Importing data by DOI
Sometimes several url go in biblio_url field. When you save, it is impossible and the message is : biblio_url cannot be longer than 255 characters but is currently 262 characters long.
It seems that the mapping between "resource", which contains urls in the crossref xml, and biblio_url, cannot be modified by the UI admin : it is written in the php parser.
Only one "resource" should go in the biblio_url, to avoid this bug.

CommentFileSizeAuthor
#4 biblio.crossref.client.php_.txt1.79 KBToki
#2 biblio-2832502-2-separate-multiple-resources.patch1.92 KBAlan D.
Support from Acquia helps fund testing for Drupal Acquia logo

Comments

St_B created an issue. See original summary.

Alan D.’s picture

Alan D. CreditAttribution: Alan D. commented
Status: Active » Needs review
FileSize
biblio-2832502-2-separate-multiple-resources.patch1.92 KB

Two issues, one is the the URLs get concatenated and the second is that the limit of 255 was exceeded and a PDO exception occurred.

So I added a separator option and also a limit check. if you cut off any of the 4 variables they become meaningless (DOI, ISSN, ISBN, URL), so all are excluded if they are too long.

I used a semi-colon (;) to separate URLs / resources, commas otherwise, albeit unlikely that these (DOI, ISSN, ISBN) would ever have multiple values...

And I've blindly used this biblio_url, so made sense to provide an option to only accept one URL :) No interface, but users can run this code once.

variable_set('biblio_crossref_allow_multiple_resources', 0);

Or update their settings.php file

$conf['biblio_crossref_allow_multiple_resources'] = 0;

Status: Needs review » Needs work

The last submitted patch, 2: biblio-2832502-2-separate-multiple-resources.patch, failed testing. View results

Toki’s picture

Toki
Primary language French
Location Caen, Normandy
CreditAttribution: Toki commented
FileSize
biblio.crossref.client.php_.txt1.79 KB

Still having this problem here, when importing reference with DOI, the URL is concatenated (e.g. https://linkinghub.elsevier.com/retrieve/pii/S0269749117320572https://ap...) and obviously the link is broken. When it is longer than 255, the import fails.
I have just tried to patch the biblio.crossref.client.php file and I got :

File to patch: biblio.crossref.client.php
patching file biblio.crossref.client.php
Hunk #1 FAILED at 279.
Hunk #2 FAILED at 311.
2 out of 2 hunks FAILED -- saving rejects to file biblio.crossref.client.php.rej

I have attached the rej file, if needed (but if I am right, everything is rejected)

Adding the conf line in settings.php (and clear caches) did not help either, unfortunately.

I am afraid this is still a major issue, my clients can not import references through DOI (if limit exceeded) and, if the import is successful, they get a long and broken link in the URL field.

Does someone have any solution here?

Toki’s picture

Toki
Primary language French
Location Caen, Normandy
CreditAttribution: Toki commented

After many tests, I have managed to make it work by adding lines manually in biblio.crossref.client.php, notably those about biblio_url (indeed, DOI, ISBN and ISSN are unique originally). The line $conf in settings.php was needed also. Finally, with both updates, the DOI import has worked and the biblio_url displays only one valid URL. Thanks for the code.