From 63fe1d017e2fad8181e3ff47185b974304957d56 Mon Sep 17 00:00:00 2001
From: Matthias Baumgartner <dev@igsor.net>
Date: Wed, 5 Apr 2023 17:16:14 +0200
Subject: IPTC tag extraction

---
 bsie/extractor/image/iptc.py | 70 ++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 70 insertions(+)
 create mode 100644 bsie/extractor/image/iptc.py

(limited to 'bsie/extractor')

diff --git a/bsie/extractor/image/iptc.py b/bsie/extractor/image/iptc.py
new file mode 100644
index 0000000..195eff7
--- /dev/null
+++ b/bsie/extractor/image/iptc.py
@@ -0,0 +1,70 @@
+
+# standard imports
+import typing
+
+# bsie imports
+from bsie.utils import bsfs, node, ns
+
+# inner-module imports
+from .. import base
+
+# exports
+__all__: typing.Sequence[str] = (
+    'Iptc',
+    )
+
+
+## code ##
+
+class Iptc(base.Extractor):
+    """Turn IPTC keywords into tags."""
+
+    CONTENT_READER = 'bsie.reader.exif.Iptc'
+
+    def __init__(self):
+        super().__init__(bsfs.schema.from_string(base.SCHEMA_PREAMBLE + '''
+            bsn:Tag rdfs:subClassOf bsfs:Node .
+
+            bse:tag rdfs:subClassOf bsfs:Predicate ;
+                rdfs:domain bsn:Entity ;
+                rdfs:range bsn:Tag .
+
+            <https://schema.bsfs.io/ie/Node/Tag#label> rdfs:subClassOf bsfs:Predicate ;
+                rdfs:domain bsn:Tag ;
+                rdfs:range xsd:string ;
+                bsfs:unique "true"^^xsd:boolean .
+
+            '''))
+        self._callmap = {
+            self.schema.predicate(ns.bse.tag): self._keywords,
+            }
+
+    def extract(
+            self,
+            subject: node.Node,
+            content: dict,
+            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
+            # produce triples
+            yield from clbk(subject, content)
+
+    def _keywords(
+            self,
+            subject: node.Node,
+            content: dict,
+            ) -> typing.Iterator[typing.Tuple[node.Node, bsfs.schema.Predicate, typing.Any]]:
+        if 'Iptc.Application2.Keywords' not in content:
+            return
+        for keyword in content['Iptc.Application2.Keywords']:
+            tag = node.Node(ns.bsn.Tag, label=keyword)
+            yield subject, self.schema.predicate(ns.bse.tag), tag
+            yield tag, self.schema.predicate(ns.bst.label), keyword
+
+
+
+## EOF ##
-- 
cgit v1.2.3


From ef50c1895c810d7a03bc139587a345273701cdbe Mon Sep 17 00:00:00 2001
From: Matthias Baumgartner <dev@igsor.net>
Date: Wed, 5 Apr 2023 17:18:02 +0200
Subject: minor fixes

---
 bsie/extractor/builder.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'bsie/extractor')

diff --git a/bsie/extractor/builder.py b/bsie/extractor/builder.py
index d691b0e..8353a93 100644
--- a/bsie/extractor/builder.py
+++ b/bsie/extractor/builder.py
@@ -67,6 +67,7 @@ class ExtractorBuilder():
             return cls(**kwargs)
 
         except Exception as err:
-            raise errors.BuilderError(f'failed to build extractor {name} due to {bsfs.typename(err)}: {err}') from err
+            raise errors.BuilderError(
+                f'failed to build extractor {name} due to {bsfs.typename(err)}: {err}') from err
 
 ## EOF ##
-- 
cgit v1.2.3