fork/cellxgene-gateway
1
0
Fork 0
mirror of https://github.com/Novartis/cellxgene-gateway.git synced 2026-10-11 07:20:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'ghall/just_gene_sets'

Browse Source
This commit is contained in:
Alok Saldanha
2024-03-02 11:09:00 -05:00
parent f046e7c5d2 19caa6cb80
commit 66cca86b51
1 changed files with 22 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
cellxgene_gateway/items/file/fileitem_source.py
+22 -5
View File
@@ -24,21 +24,21 @@ class FileItemSource(ItemSource):
h5ad_suffix=dir_util.h5ad_suffix, h5ad_suffix=dir_util.h5ad_suffix,
annotation_dir_suffix=dir_util.annotations_suffix, annotation_dir_suffix=dir_util.annotations_suffix,
annotation_file_suffix=".csv", annotation_file_suffix=".csv",
gene_set_file_suffix="_gene_sets.csv",
): ):
self._name = name self._name = name
self.base_path = base_path self.base_path = base_path
self.h5ad_suffix = h5ad_suffix self.h5ad_suffix = h5ad_suffix
self.annotation_dir_suffix = annotation_dir_suffix self.annotation_dir_suffix = annotation_dir_suffix
self.annotation_file_suffix = annotation_file_suffix self.annotation_file_suffix = annotation_file_suffix
self.gene_set_file_suffix = gene_set_file_suffix
@property @property
def name(self): def name(self):
return self._name or f"Files:{self.base_path}" return self._name or f"Files:{self.base_path}"
def is_gene_set(self, path: str) -> bool: def is_gene_set(self, path: str) -> bool:
return ("_gene_sets" in path or "-gene-sets" in path) and path.endswith( return path.endswith(self.gene_set_file_suffix)
self.annotation_file_suffix
)
def is_h5ad_file(self, path: str) -> bool: def is_h5ad_file(self, path: str) -> bool:
return path.endswith(self.h5ad_suffix) and os.path.isfile(path) return path.endswith(self.h5ad_suffix) and os.path.isfile(path)
@@ -186,12 +186,29 @@ class FileItemSource(ItemSource):
annotations_subpath = self.get_annotations_subpath(item) annotations_subpath = self.get_annotations_subpath(item)
annotations_fullpath = self.full_path(annotations_subpath) annotations_fullpath = self.full_path(annotations_subpath)
if os.path.isdir(annotations_fullpath): if os.path.isdir(annotations_fullpath):
return [ sorted_files = sorted(os.listdir(annotations_fullpath))
annotation_files = [
self.make_fileitem_from_path(annotation, annotations_subpath, True) self.make_fileitem_from_path(annotation, annotations_subpath, True)
for annotation in sorted(os.listdir(annotations_fullpath)) for annotation in sorted_files
if annotation.endswith(self.annotation_file_suffix) if annotation.endswith(self.annotation_file_suffix)
and not self.is_gene_set(annotation) and not self.is_gene_set(annotation)
and os.path.isfile(os.path.join(annotations_fullpath, annotation)) and os.path.isfile(os.path.join(annotations_fullpath, annotation))
] ]
# Catch gene sets without accompanying [annotations].csv
gene_sets_files = [
self.make_fileitem_from_path(
annotation[: -len(self.gene_set_file_suffix)] + ".csv",
annotations_subpath,
True,
)
for annotation in sorted_files
if self.is_gene_set(annotation)
and annotation[: -len(self.gene_set_file_suffix)]
not in [a.name for a in annotation_files]
and os.path.isfile(os.path.join(annotations_fullpath, annotation))
]
return sorted(annotation_files + gene_sets_files, key=lambda x: x.name)
else: else:
return None return None