Skip to content

Latest commit

 

History

History
146 lines (125 loc) · 4.54 KB

NNNN-indexer.md

File metadata and controls

146 lines (125 loc) · 4.54 KB
Raw

OCFL Community Extension NNNN: Indexer

  • Extension Name: NNNN-indexer
  • Authors: Jürgen Enge (Basel)
  • Minimum OCFL Version: 1.0
  • OCFL Community Extensions Version: 1.0
  • Obsoletes: n/a
  • Obsoleted by: n/a

Overview

This object extension integrates technical metadata extraction during the ingest process. It uses an apps for different media types, that provides the required functionality. The technical metadata is stored as newline delimited JSON format.

Usage Scenario

In order to have a better overview of the content contained in an OCFL object structure, technical metadata is very helpful. It is an aid to reporting as well as to preservation management.

Parameters

Summary

  • Name: storageType

    • Description: Location Type where the technical metadata is stored. Possible values are area, path or extension.
      • area: within an area defined by NNNN-content-subpath extension
      • path: directly within content folder
      • extension: within the extension subfolder
    • Type: string
    • Default:

  • Name: storageName

    • Description: Location within the specified Type
      • area: area name
      • path: subfolder within content folder
      • extension: subfolder within extension folder
    • Type: string
    • Default:

  • Name: compress

  • Name: actions

    • Description: actions to extract technical metadata. Capabilities depend on the indexer service.
      • siegfried: mimetype and pronom recognition using Siegfried library.
      • ffprobe: ffmpeg for audio/video metadata extraction
      • identify: Image Magick for image metadata extraction
      • tika: Tika for metadata extraction of office files, pdf etc.
      • fulltext: Tika for fulltext extraction of pdf files
      • ...
    • Type: array of strings
    • Default:

Caveat

Make sure, that there's an indexer service available to use this extension for ingest. gocfl comes with a build in service.

Procedure (tbd.)

Every entry whithin the OCFL Object Manifest is represented by a JSON line in a file called indexer_<version>.jsonl[.gz|.br]. Since this file is immutable, every version of the ocfl object gets its own indexer file.

Examples

JSON Entry for an Image

{
  "Path": "v1/content/data/together.png",
  "Indexer": {
    "mimetype": "image/png",
    "mimetypes": [
      "image/png"
    ],
    "pronom": "fmt/12",
    "pronoms": [
      "fmt/12"
    ],
    "width": 1920,
    "height": 1080,
    "size": 645553,
    "metadata": {
      "identify": {
        "magick": {
          "version": "1.0",
          "image": {
            "name": "together.png",
            "permissions": 666,
            "format": "PNG",
            "formatDescription": "Portable Network Graphics",
            "mimeType": "image/png",
            "class": "DirectClass",
            "geometry": {
              "width": 1920,
              "height": 1080
            },
            "resolution": {
              "x": 37.79,
              "y": 37.79
            },
            "printSize": {
              "x": 50.8071,
              "y": 28.579
            },
            "units": "PixelsPerCentimeter",
            "type": "TrueColor",
            "baseType": "Undefined",
            "endianness": "Undefined",
            "colorspace": "sRGB",
            "depth": 8,
            "baseDepth": 8,
            "channelDepth": {
              "blue": 1,
              "green": 8,
              "red": 8
            },
            "pixels": 6220800,
            "imageStatistics": {
              "Overall": {
                "max": 255,
                "mean": 205.498,
                "median": 219,
                "standardDeviation": 69.5403,
                "kurtosis": 3.43043,
                "skewness": -2.16984,
                "entropy": 0.369045
              }
            },
            "channelStatistics": {
            [...]

Result