Awesome

ndnp_iiif

ndnp_iiif is a command line tool for creating IIIF manifests for National Digital Newspaper Program data. This IIIF data can then be mounted on the Web and viewed using a IIIF compatible viewer.

For example, here is an example of viewing a single issue in Mirador. The manifests and generated tiles are published here on GitHub using GitHub Pages.

Model

The mapping between NDNP's METS model and IIIF has been developed by the IIIF Newspaper Interest Group.

Note: The NDNP standard does not contain Article, Section or Volume information so those are not included.

Install

Soon you'll be able to pip install ndnp_iiif but for now you'll have to:

% git clone https://github.com/umd-mith/ndnp_iiif
% cd ndnp_iiif
% pip install -r requirements.txt
% python setup.py install

Usage:

Static Images

The simplest usage is to point ndnp_iiif at a path where you have an NDNP batch stored, and a web accessible directory where you would like to build your IIIF data, as well as an absolute base URL where your data will be mounted on the web:

% ndnp_iiif /vol/ndnp/batch_mdu_kale/ /var/www/newspapers --base-url http://example.edu/newspapers/

This will cut static tiles for the page images that will be referenced in the manifests. The --base-url parameter is optional but strongly encouraged since it will result in absolute URLs being cooked into your IIIF data. Absolute URLs are preferred to relative URLs since they seem to work more predictably in IIIF viewers.

The resulting IIIF data will be laid out on the filesystem as a static site. The LCCN for the newspaper is used for the top level directory, each issue is placed into a sub-directory using the issue date, and then each page is placed in a sub-directory of the issue. For each image encountered static tiles are cut and placed inside the page directory. Here's what a four page issue would look like (some of the tile filenames are ellided):

/var/www
├── newspapers.json
└── sn83009569
    ├── 1865-10-04
    │   ├── 1
    │   │   ├── 0,0,1024,1024
    │   │   ├── 0,0,2048,2048
    │   │   ├── 0,0,4096,4096
                ...
    │   │   ├── 6144,6144,595,2048
    │   │   ├── 6144,7168,595,1024
    │   │   ├── 6144,8192,595,876
    │   │   ├── full
    │   │   └── info.json
    │   ├── 2
    │   │   ├── 0,0,1024,1024
    │   │   ├── 0,0,2048,2048
    │   │   ├── 0,0,4096,4096
                ...
    │   │   ├── 6144,6144,595,2048
    │   │   ├── 6144,7168,595,1024
    │   │   ├── 6144,8192,595,876
    │   │   ├── full
    │   │   └── info.json
    │   ├── 3
    │   │   ├── 0,0,1024,1024
    │   │   ├── 0,0,2048,2048
    │   │   ├── 0,0,4096,4096
                ...
    │   │   ├── 6144,6144,628,2048
    │   │   ├── 6144,7168,628,1024
    │   │   ├── 6144,8192,628,863
    │   │   ├── full
    │   │   └── info.json
    │   ├── 4
    │   │   ├── 0,0,1024,1024
    │   │   ├── 0,0,2048,2048
    │   │   ├── 0,0,4096,4096
                ...
    │   │   ├── 6144,6144,595,2048
    │   │   ├── 6144,7168,595,1024
    │   │   ├── 6144,8192,595,876
    │   │   ├── full
    │   │   └── info.json
    │   └── issue.json
    └── newspaper.json

IIIF Image Server

If you are using an external IIIF Image Server then ndnp_iiif will write out the IIIF Presentation API data to the filesystem and then reference the TIFF or JPEG2000 file using the given IIIF Image Server and Prefix. For example:

% ndnp_iiif /vol/ndnp/batch_mdu_kale/ /var/www/ --image-server http://images.example.edu/

will create the following files:

/var/www
├── newspapers.json
└── sn83009569
    ├── 1865-10-04
    │   └── issue.json
    └── newspaper.json

The images referenced in issue.json will target http://images.example.edu/ using the path for a TIFF file in the batch, for example:

http://images.example.edu/batch_mdu_lilac/sn83009569/00296026165/1865100401/0013.tif

You will need to make sure that the images have been made available on the IIIF image server.

Test

You can run the tests, such as they are like so:

% python setup test

The resulting IIIF data will be left in test-data/iiif. If you want you can run a simple webserver pointed at that directory, and use the Mirador viewer to display the result:

% cd test-data/iiif
% python -m SimpleHTTPServer

and then open http://localhost:8000 in your browser, at which point you should see something like this: