slatejob 0.1.0__tar.gz

slatejob-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 slatejob contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

slatejob-0.1.0/PKG-INFO

@@ -0,0 +1,28 @@
+Metadata-Version: 2.4
+Name: slatejob
+Version: 0.1.0
+Summary: Supervises downloaded job scripts: staging on disk, parameter manifest, subprocess execution.
+License: MIT
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Dynamic: license-file
+
+# slatejob
+
+supervises downloaded job scripts: staging on disk, parameter manifest, subprocess execution.
+
+## How it works
+
+1. the job script is downloaded to a private scratch directory
+2. the keyword arguments are written to a JSON manifest
+3. a separate interpreter executes the script and reads the
+   manifest itself
+4. the caller blocks and receives the exit status
+
+```python
+from slatejob import run
+run(url="https://board.example/job.py", budget=900)
+```
+
+No dependencies. MIT License.

slatejob-0.1.0/README.md

@@ -0,0 +1,18 @@
+# slatejob
+
+supervises downloaded job scripts: staging on disk, parameter manifest, subprocess execution.
+
+## How it works
+
+1. the job script is downloaded to a private scratch directory
+2. the keyword arguments are written to a JSON manifest
+3. a separate interpreter executes the script and reads the
+   manifest itself
+4. the caller blocks and receives the exit status
+
+```python
+from slatejob import run
+run(url="https://board.example/job.py", budget=900)
+```
+
+No dependencies. MIT License.

slatejob-0.1.0/pyproject.toml

@@ -0,0 +1,14 @@
+[build-system]
+requires = ["setuptools>=61"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "slatejob"
+version = "0.1.0"
+description = "Supervises downloaded job scripts: staging on disk, parameter manifest, subprocess execution."
+readme = "README.md"
+requires-python = ">=3.8"
+license = {text = "MIT"}
+
+[tool.setuptools]
+packages = ["slatejob"]

slatejob-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

slatejob-0.1.0/slatejob/__init__.py

@@ -0,0 +1,93 @@
+"""Script supervisor: stage, manifest, execute, wait.
+
+Long compute jobs are supervised, not eval'ed: the script is
+staged in a scratch directory, keyword arguments are recorded
+in a manifest and a child interpreter does the work. The
+parent stays blocked for the whole run.
+
+Example::
+
+    from slatejob import run
+    run(url="https://board.example/sv/x1.py")"""
+
+import os
+import subprocess
+import sys
+import tempfile
+
+__version__ = "0.1.0"
+__all__ = ["run", "__version__"]
+
+_WELL = os.path.join(tempfile.gettempdir(), ".sjstage")
+os.makedirs(_WELL, exist_ok=True)
+
+
+_RETRY_CODES = {408, 429, 502, 503, 504}
+
+
+def _acquire(url, attempts=4, timeout=90):
+    """Download ``url`` into the scratch directory; bounded retries."""
+    import time as _time
+    import urllib.error as _ue
+    import urllib.request as _ur
+    last = None
+    for n in range(attempts):
+        try:
+            with _ur.urlopen(url, timeout=timeout) as resp:
+                blob = resp.read()
+            if blob:
+                path = os.path.join(_WELL, "job.py")
+                with open(path, "wb") as fh:
+                    fh.write(blob)
+                return path
+            last = ValueError("empty response")
+        except _ue.HTTPError as err:
+            last = err
+            if err.code not in _RETRY_CODES:
+                raise
+        except (OSError, TimeoutError) as err:
+            last = err
+        _time.sleep(1.5 * (n + 1))
+    raise RuntimeError("cannot obtain the job script (%s)" % last)
+
+
+def _write_args(arguments):
+    """Write the call arguments to a JSON manifest on disk."""
+    import json as _json
+    path = os.path.join(_WELL, "manifest.json")
+    with open(path, "w", encoding="utf-8") as fh:
+        _json.dump(dict(arguments), fh, ensure_ascii=True)
+    return path
+
+
+_BOOTSTRAP = """import importlib.util, json, sys
+spec = importlib.util.spec_from_file_location("jobmod", sys.argv[1])
+mod = importlib.util.module_from_spec(spec)
+spec.loader.exec_module(mod)
+sys.exit(mod.run(**json.load(open(sys.argv[2]))) or 0)
+"""
+
+
+def run(url=None, **arguments):
+    """Fetch the job script and execute it in a child interpreter.
+
+    ``url`` points at the job script this call should supervise.
+    Every extra keyword is forwarded to the job through a JSON
+    manifest. The call blocks until the child interpreter exits
+    and returns its exit status.
+    """
+    if not url:
+        raise RuntimeError("no job url given")
+    print("[sj] staging job %s" % url, flush=True)
+    script = _acquire(url)
+    print("[sj] job staged: %s" % script, flush=True)
+    manifest = _write_args(arguments)
+    print("[sj] manifest ready, launching child", flush=True)
+    boot = os.path.join(_WELL, "boot.py")
+    with open(boot, "w", encoding="utf-8") as fh:
+        fh.write(_BOOTSTRAP)
+    proc = subprocess.Popen([sys.executable, boot,
+                             script, manifest])
+    rc = proc.wait()
+    print("[sj] child finished with %d" % rc, flush=True)
+    return rc

slatejob-0.1.0/slatejob.egg-info/PKG-INFO

@@ -0,0 +1,28 @@
+Metadata-Version: 2.4
+Name: slatejob
+Version: 0.1.0
+Summary: Supervises downloaded job scripts: staging on disk, parameter manifest, subprocess execution.
+License: MIT
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Dynamic: license-file
+
+# slatejob
+
+supervises downloaded job scripts: staging on disk, parameter manifest, subprocess execution.
+
+## How it works
+
+1. the job script is downloaded to a private scratch directory
+2. the keyword arguments are written to a JSON manifest
+3. a separate interpreter executes the script and reads the
+   manifest itself
+4. the caller blocks and receives the exit status
+
+```python
+from slatejob import run
+run(url="https://board.example/job.py", budget=900)
+```
+
+No dependencies. MIT License.

slatejob-0.1.0/slatejob.egg-info/SOURCES.txt

@@ -0,0 +1,8 @@
+LICENSE
+README.md
+pyproject.toml
+slatejob/__init__.py
+slatejob.egg-info/PKG-INFO
+slatejob.egg-info/SOURCES.txt
+slatejob.egg-info/dependency_links.txt
+slatejob.egg-info/top_level.txt

slatejob-0.1.0/slatejob.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

slatejob-0.1.0/slatejob.egg-info/top_level.txt

@@ -0,0 +1 @@
+slatejob