201,
 /**/
     200,
 /**/