Merge pull request #56 from sldouglas-nist/add_pdf_file_mapping

Add PDF File Mapping

case_exiftool/__init__.py

@@ -41,6 +41,7 @@
 
 _logger = logging.getLogger(os.path.basename(__file__))
 
+NS_DRAFTING = rdflib.Namespace("http://example.org/ontology/drafting/")
 NS_EXIFTOOL_COMPOSITE = rdflib.Namespace("http://ns.exiftool.org/Composite/1.0/")
 NS_EXIFTOOL_ET = rdflib.Namespace("http://ns.exiftool.org/1.0/")
 NS_EXIFTOOL_EXIFTOOL = rdflib.Namespace("http://ns.exiftool.org/ExifTool/1.0/")
@@ -50,6 +51,7 @@
 NS_EXIFTOOL_IFD0 = rdflib.Namespace("http://ns.exiftool.org/EXIF/IFD0/1.0/")
 NS_EXIFTOOL_EXIFIFD = rdflib.Namespace("http://ns.exiftool.org/EXIF/ExifIFD/1.0/")
 NS_EXIFTOOL_NIKON = rdflib.Namespace("http://ns.exiftool.org/MakerNotes/Nikon/1.0/")
+NS_EXIFTOOL_PDF_PDF = rdflib.Namespace("http://ns.exiftool.org/PDF/PDF/1.0/")
 NS_EXIFTOOL_PREVIEWIFD = rdflib.Namespace(
     "http://ns.exiftool.org/MakerNotes/PreviewIFD/1.0/"
 )
@@ -175,12 +177,13 @@ def __init__(
         self._n_exif_facet: typing.Optional[rdflib.URIRef] = None
         self._n_file_facet: typing.Optional[rdflib.URIRef] = None
         self._n_location_object: typing.Optional[rdflib.URIRef] = None
+        self._n_pdf_file_facet: typing.Optional[rdflib.URIRef] = None
         self._n_location_object_latlong_facet: typing.Optional[rdflib.URIRef] = None
         self._n_observable_object: typing.Optional[rdflib.URIRef] = None
         self._n_raster_picture_facet: typing.Optional[rdflib.URIRef] = None
         self._n_relationship_object_location: typing.Optional[rdflib.URIRef] = None
         self._n_unix_file_permissions_facet: typing.Optional[rdflib.URIRef] = None
-        self._oo_slug: typing.Optional[str] = None
+        self._oo_slug: str = "File-"
         self.ns_base = ns_base
 
     def map_raw_and_printconv_iri(self, n_exiftool_predicate: rdflib.URIRef) -> None:
@@ -369,6 +372,110 @@ def map_raw_and_printconv_iri(self, n_exiftool_predicate: rdflib.URIRef) -> None
                         rdflib.Literal(int(v_raw.toPython())),
                     )
                 )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/Author":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfAuthor,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/CreateDate":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_UCO_OBSERVABLE.pdfCreationDate,
+                        rdflib.Literal(
+                            v_raw.toPython().replace(" ", "T"), datatype=NS_XSD.dateTime
+                        ),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/Creator":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfCreator,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/Linearized":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfLinearized,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/ModifyDate":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_UCO_OBSERVABLE.pdfModDate,
+                        rdflib.Literal(
+                            v_raw.toPython().replace(" ", "T"), datatype=NS_XSD.dateTime
+                        ),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/PDFVersion":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_UCO_OBSERVABLE.version,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/PageCount":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfPageCount,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/Producer":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfProducer,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/Subject":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfSubject,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
+        elif exiftool_iri == "http://ns.exiftool.org/PDF/PDF/1.0/Title":
+            (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
+            if isinstance(v_raw, rdflib.Literal):
+                self.graph.add(
+                    (
+                        self.n_pdf_file_facet,
+                        NS_DRAFTING.pdfTitle,
+                        rdflib.Literal(v_raw.toPython()),
+                    )
+                )
         else:
             # Somewhat in the name of information preservation, somewhat as a progress marker on converting data: Attach all remaining unconverted properties directly to the ObservableObject.  Provide both values to assist with mapping decisions.
             (v_raw, v_printconv) = self.pop_n_exiftool_predicate(n_exiftool_predicate)
@@ -423,16 +530,6 @@ def _load_xml_file_into_dict(
             rdflib.URIRef("http://ns.exiftool.org/File/1.0/MIMEType")
         )
 
-        # Determine slug by MIME type.
-        self.oo_slug = "File-"  # The prefix "oo_" means generic observable object.
-        if self.mime_type == "image/jpeg":
-            self.oo_slug = "Picture-"
-        else:
-            _logger.warning("TODO - MIME type %r not yet implemented." % self.mime_type)
-
-        # Access observable object to instantiate it with the oo_slug value.
-        _ = self.n_observable_object
-
         # Finish special case MIME type processing left undone by map_raw_and_printconv_iri.
         if self.mime_type is not None:
             self.graph.add(
@@ -514,6 +611,19 @@ def mime_type(self) -> typing.Optional[str]:
     def mime_type(self, value: str) -> None:
         assert isinstance(value, str)
         self._mime_type = value
+        if value == "application/pdf":
+            self.graph.add(
+                (self.n_observable_object, NS_RDF.type, NS_UCO_OBSERVABLE.PDFFile)
+            )
+        elif self.mime_type == "image/jpeg":
+            self.graph.add(
+                (self.n_observable_object, NS_RDF.type, NS_UCO_OBSERVABLE.RasterPicture)
+            )
+
+        else:
+            _logger.warning("TODO - MIME type %r not yet implemented." % self.mime_type)
+
+        # Access observable object to instantiate it with the oo_slug value.
 
     @property
     def n_camera_object(self) -> rdflib.URIRef:
@@ -851,6 +961,28 @@ def oo_slug(self, value: str) -> None:
         assert isinstance(value, str)
         self._oo_slug = value
 
+    @property
+    def n_pdf_file_facet(self) -> rdflib.URIRef:
+        """
+        Initialized on first access.
+        """
+        if self._n_pdf_file_facet is None:
+            if self.use_deterministic_uuids:
+                self._n_pdf_file_facet = case_utils.inherent_uuid.get_facet_uriref(
+                    self.n_observable_object,
+                    NS_UCO_OBSERVABLE.PDFFileFacet,
+                    namespace=self.ns_base,
+                )
+            else:
+                self._n_pdf_file_facet = self.ns_base["PDFFileFacet-" + local_uuid()]
+            self.graph.add(
+                (self._n_pdf_file_facet, NS_RDF.type, NS_UCO_OBSERVABLE.PDFFileFacet)
+            )
+            self.graph.add(
+                (self.n_observable_object, NS_UCO_CORE.hasFacet, self._n_pdf_file_facet)
+            )
+        return self._n_pdf_file_facet
+
     @property
     def use_deterministic_uuids(self) -> bool:
         """
@@ -868,10 +1000,12 @@ def main() -> None:
     NS_BASE = rdflib.Namespace(args.base_prefix)
     out_graph = rdflib.Graph()
 
+    out_graph.namespace_manager.bind("drafting", NS_DRAFTING)
     out_graph.namespace_manager.bind("exiftool-Composite", NS_EXIFTOOL_COMPOSITE)
     out_graph.namespace_manager.bind("exiftool-et", NS_EXIFTOOL_ET)
     out_graph.namespace_manager.bind("exiftool-ExifTool", NS_EXIFTOOL_EXIFTOOL)
     out_graph.namespace_manager.bind("exiftool-System", NS_EXIFTOOL_SYSTEM)
+    out_graph.namespace_manager.bind("exiftool-PDF-PDF", NS_EXIFTOOL_PDF_PDF)
     out_graph.namespace_manager.bind("exiftool-File", NS_EXIFTOOL_FILE)
     out_graph.namespace_manager.bind("exiftool-GPS", NS_EXIFTOOL_GPS)
     out_graph.namespace_manager.bind("exiftool-IFD0", NS_EXIFTOOL_IFD0)

tests/govdocs1/Makefile

@@ -1,13 +1,16 @@
 #!/usr/bin/make -f
 
+# Portions of this file contributed by NIST are governed by the
+# following statement:
+#
 # This software was developed at the National Institute of Standards
 # and Technology by employees of the Federal Government in the course
-# of their official duties. Pursuant to title 17 Section 105 of the
-# United States Code this software is not subject to copyright
-# protection and is in the public domain. NIST assumes no
-# responsibility whatsoever for its use by other parties, and makes
-# no guarantees, expressed or implied, about its quality,
-# reliability, or any other characteristic.
+# of their official duties. Pursuant to Title 17 Section 105 of the
+# United States Code, this software is not subject to copyright
+# protection within the United States. NIST assumes no responsibility
+# whatsoever for its use by other parties, and makes no guarantees,
+# expressed or implied, about its quality, reliability, or any other
+# characteristic.
 #
 # We would appreciate acknowledgement if the software is used.
 
@@ -16,13 +19,21 @@ SHELL := /bin/bash
 all:
 	$(MAKE) \
 	  --directory files/799/987
+	$(MAKE) \
+	  --directory files/000/015
 
 check:
 	$(MAKE) \
 	  --directory files/799/987 \
 	  check
+	$(MAKE) \
+	  --directory files/000/015 \
+	  check
 
 clean:
 	@$(MAKE) \
 	  --directory files/799/987 \
 	  clean
+	@$(MAKE) \
+	  --directory files/000/015 \
+	  clean

tests/govdocs1/files/000/015/000015_printConv.xml

@@ -0,0 +1,32 @@
+<?xml version='1.0' encoding='UTF-8'?>
+<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'>
+
+<rdf:Description rdf:about='http://example.org/kb/govdocs1/000015.pdf'
+  xmlns:et='http://ns.exiftool.org/1.0/' et:toolkit='Image::ExifTool 12.76'
+  xmlns:ExifTool='http://ns.exiftool.org/ExifTool/1.0/'
+  xmlns:System='http://ns.exiftool.org/File/System/1.0/'
+  xmlns:File='http://ns.exiftool.org/File/1.0/'
+  xmlns:PDF='http://ns.exiftool.org/PDF/PDF/1.0/'>
+ <ExifTool:ExifToolVersion>12.76</ExifTool:ExifToolVersion>
+ <System:FileName>000015.pdf</System:FileName>
+ <System:Directory>.</System:Directory>
+ <System:FileSize>56 kB</System:FileSize>
+ <System:FileModifyDate>2024:06:13 17:32:40-04:00</System:FileModifyDate>
+ <System:FileAccessDate>2024:06:13 17:34:31-04:00</System:FileAccessDate>
+ <System:FileInodeChangeDate>2024:06:13 17:32:40-04:00</System:FileInodeChangeDate>
+ <System:FilePermissions>-rw-r--r--</System:FilePermissions>
+ <File:FileType>PDF</File:FileType>
+ <File:FileTypeExtension>pdf</File:FileTypeExtension>
+ <File:MIMEType>application/pdf</File:MIMEType>
+ <PDF:PDFVersion>1.5</PDF:PDFVersion>
+ <PDF:Linearized>No</PDF:Linearized>
+ <PDF:ModifyDate>2005:05:25 02:23:55</PDF:ModifyDate>
+ <PDF:Title>Document</PDF:Title>
+ <PDF:Creator>ACOMP.exe   WinVer 1b43  jul 14 2003</PDF:Creator>
+ <PDF:Producer>Acrobat Distiller 4.0 for Windows</PDF:Producer>
+ <PDF:CreateDate>2005:05:25 02:23:55</PDF:CreateDate>
+ <PDF:Author>U.S. Government Printing Office</PDF:Author>
+ <PDF:Subject>Extracted Pages</PDF:Subject>
+ <PDF:PageCount>2</PDF:PageCount>
+</rdf:Description>
+</rdf:RDF>

tests/govdocs1/files/000/015/000015_raw.xml

@@ -0,0 +1,32 @@
+<?xml version='1.0' encoding='UTF-8'?>
+<rdf:RDF xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'>
+
+<rdf:Description rdf:about='http://example.org/kb/govdocs1/000015.pdf'
+  xmlns:et='http://ns.exiftool.org/1.0/' et:toolkit='Image::ExifTool 12.76'
+  xmlns:ExifTool='http://ns.exiftool.org/ExifTool/1.0/'
+  xmlns:System='http://ns.exiftool.org/File/System/1.0/'
+  xmlns:File='http://ns.exiftool.org/File/1.0/'
+  xmlns:PDF='http://ns.exiftool.org/PDF/PDF/1.0/'>
+ <ExifTool:ExifToolVersion>12.76</ExifTool:ExifToolVersion>
+ <System:FileName>000015.pdf</System:FileName>
+ <System:Directory>.</System:Directory>
+ <System:FileSize>55964</System:FileSize>
+ <System:FileModifyDate>2024:06:13 17:32:40-04:00</System:FileModifyDate>
+ <System:FileAccessDate>2024:06:13 17:34:31-04:00</System:FileAccessDate>
+ <System:FileInodeChangeDate>2024:06:13 17:32:40-04:00</System:FileInodeChangeDate>
+ <System:FilePermissions>100644</System:FilePermissions>
+ <File:FileType>PDF</File:FileType>
+ <File:FileTypeExtension>PDF</File:FileTypeExtension>
+ <File:MIMEType>application/pdf</File:MIMEType>
+ <PDF:PDFVersion>1.5</PDF:PDFVersion>
+ <PDF:Linearized>false</PDF:Linearized>
+ <PDF:ModifyDate>2005:05:25 02:23:55</PDF:ModifyDate>
+ <PDF:Title>Document</PDF:Title>
+ <PDF:Creator>ACOMP.exe   WinVer 1b43  jul 14 2003</PDF:Creator>
+ <PDF:Producer>Acrobat Distiller 4.0 for Windows</PDF:Producer>
+ <PDF:CreateDate>2005:05:25 02:23:55</PDF:CreateDate>
+ <PDF:Author>U.S. Government Printing Office</PDF:Author>
+ <PDF:Subject>Extracted Pages</PDF:Subject>
+ <PDF:PageCount>2</PDF:PageCount>
+</rdf:Description>
+</rdf:RDF>