Commit graph

583 commits

Author SHA1 Message Date
Daniel Eklöf
033e1bd6ae
changelog: generate-alt-random-writes.py --sixel fixes 2021-04-22 11:26:59 +02:00
Daniel Eklöf
4044b6fa99
Merge branch 'releases/1.7' 2021-04-18 20:16:48 +02:00
Daniel Eklöf
7e2e9d4de8
changelog: prepare for 1.7.2 2021-04-18 20:12:26 +02:00
Daniel Eklöf
2f8ee9e61d
changelog: add cglogic to ‘contributors’ 2021-04-18 20:12:02 +02:00
Daniel Eklöf
b8f2c3b784
changelog: add cglogic to ‘contributors’ 2021-04-18 20:11:27 +02:00
Daniel Eklöf
1df48fcb33
config: bold-text-in-bright: add ‘palette-based’ as a special value
When ‘bold-text-in-bright’ is set ‘to palette-based’, colors matching
one of the 8 regular palette colors are brightened by using the
corresponding bright palette color.

Other colors, or all colors if ‘bold-text-in-bright’ is set to
‘yes|true’, are brightened by increasing the luminance.
2021-04-18 20:09:10 +02:00
Daniel Eklöf
11e8ff9dc5
render: brighten: use corresponding bright palette color for base 8 colors
When brightening one of the 8 base (“regular”) colors, use the
corresponding bright palette color, instead of increasing the
luminance.

Closes #449
2021-04-18 20:09:07 +02:00
Daniel Eklöf
8561fdb004
render: don’t let URL jump label sub-surfaces extend outside window geometry
We have no guarantee that sub-surfaces extending outside the window
geometry are rendered correctly (if at all).

For example, both Sway and River will render the window border on top
of the sub-surface.

Future versions of Sway may clip the sub-surface.

Since jump-labels are positioned slightly above, and to the left of
the URLs first character, having a label on either the top row, or on
the first column, will likely position it outside the window. This is
handled by simply setting x/y to 0 (or, to -margin, since the label
coordinate is later offsetted with the window margins).

Second, if the label is very long, it may extend outside the
window. This is very unusual for labels only showing the key, and not
the URL itself, but could happen in this case too, if e.g. the user
has configured double-width key characters.

This is handled by calculating its maximum width, and then truncating
the label.

Although very unlikely, it is possible for a label to also extend
outside the window’s vertical size. This could happen for very small
font sizes, where the label’s own margins are large, relative to the
font size. This case is currently not handled.

Closes #443
2021-04-18 20:08:41 +02:00
Daniel Eklöf
ce321e663e
config: add box-drawings-uses-font-glyphs=no|yes option
When disabled, we render box drawing characters ourselves. This is the
default.

When enabled, we instead use font glyphs. I.e. no special treatment.

Closes #430
2021-04-18 20:08:30 +02:00
Daniel Eklöf
2efd1016ce
render: draw underline cursor below text underline
This ensures the cursor is visible on cells where the text is
underlined.

Closes #415
2021-04-18 20:08:21 +02:00
Daniel Eklöf
0f792fa1ce
changelog: rewrite entry for OSC 17/117/19/119 2021-04-18 20:08:17 +02:00
Daniel Eklöf
c005c16504
changelog: auto_left_margin on cub1 only 2021-04-18 20:08:14 +02:00
Daniel Eklöf
d620a7b3b4
osc: implement OSC 17+19: change selection background/foreground colors
And of course, we also implement the corresponding reset sequences,
OSC 117+119.
2021-04-18 20:08:08 +02:00
Daniel Eklöf
660b771396
changelog: don’t use a hyphen between ‘OSC’ and the number 2021-04-18 20:08:02 +02:00
Daniel Eklöf
220fdc173e
changelog: tweak wording for URxvt’s XParseColor extension 2021-04-18 20:08:00 +02:00
Daniel Eklöf
1023f9a6e9
osc: add support for URxvt extension to include alpha in OSC-11
There are two variants, either using a ‘[percent]’ prefix for legacy
colors, or using ‘rgba:r/g/b/a’ instead of ‘rgb:r/g/b’.

Closes #436
2021-04-18 20:07:54 +02:00
Daniel Eklöf
f42e42246f
terminal: drain PTY when client terminates
This is done by:

* Not limiting the number of times we try to read from the PTY when
  we’ve have POLLHUP
* Not requiring the entire the previous read to have filled our
  buffer.
* Not erroring out on EIO.
2021-04-18 20:07:50 +02:00
Daniel Eklöf
4a2ff28850
changelog: move OSC-4/104 fix from 1.7.1 to ‘unreleased’ 2021-04-18 20:07:19 +02:00
Daniel Eklöf
4aae54e3e7
osc: verify OSC 4/104 color index is valid
Closes #434
2021-04-18 20:07:17 +02:00
Daniel Eklöf
5dbeb5ce27
config: parse_key_binding_section(): use provided action-map
Don’t call has_key_binding_collisions() with ‘binding_action_map’
unconditionally; use the provided ‘action_map’ instead.

This fixes wrong error messages for key combo collisions in key
binding sections other than the regular “key-bindings”.
2021-04-18 20:07:05 +02:00
Daniel Eklöf
6e782271ff
sixel: fix double free caused by bad free() in sixel_colors_set()
sixel_color_set() is called when the number of (sixel) color registers
is changed.

It frees the current palette, and changes the “palette size” variable.

Originally, we only had a single palette. This is the one free:d by
sixel_color_set().

Later, we added support for private vs. shared palettes. With this
change, we now have one palette that is “never” free:d (the shared
one), and a private palette that is always free:d after a sixel has
been emitted.

‘sixel.palette’ is a pointer to the palette currently in use, and
should only be accessed **while emitting a sixel**.

This is the pointer sixel_color_set() free:d. So for example, if
‘sixel.palette’ pointed to the shared palette, we’d free the shared
palette. But, we didn’t reset ‘sixel.shared_palette’, causing a double
free later on.

Closes #427
2021-04-18 20:07:02 +02:00
Daniel Eklöf
0720b3177a
revert: input: do not include consumed modifiers in the set sent to the client
There were two issues with it:

* Not all applications decode the sequence into a set of modifiers +
  key, but use a fixed sequence -> combo mapping, that we broke.

* There were unforeseen issues with e.g. F1-12, where the modifier
  were removed from combos like Ctrl+F12, or Alt+F12. The reason is
  simple; XKB tells us that Ctrl, or Alt, is a consumed modifier. Now,
  _why_ it thinks that is a different story.

This reverts 6cd72bdee6

Closes #425
2021-04-18 20:06:59 +02:00
Daniel Eklöf
6b116865f6
changelog: add new ‘unreleased’ section 2021-04-18 20:06:44 +02:00
Daniel Eklöf
cdfc864cf0
config: bold-text-in-bright: add ‘palette-based’ as a special value
When ‘bold-text-in-bright’ is set ‘to palette-based’, colors matching
one of the 8 regular palette colors are brightened by using the
corresponding bright palette color.

Other colors, or all colors if ‘bold-text-in-bright’ is set to
‘yes|true’, are brightened by increasing the luminance.
2021-04-17 22:00:48 +02:00
Daniel Eklöf
2cd9fee9c4
render: brighten: use corresponding bright palette color for base 8 colors
When brightening one of the 8 base (“regular”) colors, use the
corresponding bright palette color, instead of increasing the
luminance.

Closes #449
2021-04-17 21:02:02 +02:00
Daniel Eklöf
c6848678ad
render: don’t let URL jump label sub-surfaces extend outside window geometry
We have no guarantee that sub-surfaces extending outside the window
geometry are rendered correctly (if at all).

For example, both Sway and River will render the window border on top
of the sub-surface.

Future versions of Sway may clip the sub-surface.

Since jump-labels are positioned slightly above, and to the left of
the URLs first character, having a label on either the top row, or on
the first column, will likely position it outside the window. This is
handled by simply setting x/y to 0 (or, to -margin, since the label
coordinate is later offsetted with the window margins).

Second, if the label is very long, it may extend outside the
window. This is very unusual for labels only showing the key, and not
the URL itself, but could happen in this case too, if e.g. the user
has configured double-width key characters.

This is handled by calculating its maximum width, and then truncating
the label.

Although very unlikely, it is possible for a label to also extend
outside the window’s vertical size. This could happen for very small
font sizes, where the label’s own margins are large, relative to the
font size. This case is currently not handled.

Closes #443
2021-04-10 13:23:20 +02:00
Daniel Eklöf
f78ba6653f
Merge branch 'optionally-render-box-drawing-characters-using-font-glyphs'
Closes #430
2021-04-10 11:11:47 +02:00
Daniel Eklöf
0713ab45c6
config: add box-drawings-uses-font-glyphs=no|yes option
When disabled, we render box drawing characters ourselves. This is the
default.

When enabled, we instead use font glyphs. I.e. no special treatment.

Closes #430
2021-04-09 23:21:20 +02:00
Daniel Eklöf
20849a4183
render: draw underline cursor below text underline
This ensures the cursor is visible on cells where the text is
underlined.

Closes #415
2021-04-09 21:47:57 +02:00
Daniel Eklöf
d84341827d
changelog: rewrite entry for OSC 17/117/19/119 2021-04-08 13:50:02 +02:00
Daniel Eklöf
131db418d8
changelog: auto_left_margin on cub1 only 2021-04-08 13:12:04 +02:00
Daniel Eklöf
55b343f690
osc: implement OSC 17+19: change selection background/foreground colors
And of course, we also implement the corresponding reset sequences,
OSC 117+119.
2021-04-08 10:43:36 +02:00
Daniel Eklöf
747e5ba287
changelog: don’t use a hyphen between ‘OSC’ and the number 2021-04-08 10:41:39 +02:00
Daniel Eklöf
339aad894b
changelog: tweak wording for URxvt’s XParseColor extension 2021-04-08 10:41:38 +02:00
Daniel Eklöf
ed41bffc28
osc: add support for URxvt extension to include alpha in OSC-11
There are two variants, either using a ‘[percent]’ prefix for legacy
colors, or using ‘rgba:r/g/b/a’ instead of ‘rgb:r/g/b’.

Closes #436
2021-04-08 10:41:38 +02:00
Daniel Eklöf
fb4066f94e
terminal: drain PTY when client terminates
This is done by:

* Not limiting the number of times we try to read from the PTY when
  we’ve have POLLHUP
* Not requiring the entire the previous read to have filled our
  buffer.
* Not erroring out on EIO.
2021-04-07 19:09:31 +02:00
Daniel Eklöf
e299b1d9f4
changelog: move OSC-4/104 fix from 1.7.1 to ‘unreleased’ 2021-04-05 22:07:58 +02:00
Daniel Eklöf
4f0c0628a2
osc: verify OSC 4/104 color index is valid
Closes #434
2021-04-05 21:13:32 +02:00
Daniel Eklöf
285e3c4aa2
config: parse_key_binding_section(): use provided action-map
Don’t call has_key_binding_collisions() with ‘binding_action_map’
unconditionally; use the provided ‘action_map’ instead.

This fixes wrong error messages for key combo collisions in key
binding sections other than the regular “key-bindings”.
2021-04-05 14:36:22 +02:00
Daniel Eklöf
00083125a1
sixel: fix double free caused by bad free() in sixel_colors_set()
sixel_color_set() is called when the number of (sixel) color registers
is changed.

It frees the current palette, and changes the “palette size” variable.

Originally, we only had a single palette. This is the one free:d by
sixel_color_set().

Later, we added support for private vs. shared palettes. With this
change, we now have one palette that is “never” free:d (the shared
one), and a private palette that is always free:d after a sixel has
been emitted.

‘sixel.palette’ is a pointer to the palette currently in use, and
should only be accessed **while emitting a sixel**.

This is the pointer sixel_color_set() free:d. So for example, if
‘sixel.palette’ pointed to the shared palette, we’d free the shared
palette. But, we didn’t reset ‘sixel.shared_palette’, causing a double
free later on.

Closes #427
2021-03-30 11:08:03 +02:00
Daniel Eklöf
058eba33ec
revert: input: do not include consumed modifiers in the set sent to the client
There were two issues with it:

* Not all applications decode the sequence into a set of modifiers +
  key, but use a fixed sequence -> combo mapping, that we broke.

* There were unforeseen issues with e.g. F1-12, where the modifier
  were removed from combos like Ctrl+F12, or Alt+F12. The reason is
  simple; XKB tells us that Ctrl, or Alt, is a consumed modifier. Now,
  _why_ it thinks that is a different story.

This reverts 6cd72bdee6

Closes #425
2021-03-29 20:11:32 +02:00
Daniel Eklöf
5709630362
changelog: add new ‘unreleased’ section 2021-03-28 14:29:55 +02:00
Daniel Eklöf
3c92777dd2
changelog: prepare for 1.7.1 2021-03-28 14:25:09 +02:00
Daniel Eklöf
6ae14614a5
Merge branch 'search-treat-empty-cells-as-spaces' 2021-03-28 14:22:34 +02:00
Daniel Eklöf
efbbcf289f
changelog: empty sixels resulted in non-empty images 2021-03-28 13:22:10 +02:00
Daniel Eklöf
0a24d0e40f
changelog: spaces matches empty cells in scrollback search mode 2021-03-28 13:21:33 +02:00
Daniel Eklöf
9786197d03
changelog: updated PGO build instructions 2021-03-26 22:16:55 +01:00
Daniel Eklöf
a8de14c0bf
changelog: sub-surface sizes not a multiple of the scaling factor 2021-03-25 09:42:15 +01:00
Daniel Eklöf
8111ff4be8
render: draw sixels before taking the render worker lock
This fixes a possible deadlock; render_sixels_images() may call
render_cell(), which may need to take the worker lock (when rendering
either a blinking cell, or a box drawing glyph that isn’t yet in the
glyph cache).
2021-03-25 09:38:51 +01:00
Daniel Eklöf
6ae8ed7e6b
changelog: don’t crash upon receiving text-input::enter without keyboard focus 2021-03-25 09:36:27 +01:00