add tests

src/stdatamodels/jwst/_kwtool/_tests/conftest.py

@@ -0,0 +1,148 @@
+import json
+
+import pytest
+
+from stdatamodels.jwst._kwtool import dmd, kwd
+from stdatamodels.jwst.datamodels import ImageModel, JwstDataModel
+
+
+@pytest.fixture(scope="module")
+def fake_kwd_path(tmp_path_factory):
+    kwd_path = tmp_path_factory.mktemp("kwd")
+
+    # make 1 top file and 3 ref file
+    _top =  {
+        "type": "object",
+        "title": "root",
+        "properties": {
+            "meta": {
+                "title" : "FGS Keywords Schema Metadata",
+                "type" : "object",
+                "properties": {
+                    "program": {
+                        "title": "Programmatic information",
+                        "type": "object",
+                        "properties": {
+                            "$ref": "core.program.schema.json"
+                        },
+                    },
+                    "observation": {
+                        "title" : "Observation identifiers",
+                        "type" : "object",
+                        "properties" : {
+                            "allOf" : [
+                                {
+                                    "$ref" : "core.observation.schema.json",
+                                },
+                                {
+                                    "$ref" : "science.observation.schema.json",
+                                },
+                            ],
+                        }
+                    }
+                }
+            }
+        }
+    }
+    _core_program = {
+        "title" : {
+            "fits_keyword" : "TITLE",
+            "title" : "Proposal title",
+            "description" : "na",  # actual is longer
+            "type" : "string",
+            "units" : "",
+            "example" : "Cryo2 SIC NIRISS",
+            "default_value" : "UNKNOWN",
+            "source" : "Proposal and Planning System (PPS)",
+            "sw_source" : "PPS:dms_program_view.title",
+            "calculation" : "",
+            "destination" : ["ScienceCommon.title", "GuideStar.title"],
+            "sql_dtype" : "nvarchar(200)",
+            "si" : "Multiple",
+            "mode" : "All",
+            "level" : "1b",
+            "fits_hdu" : "PRIMARY",
+            "section" : "Programmatic information",
+            "misc" : "",
+            "comment_line" : "/ Program information"
+        },
+    }
+    _core_observation = {
+        "obs_id": {
+            "fits_keyword": "OBS_IDD",  # modified to trigger compare error
+            # modified title to trigger compare error
+            "title" : "Programmatic observation identifier modified",
+            "description": "na",  # actual is longer
+            "type" : "integer",  # modified to trigger compare error
+            "units" : "",
+            "example" : "V80600004001P0000000002101",
+            "default_value" : "",
+            "special_processing" : "VALUE_REQUIRED",
+            "source" : "Science Image Header",
+            "sw_source" : "",
+            "calculation" : "",
+            "destination" : ["ScienceCommon.obs_id","GuideStar.obs_id"],
+            "sql_dtype" : "nvarchar(26)",
+            "si" : "Multiple",
+            "mode" : "All",
+            "level" : "1a",
+            "fits_hdu" : "SCI",  # modified to trigger compare error
+            "section" : "Observation identifiers",
+            "misc" : ""
+        }
+    }
+    _science_observation = {
+        "engineering_quality" : {
+            "fits_keyword" : "ENG_QUAL",
+            "title" : "Engineering DB quality indicator",
+            "description" : "na",  # actual is longer
+            "type" : "string",
+            "enum" : ["OK", "SUSPECT", "NOT_IN_DATAMODEL"],  # modified to trigger compare error
+            "units" : "",
+            "example" : "OK",
+            "default_value" : "OK",
+            "source" : "Science Data Processing (SDP)",
+            "sw_source" : "",
+            "calculation" : "",
+            "destination" : ["ScienceCommon.eng_qual"],
+            "sql_dtype" : "nvarchar(15)",
+            "si" : "ALL",
+            "mode" : "All",
+            "level" : "1b",
+            "fits_hdu" : "PRIMARY",
+            "section" : "Observation identifiers",
+            "misc" : "",
+            "comment_line" : "/ Visit information"
+        },
+    }
+    for fn, schema in [
+        ("top.fgs.schema.json", _top),
+        ("core.program.schema.json", _core_program),
+        ("core.observation.schema.json", _core_observation),
+        ("science.observation.schema.json", _science_observation)
+    ]:
+        schema_path = kwd_path / fn
+        with open(schema_path, "w") as f:
+            json.dump(schema, f)
+    return kwd_path
+
+
+@pytest.fixture(scope="module")
+def fake_kwd(fake_kwd_path):
+    return kwd.load(fake_kwd_path)
+
+
+@pytest.fixture(scope="module")
+def limit_dmd_models():
+    # we can't use the monkeypatch fixture as it's function scoped
+    with pytest.MonkeyPatch.context() as mp:
+        mp.setattr(dmd, "_get_subclasses", lambda k: [ImageModel])
+        yield
+
+
+@pytest.fixture(scope="module")
+def fake_dmd(limit_dmd_models):
+    """
+    A fake datamodel dictionary that includes only 1 datamodel: ImageModel
+    """
+    return dmd.load()

src/stdatamodels/jwst/_kwtool/_tests/test_cli.py

@@ -0,0 +1,23 @@
+import os
+import sys
+
+from stdatamodels.jwst._kwtool import cli
+
+
+def test_cli(monkeypatch, tmp_path_factory, fake_kwd_path, limit_dmd_models):
+    # move to a new, empty directory
+    temp_cwd = tmp_path_factory.mktemp("cwd")
+    os.chdir(temp_cwd)
+
+    output_dir = tmp_path_factory.mktemp("output")
+    output_path = output_dir / "my_report.html"
+
+    monkeypatch.setattr(sys, "argv", [sys.argv[0], str(fake_kwd_path), "-o", str(output_path)])
+
+    cli._from_cmdline()
+
+    # make sure we didn't write out any local files
+    assert not os.listdir(temp_cwd)
+
+    # and that our output file exists
+    assert output_path.exists()

src/stdatamodels/jwst/_kwtool/_tests/test_compare.py

@@ -0,0 +1,135 @@
+import pytest
+
+from stdatamodels.jwst._kwtool import compare
+
+
+def test_filter_standard():
+    keywords = {
+        ("PRIMARY", "PCOUNT"): 1,
+        ("PRIMARY", "OBS_ID"): 2,
+    }
+    result = compare._filter_non_standard(keywords)
+    assert ("PRIMARY", "OBS_ID") in result
+    assert ("PRIMARY", "PCOUNT") not in result
+
+
+def test_filter_non_pattern():
+    keywords = {
+        ("PRIMARY", "P_BAND"): 1,
+        ("PRIMARY", "PBAND"): 2,
+        ("PRIMARY", "BAND"): 3,
+    }
+    result = compare._filter_non_pattern(keywords)
+    assert ("PRIMARY", "PBAND") in result
+    assert ("PRIMARY", "BAND") in result
+    assert ("PRIMARY", "P_BAND") not in result
+
+
+def test_compare_path_no_destination():
+    """
+    Ignore path differences for keywords that have
+    no destination.
+    """
+    k = [{
+        "keyword": {
+            # no destination
+            "title": "foo",
+        },
+        "path": "a.b.c".split("."),
+    }]
+    d = [{
+        "path": "d.e.f".split("."),
+    }]
+    assert compare._compare_path(k, d) is None
+
+
+def test_compare_path():
+    # make sure k has a desintation so the difference isn't ignored
+    k = [{"path": "a.b.c".split("."), "keyword": {"destination": "somewhere"}}]
+    d = [{"path": "d.e.f".split(".")}]
+    assert compare._compare_path(k, d) == {"kwd": {"a.b.c"}, "dmd": {"d.e.f"}}
+
+
+@pytest.mark.parametrize(
+    "kv, dv, expected", [
+        ("a", "a", None),
+        ("a", "z", {"kwd": {"a"}, "dmd": {"z"}}),
+        ("z", "a", {"kwd": {"z"}, "dmd": {"a"}}),
+        ("a", None, {"kwd": {"a"}, "dmd": {compare._MISSING_VALUE}}),
+        (None, "a", {"kwd": {compare._MISSING_VALUE}, "dmd": {"a"}}),
+        (None, None, None),
+    ]
+)
+def test_compare_keyword_subitem(kv, dv, expected):
+    # if kv/dv is None, don't fill title
+    if kv is None:
+        k = [{"keyword": {}}]
+    else:
+        k = [{"keyword": {"title": kv}}]
+    if dv is None:
+        d = [{"keyword": {}}]
+    else:
+        d = [{"keyword": {"title": dv}}]
+    assert compare._compare_keyword_subitem(k, d, "title") == expected
+
+
+@pytest.mark.parametrize(
+    "ktype, dtype, expected", [
+        ("string", "string", None),
+        ("integer", "string", {"kwd": {"integer"}, "dmd": {"string"}}),
+        ("string", "integer", {"kwd": {"string"}, "dmd": {"integer"}}),
+        ("float", "number", None),
+    ]
+)
+def test_compare_type(ktype, dtype, expected):
+    k = [{"keyword": {"type": ktype}}]
+    d = [{"keyword": {"type": dtype}}]
+    assert compare._compare_type(k, d) == expected
+
+
+def test_compare_enum():
+    # TODO
+    pass
+
+
+@pytest.fixture(scope="module")
+def compare_result(fake_kwd_path):
+    return compare.compare_keywords(fake_kwd_path)
+
+
+def test_obs_id(compare_result):
+    """
+    We constructed a fake keyword dictionary with OBS_IDD
+    and the datamodels have OBS_ID. Check this was found.
+    """
+    in_k, in_d, in_both, def_diff, k_keys, d_keys = compare_result
+    assert ("SCI", "OBS_IDD") in in_k
+    assert ("PRIMARY", "OBS_ID") in in_d
+
+
+def test_title(compare_result):
+    """
+    TITLE should be in both and agree
+    """
+    in_k, in_d, in_both, def_diff, k_keys, d_keys = compare_result
+    key = ("PRIMARY", "TITLE")
+    assert key in in_both
+    assert key not in def_diff
+
+
+def test_eng_qual(compare_result):
+    """
+    ENG_QUAL is in both but the enum differs.
+
+    There are other differences (due to our test setup) but we'll
+    ignore these.
+    """
+    in_k, in_d, in_both, def_diff, k_keys, d_keys = compare_result
+    key = ("PRIMARY", "ENG_QUAL")
+    assert key in in_both
+    assert key in def_diff
+    assert "enum" in def_diff[key]
+    assert def_diff[key]["enum"] == {
+        "dmd": {"OK", "SUSPECT"},
+        "kwd": {"OK", "SUSPECT", "NOT_IN_DATAMODEL"},
+    }

src/stdatamodels/jwst/_kwtool/_tests/test_dmd.py

@@ -0,0 +1,41 @@
+"""
+monkeypatch _get_subclasses to only return 1 datamodel (to make things easier)
+separately test _get_subclasses
+"""
+import pytest
+
+from stdatamodels.jwst.datamodels import JwstDataModel, FgsImgPhotomModel
+from stdatamodels.jwst._kwtool import dmd
+
+
+def test_get_subclasses():
+    # more of a smoke test
+    assert FgsImgPhotomModel in dmd._get_subclasses(JwstDataModel)
+
+
+@pytest.fixture(params=[("PRIMARY", "TITLE"), ("PRIMARY", "OBS_ID"), ("PRIMARY", "ENG_QUAL")])
+def keyword_list(request, fake_dmd):
+    """
+    There are many more keywords than the ones used here. These
+    were chosen to match the ones added in the fake keyword dictionary.
+    """
+    return fake_dmd[request.param]
+
+
+@pytest.fixture()
+def keyword(keyword_list):
+    """
+    The real datamodel dictionary will contain multiple entires for 1 "keyword"
+    (a FITS_HDU, FITS_KEYWORD pair). The test dictionary does not, this
+    fixture helps to test just the first found entry.
+    """
+    return keyword_list[0]
+
+
+def test_found(keyword_list):
+    """
+    This uses the keyword_list fixture to make sure that the test dictionary
+    only contains one entry per keyword pair (an assumption made by the
+    keyword fixture above).
+    """
+    assert len(keyword_list) == 1

src/stdatamodels/jwst/_kwtool/_tests/test_kwd.py

@@ -0,0 +1,37 @@
+"""
+Tests here use the fake_kwd fixture in conftest
+"""
+import pytest
+
+
+@pytest.fixture(params=[("PRIMARY", "TITLE"), ("SCI", "OBS_IDD"), ("PRIMARY", "ENG_QUAL")])
+def keyword_list(request, fake_kwd):
+    return fake_kwd[request.param]
+
+
+@pytest.fixture()
+def keyword(keyword_list):
+    """
+    The real keyword dictionary will contain multiple entires for 1 "keyword"
+    (a FITS_HDU, FITS_KEYWORD pair). The test dictionary does not, this
+    fixture helps to test just the first found entry.
+    """
+    return keyword_list[0]
+
+
+def test_fits_keyword_count(fake_kwd):
+    assert len(fake_kwd) == 3
+
+
+def test_found(keyword_list):
+    """
+    This uses the keyword_list fixture to make sure that the test dictionary
+    only contains one entry per keyword pair (an assumption made by the
+    keyword fixture above).
+    """
+    assert len(keyword_list) == 1
+
+
+@pytest.mark.parametrize("missing_key", ["allOf", "properties"])
+def test_path_sanitized(missing_key, keyword):
+    assert missing_key not in keyword["path"]

src/stdatamodels/jwst/_kwtool/compare.py

@@ -44,7 +44,7 @@ def _compare_path(k, d):
     # that have an archive destination TODO document this
     for i in k:
         if i['keyword'].get('destination'):
-            # since there is a destination, report the differenc
+            # since there is a destination, report the difference
             return paths
     return None