aboutsummaryrefslogtreecommitdiffstats
path: root/bsie/extractor/generic/stat.py
diff options
context:
space:
mode:
authorMatthias Baumgartner <dev@igsor.net>2022-12-18 14:22:31 +0100
committerMatthias Baumgartner <dev@igsor.net>2022-12-18 14:22:31 +0100
commit7582c280ad5324a2f0427999911c7e7abc14a6ab (patch)
tree0a59bbfe1c44d3497daad9f25ff9e7eb2bf9eb82 /bsie/extractor/generic/stat.py
parentcb49e4567a18de6851286ff672e54f9a91865fe9 (diff)
parent057e09d6537bf5c39815661a75819081e3e5fda7 (diff)
downloadbsie-7582c280ad5324a2f0427999911c7e7abc14a6ab.tar.gz
bsie-7582c280ad5324a2f0427999911c7e7abc14a6ab.tar.bz2
bsie-7582c280ad5324a2f0427999911c7e7abc14a6ab.zip
Merge branch 'develop' into main
Diffstat (limited to 'bsie/extractor/generic/stat.py')
-rw-r--r--bsie/extractor/generic/stat.py70
1 files changed, 70 insertions, 0 deletions
diff --git a/bsie/extractor/generic/stat.py b/bsie/extractor/generic/stat.py
new file mode 100644
index 0000000..0b9ce29
--- /dev/null
+++ b/bsie/extractor/generic/stat.py
@@ -0,0 +1,70 @@
+"""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 os
+import typing
+
+# bsie imports
+from bsie.base import extractor
+from bsie.utils import bsfs, node, ns
+
+# exports
+__all__: typing.Sequence[str] = (
+ 'Stat',
+ )
+
+
+## code ##
+
+class Stat(extractor.Extractor):
+ """Extract information from the file system."""
+
+ CONTENT_READER = 'bsie.reader.stat.Stat'
+
+ # mapping from predicate to handler function.
+ _callmap: typing.Dict[bsfs.schema.Predicate, typing.Callable[[os.stat_result], typing.Any]]
+
+ def __init__(self):
+ super().__init__(bsfs.schema.Schema.from_string(extractor.SCHEMA_PREAMBLE + '''
+ bse:filesize rdfs:subClassOf bsfs:Predicate ;
+ rdfs:domain bsfs:File ;
+ rdfs:range xsd:integer ;
+ rdfs:label "File size"^^xsd:string ;
+ schema:description "File size of entity in some filesystem."^^xsd:string ;
+ bsfs:unique "false"^^xsd:boolean .
+ '''))
+ self._callmap = {
+ self.schema.predicate(ns.bse.filesize): self.__filesize,
+ }
+
+ def extract(
+ self,
+ subject: node.Node,
+ content: os.stat_result,
+ 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 __filesize(self, content: os.stat_result) -> typing.Optional[int]:
+ """Return the file size."""
+ try:
+ return content.st_size
+ except Exception: # pylint: disable=broad-except # we explicitly want to catch everything
+ # FIXME: some kind of error reporting (e.g. logging)
+ return None
+
+## EOF ##