datalab-api 0.2.1__tar.gz → 0.2.3__tar.gz

{datalab_api-0.2.1/src/datalab_api.egg-info → datalab_api-0.2.3}/PKG-INFO

@@ -1,8 +1,13 @@
 Metadata-Version: 2.1
 Name: datalab-api
-Version: 0.2.1
+Version: 0.2.3
+Summary: A Python API for instances of the datalab data management platform (https://github.com/the-grey-group/datalab).
 Author-email: Matthew Evans <matthew@ml-evs.science>
 License: MIT
+Project-URL: homepage, https://github.com/datalab-org/datalab-api
+Project-URL: repository, https://github.com/datalab-org/datalab-api
+Project-URL: documentation, https://datalab-python-api.readthedocs.io
+Project-URL: changelog, https://github.com/datalab-org/datalab-api/releases
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3
@@ -21,6 +26,9 @@ Requires-Dist: bokeh~=2.4
 Requires-Dist: rich~=13.0
 Provides-Extra: dev
 Requires-Dist: pre-commit~=3.0; extra == "dev"
+Requires-Dist: pytest~=8.0; extra == "dev"
+Requires-Dist: pytest-cov~=5.0; extra == "dev"
+Requires-Dist: respx~=0.21; extra == "dev"
 Provides-Extra: docs
 Requires-Dist: mkdocs; extra == "docs"
 Requires-Dist: mkdocs-material; extra == "docs"
@@ -30,6 +38,8 @@ Requires-Dist: mkdocs-jupyter; extra == "docs"
 Provides-Extra: cli
 Requires-Dist: typer~=0.9; extra == "cli"
 Requires-Dist: click-shell~=2.1; extra == "cli"
+Provides-Extra: all
+Requires-Dist: datalab-api[cli,dev,docs]; extra == "all"
 
 # <div align="center"><i>datalab</i> Python API</div>
 
@@ -95,7 +105,7 @@ This package implements basic functionality for displaying and manipulating entr
 ```python
 from datalab_api import DatalabClient
 
-with DatalabClient("https://api.public.odbx.science") as client:
+with DatalabClient("https://public.api.odbx.science") as client:
 
     # List all items of a given type
     items = client.get_items()

{datalab_api-0.2.1 → datalab_api-0.2.3}/README.md

@@ -62,7 +62,7 @@ This package implements basic functionality for displaying and manipulating entr
 ```python
 from datalab_api import DatalabClient
 
-with DatalabClient("https://api.public.odbx.science") as client:
+with DatalabClient("https://public.api.odbx.science") as client:
 
     # List all items of a given type
     items = client.get_items()

{datalab_api-0.2.1 → datalab_api-0.2.3}/pyproject.toml

@@ -5,6 +5,7 @@ build-backend = "setuptools.build_meta"
 [project]
 name = "datalab-api"
 readme = "README.md"
+description = "A Python API for instances of the datalab data management platform (https://github.com/the-grey-group/datalab)."
 keywords = []
 license = { text = "MIT" }
 authors = [{ name = "Matthew Evans", email = "matthew@ml-evs.science" }]
@@ -31,6 +32,9 @@ dependencies =[
 [project.optional-dependencies]
 dev = [
     "pre-commit ~= 3.0",
+    "pytest ~= 8.0",
+    "pytest-cov ~= 5.0",
+    "respx ~= 0.21",
 ]
 
 docs = [
@@ -46,14 +50,18 @@ cli = [
     "click-shell ~= 2.1", # REPL-like interface
 ]
 
+all = [
+    "datalab-api[dev,docs,cli]"
+]
+
 [project.scripts]
 datalab = "datalab_api.cli:app"
 
 [project.urls]
-# homepage = "https://github.com/the-grey-group/datalab"
-# repository = "https://github.com/the-grey-group/datalab"
-# documentation = "https://the-datalab.readthedocs.io"
-# changelog = "https://github.com/the-grey-group/datalab/releases"
+homepage = "https://github.com/datalab-org/datalab-api"
+repository = "https://github.com/datalab-org/datalab-api"
+documentation = "https://datalab-python-api.readthedocs.io"
+changelog = "https://github.com/datalab-org/datalab-api/releases"
 
 [tool.versioningit.vcs]
 method = "git"

{datalab_api-0.2.1 → datalab_api-0.2.3}/src/datalab_api/__init__.py

@@ -1,3 +1,4 @@
+import warnings
 from pathlib import Path
 from typing import Any, Optional, Union
 
@@ -214,6 +215,26 @@ class DatalabClient(BaseDatalabClient):
 
         return item["item_data"]
 
+    def get_item_files(self, item_id: str) -> None:
+        """Download all the files for a given item and save them locally
+        in the current working directory.
+
+        Parameters:
+            item_id: The ID of the item to search for.
+
+        """
+
+        item_data = self.get_item(item_id)
+        for f in item_data.get("files", []):
+            file_location = f["location"]
+            url = file_location.replace("/app", self.datalab_api_url)
+            if Path(f["name"]).exists():
+                warnings.warn(f"Will not overwrite existing file {f['name']}")
+                continue
+            with open(f["name"], "wb") as file:
+                response = self.session.get(url, follow_redirects=True)
+                file.write(response.content)
+
     def get_block(self, item_id: str, block_id: str, block_data: dict[str, Any]) -> dict[str, Any]:
         """Get a block with a given ID and block data.
         Should be used in conjunction with `get_item` to load an existing block.

{datalab_api-0.2.1 → datalab_api-0.2.3}/src/datalab_api/_base.py

@@ -1,12 +1,16 @@
 import functools
 import logging
 import os
+import re
+import warnings
 from importlib.metadata import version
 from typing import Any, Optional
 
 import httpx
+from rich.console import Console
 from rich.logging import RichHandler
 from rich.pretty import pprint
+from rich.table import Table
 
 __version__ = version("datalab-api")
 
@@ -14,17 +18,46 @@ __all__ = ("__version__", "BaseDatalabClient")
 
 
 def pretty_displayer(method):
+    """A decorator which wraps a method with a 'display' kwarg, which will
+    either pretty print a JSON response, or display a Rich table.
+
+    """
+
     @functools.wraps(method)
     def rich_wrapper(self, *args, **kwargs):
         display = kwargs.pop("display", False)
+        page_limit = kwargs.pop("page_limit", 10)
         result = method(self, *args, **kwargs)
         if display:
-            if isinstance(result, dict) and "blocks_obj" in result:
-                blocks: dict[str, dict] = result["blocks_obj"]
-                for block in blocks.values():
-                    if "bokeh_plot_data" in block:
-                        bokeh_from_json(block)
-            pprint(result, max_length=None, max_string=100, max_depth=3)
+            blocks = None
+            if isinstance(result, dict):
+                if "blocks_obj" in result:
+                    blocks = result["blocks_obj"]
+                elif "blocktype" in result:
+                    blocks = {result["block_id"]: result}
+                    result = None
+                if blocks:
+                    for block in blocks.values():
+                        if "bokeh_plot_data" in block:
+                            bokeh_from_json(block)
+                if result:
+                    pprint(result, max_length=None, max_string=100, max_depth=3)
+            elif isinstance(result, list):
+                table = Table(show_lines=True)
+                table.add_column("type", overflow="crop", justify="center")
+                table.add_column("ID", style="cyan", no_wrap=True)
+                table.add_column("refcode", style="magenta", no_wrap=True)
+                table.add_column("name", width=30, overflow="ellipsis", no_wrap=True)
+                for item in result[:page_limit]:
+                    table.add_row(
+                        item["type"][0].upper(),
+                        item["item_id"],
+                        item["refcode"],
+                        item["name"],
+                        end_section=True,
+                    )
+                console = Console()
+                console.print(table)
 
         return result
 
@@ -39,16 +72,19 @@ class AutoPrettyPrint(type):
         return super().__new__(cls, name, bases, dct)
 
 
-def bokeh_from_json(block_data):
+def bokeh_from_json(block_data, show=True):
     from bokeh.io import curdoc
-    from bokeh.plotting import show
+    from bokeh.plotting import show as bokeh_show
 
     if "bokeh_plot_data" in block_data:
         bokeh_plot_data = block_data["bokeh_plot_data"]
     else:
         bokeh_plot_data = block_data
     curdoc().replace_with_json(bokeh_plot_data["doc"])
-    show(curdoc().roots[0])
+    if show:
+        bokeh_show(curdoc().roots[0])
+
+    return curdoc()
 
 
 class BaseDatalabClient(metaclass=AutoPrettyPrint):
@@ -98,6 +134,8 @@ class BaseDatalabClient(metaclass=AutoPrettyPrint):
         self._http_client = httpx.Client
         self._headers["User-Agent"] = f"Datalab Python API/{__version__}"
 
+        self._detect_api_url()
+
         info_json = self.get_info()
 
         self._datalab_api_versions: list[str] = info_json["data"]["attributes"][
@@ -110,6 +148,27 @@ class BaseDatalabClient(metaclass=AutoPrettyPrint):
 
         self._find_api_key()
 
+    def _detect_api_url(self) -> None:
+        """Perform a handshake with the chosen URL to ascertain the correct API URL.
+
+        If a datalab UI URL is passed, the client will attempt to resolve the API URL by
+        inspecting the HTML meta tags.
+
+        Do not use the session for this, so we are not passing the API key to arbitrary URLs.
+
+        """
+        response = httpx.get(self.datalab_api_url)
+        match = re.search(
+            r'<meta name="x_datalab_api_url" content="(.*?)">',
+            response.text,
+            re.IGNORECASE,
+        )
+        if match:
+            self.datalab_api_url = match.group(1)
+            warnings.warn(
+                f"Found API URL {self.datalab_api_url} in HTML meta tag. Creating client with this URL instead."
+            )
+
     def get_info(self) -> dict[str, Any]:
         raise NotImplementedError
 

{datalab_api-0.2.1 → datalab_api-0.2.3/src/datalab_api.egg-info}/PKG-INFO

@@ -1,8 +1,13 @@
 Metadata-Version: 2.1
 Name: datalab-api
-Version: 0.2.1
+Version: 0.2.3
+Summary: A Python API for instances of the datalab data management platform (https://github.com/the-grey-group/datalab).
 Author-email: Matthew Evans <matthew@ml-evs.science>
 License: MIT
+Project-URL: homepage, https://github.com/datalab-org/datalab-api
+Project-URL: repository, https://github.com/datalab-org/datalab-api
+Project-URL: documentation, https://datalab-python-api.readthedocs.io
+Project-URL: changelog, https://github.com/datalab-org/datalab-api/releases
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3
@@ -21,6 +26,9 @@ Requires-Dist: bokeh~=2.4
 Requires-Dist: rich~=13.0
 Provides-Extra: dev
 Requires-Dist: pre-commit~=3.0; extra == "dev"
+Requires-Dist: pytest~=8.0; extra == "dev"
+Requires-Dist: pytest-cov~=5.0; extra == "dev"
+Requires-Dist: respx~=0.21; extra == "dev"
 Provides-Extra: docs
 Requires-Dist: mkdocs; extra == "docs"
 Requires-Dist: mkdocs-material; extra == "docs"
@@ -30,6 +38,8 @@ Requires-Dist: mkdocs-jupyter; extra == "docs"
 Provides-Extra: cli
 Requires-Dist: typer~=0.9; extra == "cli"
 Requires-Dist: click-shell~=2.1; extra == "cli"
+Provides-Extra: all
+Requires-Dist: datalab-api[cli,dev,docs]; extra == "all"
 
 # <div align="center"><i>datalab</i> Python API</div>
 
@@ -95,7 +105,7 @@ This package implements basic functionality for displaying and manipulating entr
 ```python
 from datalab_api import DatalabClient
 
-with DatalabClient("https://api.public.odbx.science") as client:
+with DatalabClient("https://public.api.odbx.science") as client:
 
     # List all items of a given type
     items = client.get_items()

{datalab_api-0.2.1 → datalab_api-0.2.3}/src/datalab_api.egg-info/SOURCES.txt

@@ -9,4 +9,5 @@ src/datalab_api.egg-info/SOURCES.txt
 src/datalab_api.egg-info/dependency_links.txt
 src/datalab_api.egg-info/entry_points.txt
 src/datalab_api.egg-info/requires.txt
-src/datalab_api.egg-info/top_level.txt
+src/datalab_api.egg-info/top_level.txt
+tests/test_client.py

{datalab_api-0.2.1 → datalab_api-0.2.3}/src/datalab_api.egg-info/requires.txt

@@ -2,12 +2,18 @@ httpx~=0.27
 bokeh~=2.4
 rich~=13.0
 
+[all]
+datalab-api[cli,dev,docs]
+
 [cli]
 typer~=0.9
 click-shell~=2.1
 
 [dev]
 pre-commit~=3.0
+pytest~=8.0
+pytest-cov~=5.0
+respx~=0.21
 
 [docs]
 mkdocs

datalab_api-0.2.3/tests/test_client.py

@@ -0,0 +1,22 @@
+import pytest
+import respx
+from datalab_api import DatalabClient
+from httpx import Response
+
+
+@respx.mock
+def test_redirect_url(fake_ui_html, fake_info_json):
+    fake_ui = respx.get("https://ui.datalab.industries").mock(
+        return_value=Response(200, content=fake_ui_html)
+    )
+    fake_api = respx.get("https://api.datalab.industries/info").mock(
+        return_value=Response(200, json=fake_info_json)
+    )
+    with pytest.warns(
+        UserWarning,
+        match="^Found API URL https://api.datalab.industries in HTML meta tag. Creating client with this URL instead.$",
+    ):
+        client = DatalabClient("https://ui.datalab.industries")
+    assert fake_ui.called
+    assert fake_api.called
+    assert client.datalab_api_url == "https://api.datalab.industries"