#10 status bar virtual column (%v, %V) and selected chars (%n)

[Dimitar Zhekov]

As you may have noticed, when cursor is in the "virtual space" (beyond eoln), our status bar always shows the last "real" column. This small patch adds a new %v (and %V) format specification, which displays a "virtual" column number, as if the line extends to the cursor position.

sci_get_cursor_virtual_space() isn't included the plugins interface for now, though it may be useful.

Since the virtual position normally requires 3 SSM()-s, and is less likely to be used than the line and column, the calls to sci_get_cursor_virtual_space() are into add_statusbar_statistics() instead of ui_update_statusbar().

[Dimitar Zhekov]

applies against 1.22, +17 loc, 0 changed/removed

[Dimitar Zhekov]

Added %n for selected chars, since %s now shows the selected chars-or-lines.

[Dimitar Zhekov]

for 2012-12-17+

[Dimitar Zhekov]

Added %n for selected characters.

[Dimitar Zhekov]

For Geany 2013-09-26+

[Lex Trotman]

Hi Dimitar, that looks ok, can you add doco in geany.txt to the patch as well.

[Dimitar Zhekov]

Added documentation. Feel free to change it, I'm not an Englisher.

[Dimitar Zhekov]

For Geany 2015-02-22+

[Colomban Wendling]

Looks mostly good, but:

• subject in the patch is wrong;
• shouldn't we simply alter how %c and %C work so they behave like your %v and %V? I don't see no real use case where they wouldn't agree and someone wouldn't prefer the virtual space value?

[Lex Trotman]

On 25 February 2015 at 11:57, Colomban Wendling colombanw@users.sf.net wrote:

Looks mostly good, but:

subject in the patch is wrong;
shouldn't we simply alter how %c and %C work so they behave like your %v and
%V? I don't see no real use case where they wouldn't agree and someone
wouldn't prefer the virtual space value?

We have three virtual space options:

• off - %c and %v are the same

• rectangular selections only (default) - %c and %v are different, %v
  is indeed useful for ensuring your selection is wide enough if you
  can't see the first line in the range, at all other times %c and %v
  are the same

• always - %c and %v are different if the cursor is in virtual
  territory, I don't use this mode but I am not sure why you would want
  %c as well as %v

So to me just having %c being the column the cursor is in, virtual or
real is fine, and it makes better logical sense too.

And finally a warning that Scintilla gets it wrong in the case of
combining chars, compare (assuming the utf-8 gets through correctly):

Trögar
Trögar

Cheers
Lex

---

[patches:#10] status bar virtual column (%v, %V) and selected chars (%n)

Status: open
Group:
Created: Sat Aug 27, 2011 05:55 PM UTC by Dimitar Zhekov
Last Updated: Sun Feb 22, 2015 05:25 PM UTC
Owner: nobody

As you may have noticed, when cursor is in the "virtual space" (beyond
eoln), our status bar always shows the last "real" column. This small patch
adds a new %v (and %V) format specification, which displays a "virtual"
column number, as if the line extends to the cursor position.

sci_get_cursor_virtual_space() isn't included the plugins interface for now,
though it may be useful.

Since the virtual position normally requires 3 SSM()-s, and is less likely
to be used than the line and column, the calls to
sci_get_cursor_virtual_space() are into add_statusbar_statistics() instead
of ui_update_statusbar().

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/geany/patches/10/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

[Dimitar Zhekov]

I'm working under Windows now, and have trouble pasting the sample "Trogar" texts. Did you mean that the SCI column fails generally, or only when displaying position in the virtual space?

[Lex Trotman]

On 27 February 2015 at 05:34, Dimitar Zhekov sheckley@users.sf.net wrote:

I'm working under Windows now, and have trouble pasting the sample "Trogar"
texts. Did you mean that the SCI column fails generally, or only when
displaying position in the virtual space?

Thought that might happen :)

Scintilla gets the column count different between the two since the
first uses U+00F6 while the second uses U+006F U+0308 for the o with
diaresis

I don't think there is anything we can do with it, it was just a note
that Scintilla does not handle it properly. Also not sure it handles
double wide chars properly.

Cheers
Lex

---

[patches:#10] status bar virtual column (%v, %V) and selected chars (%n)

Status: open
Group:
Created: Sat Aug 27, 2011 05:55 PM UTC by Dimitar Zhekov
Last Updated: Wed Feb 25, 2015 12:57 AM UTC
Owner: nobody

As you may have noticed, when cursor is in the "virtual space" (beyond
eoln), our status bar always shows the last "real" column. This small patch
adds a new %v (and %V) format specification, which displays a "virtual"
column number, as if the line extends to the cursor position.

sci_get_cursor_virtual_space() isn't included the plugins interface for now,
though it may be useful.

Since the virtual position normally requires 3 SSM()-s, and is less likely
to be used than the line and column, the calls to
sci_get_cursor_virtual_space() are into add_statusbar_statistics() instead
of ui_update_statusbar().

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/geany/patches/10/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

[Dimitar Zhekov]

Hmm, yes, I got the subject from another patch.

I, too, do not see a case where somebody will work with virtual space, but want the current %c and %C instead.

Though whoever altered %s to display the # of selected lines if entire lines are selected must thought the same, and that's why this patch contains %n... :)

Anyway, I'll post a newer version sometime after I finish with the spawn effort...

[Dimitar Zhekov]

Fixed subject, moved %v / %V into %c / %C and updated the documentation.

[Colomban Wendling]

Looks nice, but as I'm a very annoying people, I'll ask you if you don't think that as now %v is in %c, the column should be computed only once (e.g. revert the argument you had in the 1st message here, as it's now always used):

diff --git a/src/ui_utils.c b/src/ui_utils.c
index 4600de8..512bb11 100644
--- a/src/ui_utils.c
+++ b/src/ui_utils.c
@@ -185,7 +185,7 @@ void ui_set_statusbar(gboolean log, const gchar *format, ...)

 /* note: some comments below are for translators */
 static gchar *create_statusbar_statistics(GeanyDocument *doc,
-   guint line, guint col, guint pos)
+   guint line, guint vcol, guint pos)
 {
    const gchar *cur_tag;
    const gchar *fmt;
@@ -216,12 +216,10 @@ static gchar *create_statusbar_statistics(GeanyDocument *doc,
                    sci_get_line_count(doc->editor->sci));
                break;
            case 'c':
-               g_string_append_printf(stats_str, "%d",
-                   col + sci_get_cursor_virtual_space(doc->editor->sci));
+               g_string_append_printf(stats_str, "%d", vcol);
                break;
            case 'C':
-               g_string_append_printf(stats_str, "%d",
-                   col + sci_get_cursor_virtual_space(doc->editor->sci) + 1);
+               g_string_append_printf(stats_str, "%d", vcol + 1);
                break;
            case 'p':
                g_string_append_printf(stats_str, "%u", pos);
@@ -334,7 +332,7 @@ void ui_update_statusbar(GeanyDocument *doc, gint pos)

    if (doc != NULL)
    {
-       guint line, col;
+       guint line, vcol;
        gchar *stats_str;

        if (pos == -1)
@@ -345,11 +343,12 @@ void ui_update_statusbar(GeanyDocument *doc, gint pos)
         * when current pos is beyond document end (can occur when removing
         * blocks of selected lines especially esp. brace sections near end of file). */
        if (pos <= sci_get_length(doc->editor->sci))
-           col = sci_get_col_from_position(doc->editor->sci, pos);
+           vcol = sci_get_col_from_position(doc->editor->sci, pos);
        else
-           col = 0;
+           vcol = 0;
+       vcol += sci_get_cursor_virtual_space(doc->editor->sci);

-       stats_str = create_statusbar_statistics(doc, line, col, pos);
+       stats_str = create_statusbar_statistics(doc, line, vcol, pos);

        /* can be overridden by status messages */
        set_statusbar(stats_str, TRUE);

[Dimitar Zhekov]

Looks better to me, and since and create_statusbar_statistics() is static, why not? Please drop me a note whether I should recreate the patch, or you will do so.

[Colomban Wendling]

I'll merge it in your patch, that's not a problem. (unless you're not happy with this)

[Colomban Wendling]

Committed with the small reformatting changes as [a5b99dc96dfed95bb08940e33b8ce5153f5069d5]

[Colomban Wendling]

err, of course I noticed the prototype was not inside GEANY_PRIVATE after pushing… fixed in [87af9597fc66dc87e7a5aea21d9dec91f6d4b684].

[Dimitar Zhekov]

OK, thanks.