Add client side support for the DocumentAsset type (#736)

* Add client side support for the DocumentAsset type * Allow for .md and .log files as well
kolenaIO · Jan 7, 2025 · a623c29 · a623c29
1 parent 3c8fb7e
commit a623c29
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 2 deletions.
diff --git a/docs/dataset/core-concepts/index.md b/docs/dataset/core-concepts/index.md
@@ -67,7 +67,7 @@ as text. Below table outlines what extensions are supported for optimal visualiz
 | Image          | `jpg`, `jpeg`, `png`, `gif`, `bmp` and other web browser supported image types.       |
 | Audio          | `flac`, `mp3`, `wav`, `acc`, `ogg`, `ra` and other web browser supported audio types. |
 | Video          | `mov`, `mp4`, `mpeg` and other web browser supported video types.                     |
-| Document       | `txt` and `pdf` files.                                                                |
+| Document       | `txt`, `pdf`, `log`, `md` files.                                                      |
 | Point Cloud    | `pcd` files.                                                                          |
 
 **Assets**: allow you to connect multiple referenced files to each datapoint for visualization and analysis.
@@ -81,6 +81,7 @@ Multiple assets can be attached to a single datapoint.
 | [`VideoAsset`](../../reference/asset.md#kolena.asset.VideoAsset)           | Useful if you want to attach a video file.                     | Same as above reference files |
 | [`PointCloudAsset`](../../reference/asset.md#kolena.asset.PointCloudAsset) | Useful for attaching 3D point cloud data.                      | `.pcd`, `.npy`, `.npz`        |
 | [`MeshAsset`](../../reference/asset.md#kolena.asset.MeshAsset)             | Useful for attaching and visualizing 3D mesh files.            | `.ply`                        |
+| [`DocumentAsset`](../../reference/asset.md#kolena.asset.DocumentAsset)     | Useful if you want to attach a document file.                  | `.pdf`, `.txt`, `.log`, `.md` |
 
 **Annotations**: allow you to visualize overlays on top of datapoints through the use of[`annotation`](../../reference/annotation.md).
 We currently support 10 different types of annotations each enabling a specific modality.

diff --git a/kolena/asset.py b/kolena/asset.py
@@ -24,6 +24,7 @@
 - [`VideoAsset`][kolena.asset.VideoAsset]
 - [`AudioAsset`][kolena.asset.AudioAsset]
 - [`MeshAsset`][kolena.asset.MeshAsset]
+- [`DocumentAsset`][kolena.asset.DocumentAsset]
 
 """
 from abc import ABCMeta
@@ -44,6 +45,7 @@ class _AssetType(DataType):
     VIDEO = "VIDEO"
     AUDIO = "AUDIO"
     MESH = "MESH"
+    DOCUMENT = "DOCUMENT"
 
     @staticmethod
     def _data_category() -> DataCategory:
@@ -188,6 +190,22 @@ def _data_type() -> _AssetType:
         return _AssetType.MESH
 
 
+@dataclass(frozen=True, config=ValidatorConfig)
+class DocumentAsset(Asset):
+    """
+    A document file in a cloud bucket or served at a URL.
+
+    Supported extensions include `.pdf`, `.txt`, `.log`, and `.md`.
+    """
+
+    locator: str
+    """The location of this document file in a cloud bucket, e.g. `s3://my-bucket/path/to/my-document-asset.pdf`."""
+
+    @staticmethod
+    def _data_type() -> _AssetType:
+        return _AssetType.DOCUMENT
+
+
 _ASSET_TYPES = [
     ImageAsset,
     PlainTextAsset,
@@ -197,4 +215,5 @@ def _data_type() -> _AssetType:
     VideoAsset,
     AudioAsset,
     MeshAsset,
+    DocumentAsset,
 ]
diff --git a/pyproject.toml b/pyproject.toml
@@ -39,7 +39,7 @@ dependencies = [
     "numpy >=1.26,<2; python_version >= '3.12'",
     "pandas >=1.1,<3; python_version < '3.12'",
     "pandas >=2.1.1,<3; python_version >= '3.12'",
-    "pandera>=0.9,<1",
+    "pandera>=0.22.1,<1",
     "pydantic>=1.10,<3",
     "dacite>=1.6,<2",
     "requests>=2.20,<3",