diff options
| author | Craig Jennings <c@cjennings.net> | 2026-05-12 12:53:11 -0500 |
|---|---|---|
| committer | Craig Jennings <c@cjennings.net> | 2026-05-12 12:53:11 -0500 |
| commit | c72d4abc17ae7bed792fa610c0a67b917e191f4b (patch) | |
| tree | 313988cafb0755a8116cebc499a3f60cbbee0d15 /tests/test-ai-vterm--f9-in-vterm.el | |
| parent | a15c6b3e92b8d7be6be6dc4dd6802a18ccf52326 (diff) | |
| download | dotemacs-c72d4abc17ae7bed792fa610c0a67b917e191f4b.tar.gz dotemacs-c72d4abc17ae7bed792fa610c0a67b917e191f4b.zip | |
fix(ai-vterm): make F9 toggle the agent from inside an agent buffer
vterm binds `<f1>`..`<f12>` to `vterm--self-insert`, so a plain `<f9>` typed while point is in an agent buffer goes to the terminal program instead of the global toggle. That's invisible most of the time — you press F9 from another window — but it bites when the agent buffer is the only window in the frame, because there's nowhere else to press it from.
I re-bound the F9 family in `vterm-mode-map` (via `with-eval-after-load 'vterm`) so that `<f9>`, `C-<f9>`, and `M-<f9>` reach `cj/ai-vterm`, `cj/ai-vterm-pick-project`, and `cj/ai-vterm-pick-buffer` from there too. The C-/M- variants aren't actually in vterm's intercept set, but binding them keeps things uniform. New `tests/test-ai-vterm--f9-in-vterm.el`: 4 ERT tests over the `vterm-mode-map` and global bindings. F12's `cj/vterm-toggle` has the same shape of bug and isn't touched here.
Diffstat (limited to 'tests/test-ai-vterm--f9-in-vterm.el')
| -rw-r--r-- | tests/test-ai-vterm--f9-in-vterm.el | 42 |
1 files changed, 42 insertions, 0 deletions
diff --git a/tests/test-ai-vterm--f9-in-vterm.el b/tests/test-ai-vterm--f9-in-vterm.el new file mode 100644 index 00000000..1355bd6a --- /dev/null +++ b/tests/test-ai-vterm--f9-in-vterm.el @@ -0,0 +1,42 @@ +;;; test-ai-vterm--f9-in-vterm.el --- F9 reaches Emacs from inside an agent buffer -*- lexical-binding: t; -*- + +;;; Commentary: +;; vterm binds <f1>..<f12> to `vterm--self-insert', so a plain <f9> typed +;; while point is in an agent buffer is sent to the terminal program instead +;; of toggling the agent -- which is exactly the case when the agent buffer +;; fills the frame. `ai-vterm.el' re-binds the F9 family in `vterm-mode-map'. +;; These tests load real vterm so `vterm-mode-map' exists, then confirm the +;; bindings landed (and the global ones are still there). + +;;; Code: + +(require 'ert) +(require 'package) + +(setq package-user-dir (expand-file-name "elpa" user-emacs-directory)) +(package-initialize) +(add-to-list 'load-path (expand-file-name "modules" user-emacs-directory)) +(require 'vterm) +(require 'ai-vterm) + +(ert-deftest test-ai-vterm-f9-bound-in-vterm-mode-map () + "Normal: <f9> in `vterm-mode-map' runs the agent toggle, not `vterm--self-insert'." + (should (eq (keymap-lookup vterm-mode-map "<f9>") #'cj/ai-vterm))) + +(ert-deftest test-ai-vterm-f9-family-bound-in-vterm-mode-map () + "Normal: the C-/M- F9 variants are bound in `vterm-mode-map' too." + (should (eq (keymap-lookup vterm-mode-map "C-<f9>") #'cj/ai-vterm-pick-project)) + (should (eq (keymap-lookup vterm-mode-map "M-<f9>") #'cj/ai-vterm-pick-buffer))) + +(ert-deftest test-ai-vterm-f9-not-self-insert-in-vterm () + "Boundary: vterm's default <f9> -> `vterm--self-insert' was overridden." + (should-not (eq (keymap-lookup vterm-mode-map "<f9>") 'vterm--self-insert))) + +(ert-deftest test-ai-vterm-f9-still-bound-globally () + "Normal: the global F9 family bindings are intact." + (should (eq (lookup-key (current-global-map) (kbd "<f9>")) #'cj/ai-vterm)) + (should (eq (lookup-key (current-global-map) (kbd "C-<f9>")) #'cj/ai-vterm-pick-project)) + (should (eq (lookup-key (current-global-map) (kbd "M-<f9>")) #'cj/ai-vterm-pick-buffer))) + +(provide 'test-ai-vterm--f9-in-vterm) +;;; test-ai-vterm--f9-in-vterm.el ends here |
