SDK DataLoaders 2: barebones Python support (#5328)

Exposes raw `DataLoader`s to the Python SDK through 2 new methods:
`RecordingStream.log_file_from_path` &
`RecordingStream.log_file_from_contents`.

Everything streams asynchronously, as expected.

There's no way to configure the behavior of the `DataLoader` at all,
yet. That comes next.

```python
rr.log_file_from_path(filepath)
```

![image](https://github.com/rerun-io/rerun/assets/2910679/0fe2d39c-f069-44a6-b836-e31001b3adaa)


---

Part of series of PR to expose configurable `DataLoader`s to our SDKs:
- #5327 
- #5328 
- #5330
- #5337
- #5351
- #5355

examples/python/log_file/README.md

@@ -0,0 +1,11 @@
+<!--[metadata]
+title = "Log file example"
+-->
+
+Demonstrates how to log any file from the SDK using the [`DataLoader`](https://www.rerun.io/docs/howto/open-any-file) machinery.
+
+Usage:
+```bash
+python examples/python/log_file/main.py examples/assets
+```
+

examples/python/log_file/main.py

@@ -0,0 +1,46 @@
+#!/usr/bin/env python3
+"""
+Demonstrates how to log any file from the SDK using the `DataLoader` machinery.
+
+See <https://www.rerun.io/docs/howto/open-any-file> for more information.
+
+Usage:
+```
+python examples/python/log_file/main.py -- examples/assets
+```
+"""
+from __future__ import annotations
+
+import argparse
+from pathlib import Path
+
+import rerun as rr  # pip install rerun-sdk
+
+parser = argparse.ArgumentParser(
+    description="Demonstrates how to log any file from the SDK using the `DataLoader` machinery."
+)
+rr.script_add_args(parser)
+parser.add_argument(
+    "--from-contents",
+    action="store_true",
+    default=False,
+    help="Log the contents of the file directly (files only -- not supported by external loaders).",
+)
+parser.add_argument("filepaths", nargs="+", type=Path, help="The filepaths to be loaded and logged.")
+args = parser.parse_args()
+
+rr.script_setup(args, "rerun_example_log_file")
+
+for filepath in args.filepaths:
+    if not args.from_contents:
+        # Either log the file using its path…
+        rr.log_file_from_path(filepath)
+    else:
+        # …or using its contents if you already have them loaded for some reason.
+        try:
+            with open(filepath, "rb") as file:
+                rr.log_file_from_contents(filepath, file.read())
+        except Exception:
+            pass
+
+rr.script_teardown(args)

examples/python/log_file/requirements.txt

@@ -0,0 +1 @@
+rerun-sdk

rerun_py/Cargo.toml

@@ -41,7 +41,13 @@ re_error.workspace = true
 re_log.workspace = true
 re_log_types.workspace = true
 re_memory.workspace = true
-rerun = { workspace = true, features = ["analytics", "run", "server", "sdk"] }
+rerun = { workspace = true, features = [
+  "analytics",
+  "data_loaders",
+  "run",
+  "server",
+  "sdk",
+] }
 re_web_viewer_server = { workspace = true, optional = true }
 re_ws_comms = { workspace = true, optional = true }
 

rerun_py/docs/gen_common_index.py

@@ -94,6 +94,8 @@ class Section:
         title="Logging functions",
         func_list=[
             "log",
+            "log_file_from_path",
+            "log_file_from_contents",
         ],
     ),
     Section(

rerun_py/rerun_sdk/rerun/__init__.py

@@ -77,8 +77,10 @@
     "get_recording_id",
     "get_thread_local_data_recording",
     "is_enabled",
-    "log_components",
     "log",
+    "log_components",
+    "log_file_from_contents",
+    "log_file_from_path",
     "memory_recording",
     "new_entity_path",
     "reset_time",
@@ -105,6 +107,8 @@
     escape_entity_path_part,
     log,
     log_components,
+    log_file_from_contents,
+    log_file_from_path,
     new_entity_path,
 )
 from .any_value import AnyValues

rerun_py/rerun_sdk/rerun/_log.py

@@ -1,5 +1,6 @@
 from __future__ import annotations
 
+from pathlib import Path
 from typing import Any, Iterable
 
 import pyarrow as pa
@@ -282,6 +283,72 @@ def log_components(
     )
 
 
+@catch_and_log_exceptions()
+def log_file_from_path(
+    file_path: str | Path,
+    *,
+    recording: RecordingStream | None = None,
+) -> None:
+    r"""
+    Logs the file at the given `path` using all `DataLoader`s available.
+
+    A single `path` might be handled by more than one loader.
+
+    This method blocks until either at least one `DataLoader` starts
+    streaming data in or all of them fail.
+
+    See <https://www.rerun.io/docs/howto/open-any-file> for more information.
+
+    Parameters
+    ----------
+    file_path:
+        Path to the file to be logged.
+
+    recording:
+        Specifies the [`rerun.RecordingStream`][] to use. If left unspecified,
+        defaults to the current active data recording, if there is one. See
+        also: [`rerun.init`][], [`rerun.set_global_data_recording`][].
+
+    """
+
+    bindings.log_file_from_path(Path(file_path), recording=recording)
+
+
+@catch_and_log_exceptions()
+def log_file_from_contents(
+    file_path: str | Path,
+    file_contents: bytes,
+    *,
+    recording: RecordingStream | None = None,
+) -> None:
+    r"""
+    Logs the given `file_contents` using all `DataLoader`s available.
+
+    A single `path` might be handled by more than one loader.
+
+    This method blocks until either at least one `DataLoader` starts
+    streaming data in or all of them fail.
+
+    See <https://www.rerun.io/docs/howto/open-any-file> for more information.
+
+    Parameters
+    ----------
+    file_path:
+        Path to the file that the `file_contents` belong to.
+
+    file_contents:
+        Contents to be logged.
+
+    recording:
+        Specifies the [`rerun.RecordingStream`][] to use. If left unspecified,
+        defaults to the current active data recording, if there is one. See
+        also: [`rerun.init`][], [`rerun.set_global_data_recording`][].
+
+    """
+
+    bindings.log_file_from_contents(Path(file_path), file_contents, recording=recording)
+
+
 def escape_entity_path_part(part: str) -> str:
     r"""
     Escape an individual part of an entity path.

rerun_py/src/python_bridge.rs

@@ -180,6 +180,8 @@ fn rerun_bindings(_py: Python<'_>, m: &PyModule) -> PyResult<()> {
 
     // log any
     m.add_function(wrap_pyfunction!(log_arrow_msg, m)?)?;
+    m.add_function(wrap_pyfunction!(log_file_from_path, m)?)?;
+    m.add_function(wrap_pyfunction!(log_file_from_contents, m)?)?;
 
     // misc
     m.add_function(wrap_pyfunction!(version, m)?)?;
@@ -962,6 +964,54 @@ fn log_arrow_msg(
     Ok(())
 }
 
+#[pyfunction]
+#[pyo3(signature = (
+    file_path,
+    recording=None,
+))]
+fn log_file_from_path(
+    py: Python<'_>,
+    file_path: std::path::PathBuf,
+    recording: Option<&PyRecordingStream>,
+) -> PyResult<()> {
+    let Some(recording) = get_data_recording(recording) else {
+        return Ok(());
+    };
+
+    recording
+        .log_file_from_path(file_path)
+        .map_err(|err| PyRuntimeError::new_err(err.to_string()))?;
+
+    py.allow_threads(flush_garbage_queue);
+
+    Ok(())
+}
+
+#[pyfunction]
+#[pyo3(signature = (
+    file_path,
+    file_contents,
+    recording=None,
+))]
+fn log_file_from_contents(
+    py: Python<'_>,
+    file_path: std::path::PathBuf,
+    file_contents: &[u8],
+    recording: Option<&PyRecordingStream>,
+) -> PyResult<()> {
+    let Some(recording) = get_data_recording(recording) else {
+        return Ok(());
+    };
+
+    recording
+        .log_file_from_contents(file_path, std::borrow::Cow::Borrowed(file_contents))
+        .map_err(|err| PyRuntimeError::new_err(err.to_string()))?;
+
+    py.allow_threads(flush_garbage_queue);
+
+    Ok(())
+}
+
 // --- Misc ---
 
 /// Return a verbose version string