Map Of London

Tagging of marginal labels and other features was broken, and the <supplied> element was not encoded correctly. Fixed all the problems.

As previously requested, I've created a new document category, and added a bunch of quays, wharfs and stairs to it. Other locations might well have this category replace their current category, or have it added (buildings adjacent to the river, for instance); created an Asana task for JJ to provide instructions on those.

Someone will now have to add documents to this category.

Implemented in dev; awaiting approval before migrating to live.

A preliminary implementation of this is working on dev, and will be ported to live on approval from the team. There have been changes to the schemas to support this, and documentation will be added as soon as it's all finalized.

We have excludes now tested and working, and I've created some more useful sets of update scripts.

Adapted some RVDB code for listing collections and resources for the purpose of chowning, chgrping and setting permissions on them. I've done this before, in a way which was less elegant but more flexible; I now need to merge the two approaches to get something really good.

Also tried to modify the upyerdb script to enable pre-specified folder exclusions. This involved using GN's latest version, which assumes permissions it doesn't have when accessing my data (which lives in the root db collection). I need to take the old version I use, which depends on a last-mod library on the server, and merge the new exclusions functionality into it (or the reverse), to get something that works for me but enables me to block upload of stuff that shouldn't be in the db (but which people keeps storing in /data/).

Tried to install my local copy of MoEML on JJ's laptop for a conference, but it turned out to be broken. Had to download a fresh copy from the server. Eventually it did work, though. Tried initially with Tomcat 8 and had problems there. We must test carefully with Tomcat 8 before we think about deploying it.

Per KMF's instructions, made the following changes to Stow rendering:

• Fixed a bug where editorial notes were not being included at the bottom of the page, although they were clearly there in the XML. This is rather a strange thing; it seems to recur sporadically for no good reason in the live site, but not in dev. I'm at a loss to explain what makes it come and go, but at the last test it was working fine, and the live and dev trees (XSLT and XQuery) were identical. We'll have to watch this. It appears this is due to a bug in eXist's parsing of a generated in-memory XML document; I seem to have worked around it by making paths like this: $dataDoc/descendant::text/descendant::note into paths like this: $dataDoc//text//note.
• Increased default line-height in prose paras.
• Added a default top margin for line groups.
• Set default rendering for head elements to 200% font size and centred.
• Tweaked the CSS for the page navigation at the bottom so that it doesn't overlap with forme works. Did the same for the footnote block.
• Fixed a couple of encoding errors in Stow.

All of the default rendering settings will of course be overridden by any @style or @rendition settings in the XML itself.

A couple of minor fixes today:

• A new news item was structured wrongly, and had some superfluous <hi> elements in <figDesc>s.
• The code rendering title attributes on images was not very intelligent; I've added explicit string casting.
• The code for harvesting variant spellings for a location was erroneously pulling values from a specific context in the ISE linking code which resulted in mol:ABCD1 values ending up in the name list. I've fixed that.