ydeng/autoBIGS.engine
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Changed FASTA reading to group based on file for merging partial targets

Browse Source
This commit is contained in:
Harrison Deng 2025-02-14 14:35:53 +00:00
parent 765cf9d418
commit d0318536b2
2 changed files with 8 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
src/autobigs/engine/reading.py
View File

@ -5,12 +5,13 @@ from Bio import SeqIO
from autobigs.engine.structures.genomics import NamedString from autobigs.engine.structures.genomics import NamedString
async def read_fasta(handle: Union[str, TextIOWrapper]) -> AsyncGenerator[NamedString, Any]: async def read_fasta(handle: Union[str, TextIOWrapper]) -> Iterable[NamedString]:
fasta_sequences = asyncio.to_thread(SeqIO.parse, handle=handle, format="fasta") fasta_sequences = asyncio.to_thread(SeqIO.parse, handle=handle, format="fasta")
results = []
for fasta_sequence in await fasta_sequences: for fasta_sequence in await fasta_sequences:
yield NamedString(fasta_sequence.id, str(fasta_sequence.seq)) results.append(NamedString(fasta_sequence.id, str(fasta_sequence.seq)))
return results
async def read_multiple_fastas(handles: Iterable[Union[str, TextIOWrapper]]) -> AsyncGenerator[NamedString, Any]: async def read_multiple_fastas(handles: Iterable[Union[str, TextIOWrapper]]) -> AsyncGenerator[Iterable[NamedString], Any]:
for handle in handles: for handle in handles:
async for named_seq in read_fasta(handle): yield await read_fasta(handle)
yield named_seq

4
tests/autobigs/engine/test_reading.py
View File

@ -2,6 +2,6 @@ from autobigs.engine.reading import read_fasta
async def test_fasta_reader_not_none(): async def test_fasta_reader_not_none():
named_strings = read_fasta("tests/resources/tohama_I_bpertussis.fasta") named_strings = await read_fasta("tests/resources/tohama_I_bpertussis.fasta")
async for named_string in named_strings: for named_string in named_strings:
assert named_string.name == "BX470248.1" assert named_string.name == "BX470248.1"