made documents flake8 compliant and added tests

Author: notactuallyfinn

src/hermes_toml/harvest.py

@@ -90,10 +90,9 @@ def read_from_toml(cls, file, data):
 
         # load the toml file as a dictionary
         try:
-            if not isinstance(toml_data := toml.load(file), dict):
-                return
-        except Exception as exc:
-            raise type(exc)(f"Something went wrong while reading the given file {file}") from exc
+            toml_data = toml.load(file)
+        except Exception:
+            return
 
         # harvest project table
         project_data = toml_data.get("project")
@@ -373,12 +372,11 @@ def handle_pypi_classifieres(cls, classifiers: str | list[str], data):
 
         if not isinstance(classifiers, (str, list)):
             return
-        if isinstance(classifiers, str):
+        if isinstance(classifiers, str) and len(classifiers.split(" :: ")) > 1:
             classifiers = [classifiers]
         else:
-            classifiers = [classifier for classifier in classifiers if isinstance(classifier, str)]
-            if len(classifiers) == 0:
-                return
+            # remove non strings and to short classifiers
+            classifiers = [clsf for clsf in classifiers if isinstance(clsf, str) and len(clsf.split(" :: ")) > 1]
 
         # remove duplicates
         classifiers = list(set(classifiers))
@@ -390,15 +388,13 @@ def handle_pypi_classifieres(cls, classifiers: str | list[str], data):
         # iterate over all classifiers and put them into the correct buckets
         for classifier in classifiers:
             classifier = classifier.split(" :: ")
-            if len(classifier) < 2:
-                continue
-            if (classifier[0] == "Operating System" and not (len(classifier) == 2 and classifier[1] == "Microsoft")):
+            if classifier[0] == "Operating System" and not (len(classifier) == 2 and classifier[1] == "Microsoft"):
                 temp = {"@type": "schema:SoftwareApplication", "schema:name": classifier[-1]}
                 sorted_classifiers["schema:targetProduct"].append(temp)
             elif classifier[0] == "Intended Audience":
                 temp = {"@type": "schema:Audience", "schema:name": classifier[-1]}
                 sorted_classifiers["schema:audience"].append(temp)
-            elif (classifier[0] == "License" and not (classifier[1] == "OSI Approved" and len(classifier) == 2)):
+            elif classifier[0] == "License" and not (classifier[1] == "OSI Approved" and len(classifier) == 2):
                 temp = {"@type": "schema:CreativeWork", "schema:name": classifier[-1]}
                 sorted_classifiers["schema:license"].append(temp)
             elif classifier[0] == "Natural Language":
@@ -421,10 +417,9 @@ def handle_pypi_classifieres(cls, classifiers: str | list[str], data):
 
         # add everything to the SoftwareMetadata object
         for key, value in sorted_classifiers.items():
-            if len(value) > 1:
-                data[key] = value
-            elif len(value) == 1:
-                data[key] = value[0]
+            if len(value) == 0:
+                continue
+            data[key] = value if len(value) > 1 else value[0]
 
     @classmethod
     def handle_urls(cls, urls: dict[str, str], data):

test/hermes_toml_test/test_harvest.py

@@ -5,6 +5,7 @@
 # SPDX-FileContributor: Michael Meinel
 # SPDX-FileContributor: Michael Fritzsche
 
+import os
 import pytest
 from pytest_unordered import unordered
 import toml
@@ -279,6 +280,7 @@ def toml_file(tmp_path_factory):
 
 
 @pytest.mark.parametrize("in_data, out_data", [
+    ("", {}),
     ({"project": {"name": "a", "version": "x.x.x", "description": "abc", "keywords": ["a", "b", "c"],
                   "authors": [{"name": "ab", "email": "ab@ab.ab"}, {"name": "a", "email": "a@a.a"}],
                   "maintainers": [{"name": "ab", "email": "ab@ab.ab"}, {"name": "a", "email": "a@a.a"}],
@@ -328,12 +330,32 @@ def test_read_from_toml_success(in_data, out_data, toml_file):
     assert data == out_data
 
 
+@pytest.fixture(scope="session")
+def txt_file(tmp_path_factory):
+    fn = tmp_path_factory.mktemp("data") / "test.txt"
+    return fn
+
+
+def test_read_from_toml_error(txt_file):
+    pre = os.path.splitext(txt_file)[0]
+    toml_file = pre + ".toml"
+    with open(txt_file, "w") as temp:
+        temp.write("a")
+    # try non toml file
+    TomlHarvestPlugin.read_from_toml(txt_file, data := {})
+    assert data == {}
+    # try non existing file
+    TomlHarvestPlugin.read_from_toml(toml_file, data := {})
+    assert data == {}
+    os.rename(txt_file, toml_file)
+    # try malformed file
+    TomlHarvestPlugin.read_from_toml(toml_file, data := {})
+    assert data == {}
+
+
 """
-# create different types of invalid files and save them somehow then run read_from_toml and validate the raised errors
-@pytest.mark.parametrize("in_data", [
-])
-def test_read_from_toml_error(in_data, out_data, toml_file):
+test files with multiple tables e.g. project and poetry or with different data in fileds with the same meaning
 
-# test files with multiple tables e.g. project and poetry or with different data in fileds with the same meaning
-# not tested because it is unknown how these cases should be treated
+not tested because it is unknown how these cases should be treated
+as the SoftwareMetadata object is not yet implemented nor its behaivior documented
 """