maybe up

.gitignore

@@ -159,3 +159,5 @@ cython_debug/
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
 
+boldsystems.db
+**/boldsystems.db

src/python/dpf-sanger-sequencing/README.md

@@ -0,0 +1,5 @@
+
+
+
+# Citations
+Ratnasingham, Sujeevan, and Paul D N Hebert. “bold: The Barcode of Life Data System (http://www.barcodinglife.org).” Molecular ecology notes vol. 7,3 (2007): 355-364. doi:10.1111/j.1471-8286.2007.01678.x

src/python/dpf-sanger-sequencing/pyproject.toml

@@ -0,0 +1,7 @@
+[project]
+name = "dpf_sanger_sequencing"
+description="A library for interpreting sanger sequencing data"
+requires-python = ">=3.11"
+authors = [{name = "Danny Farrell", email = "[email protected]"}]
+dependencies = ["biopython", "requests", "sqlalchemy"]
+version="1.0.0"

src/python/dpf-sanger-sequencing/scripts/TRACE_FILE_INFO.txt

@@ -0,0 +1,2 @@
+PROCESSID	TAXON	MARKER	GENBANK_ACCESSION	TRACEFILE
+

src/python/dpf-sanger-sequencing/scripts/run_bs_dl.py

@@ -0,0 +1,76 @@
+from pathlib import Path
+from dpf_sanger_sequencing.boldsystems import BoldSystemsAPI, Session, CombinedData
+from dpf_sanger_sequencing.boldsystems.bold_api_access import parse_tsv_weird_format
+import asyncio
+
+
+def store_combined_data_for_taxonomies(taxonomies: list[str]):
+    api = BoldSystemsAPI()
+    # api.get_data_package_data(Path("~/Downloads/bold_pub/BOLD_Public.05-May-2023.tsv").expanduser())
+    # print("x1?")
+    # asyncio.run(api.get_combined_data_from_local_package_data())
+    # raise
+
+    # for i, taxonomy in enumerate(taxonomies):
+    #     print(f"getting {i} / {len(taxonomies)} {taxonomy=}")
+    #     api.get_combined_data(dict(taxon=taxonomy))
+    session = Session()
+    # rows = session.query(CombinedData).all()
+    asyncio.run(api.download_all_trace_data(session))
+
+    # for row in rows:
+        # print(row.processid)
+
+
+if __name__ == "__main__":
+    taxonomies = [
+        "Acanthocephala",
+        # "Acoelomorpha",  # Missing data?
+        "Annelida",
+        "Arthropoda",
+        "Brachiopoda",
+        "Bryozoa",
+        "Chaetognatha",
+        "Chordata",
+        "Cnidaria",
+        "Ctenophora",
+        "Cycliophora",
+        "Echinodermata",
+        "Entoprocta",
+        "Gastrotricha",
+        "Gnathostomulida",
+        "Hemichordata",
+        "Kinorhyncha",
+        "Mollusca",
+        "Nematoda",
+        "Nematomorpha",
+        "Nemertea",
+        "Onychophora",
+        "Phoronida",
+        "Placozoa",
+        "Platyhelminthes",
+        "Porifera",
+        "Priapulida",
+        "Rhombozoa",
+        "Rotifera",
+        "Tardigrada",
+        "Xenacoelomorpha",
+        "Bryophyta",
+        "Chlorophyta",
+        "Lycopodiophyta",
+        "Magnoliophyta",
+        "Pinophyta",
+        "Pteridophyta",
+        "Ascomycota",
+        "Basidiomycota",
+        "Chytridiomycota",
+        "Glomeromycota",
+        "Myxomycota",
+        "Zygomycota",
+        "Chlorarachniophyta",
+        "Ciliophora",
+        "Heterokontophyta",
+        "Pyrrophycophyta",
+        "Rhodophyta",
+    ]
+    store_combined_data_for_taxonomies(taxonomies)

src/python/dpf-sanger-sequencing/scripts/test_bs_api.py

@@ -0,0 +1,8 @@
+
+
+from dpf_sanger_sequencing.boldsystems.bold_api_access import BoldSystemsAPI
+
+api = BoldSystemsAPI()
+search_params = {"taxon": "Brachiopoda"}
+# api.get_combined_data(search_params, "tsv")
+api.get_trace_data(search_params)

src/python/dpf-sanger-sequencing/src/dpf_sanger_sequencing/boldsystems/__init__.py

@@ -0,0 +1,2 @@
+from .models import Base as Base, CombinedData as CombinedData, TraceData as TraceData, UrlQueryData as UrlQueryData
+from .bold_api_access import BoldSystemsAPI as BoldSystemsAPI, Session as Session