Merge pull request #44 from aarhusstadsarkiv/aca-fmt-23

aca-fmt/23 - Microsoft Word XML Document

custom_signatures.yml

@@ -112,3 +112,10 @@
   puid: aca-fmt/22
   signature: Windows Compressed Enhanced Metafile, usually image file
   description: .emz files are actually .gz files, which are identified with a 10-byte header, containing a magic number (1f 8b), a compression ID (08 for DEFLATE which is normal), and a variety of timestamps and flags. .emz files are compressed image files that we can convert directly with libreoffice.
+- bof: (?i)^3c3f786d6c(3[^e]|[^3].)*3e(0a|20)*3c3f6d736f2d6170706c69636174696f6e2070726f6769643d22576f72642e446f63756d656e74223f3e
+  plain_bof: |
+    <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+    <?mso-application progid="Word.Document"?>
+  puid: aca-fmt/23
+  signature: Microsoft Word XML Document
+  description: Microsoft Word allows exporting document as standalone XML files, which PRONOM incorrectly identifies as plain XML (fmt/101)

fileformats.yml

@@ -115,6 +115,15 @@ aca-fmt/22:
       converter_type: master
       outputs:
         - jpg
+aca-fmt/23:
+  name: Microsoft Word XML Document
+  description: A Microsoft Word document saved as a standalone XML file
+  action: convert
+  convert:
+    - converter: libre
+      converter_type: master
+      outputs:
+        - odt
 fmt/3:
   name: Graphics Interchange Format 87a
   action: convert
@@ -858,7 +867,7 @@ fmt/100:
         - pdf
 fmt/101:
   name: Extensible Markup Language 1.0
-  action: convert
+  action: reidentify
   convert:
     - converter: text
       converter_type: statutory
@@ -868,6 +877,9 @@ fmt/101:
       converter_type: master
       outputs:
         - _
+  reidentify:
+    reason: Some applications allow saving documents as XML and can re-open them
+    onfail: convert
 fmt/102:
   name: Extensible Hypertext Markup Language 1.0
   action: convert