bed-reader 1.0.5b1__tar.gz → 1.0.5b2__tar.gz

{bed_reader-1.0.5b1 → bed_reader-1.0.5b2}/Cargo.lock

@@ -142,7 +142,7 @@ checksum = "72b3254f16251a8381aa12e40e3c4d2f0199f8c6508fbecb9d91f575e0fbb8c6"
 
 [[package]]
 name = "bed-reader"
-version = "1.0.5-beta.1"
+version = "1.0.5-beta.2"
 dependencies = [
  "anyhow",
  "anyinput",

{bed_reader-1.0.5b1 → bed_reader-1.0.5b2}/Cargo.toml

@@ -1,5 +1,5 @@
 [package]
-version = "1.0.5-beta.1"
+version = "1.0.5-beta.2"
 name = "bed-reader"
 description = "Read and write the PLINK BED format, simply and efficiently."
 repository = "https://github.com/fastlmm/bed-reader"

{bed_reader-1.0.5b1 → bed_reader-1.0.5b2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: bed-reader
-Version: 1.0.5b1
+Version: 1.0.5b2
 Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10

{bed_reader-1.0.5b1 → bed_reader-1.0.5b2}/bed_reader/_open_bed.py

@@ -1,6 +1,8 @@
 import logging
 import multiprocessing
 import os
+import re
+import sys
 from dataclasses import dataclass
 from io import BytesIO
 from itertools import repeat, takewhile
@@ -1792,6 +1794,8 @@ class open_bed:
 
 
 def _read_csv(filepath, delimiter=None, dtype=None, usecols=None):
+    pattern = re.compile(r"^np\.\w+\((.+?)\)$")
+
     # Prepare the usecols by ensuring it is a list of indices
     usecols_indices = list(usecols)
     transposed = np.loadtxt(
@@ -1809,6 +1813,16 @@ def _read_csv(filepath, delimiter=None, dtype=None, usecols=None):
     columns = []
     for output_index, input_index in enumerate(usecols_indices):
         col = transposed[output_index]
+
+        # work around numpy/python bug
+        if (
+            (sys.version_info.major, sys.version_info.minor) <= (3, 9)
+            and int(np.__version__.split(".")[0]) >= 2
+            and len(col) > 0
+            and pattern.fullmatch(col[0])
+        ):
+            col = np.array([pattern.fullmatch(x).group(1) for x in col])
+
         # Find the dtype for this column
         col_dtype = dtype.get(input_index, np.str_)
         # Convert the column list to a numpy array with the specified dtype