Merge branch 'develop' into main

1 files changed, 74 insertions, 0 deletions

diff --git a/bsie/extractor/generic/path.py b/bsie/extractor/generic/path.py
new file mode 100644
index 0000000..7018e12
--- /dev/null
+++ b/bsie/extractor/generic/path.py
@@ -0,0 +1,74 @@
+"""
+
+Part of the bsie module.
+A copy of the license is provided with the project.
+Author: Matthias Baumgartner, 2022
+"""
+# imports
+import os
+import typing
+
+# bsie imports
+from bsie.base import extractor
+from bsie.utils import bsfs, node, ns
+
+# exports
+__all__: typing.Sequence[str] = (
+    'Path',
+    )
+
+
+## code ##
+
+class Path(extractor.Extractor):
+    """Extract information from file's path."""
+
+    CONTENT_READER = 'bsie.reader.path.Path'
+
+    # mapping from predicate to handler function.
+    _callmap: typing.Dict[bsfs.schema.Predicate, typing.Callable[[str], typing.Any]]
+
+    def __init__(self):
+        super().__init__(bsfs.schema.Schema.from_string(extractor.SCHEMA_PREAMBLE + '''
+            bse:filename rdfs:subClassOf bsfs:Predicate ;
+                rdfs:domain bsfs:File ;
+                rdfs:range xsd:string ;
+                rdfs:label "File name"^^xsd:string ;
+                schema:description "Filename of entity in some filesystem."^^xsd:string ;
+                bsfs:unique "false"^^xsd:boolean .
+            '''))
+        self._callmap = {
+            self.schema.predicate(ns.bse.filename): self.__filename,
+            }
+
+    def extract(
+            self,
+            subject: node.Node,
+            content: str,
+            principals: typing.Iterable[bsfs.schema.Predicate],
+            ) -> typing.Iterator[typing.Tuple[node.Node, bsfs.schema.Predicate, typing.Any]]:
+        for pred in principals:
+            # find callback
+            clbk = self._callmap.get(pred)
+            if clbk is None:
+                continue
+            # get value
+            value = clbk(content)
+            if value is None:
+                continue
+            # produce triple
+            yield subject, pred, value
+
+    def __filename(self, path: str) -> typing.Optional[str]:
+        try:
+            return os.path.basename(path)
+        except Exception: # pylint: disable=broad-except # we explicitly want to catch everything
+            # some error, skip
+            # FIXME: some kind of error reporting (e.g. logging)?
+            # Options: (a) Fail silently (current); (b) Skip and report to log;
+            # (c) Raise ExtractorError (aborts extraction); (d) separate content type
+            # checks from basename errors (report content type errors, skip basename
+            # errors)
+            return None
+
+## EOF ##