npm, maven: ignore blatantly invalid licenses

They cause noisy logs

swh/indexer/metadata_dictionary/maven.py

@@ -75,6 +75,9 @@ class MavenMapping(XmlMapping, SingleFileIntrinsicMapping):
             and isinstance(artifact_id, str)
         ):
             repo = os.path.join(url, *group_id.split("."), artifact_id)
+            if "${" in repo:
+                # Often use as templating in pom.xml files collected from VCSs
+                return
             graph.add((root, SCHEMA.codeRepository, URIRef(repo)))
 
     def normalize_groupId(self, id_):

swh/indexer/metadata_dictionary/npm.py

@@ -270,6 +270,12 @@ class NpmMapping(JsonMapping, SingleFileIntrinsicMapping):
         rdflib.term.URIRef('https://spdx.org/licenses/MIT')
         """
         if isinstance(s, str):
+            if s.startswith("SEE LICENSE IN "):
+                # Very common pattern, because it is an example in the specification.
+                # It is followed by the filename; and the indexer architecture currently
+                # does not allow accessing that from metadata mappings.
+                # (Plus, an hypothetical license mapping would eventually pick it up)
+                return
             return SPDX + s
 
     def normalize_keywords(self, lst):

swh/indexer/tests/metadata_dictionary/test_maven.py

@@ -353,6 +353,47 @@ def test_compute_metadata_maven_multiple():
     }
 
 
+def test_compute_metadata_maven_invalid_repository():
+    raw_content = b"""
+    <project>
+      <name>Maven Default Project</name>
+      <modelVersion>4.0.0</modelVersion>
+      <groupId>com.mycompany.app</groupId>
+      <artifactId>my-app</artifactId>
+      <version>1.2.3</version>
+      <repositories>
+        <repository>
+          <id>tcc-transaction-internal-releases</id>
+          <name>internal repository for released artifacts</name>
+          <url>${repo.internal.releases.url}</url>
+          <snapshots>
+              <enabled>false</enabled>
+          </snapshots>
+          <releases>
+              <enabled>true</enabled>
+          </releases>
+        </repository>
+      </repositories>
+      <licenses>
+        <license>
+          <name>Apache License, Version 2.0</name>
+          <url>https://www.apache.org/licenses/LICENSE-2.0.txt</url>
+          <distribution>repo</distribution>
+          <comments>A business-friendly OSS license</comments>
+        </license>
+      </licenses>
+    </project>"""
+    result = MAPPINGS["MavenMapping"]().translate(raw_content)
+    assert result == {
+        "@context": "https://doi.org/10.5063/schema/codemeta-2.0",
+        "type": "SoftwareSourceCode",
+        "name": "Maven Default Project",
+        "schema:identifier": "com.mycompany.app",
+        "version": "1.2.3",
+        "license": "https://www.apache.org/licenses/LICENSE-2.0.txt",
+    }
+
+
 @settings(suppress_health_check=[HealthCheck.too_slow])
 @given(
     xml_document_strategy(

swh/indexer/tests/metadata_dictionary/test_npm.py

@@ -361,6 +361,24 @@ def test_npm_invalid_uris():
     }
 
 
+def test_npm_invalid_licenses():
+    package_json = rb"""{
+  "version": "1.0.0",
+  "license": "SEE LICENSE IN LICENSE.md",
+  "author": {
+    "name": "foo",
+    "url": "http://example.org"
+  }
+}"""
+    result = MAPPINGS["NpmMapping"]().translate(package_json)
+    assert result == {
+        "@context": "https://doi.org/10.5063/schema/codemeta-2.0",
+        "type": "SoftwareSourceCode",
+        "author": [{"name": "foo", "type": "Person", "url": "http://example.org"}],
+        "version": "1.0.0",
+    }
+
+
 @settings(suppress_health_check=[HealthCheck.too_slow])
 @given(json_document_strategy(keys=list(MAPPINGS["NpmMapping"].mapping)))  # type: ignore
 def test_npm_adversarial(doc):