Adds README for environments

dockerfiles/Makefile

@@ -2,7 +2,7 @@
 
 NAMESPACE ?= ghcr.io/esgf-nimbus
 
-build push run: REPOSITORY = $(lastword $(subst /, ,$(PWD)))
+build push run readme: REPOSITORY = $(lastword $(subst /, ,$(PWD)))
 
 build:
 	docker build $(ARGS) -t $(NAMESPACE)/$(REPOSITORY):$(VERSION) . $(if $(POST_BUILD),&& $(POST_BUILD),)
@@ -18,3 +18,6 @@ changelog:
 
 bump-%:
 	tbump --no-tag --no-push $(shell pysemver bump $* $(shell tbump current-version)) && make changelog && git add changelog.md && git commit -m "Updates changelog"
+
+readme:
+	docker run -it --rm --entrypoint python -v $(PWD):/host -w /host $(NAMESPACE)/$(REPOSITORY):$(VERSION) /host/generate.py $(ENVIRONMENT_FILE)

dockerfiles/climate-notebook-gpu/Makefile

@@ -4,4 +4,6 @@ VERSION = 0.1.5
 
 run: ARGS = -p 8888:8888
 
+ENVIRONMENT_FILE = base.yaml
+
 include ../Makefile

dockerfiles/climate-notebook-gpu/README.md

@@ -0,0 +1,6 @@
+# Packages
+
+| ML packages |
+| ----------- |
+| rapids 0.0.1 |
+| cupy 13.2.0 |

dockerfiles/climate-notebook-gpu/generate.py

@@ -0,0 +1,104 @@
+import json
+import re
+import subprocess
+import itertools
+import argparse
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("file", help="Conda environment file")
+
+    kwargs = vars(parser.parse_args())
+
+    categories = get_categories(**kwargs)
+
+    search_args = "|".join(
+        [f"^{x}$" for packages in categories.values() for x in packages]
+    )
+
+    output = subprocess.run(
+        f"mamba list {search_args!r} --json",
+        capture_output=True,
+        shell=True
+    )
+
+    output_json = json.loads(output.stdout)
+
+    versions = {
+        package["name"]: package["version"] for package in output_json
+    }
+
+    table = []
+    categories = list(categories.items())
+
+    for i in range(0, len(categories), 3):
+        header = " | ".join([x[0] for x in categories[i:i+3]])
+        separator = " | ".join([f"{'-'*len(x[0])}" for x in categories[i:i+3]])
+        header = f"| {header} |\n| {separator} |"
+
+        rows = itertools.zip_longest(*[x[1] for x in categories[i:i+3]])
+        rows = [
+            " | ".join(
+                [
+                    "" if y is None else f"{y} {get_version(y, versions)}"
+                    for y in x
+                ]
+            ) for x in rows
+        ]
+        rows = "\n".join(f"| {x} |" for x in rows)
+
+        table.append(f"{header}\n{rows}")
+
+    table = "\n".join(table)
+
+    data = f"""# Packages
+
+{table}"""
+
+    with open("README.md", "w") as fd:
+        fd.write(data)
+
+
+def get_version(package, versions):
+    m = re.search(r"([^<>=]*)(?:(?:<|>|<=|>=|==).*)?", package)
+
+    if m is None:
+        raise Exception("Could not parse ", package)
+
+    name = m.group(1)
+
+    return versions[name]
+
+
+def get_categories(file, **_):
+    with open(file) as fd:
+        lines = fd.readlines()
+
+    category = None
+    packages = {}
+
+    for line in lines:
+        if "#" in line:
+            m = re.search("# (.*)(?: packages)?", line)
+
+            if m is None:
+                continue
+
+            category = m.group(1)
+        elif category != "ignore" and category is not None:
+            m = re.search("- (.*)", line)
+
+            if m is None:
+                continue
+
+            if category in packages:
+                packages[category].append(m.group(1))
+            else:
+                packages[category] = [m.group(1)]
+
+    return packages
+
+
+if __name__ == "__main__":
+    main()

dockerfiles/climate-notebook/Makefile

@@ -12,4 +12,6 @@ export-env: ENTRYPOINT = /bin/bash
 export-env: COMMAND = -c "cp /base.yaml /conda-envs/"
 export-env: run
 
+ENVIRONMENT_FILE = base.yaml
+
 include ../Makefile

dockerfiles/climate-notebook/README.md

@@ -0,0 +1,16 @@
+# Packages
+
+| Data packages | Climate packages | Plotting packages |
+| ------------- | ---------------- | ----------------- |
+| xarray 2024.6.0 | xcdat 0.7.0 | matplotlib 3.8.4 |
+| xarray-datatree 0.0.14 | xesmf 0.8.5 | holoviews 1.17.1 |
+| pandas 2.2.2 | xgcm 0.8.1 | hvplot 0.10.0 |
+| flox 0.9.8 | pcmdi_metrics 3.4.1 | geoviews 1.9.6 |
+| netcdf4 1.6.5 | gsw-xarray 0.4.0 | seaborn 0.13.2 |
+| s3fs 2024.6.0 |  | cartopy 0.23.0 |
+| zarr 2.18.2 |  | datashader 0.16.2 |
+| Computing packages | Discovery packages |
+| ------------------ | ------------------ |
+| dask-gateway 2024.1.0 | intake-esm 2024.2.6 |
+|  | intake-stac 0.4.0 |
+|  | intake-xarray 0.7.0 |

dockerfiles/climate-notebook/generate.py

@@ -0,0 +1,104 @@
+import json
+import re
+import subprocess
+import itertools
+import argparse
+
+
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("file", help="Conda environment file")
+
+    kwargs = vars(parser.parse_args())
+
+    categories = get_categories(**kwargs)
+
+    search_args = "|".join(
+        [f"^{x}$" for packages in categories.values() for x in packages]
+    )
+
+    output = subprocess.run(
+        f"mamba list {search_args!r} --json",
+        capture_output=True,
+        shell=True
+    )
+
+    output_json = json.loads(output.stdout)
+
+    versions = {
+        package["name"]: package["version"] for package in output_json
+    }
+
+    table = []
+    categories = list(categories.items())
+
+    for i in range(0, len(categories), 3):
+        header = " | ".join([x[0] for x in categories[i:i+3]])
+        separator = " | ".join([f"{'-'*len(x[0])}" for x in categories[i:i+3]])
+        header = f"| {header} |\n| {separator} |"
+
+        rows = itertools.zip_longest(*[x[1] for x in categories[i:i+3]])
+        rows = [
+            " | ".join(
+                [
+                    "" if y is None else f"{y} {get_version(y, versions)}"
+                    for y in x
+                ]
+            ) for x in rows
+        ]
+        rows = "\n".join(f"| {x} |" for x in rows)
+
+        table.append(f"{header}\n{rows}")
+
+    table = "\n".join(table)
+
+    data = f"""# Packages
+
+{table}"""
+
+    with open("README.md", "w") as fd:
+        fd.write(data)
+
+
+def get_version(package, versions):
+    m = re.search(r"([^<>=]*)(?:(?:<|>|<=|>=|==).*)?", package)
+
+    if m is None:
+        raise Exception("Could not parse ", package)
+
+    name = m.group(1)
+
+    return versions[name]
+
+
+def get_categories(file, **_):
+    with open(file) as fd:
+        lines = fd.readlines()
+
+    category = None
+    packages = {}
+
+    for line in lines:
+        if "#" in line:
+            m = re.search("# (.*)(?: packages)?", line)
+
+            if m is None:
+                continue
+
+            category = m.group(1)
+        elif category != "ignore" and category is not None:
+            m = re.search("- (.*)", line)
+
+            if m is None:
+                continue
+
+            if category in packages:
+                packages[category].append(m.group(1))
+            else:
+                packages[category] = [m.group(1)]
+
+    return packages
+
+
+if __name__ == "__main__":
+    main()