Open glenrobson opened 5 months ago
@digitaldogsbody you were right. The book reader returns
{
"error": {
"string": "Couldn't find image stack.",
"itemid": "arkivkopia.se-lms-G70-48.3",
"server": "ia800802.us.archive.org",
"request_uri": "/BookReader/BookReaderJSIA.php?id=arkivkopia.se-lms-G70-48.3&itemPath=/1/items/arkivkopia.se-lms-G70-48.3&server=ia800802.us.archive.org&format=jsonp&subPrefix=arkivkopia.se-lms-G70-48.3"
}
}
Original files in object:
__ia_thumb.jpg format: 'Item Tile'
arkivkopia.se-lms-G70-48.3_files.xml format: 'Metadata'
arkivkopia.se-lms-G70-48.3_meta.sqlite format: 'Metadata'
arkivkopia.se-lms-G70-48.3_meta.xml format: 'Metadata'
original_lms2_4c4d535f4737302d34383a33.index.1.tiff.jpg format: 'JPEG'
original_lms2_4c4d535f4737302d34383a33.index.2.tiff.jpg format: 'JPEG'
original_lms2_4c4d535f4737302d34383a33.index.3.tiff.jpg format: 'JPEG'
Maybe use count(format in valid_filetypes
) > 1
@digitaldogsbody you were right. The book reader returns
{ "error": { "string": "Couldn't find image stack.", "itemid": "arkivkopia.se-lms-G70-48.3", "server": "ia800802.us.archive.org", "request_uri": "/BookReader/BookReaderJSIA.php?id=arkivkopia.se-lms-G70-48.3&itemPath=/1/items/arkivkopia.se-lms-G70-48.3&server=ia800802.us.archive.org&format=jsonp&subPrefix=arkivkopia.se-lms-G70-48.3" } }
It occurs to me there is a situation where this might not occur - if there is a file called <identifier>_images.zip
then the BookReader may well find an "image stack" regardless of the mediatype
in the metadata (based on Hank's comment here rather than any testing).
I think it won't affect us if we go with the counting of originals with valid formats you outlined above, but we should be aware of it for items such as mareful
which have multiple book stacks (#12).
Can this be closed?
Transfering from: https://github.com/ArchiveLabs/iiif.archivelab.org/issues/137
The following item should have three images but only has one:
The logic in the app resolve right now only takes the last image in the file list and creates a manifest for that single image.
We'll need a way to check for multiple valid images, and then add those as canvases individually.