T59: a click goes to a shell that takes clicks only past its prompt's last cell before the cursor, never on a prompt cell, so a click on fish's $ leaves its cursor where it stood

AGENTS.md

@@ -114,4 +114,4 @@ ELYTerminal. A terminal on GPUI and Ely components. `PRD.md` says what it does;
 - 2026-10-19: a drag moves Ely along the edge she stands on, which the owner asked for after 0.1.5. On macOS, Windows and X11 her glue moves her window after the pointer's place on the screen, `NSEvent`'s `mouseLocation`, `GetCursorPos` and `query_pointer`, since her window moves under the pointer, so a place in it would lag; her window keeps the pointer while the button is held, inside it or not, as AppKit, gpui's `SetCapture` and X's implicit grab give it. A Wayland popup cannot move, so there she moves once let go: the settings keep the place, and the twice-a-second check opens her popup there. She moves by whole points inside the pane, and `ely_left` keeps where she stands, 140 by default, where she stood before; a narrower pane holds her at its right end, and the log names where she is let go. Under Xorg's dummy screen xdotool dragged her 200 pixels and her window followed them; under Sway nested in Xorg, whose X input gives its seat a pointer, xdotool's drag reopened her popup 200 points on. Sway's own `seat cursor` commands send no motion while their press is held, and Sway draws no popup opened before its window settled, so a probe there reopens her before it drags.
 - 2026-10-19: Settings › Terminal checks for updates, which the owner asked for after 0.1.5: under Ely, a switch, on by default, checks at start and once a day, and Check now checks at once. The check runs `curl`, as git runs through its command line, so no HTTP or TLS crate joins the build; macOS and Windows 10 ship it, and a missing curl shows as a failed check. It asks Gitea's API for the latest published release, compares its tag's three numbers with the running version, and the row says what it found; a newer one brings Download, which opens the release's page in the browser, since replacing a running, signed app is a job of its own. An automatic check that finds one posts a system notification, at most once a day, whose press opens the page. Checks run off the main thread, one at a time, each start and finding logged; a failure shows its error in the row. Under Xorg's dummy screen the start's check found 0.1.5 the latest, Check now and the switch checked again, and a build named 0.1.4 showed 0.1.5 out with Download, which opened its page.
 - 2026-10-19: the terminal shows a scrollbar, which a user asked for in issue #2 after 0.1.5: a program that tracks the mouse takes the wheel, so the wheel alone could not reach the history. A thumb rides the view's right edge, over the padding, in Ely's scrollbar sizes and colors, its thickness, its least length and the foreground at 22 and 42 percent; its length is the screen's share of the history, and its place where the view stands. A drag moves the view through the history, whatever mouse mode the program asked for; the bar covers the grid's hitbox, so the terminal sees none of its presses, and a release with no press of its own reports nothing. Ely's `Scrollbar` reads a gpui `ScrollHandle`, and the terminal's place is alacritty's display offset, so the bar is the terminal's own, drawn as Ely draws hers and dragged through gpui's drag as hers is; gpui sets a drag's anchor at its first move past the threshold, so a pointer that jumps far at once loses that jump. The alternate screen holds no history, so it shows none. Under Xorg's dummy screen, 500 lines in, a drag to the top showed the first prompt and lines 1 to 39, and with mouse tracking on, which sent the wheel to the program, the bar still scrolled.
-- 2026-10-19: a plain click on the line the shell reads moves its cursor there, which a user asked for after 0.1.6, as Ghostty's `cursor-click-to-move`: a press with no modifier, one click, let go in its own cell with no selection, while OSC 133 says the shell reads a line, B since the last A or C, and the program tracks no mouse at the release. A shell whose prompt start says `click_events=1` or `2`, as fish from 4.1 says on its own, hears the click itself, a left press in SGR, its row the screen's or counted from the prompt's first, as kitty and Ghostty send it, and moves its own cursor, combining marks and its own redraws included: fish moves by code point, and its redraws drop the wrap flags a count would follow. Any other shell gets Left or Right arrows, one per cell, each pressed and let go as the keyboard would, through the key encoder, so DECCKM, kitty's flags and win32-input-mode shape them; the count runs only on the cursor's line, soft wraps joined, past the prompt's cells, which the prompt's link marks, so bash 3.2, which sends no C, moves nothing once its command left the prompt's line. Cells after the last character typed count only up to the cursor, so a click past the end goes to the end; a wide character takes one arrow. Never on the alternate screen. One line only, as Ghostty's `cl=line`. A right prompt or an autosuggestion past the typed text counts as text, as in Ghostty, so a click on it may take the suggestion. The first click of a double click moves the cursor before the word selects, as in Ghostty.
+- 2026-10-19: a plain click on the line the shell reads moves its cursor there, which a user asked for after 0.1.6, as Ghostty's `cursor-click-to-move`: a press with no modifier, one click, let go in its own cell with no selection, while OSC 133 says the shell reads a line, B since the last A or C, and the program tracks no mouse at the release. A shell whose prompt start says `click_events=1` or `2`, as fish from 4.1 says on its own, hears a click past the prompt's cells, none on them or on a prompt cell after the cursor, itself, a left press in SGR, its row the screen's or counted from the prompt's first, as kitty and Ghostty send it, and moves its own cursor, combining marks and its own redraws included: fish moves by code point, and its redraws drop the wrap flags a count would follow. Any other shell gets Left or Right arrows, one per cell, each pressed and let go as the keyboard would, through the key encoder, so DECCKM, kitty's flags and win32-input-mode shape them; the count runs only on the cursor's line, soft wraps joined, past the prompt's cells, which the prompt's link marks, so bash 3.2, which sends no C, moves nothing once its command left the prompt's line. Cells after the last character typed count only up to the cursor, so a click past the end goes to the end; a wide character takes one arrow. Never on the alternate screen. One line only, as Ghostty's `cl=line`. A right prompt or an autosuggestion past the typed text counts as text, as in Ghostty, so a click on it may take the suggestion. The first click of a double click moves the cursor before the word selects, as in Ghostty.

src/term/click.rs

@@ -92,16 +92,30 @@ pub fn arrows<T>(term: &Term<T>, click: Point) -> Option<i32> {
     })
 }
 
-/// The left press, SGR, a shell that takes clicks hears for `click`, its row the screen's or counted from its prompt's first, as kitty and Ghostty send it; none before that prompt.
+/// The left press, SGR, a shell that takes clicks hears for `click`, its row the screen's or counted from its prompt's first, as kitty and Ghostty send it; none on the prompt or before its end.
 pub fn press<T>(term: &Term<T>, click: Point, relative: bool) -> Option<Vec<u8>> {
     if term.mode().contains(TermMode::ALT_SCREEN) {
         return None;
     }
-    let first = start(term, term.grid().cursor.point.line.0)?;
-    let opens = (0..term.columns())
-        .position(|column| prompted(&term.grid()[Line(first)][Column(column)]))
+    let grid = term.grid();
+    let cursor = grid.cursor.point;
+    let first = start(term, cursor.line.0)?;
+    let cell = |line: i32, column: usize| &grid[Line(line)][Column(column)];
+    // The prompt ends at its last cell before the cursor.
+    let end = (first..=cursor.line.0)
+        .flat_map(|line| {
+            let width = if line == cursor.line.0 {
+                cursor.column.0
+            } else {
+                term.columns()
+            };
+            (0..width).map(move |column| (line, column))
+        })
+        .filter(|&(line, column)| prompted(cell(line, column)))
+        .last()
         .expect("a prompt's first row holds its cells");
-    if (click.line.0, click.column.0) < (first, opens) {
+    let at = (click.line.0, click.column.0);
+    if at <= end || prompted(cell(at.0, at.1)) {
         return None;
     }
     let row = if relative {
@@ -129,7 +143,7 @@ impl TerminalView {
         }
         if takes != Click::Arrows {
             let Some(bytes) = press(&self.pty.term.lock(), click, takes == Click::Relative) else {
-                log::debug!("terminal {pid}: a click at {click:?}, before the prompt");
+                log::debug!("terminal {pid}: a click at {click:?}, off the input");
                 return;
             };
             log::debug!("terminal {pid}: a click at {click:?} goes to the shell as {bytes:?}");

src/term/click/tests.rs

@@ -114,6 +114,31 @@ fn a_shell_that_takes_clicks_hears_the_press_where_it_landed() {
         "a row below goes too"
     );
     assert_eq!(sent(&typed, click(0, 1), false), None, "before the prompt");
+    assert_eq!(sent(&typed, click(1, 0), false), None, "on the prompt");
+    assert_eq!(
+        sent(&typed, click(1, 1), false),
+        None,
+        "on the prompt's space"
+    );
+    let past = sent(&typed, click(1, 15), false);
+    assert_eq!(
+        past.as_deref(),
+        Some("\x1b[<0;16;2M"),
+        "past the end, the shell decides"
+    );
+    let right = term(
+        20,
+        &format!("{marked}ls\x1b7\x1b[1;17H\x1b]133;A\x07~/a\x1b]133;B\x07\x1b8"),
+    );
+    assert_eq!(
+        sent(&right, click(0, 18), false),
+        None,
+        "a right prompt after the cursor"
+    );
+    assert_eq!(
+        sent(&right, click(0, 3), false).as_deref(),
+        Some("\x1b[<0;4;1M")
+    );
     let two = term(
         20,
         "\x1b]133;A;click_events=2\x07~/src\r\n> \x1b]133;B\x07ls",

tasks/acceptance.md

@@ -71,7 +71,7 @@ What no check reached: no real KWin drew the blur; of Wayland desktops, only Hyp
 | OSC 8 and plain links on Cmd-press, Ctrl-press off macOS; inside herdr, herdr opens them whole | T09, T45 | `term::links` tests; on macOS with real herdr 0.9.5 and a stand-in `open` on its PATH, Cmd-press on a URL herdr wraps, on its wrapped tail, on an OSC 8 link and on a URL over three rows opened the whole URL and the OSC 8 target, where before the app opened `.../ELYTermin` and nothing; Cmd-hover underlined the long URL across its rows; outside herdr the app opened the URL itself; an OSC 8 link whose label holds `file:///etc/hosts` and whose target is a web page opened the page through herdr, which keeps the target from the terminal while it takes the mouse; Shift-Cmd-press on a plain `file:///etc/hosts` inside herdr the app opened itself, caught by an `NSWorkspace` stand-in | Pass |
 | Focus reporting, title, scrollback limit, find | T10 | `term::find` tests; round 3 failed at the cap, its fix stands on a red-green test | Pass |
 | A scrollbar over the history: a drag on its thumb scrolls, whatever the program does with the wheel | T58 | `term::scrollbar` test: the thumb's share and place follow the history and the view's offset, and a place past either end of the track holds at it; under Xorg's dummy screen, after `seq 1 500`, the thumb stood at the bottom right, a drag to the top showed the first prompt and lines 1 to 39, and one back down lines 220 to 259; with mouse tracking on, the wheel went to the program as reports and did not scroll, while a drag on the thumb showed lines 245 to 284 | Pass |
-| Click to move: a plain click on the line the shell reads, past its prompt, moves its cursor there, the press itself to a shell that takes clicks, else arrows | T59 | `term::marks` test: A's `click_events=1` and `2` read as a press at the screen's row or counted from the prompt; `term::click` tests: a shell that takes clicks hears an SGR left press at the clicked cell, a row below the prompt too, none before the prompt, in the history or on the alternate screen; B reads the line its prompt started; a click on the typed line counts the characters to it, left or right, past the end means the end, a wide character takes one arrow, soft-wrapped rows are one line, trailing spaces count up to the cursor, and a click on the prompt, another line, a line no prompt shows, output without marks or the alternate screen moves nothing; on macOS `shot` ran zsh with a marked prompt: after `echo hello` a click on the `h` logged `moves the cursor -5`, and `X` typed then ran `echo Xhello`; a drag sent nothing, and clicks on the prompt and on output logged `off the input`; round 1 of T59's review found fish 4.9 a code point per Left, so a count put `X` inside `é`, and its redraw dropped the wrap flag a count followed: with fish's own `click_events=1` the press goes to fish, which then typed `echo Xéabc`, and a second click on the redrawn wrapped line put `Y` at row 2, column 3; zsh still counted 4 arrows; a press let go in another cell and one whose program turned on tracking before the release sent nothing | Pass |
+| Click to move: a plain click on the line the shell reads, past its prompt, moves its cursor there, the press itself to a shell that takes clicks, else arrows | T59 | `term::marks` test: A's `click_events=1` and `2` read as a press at the screen's row or counted from the prompt; `term::click` tests: a shell that takes clicks hears an SGR left press at the clicked cell, a row below the prompt too, none before the prompt, in the history or on the alternate screen; B reads the line its prompt started; a click on the typed line counts the characters to it, left or right, past the end means the end, a wide character takes one arrow, soft-wrapped rows are one line, trailing spaces count up to the cursor, and a click on the prompt, another line, a line no prompt shows, output without marks or the alternate screen moves nothing; on macOS `shot` ran zsh with a marked prompt: after `echo hello` a click on the `h` logged `moves the cursor -5`, and `X` typed then ran `echo Xhello`; a drag sent nothing, and clicks on the prompt and on output logged `off the input`; round 1 of T59's review found fish 4.9 a code point per Left, so a count put `X` inside `é`, and its redraw dropped the wrap flag a count followed: with fish's own `click_events=1` the press goes to fish, which then typed `echo Xéabc`, and a second click on the redrawn wrapped line put `Y` at row 2, column 3; zsh still counted 4 arrows; a press let go in another cell and one whose program turned on tracking before the release sent nothing; round 2 found a click on fish's `$` sent to fish, which moved its cursor to the start: a press now goes only past the prompt's last cell before the cursor and never on a prompt cell, so clicks on `$` and its space logged `off the input` and `X` still ended `echo helloX` | Pass |
 | A bell makes no sound; the row marks it | T10, T12 | T12 review | Pass |
 
 ## 6. Main window and layout