Use environment variables for configuration.

Author: meier-rene

api-similarity-score/similarity-score-impl/db.py

@@ -1,35 +1,37 @@
+import os
 import psycopg
-
 from matchms import set_matchms_logger_level, Spectrum
 import numpy as np
 
-# TODO: Move either consts or postgresql url to env file
-DB_PORT = 5432
-DB_USER = "massbank3"
-DB_PASSWORD = "massbank3password"
-DB_NAME = "massbank3"
-DB_HOST = "localhost"
-
+DB_PORT = os.environ.get('DB_PORT', 5432)
+DB_USER = os.environ.get('DB_USER', "massbank3")
+DB_PASSWORD = os.environ.get('DB_PASSWORD', "massbank3password")
+DB_HOST = os.environ.get('DB_HOST', "localhost")
+DB_NAME = os.environ.get('DB_NAME', "massbank3")
 spectra = []
 
+
 # Load all (non-deprecated) spectra from the database for faster lookup
 def load_spectra():
     global spectra
-    
+
     spectra = []
 
     # Prevent matchms from complaining about spectra not having a precursor_mz
     set_matchms_logger_level("ERROR")
 
     with psycopg.connect(f"postgresql://{DB_NAME}:{DB_PASSWORD}@{DB_HOST}:{DB_PORT}/{DB_NAME}") as conn:
         with conn.cursor() as cur:
-            cur.execute("select document->'accession', document->'peak' from massbank where document->>'deprecated' is null;")
+            cur.execute(
+                "select document->'accession', document->'peak' from massbank where document->>'deprecated' is null;")
 
             for spectrum in cur:
-                id = spectrum[0]
+                accession = spectrum[0]
                 peak = spectrum[1]["peak"]
 
                 mz = peak["mz"]
                 intensities = peak["rel"]
 
-                spectra += [Spectrum(mz=np.array(mz).astype(float), intensities=np.array(intensities).astype(float), metadata={"id": id})]
+                #metadata key "accession" gets silently converted to spectrum_id, so we can use spectrum_id right away
+                spectra += [Spectrum(mz=np.array(mz).astype(float), intensities=np.array(intensities).astype(float),
+                                     metadata={'spectrum_id': accession})]

api-similarity-score/similarity-score-impl/similarity-score-controller.py

@@ -33,7 +33,7 @@ def similarity_post(similarity_calculation):  # noqa: E501
 
         if len(request.reference_spectra_list) > 0:
             def filter_fn(spectrum):
-                return spectrum.metadata['id'] in request.reference_spectra_list
+                return spectrum.metadata['spectrum_id'] in request.reference_spectra_list
 
             references = list(filter(filter_fn, spectra))
 
@@ -44,7 +44,7 @@ def filter_fn(spectrum):
         match_list = SimilarityScoreList([])
 
         for match in matches:
-            match_list.similarity_score_list.append(SimilarityScore(match[0].metadata['id'], match[1][0]))
+            match_list.similarity_score_list.append(SimilarityScore(match[0].metadata['spectrum_id'], match[1][0]))
 
         return match_list
 
@@ -58,13 +58,3 @@ def version_get():  # noqa: E501
     :rtype: Union[str, Tuple[str, int], Tuple[str, int, Dict[str, str]]
     """
     return 'cosine similarity 1.0.0'
-
-import db
-
-def main():
-    print("Loading spectra... ", end='', flush=True)
-    db.load_spectra()
-    print("Finished")
-
-if __name__ == '__main__':
-    main()