API: Allow [] for array query parameters

(cherry picked from commit 467e041)

Author: Harald Wilhelmi

server/src/scimodom/api/helpers.py

@@ -98,7 +98,7 @@ def get_valid_bam_file(dataset, name) -> BamFile:
 
 
 def get_valid_dataset_id_list_from_request_parameter(parameter: str) -> list[str]:
-    as_list = request.args.getlist(parameter, type=str)
+    as_list = get_unique_list_from_query_parameter(parameter, str)
     if len(as_list) > MAX_DATASET_IDS_IN_LIST:
         raise ClientResponseException(
             400,
@@ -240,7 +240,7 @@ def get_valid_coords(taxa_id: int, context: int = 0) -> tuple[str, int, int, Str
         if end > chrom_size:
             end = chrom_size
 
-    return (chrom, start, end, strand_dto)
+    return chrom, start, end, strand_dto
 
 
 def get_valid_logo(motif: str) -> Path:
@@ -287,7 +287,9 @@ def get_optional_positive_int(field: str) -> int | None:
 
 
 def get_response_from_pydantic_object(obj: BaseModel):
-    return Response(response=obj.json(), status=200, mimetype="application/json")
+    return Response(
+        response=obj.model_dump_json(), status=200, mimetype="application/json"
+    )
 
 
 def _is_valid_identifier(identifier, length):
@@ -303,3 +305,25 @@ def _validate_taxa_id(taxa_id: int) -> None:
     taxa_ids = [d["id"] for d in utilities_service.get_taxa()]
     if taxa_id not in taxa_ids:
         raise ClientResponseException(404, "Unrecognized Taxa ID")
+
+
+def get_unique_list_from_query_parameter(name: str, list_type):
+    """
+    There seems to be some confusion how arrays should be transmitted as query parameters.
+    While most people seem to agree that the values are packed into multiple query parameters,
+    some (older?) implementations leave the original name, while newer ones insist on
+    adding square brackets '[]' at the end of the name, e.g. my_array = ['x', 'y'] may be
+    transmitted like this:
+
+        Old: ?my_array=x&my_array=y
+        New: ?my_array[]=x&my_array[]=y
+
+    Flask seems not to be aware of this. We don't care and allow both. Also, don't want
+    that our code breaks if Flask fixes this - so we ignore double results. So don't use this
+    function for lists, which are allowed to contain the same value multiple times.
+    """
+    result_as_set = {
+        *request.args.getlist(name, type=list_type),
+        *request.args.getlist(f"{name}[]", type=list_type),
+    }
+    return list(result_as_set)

server/src/scimodom/api/modification.py

@@ -22,6 +22,7 @@
     get_optional_positive_int,
     get_optional_non_negative_int,
     validate_rna_type,
+    get_unique_list_from_query_parameter,
 )
 from scimodom.services.bedtools import BedToolsService, get_bedtools_service
 from scimodom.utils.dtos.bedtools import Bed6Record
@@ -269,7 +270,7 @@ def _get_annotation_source():
 
 # TODO: for mod, org, and tech, we should in fact check that they exists in the DB...
 def _get_technology_ids():
-    raw = request.args.getlist("technology", type=int)
+    raw = get_unique_list_from_query_parameter("technology", int)
     if raw is None:
         return []
     for i in raw:
@@ -279,14 +280,14 @@ def _get_technology_ids():
 
 
 def _get_gene_filters():
-    raw = request.args.getlist("geneFilter", type=str)
+    raw = get_unique_list_from_query_parameter("geneFilter", str)
     if raw is None:
         return []
     return raw
 
 
 def _get_gene_or_chrom_required() -> GeneSearch:
-    gene = request.args.getlist("geneFilter", type=str)
+    gene = get_unique_list_from_query_parameter("geneFilter", str)
     if gene:
         return GeneSearch(gene_filter=gene)
     else:
@@ -302,7 +303,7 @@ def _get_gene_or_chrom_required() -> GeneSearch:
 
 
 def _get_multi_sort(url_split: str = "%2B"):
-    raw = request.args.getlist("multiSort", type=str)
+    raw = get_unique_list_from_query_parameter("multiSort", str)
     if raw is None or (len(raw) == 1 and raw[0] == ""):
         return []
     for i in raw:

server/src/scimodom/api/utilities.py

@@ -1,4 +1,4 @@
-from flask import Blueprint, request, Response
+from flask import Blueprint, Response
 from flask_cors import cross_origin
 from sqlalchemy.exc import NoResultFound
 
@@ -7,6 +7,7 @@
     get_valid_taxa_id_from_string,
     ClientResponseException,
     validate_rna_type,
+    get_unique_list_from_query_parameter,
 )
 from scimodom.services.annotation import get_annotation_service, BIOTYPES
 from scimodom.services.assembly import get_assembly_service
@@ -61,7 +62,7 @@ def get_selections():
 @cross_origin(supports_credentials=True)
 def get_genes():
     file_service = get_file_service()
-    selection_ids = request.args.getlist("selection[]", type=int)
+    selection_ids = get_unique_list_from_query_parameter("selection", int)
     try:
         return file_service.get_gene_cache(selection_ids)
     except FileNotFoundError:
@@ -70,7 +71,7 @@ def get_genes():
 
 @api.route("/biotypes/<rna_type>", methods=["GET"])
 @cross_origin(supports_credentials=True)
-def get_biotypes(rna_type):
+def get_biotypes(rna_type):  # noqa
     # TODO: do biotypes also depend on RNA type/annotation?
     return {"biotypes": MAPPED_BIOTYPES}