Map Of London

No time for details here, but a lot of progress made. In particular, I've forced the presence of id attributes in the output which begin with the primary xml:id of the document, and that means I can now write CSS rules based on any element ancestored by an element with an id beginning with the primary id, so I can write e.g. Stow 1598 rules into the primary_source.css stylesheet. That's better than encoding them in the document itself, as rendition elements, because we're rendering in ways that are not true to the original document.

Today I did some research on the structure of Stow and the size of sections, and sent email to the team laying out some issues we need to discuss, and some suggestions, for how we might chunk and render Stow.

I then went on to make a number of changes to the Stow encoding and to rendering rules in preparation for serious work on this. All the changes below have been made identically in the main Stow file and the Meredith file:

• I've added some <rendition> elements to the header for display of marginal labels, page numbers, running headers, catchwords etc. I still have to handle signatures and default paragraph indents.
• I've added some basic page dimension CSS to the @style attribute on the main <text> element.
• I've changed all marginal "notes" to labels (that's what they actually are in almost all cases I can see), and linked them to their rendition elements.
• I've linked all page numbers and running headers to rendition elements.
• I've added handling for choice/sic/corr to the rendering.
• I've fixed up the page-by-page rendering so you can use the showDraft flag in the URL and it will be preserved as you move from page to page. This should make it easier to page through the rendered document while it's still in draft.
• I've eliminated spaces following <lb/> tags, since they really shouldn't be there (the idea is that the word is unbroken when indexed).

Last week I started work on this while I was away, and this afternoon I completed it. Both the on-the-fly list of variant names shown on a location page and the pre-generated gazetteer now correctly include names reconstituted from split tags linked through @next and @prev. In the process of doing this, I found that most of the @next and @prev values were wrong (they lacked a leading hash), so I've fixed them all, and also added Schematron to enforce the hash (it's not plausible that @next and @prev would ever be used to link to something in a separate file).

At the last meeting, we discussed the placement and display of the map links, listed containing documents, and placename variants at the top of location files. The consensus was that they should appear on the left, as they do now, but each item should be preceded by a little rose. I've now done that in the dev site. Also, the placename variants are now separated by pipes, as KMF requested.

Originally, the two map-related links were together at the top, and the other two were separated. Per JJ, made them all equidistant like the bottom two.

The generation of the gazetteer was allowing the content of note elements that appeared within placenames to be included in the list of variant names. This is now fixed.

Investigating an oddity in someone's bio, I discovered it was caused by an empty ref tag. I've now added Schematron to catch empty refs, along with names and other similar things that will be turned into links, and this new rule found a large number of errors in existing files which I've fixed.

SM asked about tagging events, so we've added this type attribute to the schema and I've added XSLT to process it just like bibl. It works out of the box, so I must have done the JS and XQuery already before, in anticipation.

The two key files from the video of the presentation at Ideafest are corrupted on the SD card in the ETCL camera. I've spent a fair amount of time trying to rescue them without success. I have a complete copy of the SD card as an ISO on Greg's machine in /home/hcmc, and the file sizes look right, but even trying to copy the files out of the ISO fails. I've tried tsmuxer and photorec without success. Maybe Greg knows some better utilities.

• DONE: Location document lists of mentions and variant names: move them back to the left, use a pipe instead of a comma as a divider, and try putting a tiny daisy before each item. Do that in dev, then publish if/when everyone's happy.
• DONE: Look at P Herman's person entry -- the popup seems to show an empty bullet point.
• Work with TL on a proposal for handling classes of pedagogical partner contributors, using <personGrp>, to be assigned as if they were a single person to e.g. <respStmt>s, and to display as a single popup with a list of names.
• DONE 2014-04-15: Add a new category to the taxonomy: Wharfs, Quays, Landings and Stairs (mdtEncyclopediaLocationWaterside).
• DONE: Look at Graſſe church entry in gazetteer: unwanted editorial text.
• In May, work on a more sophisticated plan for bibl/@type and @subtype.

Per JJ's instructions, replaced the document status in the final column of the table with a category descriptor.