DEBUGGING: 7-Jun-2022. Today, I changed the code in .htaccess to make it redirect correctly for free-issue documents, where the DOI ends in .f. However, although that seems to work for URLs typed in the address box, it doesnt seem to be working for trhe URLs on this page, so try it explicitly ...

• https://doi.bcra.org.uk/j117002.f
• https://doi.bcra.org.uk/cks140055.f

Aha - part of an explanation. It seems to work in Firefox. In Chrome, the download counter is getting incremented to something is happening, and I think Chrome is probably downloading the file automatically but where is it putting it? What happens with...

• A bad DOI https://doi.bcra.org.uk/j117xxx.f
• A free PDF from the database https://bcra.org.uk/pub/docs/downloads.html?f=inf002001.f
• A direct PDF https://bcra.org.uk/pub/cregj/_pdf/j113015.pdf
• A non-free PDF from the database https://bcra.org.uk/pub/docs/downloads.html?f=j113015

That second set works correctly, and the first set works at localhost - so is the problem to do with htaccess? Or a strange Chrome problem.

SOLVEDI needed to redirect to https. Not sure what was happening but it seems to be working now.

The non-free-issue DOIs ought - apart from the above problem - to work via htaccess without needing the code at index.php. But do I need additiona lshortcuts to the cloud for ckc, ckt and any new catalogues? Point to check: I should have a master list of catalogues and a list of places where that needs updating. htaccess obviously, but also listing.php and where else?

BCRA's Digital Object Identifiers: Index of all DOIs

Please note that the list is long and contains only the DOIs and not the document titles. It is intended primarily for the use of indexing services. For preference, you should view the catalogued lists of BCRA documents by visiting one of our Bibliography pages. (NB - these lists are still under development)

• This list is generated on-the-fly by inspecting BCRA's web server(s) for downloadable content. It may take a few seconds to appear.
• Because this list results from a dynamic search, it returns the documents actually on the server and not those listed in the database files. The two lists ought be the same but could easily get out of step.
• Documents in some categories (e.g. DOI prefixes cts and ckm) are not yet processed by the DOI handler. You can identify these, in the list below, because the first column of the table does not contain a hyperlink.
• DEVELOPMENT NOTE: This list might include items that have not been released for publication. For example, a 'sandboxed' bibliography page might not display items that this scanner considers to be legitimate downloads.

SOFTWARE DEVELOPMENT NOTE: This table cannot be publicly exposed because it contains direct hyperlinks to the Cloud, which would bypass the controls we have to stop access by robots. (The problem being that we pay for the downloads). The existing catalogue page now has a software mod that says If HTTP_USER_AGENT contains 'google'... Construct a direct link to BCL instead of going via "onmouseover". However, the similar mod for adjusting the download link has not yet been implemented because it is not sufficient to merely do the above - Google would simply record the given link. What is needed is for the server to serve a different page when the User Agent is Google. That's also the case for this page - and for any page that quotes a DOI that points to the cloud server. Therefore further work is needed both for the links of the catalogue page and the links here.

HOW TO DECODE A DOI: The DOI codes follow a strict set of rules, and are identical to the filename on our server. Note, however, that when you download a file, the name might be amended.

• The first two or three alphabetic characters describe the catalogue of the file - cks for Cave and Karst Science, sp for Speleology and so on.
• The following three digits give the sequence number in the catalogue. This is often the same as the issue number, but not always.
• For C&KS, the sequence number N can be converted (E&OE) into a volume and issue as follows. Vol = int((N+2)/3); Iss = (N+2) mod 3 + 1. But please note that that only applies to C&KS and not to its predecessors, which have a variable number of parts per volume.
• The next set of characters in the DOI describe the starting page number of the paper or article or, for books, the chapter number. The string is either a three digit number, or one of the character strings i, ii, iii or iv. If the DOI refers to an entire publication, the starting page is given as 000. For items that do not yet show a hyperlink to the DOI scanner these digits might be omitted. DOIs for these items should not be quoted.
• The DOI might finish with an optional .f which means that the file is a free download. (Note that this is part of the filename on our server, so you cannot simply add .f to a DOI and expect to download the file for free!).