Update pysam depedency to 0.20.0

Fix `pysam.index()` calls which now use samtools 1.16 where
`samtools index input output` doesn't work any more if `output` is the path
of an already existing file, but `samtools index -o output input` works
fine.

Also:
- Pre-create the (empty) index file for mock test datasets as a unit test
  reproducer of the above issue.
- Fix `test_cram` unit test that wasn't checking the return value of
  `CRAM.set_index_file()`, which was in fact returning False because
  `test-data/2.cram` was not sorted (updated now).

lib/galaxy/datatypes/binary.py

@@ -749,13 +749,13 @@ def dataset_content_needs_grooming(self, file_name: str) -> bool:
                 cmd = [
                     "python",
                     "-c",
-                    f"import pysam; pysam.set_verbosity(0); pysam.index('{file_name}', '{index_name}')",
+                    f"import pysam; pysam.set_verbosity(0); pysam.index('-o', '{index_name}', '{file_name}')",
                 ]
             else:
                 cmd = [
                     "python",
                     "-c",
-                    f"import pysam; pysam.set_verbosity(0); pysam.index('{index_flag}', '{file_name}', '{index_name}')",
+                    f"import pysam; pysam.set_verbosity(0); pysam.index('{index_flag}', '-o', '{index_name}', '{file_name}')",
                 ]
             with open(os.devnull, "w") as devnull:
                 subprocess.check_call(cmd, stderr=devnull, shell=False)
@@ -786,9 +786,9 @@ def set_meta(
             )
         if index_flag == "-b":
             # IOError: No such file or directory: '-b' if index_flag is set to -b (pysam 0.15.4)
-            pysam.index(dataset.file_name, index_file.file_name)  # type: ignore [attr-defined]
+            pysam.index("-o", index_file.file_name, dataset.file_name)  # type: ignore [attr-defined]
         else:
-            pysam.index(index_flag, dataset.file_name, index_file.file_name)  # type: ignore [attr-defined]
+            pysam.index(index_flag, "-o", index_file.file_name, dataset.file_name)  # type: ignore [attr-defined]
         dataset.metadata.bam_index = index_file
 
     def sniff(self, filename: str) -> bool:
@@ -979,7 +979,7 @@ def get_cram_version(self, filename: str) -> Tuple[int, int]:
 
     def set_index_file(self, dataset: "DatasetInstance", index_file) -> bool:
         try:
-            pysam.index(dataset.file_name, index_file.file_name)  # type: ignore [attr-defined]
+            pysam.index("-o", index_file.file_name, dataset.file_name)  # type: ignore [attr-defined]
             return True
         except Exception as exc:
             log.warning("%s, set_index_file Exception: %s", self, exc)

lib/galaxy/datatypes/converters/cram_to_bam_converter.xml

@@ -16,7 +16,7 @@ python '$__tool_directory__/cram_to_bam.py' '$input' '$output'
             <param name="input" ftype="cram" value="2.cram"/>
             <output name="output" ftype="bam">
                 <assert_contents>
-                    <has_size value="57232"/>
+                    <has_size value="60331"/>
                 </assert_contents>
             </output>
         </test>

lib/galaxy/dependencies/pinned-requirements.txt

@@ -132,7 +132,7 @@ pyparsing==3.0.9 ; python_version >= "3.7" and python_version < "3.11"
 pyreadline3==3.4.1 ; sys_platform == "win32" and python_version >= "3.8" and python_version < "3.11"
 pyreadline==2.1 ; sys_platform == "win32" and python_version < "3.8" and python_version >= "3.7"
 pyrsistent==0.19.3 ; python_version >= "3.7" and python_version < "3.11"
-pysam==0.19.1 ; python_version >= "3.7" and python_version < "3.11"
+pysam==0.20.0 ; python_version >= "3.7" and python_version < "3.11"
 python-dateutil==2.8.2 ; python_version >= "3.7" and python_version < "3.11"
 python-jose==3.3.0 ; python_version >= "3.7" and python_version < "3.11"
 python-magic==0.4.27 ; python_version >= "3.7" and python_version < "3.11"

packages/app/setup.cfg

@@ -60,7 +60,7 @@ install_requires =
     pebble
     pulsar-galaxy-lib>=0.15.0.dev0
     pydantic
-    pysam
+    pysam>=0.20
     PyJWT
     PyYAML
     refgenconf>=0.12.0

packages/data/setup.cfg

@@ -49,7 +49,7 @@ install_requires =
     pydantic[email]
     pylibmagic
     python-magic
-    pysam
+    pysam>=0.20
     rocrate
     social-auth-core[openidconnect]==4.0.3
     SQLAlchemy>=1.4.25,<2

pyproject.toml

@@ -83,7 +83,7 @@ PyJWT = "*"
 pykwalify = "*"
 pylibmagic = "*"
 pyparsing = "*"
-pysam = "*"
+pysam = ">=0.20"
 python = ">=3.7,<3.11"  # Keep maximum version strict to appease numpy and scipy
 python-dateutil = "*"
 python-magic = "*"

test/unit/data/datatypes/test_cram.py

@@ -1,5 +1,3 @@
-import os
-
 import pysam
 
 from galaxy.datatypes.binary import CRAM
@@ -12,9 +10,7 @@
 def test_cram():
     c = CRAM()
     with get_input_files("2.cram") as input_files, get_dataset(input_files[0], index_attr="cram_index") as dataset:
-        assert os.path.exists(dataset.metadata.cram_index.file_name) is False
-        c.set_index_file(dataset=dataset, index_file=dataset.metadata.cram_index)
-        assert os.path.exists(dataset.metadata.cram_index.file_name) is True
+        assert c.set_index_file(dataset=dataset, index_file=dataset.metadata.cram_index) is True
         c.set_meta(dataset)
         pysam.AlignmentFile(dataset.file_name, index_filename=dataset.metadata.cram_index.file_name)
     assert dataset.metadata.cram_version == "3.0"

test/unit/data/datatypes/util.py

@@ -36,7 +36,7 @@ def get_size(self):
 @contextmanager
 def get_dataset(filename, index_attr="bam_index", dataset_id=1, has_data=True):
     dataset = MockDataset(dataset_id)
-    with get_input_files(filename) as input_files, get_tmp_path() as index_path:
+    with get_input_files(filename) as input_files, get_tmp_path(should_exist=True) as index_path:
         dataset.file_name = input_files[0]
         index = MockMetadata()
         index.file_name = index_path
@@ -64,6 +64,6 @@ def get_input_files(*args):
         yield test_files
         new_md5_sums = [md5_hash_file(f) for f in test_files]
         for old_hash, new_hash, f in zip(md5_sums, new_md5_sums, test_files):
-            assert old_hash == new_hash, "Unexpected change of content for file %s" % f
+            assert old_hash == new_hash, f"Unexpected change of content for file {f}"
     finally:
         shutil.rmtree(temp_dir, ignore_errors=True)