feat(IPVC-2229): add utility to download from eutils (#11)

src/uta/exceptions.py

@@ -17,6 +17,10 @@ class InvalidIntervalError(UTAError):
 class InvalidHGVSVariantError(UTAError):
     pass
 
+
+class EutilsDownloadError(Exception):
+    pass
+
 # <LICENSE>
 # Copyright 2014 UTA Contributors (https://bitbucket.org/biocommons/uta)
 ##

src/uta/tools/eutils.py

@@ -0,0 +1,35 @@
+from enum import Enum
+
+import requests
+
+from uta import EutilsDownloadError
+
+
+class NcbiFileFormatEnum(str, Enum):
+    FASTA = "fasta"
+    GENBANK = "gb"
+
+
+def download_from_eutils(accession: str, file_format: NcbiFileFormatEnum, output_file: str) -> None:
+    """
+    Download a file from NCBI using the eutils endpoint.
+    Args:
+    - accession: NCBI accession ID
+    - file_format: File format to download ("fasta" or "gb")
+    - output_file: Path to the file where the downloaded content will be saved
+    """
+
+    base_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi"
+    params = {
+        "db": "nuccore",
+        "id": accession,
+        "retmode": "text",
+        "rettype": file_format
+    }
+    response = requests.get(base_url, params=params)
+
+    if response.status_code == 200:
+        with open(output_file, 'w') as file:
+            file.write(response.text)
+    else:
+        raise EutilsDownloadError(f"Failed to download {file_format} file for {accession}. HTTP status code: {response.status_code}")

tests/test_uta_tools_eutils.py

@@ -0,0 +1,47 @@
+import os
+import unittest
+from unittest.mock import Mock, patch
+
+from uta import EutilsDownloadError
+from uta.tools.eutils import download_from_eutils, NcbiFileFormatEnum
+
+
+class TestEutils(unittest.TestCase):
+    URL = 'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi'
+
+    def setUp(self):
+        self.output_file = 'test_output.fa'
+
+    def tearDown(self):
+        if os.path.exists(self.output_file):
+            os.remove(self.output_file)
+
+    @patch('requests.get')
+    def test_successful_download(self, mock_get):
+        mock_response = Mock()
+        mock_response.status_code = 200
+        mock_response.text = 'file content'
+        mock_get.return_value = mock_response
+
+        download_from_eutils('accession', NcbiFileFormatEnum.FASTA, self.output_file)
+
+        mock_get.assert_called_once_with(
+            self.URL,
+            params={
+                'db': 'nuccore',
+                'id': 'accession',
+                'retmode': 'text',
+                'rettype': 'fasta'
+            }
+        )
+
+        with open(self.output_file, 'r') as file:
+            self.assertEqual(file.read(), 'file content')
+
+    @patch('requests.get')
+    def test_unsuccessful_download(self, mock_get):
+        mock_response = Mock()
+        mock_response.status_code = 404
+        mock_get.return_value = mock_response
+        with self.assertRaises(EutilsDownloadError):
+            download_from_eutils('accession', NcbiFileFormatEnum.FASTA, self.output_file)