Minor code cleanup #826
Minor code cleanup #826
Conversation
|
|
||
| /** | ||
| * Get a new TransformationFactory with the default classes | ||
| * @throws SecurityException if security exception occurred | ||
| */ | ||
| public TransformationFactory() { | ||
| mimeToTransform.put(contentTypeSPARQLQuery, new SparqlQueryTransform(null)); | ||
| mimeToTransform.put(APPLICATION_RDF_LDPATH, new LDPathTransform(null)); | ||
| mimeToTransform.put(contentTypeSPARQLQuery, is -> new SparqlQueryTransform(is)); |
There was a problem hiding this comment.
Same comment as in earlier PR:
mimeToTransform.put(contentTypeSPARQLQuery, SparqlQueryTransform::new);
| @@ -56,7 +59,7 @@ public TransformationFactory() { | |||
| public <T> Transformation<T> getTransform(final MediaType contentType, final InputStream inputStream) { | |||
| final String mimeType = contentType.toString(); | |||
| if (mimeToTransform.containsKey(mimeType)) { | |||
| return (Transformation<T>) mimeToTransform.get(contentType.toString()).newTransform(inputStream); | |||
| return mimeToTransform.get(contentType.toString()).apply(inputStream); | |||
There was a problem hiding this comment.
This seems incorrect to me. The previous logic returned a new transformation with the arg inputStream query. The new logic returns a new transformation trying to apply the arg inputStream as if it were an RdfStream.
There was a problem hiding this comment.
No, it doesn't. The tests wouldn't even pass if it did that. RdfStream has no relevance here. Look at the new type of mimeToTransform: Map<String, Function<InputStream, Transformation>>. Give it a String, it will give you a Function<InputStream, Transformation>. The old type let you give it a String and gave you a Transform which was only used to produce a new Transform via newTransform(InputStream) (which no longer exists). newTransform(InputStream) was semantically equivalent to Function<InputStream, Transformation>, but the new way is not idiosyncratic. The semantics are the same.
There was a problem hiding this comment.
What was perplexing me is the fact that Transformation has an apply(RdfStream) method that overrides Function::apply. However, there is another lambda function at play that has its own apply method.
I see the layers now.
There was a problem hiding this comment.
It's currying. The whole type is: String -> InputStream -> RdfStream -> Something (whatever the Transformation goes to).
|
Resolved with: 1a4cf67 |
https://jira.duraspace.org/browse/FCREPO-1606