deriva-ml 1.14.35__py3-none-any.whl → 1.14.37__py3-none-any.whl

deriva_ml/bump_version.py

@@ -0,0 +1,142 @@
+#!/usr/bin/env python3
+"""
+Release helper: seed initial semver tag if none exists, otherwise bump with bump-my-version.
+
+Usage:
+  python release.py [patch|minor|major]
+
+Env vars:
+  START   - initial version if no tag exists (default: 0.1.0)
+  PREFIX  - tag prefix (default: v)
+"""
+
+from __future__ import annotations
+
+import argparse
+import os
+import shutil
+import subprocess
+import sys
+from typing import Sequence
+
+
+def run(
+    cmd: Sequence[str], check: bool = True, capture: bool = False, quiet: bool = False
+) -> subprocess.CompletedProcess:
+    if not quiet:
+        print(f"$ {' '.join(cmd)}")
+    return subprocess.run(
+        cmd,
+        check=check,
+        capture_output=capture,
+        text=True,
+    )
+
+
+def in_git_repo() -> bool:
+    try:
+        run(["git", "rev-parse", "--is-inside-work-tree"], capture=True)
+        return True
+    except subprocess.CalledProcessError:
+        return False
+
+
+def has_commits() -> bool:
+    try:
+        run(["git", "log", "-1"], capture=True)
+        return True
+    except subprocess.CalledProcessError:
+        return False
+
+
+def latest_semver_tag(prefix: str) -> str | None:
+    # Use git's matcher to keep parity with Bash: prefix + x.y.z
+    pattern = f"{prefix}[0-9]*.[0-9]*.[0-9]*"
+    try:
+        cp = run(["git", "describe", "--tags", "--abbrev=0", "--match", pattern], capture=True, quiet=True)
+        tag = cp.stdout.strip()
+        return tag or None
+    except subprocess.CalledProcessError:
+        return None
+
+
+def seed_initial_tag(tag: str) -> None:
+    print(f"No existing semver tag found. Seeding initial tag: {tag}")
+    run(["git", "tag", tag, "-m", f"Initial release {tag}"])
+    # Push tags (ignore failure to keep parity with bash's simple flow)
+    run(["git", "push", "--tags"])
+
+
+def require_tool(name: str) -> None:
+    if shutil.which(name) is None:
+        print(f"Error: required tool '{name}' not found on PATH.", file=sys.stderr)
+        sys.exit(1)
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(
+        description="Set a new version tag for the current repository, and push to remote."
+    )
+    parser.add_argument(
+        "bump", nargs="?", default="patch", choices=["patch", "minor", "major"], help="Which semver part to bump."
+    )
+    args = parser.parse_args()
+
+    start = os.environ.get("START", "0.1.0")
+    prefix = os.environ.get("PREFIX", "v")
+
+    # Sanity checks
+    require_tool("git")
+    require_tool("uv")
+
+    if not in_git_repo():
+        print("Not a git repo.", file=sys.stderr)
+        return 1
+
+    # Ensure tags visible in shallow clones
+    try:
+        run(["git", "fetch", "--tags", "--quiet"], check=False)
+    except Exception:
+        pass  # non-fatal
+
+    if not has_commits():
+        print("No commits found. Commit something before tagging.", file=sys.stderr)
+        return 1
+
+    # Find latest semver tag with prefix
+    tag = latest_semver_tag(prefix)
+
+    if not tag:
+        seed_initial_tag(f"{prefix}{start}")
+        print(f"Seeded {prefix}{start}. Done.")
+        return 0
+
+    print(f"Bumping version: {args.bump}")
+
+    # Bump using bump-my-version via uv
+    # Mirrors: uv run bump-my-version bump $BUMP --verbose
+    try:
+        run(["uv", "run", "bump-my-version", "bump", args.bump, "--verbose"])
+    except subprocess.CalledProcessError as e:
+        print(e.stdout or "", end="")
+        print(e.stderr or "", end="", file=sys.stderr)
+        return e.returncode
+
+    # Push commits and tags
+    print("Pushing changes to remote repository...")
+    run(["git", "push", "--follow-tags"])
+
+    # Retrieve new version tag
+    try:
+        cp = run(["git", "describe", "--tags", "--abbrev=0"], capture=True)
+        new_tag = cp.stdout.strip()
+        print(f"New version tag: {new_tag}")
+    except subprocess.CalledProcessError:
+        print("Warning: unable to determine new tag via git describe.", file=sys.stderr)
+
+    print("Release process complete!")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())

deriva_ml/execution/execution.py

@@ -360,6 +360,14 @@ class Execution:
             with Path(cfile).open("w", encoding="utf-8") as config_file:
                 json.dump(self.configuration.model_dump(), config_file)
 
+            lock_file = Path(self.configuration.workflow.git_root) / "uv.lock"
+            if lock_file.exists():
+                _ = self.asset_file_path(
+                    MLAsset.execution_metadata,
+                    lock_file,
+                    ExecMetadataType.execution_config.value,
+                )
+
             for parameter_file in self.configuration.parameters:
                 self.asset_file_path(
                     MLAsset.execution_asset,
@@ -896,9 +904,7 @@ class Execution:
                     asset_path.write_bytes(file_name.read_bytes())
 
         # Persist the asset types into a file
-        with Path(
-            asset_type_path(self._working_dir, self.execution_rid, asset_table)
-        ).open("a") as asset_type_file:
+        with Path(asset_type_path(self._working_dir, self.execution_rid, asset_table)).open("a") as asset_type_file:
             asset_type_file.write(json.dumps({file_name.name: asset_types}) + "\n")
 
         return AssetFilePath(

deriva_ml/execution/workflow.py

@@ -84,6 +84,7 @@ class Workflow(BaseModel):
     rid: RID | None = None
     checksum: str | None = None
     is_notebook: bool = False
+    git_root: Path | None = None
 
     _logger: logging.Logger = PrivateAttr(default=10)
 
@@ -125,6 +126,7 @@ class Workflow(BaseModel):
         if not self.url:
             path, self.is_notebook = Workflow._get_python_script()
             self.url, self.checksum = Workflow.get_url_and_checksum(path)
+            self.git_root = Workflow._get_git_root(path)
 
         self._logger = logging.getLogger("deriva_ml")
         return self
@@ -315,12 +317,9 @@ class Workflow(BaseModel):
             ]
             # Get the caller's filename, which is two up the stack from here.
             filename = Path(stack[-1])
-            if not filename.exists() or Workflow._in_repl():
+            if not (filename.exists() or Workflow._in_repl()):
                 # Being called from the command line interpreter.
                 filename = Path.cwd() / Path("REPL")
-            # Get the caller's filename, which is two up the stack from here.
-            else:
-                raise DerivaMLException("Looking for caller failed")  # Stack is too shallow
         return filename, is_notebook
 
     @staticmethod

deriva_ml/install_kernel.py

@@ -1,9 +1,13 @@
 # your_pkg/install_kernel.py
-import sys
 import re
+import sys
+from argparse import ArgumentParser
 from importlib import metadata
+from pathlib import Path
+
 from ipykernel.kernelspec import install as install_kernel
 
+
 def _dist_name_for_this_package() -> str:
     """
     Try to resolve the distribution name that provides this package.
@@ -20,6 +24,7 @@ def _dist_name_for_this_package() -> str:
     dist_name = dists[0] if dists else metadata.metadata(top_pkg).get("Name", top_pkg)
     return dist_name
 
+
 def _normalize_kernel_name(name: str) -> str:
     """
     Jupyter kernel directory names should be simple: lowercase, [-a-z0-9_].
@@ -28,19 +33,40 @@ def _normalize_kernel_name(name: str) -> str:
     name = re.sub(r"[^a-z0-9._-]+", "-", name)
     return name
 
+
+def _name_for_this_venv() -> str:
+    config_path = Path(sys.prefix) / "pyvenv.cfg"
+    with config_path.open() as f:
+        m = re.search("prompt *= *(?P<prompt>.*)", f.read())
+    return m["prompt"] if m else ""
+
+
 def main() -> None:
-    dist_name = _dist_name_for_this_package()               # e.g., "deriva-model-template"
-    kernel_name = _normalize_kernel_name(dist_name)         # e.g., "deriva-model-template"
+    parser = ArgumentParser()
+    parser.add_argument(
+        "--install-local",
+        action="store_true",
+        help="Create kernal in local venv directory instead of sys.prefix.",
+    )
+
+    dist_name = _name_for_this_venv()  # e.g., "deriva-model-template"
+    kernel_name = _normalize_kernel_name(dist_name)  # e.g., "deriva-model-template"
     display_name = f"Python ({dist_name})"
 
     # Install into the current environment's prefix (e.g., .venv/share/jupyter/kernels/..)
+    prefix_arg = {}
+    install_local = False
+    if install_local:
+        prefix_arg = {"prefix": sys.prefix}
+
     install_kernel(
-        user=False,                 # write under sys.prefix (the active env)
+        user=False,  # write under sys.prefix (the active env)
         kernel_name=kernel_name,
         display_name=display_name,
-        prefix=sys.prefix,
+        **prefix_arg,
     )
     print(f"Installed Jupyter kernel '{kernel_name}' with display name '{display_name}' under {sys.prefix!s}")
 
+
 if __name__ == "__main__":
-    main()
+    main()

deriva_ml/schema/create_schema.py

@@ -298,7 +298,7 @@ def initialize_ml_schema(model: Model, schema_name: str = "deriva-ml"):
             },
             {
                 "Name": "Runtime_Env",
-                "Description": "Information about the execution environment",
+                "Description": "Information about the runtime environment",
             },
             {
                 "Name": "Execution_Metadata",

{deriva_ml-1.14.35.dist-info → deriva_ml-1.14.37.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: deriva-ml
-Version: 1.14.35
+Version: 1.14.37
 Summary: Utilities to simplify use of Dervia and Pandas to create reproducable ML pipelines
 Author-email: ISRD <isrd-dev@isi.edu>
 Requires-Python: >=3.10

{deriva_ml-1.14.35.dist-info → deriva_ml-1.14.37.dist-info}/RECORD

@@ -1,7 +1,8 @@
 deriva_ml/__init__.py,sha256=_aMdxGG4mRTcXodLZLNpXqH8v5uqMbqFUryE9KqNSB8,1158
+deriva_ml/bump_version.py,sha256=KpHmkpEztly2QHYL4dyaIGdEMyP4F0D89rawyh5EDTs,3982
 deriva_ml/demo_catalog.py,sha256=JjPAIac_hKPh5krEhGJydjXquRnivi7kQoR8W4Khp-s,14928
 deriva_ml/feature.py,sha256=6-aphkxdKjWa9oPSGFWxHcwAc_8hmWj-7I4M178YG5Y,8470
-deriva_ml/install_kernel.py,sha256=LHAdy7s-m2jIvzptMpxLdHHkHc62Tn16e9MUWSGIznA,1651
+deriva_ml/install_kernel.py,sha256=Lzy7gYGOpPU4vtQGh-Buo2TFqPRXWeGOvCoINJzELZ0,2195
 deriva_ml/run_notebook.py,sha256=XPZPtGDLv0Ej6dcqIwA1dx_ya1Z9kzsP7um473WtLpU,6519
 deriva_ml/core/__init__.py,sha256=V_i90pc5PB1F4UdOO6DZWzpEFaZDTaPRU-EzKXQ19eI,787
 deriva_ml/core/base.py,sha256=uG7xyVk12fXXWipusRmaBGplxbdlnN9sIOGAEHWD8i8,61090
@@ -19,9 +20,9 @@ deriva_ml/dataset/history.py,sha256=FK5AYYz11p4E4FWMVg4r7UPWOD4eobrq3b3xMjWF59g,
 deriva_ml/dataset/upload.py,sha256=Ad5JDfGvkIvefE-plP8SN9pNAxHzYrBoid5isz_bnNs,16411
 deriva_ml/execution/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 deriva_ml/execution/environment.py,sha256=B7nywqxFTRUWgyu8n7rFoKcVC9on422kjeFG2FPQfvg,9302
-deriva_ml/execution/execution.py,sha256=tXWkFLDoSre836x6MMkcmhtmr3zP5_VoSioQ72-XmvE,44298
+deriva_ml/execution/execution.py,sha256=NJT4qzZvvBXAlh73NVM39VE-uinSVBu2mHtuZD35G1M,44591
 deriva_ml/execution/execution_configuration.py,sha256=Rw4VWkBCZN9yatvSKdTqEWTfu470lpcVKfHFR0uN0jI,6248
-deriva_ml/execution/workflow.py,sha256=AeCT4eYAnYvBIx-Ar-Cyg64YoPyIB8IJdg3lKkCiKCU,13390
+deriva_ml/execution/workflow.py,sha256=-dX2Yyd9PWRO0Lcr-ji-e7KmlOShE795B5PvJKacz2Q,13293
 deriva_ml/model/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 deriva_ml/model/catalog.py,sha256=dzTBcRlqgEVkPY32AUax_iu75RgFiT4Pu5au7rmrv8k,14068
 deriva_ml/model/database.py,sha256=MlXQQFgFmGxZbRx-unRFoRttXwpJspV4v2AIgppttCU,14805
@@ -29,13 +30,13 @@ deriva_ml/model/sql_mapper.py,sha256=_0QsJEVSgSPtxrWKSgjfPZCQ1aMVcjR_Tk2OxLhWEvY
 deriva_ml/schema/__init__.py,sha256=yV-MfzCF3FA4OOz7mZwMM2q6-x1vgOJ057kUvikFF6E,130
 deriva_ml/schema/annotations.py,sha256=TuQ3vWFnK0160fRmtvsCkHx9qAcRa63MSyERB4x5a98,18197
 deriva_ml/schema/check_schema.py,sha256=6dadLYHPqRex6AYVClmsESI8WhC7-rb-XnGf2G298xw,3609
-deriva_ml/schema/create_schema.py,sha256=0ydJSZEg3C3-m8hWPN6k2MoUvm-RWxAlKFzVChxcx3I,12791
+deriva_ml/schema/create_schema.py,sha256=IrnSfN0ufS3M31MD8M6ZWyfJidKllLPqDFBUDAIDPY0,12789
 deriva_ml/schema/deriva-ml-reference.json,sha256=AEOMIgwKO3dNMMWHb0lxaXyamvfAEbUPh8qw0aAtsUQ,242460
 deriva_ml/schema/policy.json,sha256=5ykB8nnZFl-oCHzlAwppCFKJHWJFIkYognUMVEanfY8,1826
 deriva_ml/schema/table_comments_utils.py,sha256=4flCqnZAaqg_uSZ9I18pNUWAZoLfmMCXbmI5uERY5vM,2007
-deriva_ml-1.14.35.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-deriva_ml-1.14.35.dist-info/METADATA,sha256=ozEBGEehEvA6IHHD2u_vf8cupR-tJIvNdr9QDHjObV0,1122
-deriva_ml-1.14.35.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-deriva_ml-1.14.35.dist-info/entry_points.txt,sha256=Fg8f56jlXzbxzAqnIix7ntqZWmGdqmD11m1OOgiCXDM,420
-deriva_ml-1.14.35.dist-info/top_level.txt,sha256=I1Q1dkH96cRghdsFRVqwpa2M7IqJpR2QPUNNc5-Bnpw,10
-deriva_ml-1.14.35.dist-info/RECORD,,
+deriva_ml-1.14.37.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+deriva_ml-1.14.37.dist-info/METADATA,sha256=E_BHOiZCxJcYC6UBwuanetObG6g4WJNyoHh-tedBYmk,1122
+deriva_ml-1.14.37.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+deriva_ml-1.14.37.dist-info/entry_points.txt,sha256=XsHSbfp7S1cKMjHoPUdFIaFcp9lHXHS6CV1zb_MEXkg,463
+deriva_ml-1.14.37.dist-info/top_level.txt,sha256=I1Q1dkH96cRghdsFRVqwpa2M7IqJpR2QPUNNc5-Bnpw,10
+deriva_ml-1.14.37.dist-info/RECORD,,

{deriva_ml-1.14.35.dist-info → deriva_ml-1.14.37.dist-info}/entry_points.txt

@@ -1,4 +1,5 @@
 [console_scripts]
+bump-version = deriva_ml.bump_version:main
 deriva-ml-alter-annotation = deriva_ml.schema_setup.alter_annotation:main
 deriva-ml-check-catalog-schema = deriva_ml.schema.check_schema:main
 deriva-ml-create-schema = deriva_ml.schema_setup.create_schema:main