add sort of integration test

Author: florian-huber

tests/test_library_io.py

@@ -0,0 +1,92 @@
+# tests/test_library_workflow.py
+from pathlib import Path
+import numpy as np
+import pytest
+from matchms.importing import load_spectra
+from ms2query import MS2QueryDatabase
+from ms2query.library_io import create_new_library, load_created_library
+
+
+TEST_COMP_ID = "ZBSGKPYXQINNGF"   # expected InChIKey14 present in the test data
+EXPECTED_METADATA_SHAPE = (5, 11)
+EXPECTED_METADATA_FIELDS = [
+    "precursor_mz", "ionmode", "smiles", "inchikey", "inchi", "name",
+    "charge", "instrument_type", "adduct", "collision_energy",
+]
+
+
+def _data_dir() -> Path:
+    return Path(__file__).parent / "test_data"
+
+
+def _paths():
+    data_dir = _data_dir()
+    spectra_file = data_dir / "10_spectra.mgf"
+    model_path = data_dir / "ms2deepscore_testmodel_v1.pt"
+    assert spectra_file.exists(), f"Missing test spectra file: {spectra_file}"
+    assert model_path.exists(), f"Missing test model file: {model_path}"
+    return spectra_file, model_path
+
+
+@pytest.mark.filterwarnings("ignore::UserWarning")
+def test_create_and_load_library(tmp_path: Path):
+    spectra_file, model_path = _paths()
+
+    # ---------- Create ----------
+    outdir = tmp_path / "results"
+    outdir.mkdir(parents=True, exist_ok=True)
+
+    lib = create_new_library(
+        spectra_files=[str(spectra_file)],
+        annotation_files=[],                 # currently unused in the workflow
+        output_folder=str(outdir),
+        model_path=str(model_path),
+        # Keep the index small/fast for CI:
+        embedding_index_params={"M": 8, "ef_construction": 50, "post_init_ef": 50, "batch_rows": 100_000},
+    )
+    # basic sanity
+    assert lib is not None
+    assert isinstance(lib.db, MS2QueryDatabase)
+
+    # SQLite must exist
+    db_path = outdir / "ms2query_library.sqlite"
+    assert db_path.exists(), "MS2Query database file was not created."
+
+    # ---------- DB content checks ----------
+    ms2query_db = lib.db
+
+    # Metadata query by compound id (expected shape from your snippet)
+    df_meta = ms2query_db.metadata_by_comp_id(TEST_COMP_ID)
+    assert tuple(df_meta.shape) == EXPECTED_METADATA_SHAPE
+
+    # Metadata fields presence both in db wrapper and in returned dataframe
+    md_fields = ms2query_db.metadata_fields
+    for f in EXPECTED_METADATA_FIELDS:
+        assert f in md_fields, f"Metadata field '{f}' is missing in the database."
+        assert f in df_meta.columns, f"Metadata field '{f}' missing in metadata_by_comp_id result."
+
+    # ---------- Embedding index artifacts ----------
+    # The workflow saves with a base prefix (no extension). NMSLIB writes two files: <prefix> and <prefix>.dat
+    emb_prefix = outdir / "embedding_index"
+    pair_exists = emb_prefix.exists() and (emb_prefix.with_suffix(".dat")).exists()
+    assert pair_exists, "EmbeddingIndex files not found."
+
+    # ---------- Load ----------
+    lib2 = load_created_library(str(outdir))
+    assert lib2 is not None
+    assert isinstance(lib2.db, MS2QueryDatabase)
+
+    # ---------- Tiny embedding-index query ----------
+    # Take one spectrum from the same file and try a small top-k query
+    spectra = list(load_spectra(spectra_file))
+    assert len(spectra) > 0, "No spectra parsed from test file."
+
+    # queries should return non-empty hits with spec_ids
+    results = lib2.query_embedding_index(spectra[0], k=3, return_dataframe=False)
+    assert isinstance(results, list)
+    assert len(results) == 1
+
+    hits = results[0]
+    assert len(hits) == 3, "k was set to 3, but returned different number of hits."
+    assert set(hits[0].keys()) >= {"rank", "spec_id", "score"}
+    assert isinstance(hits[0]["spec_id"], (str, np.str_))