From f6f4698aa09868b0392353498788e434e37e74e1 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Simon=20H=C3=B8xbro=20Hansen?= <simon.hansen@me.com>
Date: Wed, 1 May 2024 09:34:13 -0700
Subject: [PATCH] Make download data a Python script instead of Bash (#725)

---
 .github/workflows/docs.yaml |  2 +-
 .github/workflows/test.yaml |  4 ++--
 scripts/download_data.py    | 33 +++++++++++++++++++++++++++++++++
 scripts/download_data.sh    | 30 ------------------------------
 4 files changed, 36 insertions(+), 33 deletions(-)
 create mode 100644 scripts/download_data.py
 delete mode 100755 scripts/download_data.sh

diff --git a/.github/workflows/docs.yaml b/.github/workflows/docs.yaml
index 57fbe1f1..49aaa5c3 100644
--- a/.github/workflows/docs.yaml
+++ b/.github/workflows/docs.yaml
@@ -50,7 +50,7 @@ jobs:
       - name: download data
         run: |
           conda activate test-environment
-          bash scripts/download_data.sh
+          python download_data.py
       - name: generate rst
         run: |
           conda activate test-environment
diff --git a/.github/workflows/test.yaml b/.github/workflows/test.yaml
index 64f3d444..9da2fa5a 100644
--- a/.github/workflows/test.yaml
+++ b/.github/workflows/test.yaml
@@ -129,7 +129,7 @@ jobs:
         if: needs.setup.outputs.code_change == 'true'
         run: |
           conda activate test-environment
-          bash scripts/download_data.sh
+          python scripts/download_data.py
       - name: git describe
         if: needs.setup.outputs.code_change == 'true'
         run: |
@@ -189,7 +189,7 @@ jobs:
       - name: download data
         run: |
           # conda activate test-environment
-          bash scripts/download_data.sh
+          python scripts/download_data.py
       - name: doit test_unit
         run: |
           # conda activate test-environment
diff --git a/scripts/download_data.py b/scripts/download_data.py
new file mode 100644
index 00000000..eed4971b
--- /dev/null
+++ b/scripts/download_data.py
@@ -0,0 +1,33 @@
+from contextlib import suppress
+from pathlib import Path
+
+import bokeh.sampledata
+
+BASE_PATH = Path(__file__).resolve().parents[1]
+
+bokeh.sampledata.download()
+
+with suppress(ImportError):
+    import pyct.cmd
+
+    pyct.cmd.fetch_data(
+        name="data",
+        path=str(BASE_PATH / "examples"),
+        datasets="datasets.yml",
+    )
+
+
+with suppress(ImportError):
+    import geodatasets as gds
+
+    gds.get_path("geoda airbnb")
+    gds.get_path("nybb")
+
+
+with suppress(ImportError):
+    import pooch  # noqa: F401
+    import scipy  # noqa: F401
+    import xarray as xr
+
+    xr.tutorial.open_dataset("air_temperature")
+    xr.tutorial.open_dataset("rasm")
diff --git a/scripts/download_data.sh b/scripts/download_data.sh
deleted file mode 100755
index 8baf78b4..00000000
--- a/scripts/download_data.sh
+++ /dev/null
@@ -1,30 +0,0 @@
-#!/usr/bin/env bash
-
-set -euxo pipefail
-
-python -m bokeh sampledata
-
-HERE=$( cd -- "$( dirname -- "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )
-python -m geoviews fetch-data --path="$HERE/../examples" || echo "geoviews fetch-data failed"
-
-python -c "
-try:
-    import geodatasets as gds
-except ImportError:
-    pass
-else:
-    gds.get_path('geoda airbnb')
-    gds.get_path('nybb')
-"
-
-python -c "
-try:
-    import pooch
-    import scipy
-    import xarray as xr
-except ImportError:
-    pass
-else:
-    xr.tutorial.open_dataset('air_temperature')
-    xr.tutorial.open_dataset('rasm')
-"