yaml peek to quickly find the value of a key

nwb_linkml/src/nwb_linkml/io/yaml.py

@@ -0,0 +1,63 @@
+"""
+Utility functions for dealing with yaml files.
+
+No we are not going to implement a yaml parser here
+"""
+import re
+from pathlib import Path
+from typing import Literal, List, Union, overload
+
+
+@overload
+def yaml_peek(key: str, path: Union[str, Path], root:bool = True, first:Literal[True]=True) -> str: ...
+
+@overload
+def yaml_peek(key: str, path: Union[str, Path], root:bool = True, first:Literal[False]=False) -> List[str]: ...
+
+@overload
+def yaml_peek(key: str, path: Union[str, Path], root:bool = True, first:bool=True) -> Union[str, List[str]]: ...
+
+def yaml_peek(key: str, path: Union[str, Path], root:bool = True, first:bool=True) -> Union[str, List[str]]:
+    """
+    Peek into a yaml file without parsing the whole file to retrieve the value of a single key.
+
+    This function is _not_ designed for robustness to the yaml spec, it is for simple key: value
+    pairs, not fancy shit like multiline strings, tagged values, etc. If you want it to be,
+    then i'm afraid you'll have to make a PR about it.
+
+    Returns a string no matter what the yaml type is so ya have to do your own casting if you want
+
+    Args:
+        key (str): The key to peek for
+        path (:class:`pathlib.Path` , str): The yaml file to peek into
+        root (bool): Only find keys at the root of the document (default ``True`` ), otherwise
+            find keys at any level of nesting.
+        first (bool): Only return the first appearance of the key (default). Otherwise return a
+            list of values (not implemented lol)
+
+    Returns:
+        str
+    """
+    if root:
+        pattern = re.compile(rf'^(?P<key>{key}):\s*(?P<value>\S.*)')
+    else:
+        pattern = re.compile(rf'^\s*(?P<key>{key}):\s*(?P<value>\S.*)')
+
+    res = None
+    if first:
+        with open(path, 'r') as yfile:
+            for l in yfile:
+                res = pattern.match(l)
+                if res:
+                    break
+        if res:
+            return res.groupdict()['value']
+    else:
+        with open(path, 'r') as yfile:
+            text = yfile.read()
+        res = [match.groupdict()['value'] for match in pattern.finditer(text)]
+        if res:
+            return res
+
+    raise KeyError(f'Key {key} not found in {path}')
+

nwb_linkml/src/nwb_linkml/providers/pydantic.py

@@ -13,6 +13,7 @@ from typing import List, Optional, Type
 from pydantic import BaseModel
 
 from nwb_linkml import io
+from nwb_linkml.io.yaml import yaml_peek
 from nwb_linkml.generators.pydantic import NWBPydanticGenerator
 from nwb_linkml.maps.naming import module_case, version_module_case
 from nwb_linkml.providers import LinkMLProvider, Provider
@@ -36,9 +37,6 @@ class PydanticProvider(Provider):
 
     def __init__(self, path: Optional[Path] = None, verbose: bool = True):
         super().__init__(path, verbose)
-        # create a metapathfinder to find module we might create
-        pathfinder = EctopicModelFinder(self.path)
-        sys.meta_path.append(pathfinder)
 
     @property
     def path(self) -> Path:
@@ -50,7 +48,6 @@ class PydanticProvider(Provider):
         namespace: str | Path,
         out_file: Optional[Path] = None,
         version: Optional[str] = None,
-        versions: Optional[dict] = None,
         split: bool = True,
         dump: bool = True,
         force: bool = False,
@@ -75,13 +72,6 @@ class PydanticProvider(Provider):
             version (Optional[str]): The version of the schema to build, if present.
                 Works similarly to ``version`` in :class:`.LinkMLProvider`.
                 Ignored if ``namespace`` is a Path.
-            versions (Optional[dict]): An explicit mapping of namespaces and versions to use when
-                building the combined pydantic `namespace.py` file.
-                Since NWB doesn't have an explicit version dependency system between schema,
-                there is intrinsic ambiguity between which version
-                of which schema should be used when imported from another.
-                This mapping allows those ambiguities to be resolved.
-                See :class:`.NWBPydanticGenerator` 's ``versions`` argument for more information.
             split (bool): If ``False`` (default), generate a single ``namespace.py`` file,
                 otherwise generate a python file for each schema in the namespace
                 in addition to a ``namespace.py`` that imports from them
@@ -107,19 +97,15 @@ class PydanticProvider(Provider):
             if version is None:
                 # Get the most recently built version
                 version = LinkMLProvider(path=self.config.cache_dir).available_versions[name][-1]
-            fn = path.parts[-1]
+            fn = path.name
         else:
             # given a path to a namespace linkml yaml file
             path = Path(namespace)
-            # FIXME: this is extremely fragile, but get the details from the path.
+            name = yaml_peek('name', path)
-            # this is faster than reading yaml for now
+            version = yaml_peek('version', path)
-            name = path.parts[-3]
+            fn = path.name
-            version = path.parts[-2]
-            fn = path.parts[-1]
 
         version = version_module_case(version)
-        # this is extremely fragile, we should not be inferring version number from paths...
-        # TODO: we need an efficient peek for specific keys within a yaml file
         if out_file is None:
             fn = fn.removesuffix(".yaml")
             fn = module_case(fn) + ".py"
@@ -137,10 +123,14 @@ class PydanticProvider(Provider):
         if versions is None:
             versions = self._get_dependent_versions(path)
 
+
         if split:
-            return self._build_split(path, versions, default_kwargs, dump, out_file, force)
+            result = self._build_split(path, versions, default_kwargs, dump, out_file, force)
         else:
-            return self._build_unsplit(path, versions, default_kwargs, dump, out_file, force)
+            result = self._build_unsplit(path, versions, default_kwargs, dump, out_file, force)
+
+        self.install_pathfinder()
+        return result
 
     def _build_unsplit(
         self,
@@ -406,6 +396,19 @@ class PydanticProvider(Provider):
         mod = self.get(namespace, version)
         return getattr(mod, class_)
 
+    def install_pathfinder(self):
+        """
+        Add a :class:`.EctopicModelFinder` instance that allows us to import from
+        the directory that we are generating models into
+        """
+        # check if one already exists
+        matches = [finder for finder in sys.meta_path if isinstance(finder, EctopicModelFinder) and finder.path == self.path]
+        if len(matches) > 0:
+            return
+
+        pathfinder = EctopicModelFinder(self.path)
+        sys.meta_path.append(pathfinder)
+
 
 class EctopicModelFinder(MetaPathFinder):
     """

nwb_linkml/tests/test_io/test_io_yaml.py

@@ -0,0 +1,44 @@
+import pytest
+import yaml
+
+from nwb_linkml.io.yaml import yaml_peek
+
+@pytest.fixture()
+def yaml_file(tmp_path):
+    data = {
+        'key1': 'val1',
+        'key2': 'val2',
+        'key3': {
+            'key1': 'val3',
+            'key4': 'val4'
+        }
+    }
+    out_file = tmp_path / 'test.yaml'
+    with open(out_file, 'w') as yfile:
+        yaml.dump(data, yfile)
+
+    yield out_file
+
+    out_file.unlink()
+
+
+
+@pytest.mark.parametrize(
+    'key,expected,root,first',
+    [
+        ('key1', 'val1', True, True),
+        ('key1', 'val1', False, True),
+        ('key1', ['val1'], True, False),
+        ('key1', ['val1', 'val3'], False, False),
+        ('key2', 'val2', True, True),
+        ('key3', False, True, True),
+        ('key4', False, True, True),
+        ('key4', 'val4', False, True)
+    ]
+)
+def test_peek_yaml(key, expected, root, first, yaml_file):
+    if not expected:
+        with pytest.raises(KeyError):
+            _ = yaml_peek(key, yaml_file, root=root, first=first)
+    else:
+        assert yaml_peek(key, yaml_file, root=root, first=first)