Add button to check and install Meshroom dependency#580
Conversation
| MESHROOM_VERSION = "2025.1.0" | ||
| MESHROOM_WINDOWS_URL = "https://zenodo.org/records/16887472/files/Meshroom-2025.1.0-Windows.zip" | ||
| MESHROOM_LINUX_URL = "https://zenodo.org/records/16887472/files/Meshroom-2025.1.0-Linux.tar.gz" | ||
| MESHROOM_EXTRACTED_DIR_NAME = "Meshroom-2025.1.0" |
There was a problem hiding this comment.
The meshroom version is defined by the MESHROOM_VERSION variable and these download links. So if updating meshroom in the future, only these variables would need to be updated.
sadhana-r
force-pushed
the
542_meshroom_installer
branch
from
a4714e8 to
e5881e3
Compare
sadhana-r
force-pushed
the
542_meshroom_installer
branch
from
e5881e3 to
e3511f7
Compare
There was a problem hiding this comment.
Pull request overview
Adds first-class Meshroom dependency detection and installation flow to the Login module so users can verify/install Meshroom (required by openlifu photogrammetry tooling) from within the application.
Changes:
- Adds a Meshroom status indicator and “Install Meshroom 2025” button to the Login module’s Optional dependencies UI.
- Implements a background (QProcess) Meshroom download/extract workflow with progress + cancel, plus PATH persistence (Windows registry).
- Adds Login-module tests covering Meshroom install CLI extraction/cancel behavior and GUI controller behavior.
Reviewed changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 4 comments.
Show a summary per file
| File | Description |
|---|---|
| OpenLIFULogin/Resources/UI/OpenLIFULogin.ui | Adds Meshroom status icon + install button to the “Optional” dependencies section. |
| OpenLIFULogin/OpenLIFULogin.py | Wires UI to Meshroom status detection, platform-specific install prompts, and controller kickoff; adds Meshroom install tests. |
| OpenLIFULib/OpenLIFULib/meshroom_install_gui.py | New GUI-side controller to run the install helper via QProcess with progress dialog and cancel handling. |
| OpenLIFULib/OpenLIFULib/meshroom_install_cli.py | New CLI helper to download and securely extract Meshroom archives while emitting progress events. |
| OpenLIFULib/CMakeLists.txt | Ensures the new Meshroom installer scripts are packaged with the extension. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| destination = Path(args.destination) | ||
| work_dir = Path(args.work_dir) | ||
| url = args.archive_url |
| self._meshroom_install_controller = None | ||
| self._checkMeshroomStatus() | ||
|
|
|
To programtically enable GPU acceleration, it seems like the following registry key: |
| self.ui.installMeshroomPushButton.setText(f"Meshroom {version_str}") | ||
| else: | ||
| self.ui.installMeshroomPushButton.setEnabled(True) | ||
| self.ui.installMeshroomPushButton.setText("Install Meshroom 2025") |
There was a problem hiding this comment.
The button text would also need to be updated if updating meshroom.
Addresses #542
Adds a button to the login module's "Optional" dependencies section for checking and installing Meshroom 2025.1.0.
Detects Meshroom by looking for meshroom_batch on PATH (the executable required by openlifu). Since meshroom_batch has no --version flag, the version is inferred from the parent directory name (e.g., Meshroom-2025.1.0).
The Meshroom archive is large (~10 GB on Windows, ~13 GB on Linux), so the download runs in a background QProcess with a progress dialog and cancel support - the same approach used for the sample database download. After extraction, the Meshroom directory is added to PATH using the same registry-based approach used for Android Platform Tools. On Windows, a post-install dialog provides step-by-step instructions for enabling NVIDIA GPU acceleration, based on the instructions provided here: https://github.com/OpenwaterHealth/OpenLIFU-python?tab=readme-ov-file#enable-gpu-acceleration
I also added some unit tests to the Login module.
For Review
Test it out, and review the overall design. The specifics of the progress bar/download code is mostly borrowed from the sample data functionality