datatrail-cli 0.10.1__tar.gz → 0.10.2__tar.gz

{datatrail_cli-0.10.1 → datatrail_cli-0.10.2}/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Changelog
 
+## [0.10.2](https://github.com/CHIMEFRB/datatrail-cli/compare/v0.10.1...v0.10.2) (2026-05-07)
+
+
+### Bug Fixes
+
+* **pull:** more processors than files bug ([e44bbe1](https://github.com/CHIMEFRB/datatrail-cli/commit/e44bbe1a14900f034f998b664c1d394fcf31c7a7)), closes [#147](https://github.com/CHIMEFRB/datatrail-cli/issues/147)
+
 ## [0.10.1](https://github.com/CHIMEFRB/datatrail-cli/compare/v0.10.0...v0.10.1) (2026-03-02)
 
 

{datatrail_cli-0.10.1 → datatrail_cli-0.10.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: datatrail-cli
-Version: 0.10.1
+Version: 0.10.2
 Summary: CHIME/FRB Datatrail CLI
 Author: CHIME FRB Project Office
 Author-email: Tarik Zegmott <tarik.zegmott@mcgill.ca>

{datatrail_cli-0.10.1 → datatrail_cli-0.10.2}/dtcli/utilities/cadcclient.py

@@ -161,6 +161,9 @@ def pget(
     elif verbose > 1:
         logger.setLevel("DEBUG")
 
+    # Cap processors to the number of files so we never spawn more workers than
+    # there are files to download (avoids IndexError when -c > number of files).
+    processors = min(processors, len(source))
     sources: List[List[Any]] = split(source, processors)
     destinations: List[List[Any]] = split(destination, processors)
     logger.info(f"Starting {processors} processes.")

{datatrail_cli-0.10.1 → datatrail_cli-0.10.2}/dtcli/utilities/utilities.py

@@ -8,8 +8,6 @@ import requests
 from requests.models import Response
 from rich.console import Console
 
-from dtcli.utilities import cadcclient
-
 try:
     from packaging.version import parse
 except ImportError:
@@ -148,6 +146,8 @@ def check_canfar_status(console: Console) -> Tuple[bool, bool]:
         Tuple[bool, bool]: Status of Minoc and Luskan, respectively.
     """
     # Check Canfar status.
+    from dtcli.utilities import cadcclient  # local import to avoid circular dependency
+
     minoc_up, luskan_up = cadcclient.status()
     if not minoc_up:
         console.print(":warning: Minoc is down!", style="bold yellow")

{datatrail_cli-0.10.1 → datatrail_cli-0.10.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "datatrail-cli"
-version = "0.10.1"
+version = "0.10.2"
 description = "CHIME/FRB Datatrail CLI"
 authors = [
   { name = "CHIME FRB Project Office" },

datatrail_cli-0.10.2/tests/test_utils.py

@@ -0,0 +1,27 @@
+from dtcli.utilities import utilities
+
+
+def test_split():
+    test_array = ["a", "b"]
+    split_by = 2
+    result = utilities.split(test_array, split_by)
+    assert result == [["a"], ["b"]]
+
+
+def test_split_more_batches_than_items():
+    """split() must not return more batches than there are items.
+
+    Regression guard for GitHub issue #147: when the user requests more
+    parallel workers (-c flag) than there are files to download, the
+    number of sublists returned by split() must equal len(data), not
+    the requested count.  The pget() caller now caps processors to
+    min(processors, len(source)) before iterating, so this test documents
+    the invariant that split() never produces empty sublists.
+    """
+    test_array = ["a", "b", "c", "d", "e", "f"]  # 6 files
+    split_by = 8  # user requested 8 workers
+    result = utilities.split(test_array, split_by)
+    # split() drops empty batches, so we get at most len(data) sublists
+    assert len(result) <= len(test_array)
+    # every element appears exactly once
+    assert sorted(sum(result, [])) == sorted(test_array)

datatrail_cli-0.10.1/tests/test_utils.py

@@ -1,8 +0,0 @@
-from dtcli.utilities import utilities
-
-
-def test_split():
-    test_array = ["a", "b"]
-    split_by = 2
-    result = utilities.split(test_array, split_by)
-    assert result == [["a"], ["b"]]