search

Conal-Tuohy / swinburne

Algernon Charles Swinburne website
4 stars 0 forks source link

choice[@n = 'eol'] processing #11

Closed jawalsh closed 3 years ago

jawalsh commented 3 years ago

Some files have this structure to regularize end-of-line hyphenation found in the original source:

<choice n="eol">
      <reg>undramatic</reg>
      <orig>undra-matic</orig>
</choice>

The above example is found in acs0000001-01.xml in the component div with @xml:id acs0000001-01-i000.

In the case of choice[@n = 'eol'] we want to display the regularized (reg) option, but currently the original (orig) is being displayed. I think this is probably the case for all choice[reg and orig] elements; we want to display reg.

Conal-Tuohy commented 3 years ago

For the record the issue is here was related to https://github.com/Conal-Tuohy/swinburne/commit/930c11257aaecde73ac944eb9fa38d6d18d4e768 in which the p5-to-html.xsl stylesheet output the content of a choice/reg in the HTML text, while capturing the choice/orig in a data-orig attribute in the HTML, so that the regularised form was indexed, but then when the HTML was fed to Solr an additional pipeline step would remove the regularised form and replace it with the data-orig step. The point of that was so as to be able to search for Aeschylean and find Æschylean.

However, that's not the behaviour we want with the "end of line" choices. I modified the choice template so that it didn't apply when @n='eol', and hence didn't create the data-orig attribute.