Skip to content

plaid.viewer.cli

plaid.viewer.cli

Command-line entry point for the dataset viewer.

Starts a single self-contained trame server. There is no FastAPI backend and no separate port: dataset discovery, sample loading, CGNS export and the 3D view are all served by the same trame process.

plaid.viewer.cli.main 

main(argv=None)

Run the viewer until interrupted.

Parameters:

  • argv (list[str] | None, default: None ) –

    Optional override of sys.argv[1:] for tests.

Returns:

  • int –

    Process exit code.

Source code in plaid/viewer/cli.py 
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
def main(argv: list[str] | None = None) -> int:
    """Run the viewer until interrupted.

    Args:
        argv: Optional override of ``sys.argv[1:]`` for tests.

    Returns:
        Process exit code.
    """
    args = _build_parser().parse_args(argv)
    logging.basicConfig(
        level=logging.INFO, format="%(asctime)s %(levelname)s %(name)s: %(message)s"
    )

    # Permanently silence the process's file-descriptor 2 so the HDF5 /
    # CGNS C libraries (used by both VTK's ``vtkCGNSReader`` and PLAID's
    # pyCGNS loader) cannot pollute the console with messages like
    # ``Mismatch in number of children and child IDs read``. Python's
    # ``sys.stderr`` is preserved so tracebacks and the logger keep
    # working. See ``_reroute_c_stderr`` for the details.
    from plaid.viewer.trame_app.server import (  # noqa: PLC0415
        _reroute_c_stderr,
    )

    _reroute_c_stderr()

    # When no explicit ``--datasets-root`` is passed, fall back to the
    # last local root the user selected in a previous session (persisted
    # under ``$XDG_CONFIG_HOME/plaid/viewer.json``). This makes the
    # viewer "remember" the last dataset directory without requiring the
    # CLI flag on every launch.
    effective_datasets_root = args.datasets_root
    if effective_datasets_root is None:
        effective_datasets_root = get_last_datasets_root()
        if effective_datasets_root is not None:
            logger.info("Using persisted datasets root: %s", effective_datasets_root)
    browse_roots = tuple(args.browse_roots) if args.browse_roots else ()
    config = ViewerConfig(
        datasets_root=effective_datasets_root,
        browse_roots=browse_roots,
        allow_root_change=not args.disable_root_change,
        initial_dataset_id=args.dataset_id,
        allow_dataset_change=not args.disable_dataset_change,
    )

    with CacheRoot() as cache:
        dataset_service = PlaidDatasetService(config)
        for repo_id in args.hub_repo or []:
            try:
                dataset_service.add_hub_dataset(repo_id)
            except ValueError as exc:
                logger.warning("Ignoring --hub-repo %r: %s", repo_id, exc)
        artifact_service = ParaviewArtifactService(dataset_service, cache.path)

        # Deferred import so ``--help`` works without trame installed.
        from plaid.viewer.trame_app.server import build_server  # noqa: PLC0415

        server = build_server(dataset_service, artifact_service)
        server.start(host=args.host, port=args.port, open_browser=False)
    return 0