Support id maps in preprocessing.

Author: Peter Amstutz

schema_salad/ref_resolver.py

@@ -171,6 +171,7 @@ def add_context(self, newcontext, baseuri=""):
         self.identity_links = set()
         self.standalone = set()
         self.nolinkcheck = set()
+        self.idmap = set()
         self.vocab = {}
         self.rvocab = {}
 
@@ -193,6 +194,9 @@ def add_context(self, newcontext, baseuri=""):
             if isinstance(self.ctx[c], dict) and self.ctx[c].get("noLinkCheck"):
                 self.nolinkcheck.add(c)
 
+            if isinstance(self.ctx[c], dict) and self.ctx[c].get("idMap"):
+                self.idmap.add(c)
+
             if isinstance(self.ctx[c], dict) and "@id" in self.ctx[c]:
                 self.vocab[c] = self.ctx[c]["@id"]
             elif isinstance(self.ctx[c], basestring):
@@ -330,6 +334,14 @@ def resolve_all(self, document, base_url, file_base=None):
                 metadata, _ = loader.resolve_all(metadata, base_url, file_base)
 
         if isinstance(document, dict):
+            for idmapField in loader.idmap:
+                if idmapField in document and isinstance(document[idmapField], dict):
+                    ls = []
+                    for k,v in document[idmapField].items():
+                        v["id"] = k
+                        ls.append(v)
+                    document[idmapField] = ls
+
             for identifer in loader.identity_links:
                 if identifer in document:
                     if isinstance(document[identifer], basestring):

setup.py

@@ -38,7 +38,7 @@
     install_requires.append("avro-python3")
 
 setup(name='schema-salad',
-      version='1.7',
+      version='1.8',
       description='Schema Annotations for Linked Avro Data (SALAD)',
       long_description=open(README).read(),
       author='Common workflow language working group',

tests/test_examples.py

@@ -66,6 +66,43 @@ def test_jsonld_ctx(self):
             'http://example.com/foo#bar': 'asym'
         })
 
+
+    def test_idmap(self):
+        ldr = schema_salad.ref_resolver.Loader({})
+        ldr.add_context({
+            "inputs": {
+                "@id": "http://example.com/inputs",
+                "idMap": True
+            },
+            "outputs": {
+                "@id": "http://example.com/outputs",
+                "idMap": True
+            },
+                         "id": "@id"})
+
+        ra, _ = ldr.resolve_all({
+            "inputs": {
+                "zip": {"a": 1},
+                "zing": {"a": 2}
+            },
+            "outputs": {
+                "out": {"b": 3},
+            },
+            "other": {
+                'n': 9
+            }
+        }, "http://example2.com/")
+
+        self.assertEqual(ra,
+                         {'inputs': [{'a': 2, 'id': 'http://example2.com/#zing'},
+                                     {'a': 1, 'id': 'http://example2.com/#zip'}],
+                          'outputs': [{'b': 3, 'id': 'http://example2.com/#out'}],
+                          'other': {
+                              'n': 9
+                          }
+                      }
+                     )
+
     def test_examples(self):
         self.maxDiff = None
         for a in ["field_name", "ident_res", "link_res", "vocab_res"]: