learningfoundry 0.83.2__tar.gz → 0.83.3__tar.gz

{learningfoundry-0.83.2 → learningfoundry-0.83.3}/CHANGELOG.md

@@ -7,6 +7,15 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [0.83.3] - 2026-06-18
+
+**`learningfoundry launch`/`stop` fixes from first real-world use** (Story K.l).
+
+### Fixed
+
+- **`launch`/`stop` auto-detect the build in `dist/`.** The learner runs these from the curriculum project root, where `build` writes the app into `dist/` — so the manifest is at `dist/exercises-manifest.json` and the bare command previously failed (`No exercises-manifest.json found in .`), forcing `--dir dist` every time. The `--dir` value now resolves to `<dir>/exercises-manifest.json`, else `<dir>/dist/exercises-manifest.json`; an explicit `--dir` still overrides.
+- **`stop` no longer leaves marimo children hanging.** `terminate_pid` signalled only the parent marimo pid, orphaning its multiprocessing children — which printed marimo's goodbye banner + `resource_tracker` leaked-semaphore + `parent_poller` warnings *after* the shell prompt returned. It now `SIGTERM`s the whole process group (`os.killpg(os.getpgid(pid), …)`; marimo is its own group leader via `start_new_session=True`) so children shut down cleanly; Windows uses `taskkill /T` (tree-kill).
+
 ## [0.83.2] - 2026-06-18
 
 **Single-source the package version** (Story K.k) — build hygiene.

{learningfoundry-0.83.2 → learningfoundry-0.83.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: learningfoundry
-Version: 0.83.2
+Version: 0.83.3
 Summary: A curriculum engine that turns a YAML curriculum definition into a deployable SvelteKit learning application.
 Project-URL: Homepage, https://github.com/pointmatic/learningfoundry
 Project-URL: Repository, https://github.com/pointmatic/learningfoundry
@@ -213,15 +213,16 @@ This serves the SvelteKit project from source via Vite's dev server; it does **n
 
 ### `learningfoundry launch` / `learningfoundry stop`
 
-Run a `ready` exercise's marimo notebook locally. These are **learner-side** commands — run from inside the generated app (where `build` writes `exercises-manifest.json`), not from the author's repo. The generated app is static and a browser page can't spawn a process, so the exercise's banner asks the learner to copy and run `learningfoundry launch <id>`.
+Run a `ready` exercise's marimo notebook locally. These are **learner-side** commands — run them from the curriculum project root (they auto-detect the build in `dist/`) or from inside the generated app itself. The generated app is static and a browser page can't spawn a process, so the exercise's banner asks the learner to copy and run `learningfoundry launch <id>`.
 
 ```
 Usage: learningfoundry launch [OPTIONS] EXERCISE_ID
        learningfoundry stop [OPTIONS] [EXERCISE_ID]
 
 Options:
-  --dir PATH              Directory holding exercises-manifest.json (the
-                          generated app's root).  [default: .]
+  --dir PATH              Where to find exercises-manifest.json. Auto-detects
+                          `<dir>/exercises-manifest.json`, else
+                          `<dir>/dist/exercises-manifest.json`.  [default: .]
   --log-level LEVEL       Logging verbosity.  [default: INFO]
   --help                  Show this message and exit.
 ```

{learningfoundry-0.83.2 → learningfoundry-0.83.3}/README.md

@@ -182,15 +182,16 @@ This serves the SvelteKit project from source via Vite's dev server; it does **n
 
 ### `learningfoundry launch` / `learningfoundry stop`
 
-Run a `ready` exercise's marimo notebook locally. These are **learner-side** commands — run from inside the generated app (where `build` writes `exercises-manifest.json`), not from the author's repo. The generated app is static and a browser page can't spawn a process, so the exercise's banner asks the learner to copy and run `learningfoundry launch <id>`.
+Run a `ready` exercise's marimo notebook locally. These are **learner-side** commands — run them from the curriculum project root (they auto-detect the build in `dist/`) or from inside the generated app itself. The generated app is static and a browser page can't spawn a process, so the exercise's banner asks the learner to copy and run `learningfoundry launch <id>`.
 
 ```
 Usage: learningfoundry launch [OPTIONS] EXERCISE_ID
        learningfoundry stop [OPTIONS] [EXERCISE_ID]
 
 Options:
-  --dir PATH              Directory holding exercises-manifest.json (the
-                          generated app's root).  [default: .]
+  --dir PATH              Where to find exercises-manifest.json. Auto-detects
+                          `<dir>/exercises-manifest.json`, else
+                          `<dir>/dist/exercises-manifest.json`.  [default: .]
   --log-level LEVEL       Logging verbosity.  [default: INFO]
   --help                  Show this message and exit.
 ```

{learningfoundry-0.83.2 → learningfoundry-0.83.3}/src/learningfoundry/__init__.py

@@ -1,4 +1,4 @@
 # Copyright 2026 Pointmatic
 # SPDX-License-Identifier: Apache-2.0
 
-__version__ = "0.83.2"
+__version__ = "0.83.3"

{learningfoundry-0.83.2 → learningfoundry-0.83.3}/src/learningfoundry/cli.py

@@ -286,8 +286,9 @@ _launch_dir_option = click.option(
     default=Path("."),
     show_default=True,
     help=(
-        "Directory holding `exercises-manifest.json` (the generated app's "
-        "root). Defaults to the current directory."
+        "Where to find `exercises-manifest.json`. Auto-detects "
+        "`<dir>/exercises-manifest.json`, else `<dir>/dist/...`. "
+        "Defaults to the current directory."
     ),
 )
 
@@ -302,6 +303,8 @@ def launch(exercise_id: str, launch_dir: Path, log_level: str) -> None:
 
     from learningfoundry import launch as _launch
 
+    launch_dir = _launch.resolve_manifest_dir(launch_dir)
+
     try:
         spec = _launch.resolve_launch_spec(launch_dir, exercise_id)
     except LaunchError as exc:
@@ -362,6 +365,8 @@ def stop(exercise_id: str | None, launch_dir: Path, log_level: str) -> None:
 
     from learningfoundry import launch as _launch
 
+    launch_dir = _launch.resolve_manifest_dir(launch_dir)
+
     if exercise_id is not None:
         try:
             spec = _launch.resolve_launch_spec(launch_dir, exercise_id)

{learningfoundry-0.83.2 → learningfoundry-0.83.3}/src/learningfoundry/launch.py

@@ -58,6 +58,24 @@ class LaunchSpec:
     port: int
 
 
+def resolve_manifest_dir(start: Path) -> Path:
+    """Locate the directory that holds ``exercises-manifest.json``.
+
+    The learner typically runs ``launch``/``stop`` from the curriculum project
+    root, where ``build`` writes the app into ``dist/`` — so the manifest is at
+    ``dist/exercises-manifest.json``, not in the cwd. Prefer ``start`` itself
+    (running from inside the app), then fall back to ``start/dist``. If neither
+    holds the manifest, return ``start`` unchanged so the downstream
+    :class:`ManifestNotFoundError` names the directory the user actually gave.
+    """
+    if (start / MANIFEST_FILENAME).is_file():
+        return start
+    nested = start / "dist"
+    if (nested / MANIFEST_FILENAME).is_file():
+        return nested
+    return start
+
+
 def resolve_launch_spec(manifest_dir: Path, exercise_id: str) -> LaunchSpec:
     """Resolve an exercise id to its :class:`LaunchSpec` from the manifest.
 
@@ -295,20 +313,25 @@ def spawn_detached(argv: list[str], cwd: Path) -> int:
 
 
 def terminate_pid(pid: int) -> None:
-    """Ask the process ``pid`` to terminate (idempotent if already gone).
-
-    POSIX sends ``SIGTERM`` (graceful); Windows shells out to ``taskkill``.
+    """Ask the launch-owned marimo at ``pid`` to terminate (idempotent).
+
+    Signals the whole **process group**, not just ``pid``: ``launch`` spawns
+    marimo with ``start_new_session=True``, so marimo is its own group leader
+    and its multiprocessing children share the group. Killing only the parent
+    orphans those children, which then print marimo's goodbye banner and
+    ``resource_tracker`` leaked-semaphore warnings *after* the shell returns.
+    POSIX `SIGTERM`s the group (graceful); Windows uses ``taskkill /T`` (tree).
     A process that has already exited is treated as success.
     """
     if sys.platform == "win32":  # pragma: no cover - Windows only
         subprocess.run(
-            ["taskkill", "/F", "/PID", str(pid)],
+            ["taskkill", "/F", "/T", "/PID", str(pid)],
             check=False,
             capture_output=True,
         )
         return
     try:
-        os.kill(pid, signal.SIGTERM)
+        os.killpg(os.getpgid(pid), signal.SIGTERM)
     except ProcessLookupError:
         pass  # already gone — nothing to do