Merge pull request #1463 from joshmoore/python-graph

RFC: Add PythonInputSource to create py-based graphs

rdflib/parser.py

@@ -20,7 +20,6 @@
 
 from urllib.request import Request
 from urllib.request import url2pathname
-from urllib.parse import urljoin
 from urllib.request import urlopen
 from urllib.error import HTTPError
 
@@ -36,6 +35,7 @@
     "StringInputSource",
     "URLInputSource",
     "FileInputSource",
+    "PythonInputSource",
 ]
 
 
@@ -105,6 +105,45 @@ def close(self):
                 pass
 
 
+class PythonInputSource(InputSource):
+    """
+    Constructs an RDFLib Parser InputSource from a Python data structure,
+    for example, loaded from JSON with json.load or json.loads:
+
+    >>> import json
+    >>> as_string = \"\"\"{
+    ...   "@context" : {"ex" : "http://example.com/ns#"},
+    ...   "@graph": [{"@type": "ex:item", "@id": "#example"}]
+    ... }\"\"\"
+    >>> as_python = json.loads(as_string)
+    >>> source = create_input_source(data=as_python)
+    >>> isinstance(source, PythonInputSource)
+    True
+    """
+
+    def __init__(self, data, system_id=None):
+        self.content_type = None
+        self.auto_close = False  # see Graph.parse(), true if opened by us
+        self.public_id = None
+        self.system_id = system_id
+        self.data = data
+
+    def getPublicId(self):
+        return self.public_id
+
+    def setPublicId(self, public_id):
+        self.public_id = public_id
+
+    def getSystemId(self):
+        return self.system_id
+
+    def setSystemId(self, system_id):
+        self.system_id = system_id
+
+    def close(self):
+        self.data = None
+
+
 class StringInputSource(InputSource):
     """
     Constructs an RDFLib Parser InputSource from a Python String or Bytes
@@ -295,10 +334,15 @@ def create_input_source(
         input_source = FileInputSource(file)
 
     if data is not None:
-        if not isinstance(data, (str, bytes, bytearray)):
-            raise RuntimeError("parse data can only str, or bytes.")
-        input_source = StringInputSource(data)
-        auto_close = True
+        if isinstance(data, dict):
+            input_source = PythonInputSource(data)
+            auto_close = True
+        elif isinstance(data, (str, bytes, bytearray)):
+            input_source = StringInputSource(data)
+            auto_close = True
+        else:
+            raise RuntimeError(
+                f"parse data can only str, or bytes. not: {type(data)}")
 
     if input_source is None:
         raise Exception("could not create InputSource")

rdflib/plugins/shared/jsonld/util.py

@@ -17,12 +17,16 @@
 
 from urllib.parse import urljoin, urlsplit, urlunsplit
 
-from rdflib.parser import create_input_source
+from rdflib.parser import create_input_source, PythonInputSource
 
 from io import StringIO
 
 
 def source_to_json(source):
+
+    if isinstance(source, PythonInputSource):
+        return source.data
+
     # TODO: conneg for JSON (fix support in rdflib's URLInputSource!)
     source = create_input_source(source, format="json-ld")
 

test/jsonld/test_pythonparse.py

@@ -0,0 +1,76 @@
+from rdflib import Graph
+from rdflib.compare import isomorphic
+import json
+
+
+def test_wrap():
+    """
+    Example of intercepting a JSON-LD structure and performing some
+    in-memory manipulation and then passing that structure to Graph.parse
+    lists in the shacl graph.
+    """
+
+    _data = """
+    {
+        "@context" : {
+            "ngff" : "http://example.com/ns#"
+        },
+        "@graph": [{
+            "@type": "ngff:ItemList",
+            "ngff:collectionType": {"@type": "ngff:Image"},
+            "ngff:itemListElement": [
+                {
+                    "@type": "ngff:Image",
+                    "path": "image1",
+                    "name": "Image 1"
+                },
+                {
+                    "@type": "ngff:Image",
+                    "path": "something-else",
+                    "name": "bob"
+                }
+            ]
+        }]
+    }
+    """
+
+    # Current workaround
+    data = json.loads(_data)
+    data = walk(data)
+    data = json.dumps(data)  # wasteful
+    g1 = Graph()
+    g1.parse(data=data, format="json-ld")
+
+    # Desired behavior
+    data = json.loads(_data)
+    data = walk(data)
+    g2 = Graph()
+    g2.parse(data=data, format="json-ld")
+
+    assert isomorphic(g1, g2)
+
+
+def walk(data, path=None):
+    """
+    Some arbitrary operation on a Python data structure.
+    """
+
+    if path is None:
+        path = []
+
+    if isinstance(data, dict):
+        for k, v in data.items():
+            data[k] = walk(v, path + [k])
+
+    elif isinstance(data, list):
+        replacement = list()
+        for idx, item in enumerate(data):
+            if path[-1] == "@graph":
+                replacement.append(walk(item, path))
+            else:
+                wrapper = {"@type": "ListItemWrapper", "ngff:position": idx}
+                wrapper["ngff:item"] = walk(item, path + [idx])
+                replacement.append(wrapper)
+        data = replacement
+
+    return data