aboutsummaryrefslogtreecommitdiffstats
path: root/bsie/extractor/generic/stat.py
diff options
context:
space:
mode:
authorMatthias Baumgartner <dev@igsor.net>2022-10-31 14:14:57 +0100
committerMatthias Baumgartner <dev@igsor.net>2022-10-31 14:14:57 +0100
commit9389c741bdbbca9adbff6099d440706cd63deac4 (patch)
tree48ee0e912e2f19f51bd684d790f0bcc2d906e887 /bsie/extractor/generic/stat.py
parentd4023fa972af379a4235f51783954671de974372 (diff)
parent2da348c638ac5058d5acf09ab5df323ee04503d5 (diff)
downloadbsie-9389c741bdbbca9adbff6099d440706cd63deac4.tar.gz
bsie-9389c741bdbbca9adbff6099d440706cd63deac4.tar.bz2
bsie-9389c741bdbbca9adbff6099d440706cd63deac4.zip
Merge branch 'mb/extractors' into develop
Diffstat (limited to 'bsie/extractor/generic/stat.py')
-rw-r--r--bsie/extractor/generic/stat.py71
1 files changed, 71 insertions, 0 deletions
diff --git a/bsie/extractor/generic/stat.py b/bsie/extractor/generic/stat.py
new file mode 100644
index 0000000..d74369c
--- /dev/null
+++ b/bsie/extractor/generic/stat.py
@@ -0,0 +1,71 @@
+"""Extract information from the file system, such as filesize.
+
+Part of the bsie module.
+A copy of the license is provided with the project.
+Author: Matthias Baumgartner, 2022
+"""
+# imports
+import typing
+
+# inner-module imports
+from bsie.base import extractor
+from bsie.utils import node, ns
+from bsie.utils.bsfs import URI
+import bsie.reader.stat
+
+
+# exports
+__all__: typing.Sequence[str] = (
+ 'Stat',
+ )
+
+
+## code ##
+
+class Stat(extractor.Extractor):
+ """Extract information from the file system."""
+
+ CONTENT_READER = bsie.reader.stat.Stat
+
+ def __init__(self):
+ self.__callmap = {
+ ns.bse.filesize: self.__filesize,
+ }
+
+ def schema(self) -> str:
+ return '''
+ bse:filesize a bsfs:Predicate ;
+ rdfs:domain bsfs:Entity ;
+ rdfs:range xsd:integer ;
+ rdf:label "File size"^^xsd:string ;
+ schema:description "File size of entity in some filesystem."^^xsd:string ;
+ owl:maxCardinality "INF"^^xsd:number .
+ '''
+
+ def extract(
+ self,
+ subject: node.Node,
+ content: CONTENT_READER.CONTENT_TYPE,
+ predicates: typing.Iterable[URI],
+ ) -> typing.Iterator[typing.Tuple[node.Node, URI, typing.Any]]:
+ for pred in predicates:
+ # 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 __filesize(self, content: CONTENT_READER.CONTENT_TYPE) -> int:
+ """Return the file size."""
+ try:
+ return content.st_size
+ except Exception:
+ # FIXME: some kind of error reporting (e.g. logging)
+ return None
+
+## EOF ##