Browse results not forcing Syriac font choice

[wlpotter]

The Syriac text in browse results (e.g., contents list) are not forcing the Syriac font choice (compare how we did this by adding a span, I think, to explicitly tag Syriac text sequences).

Syriac text in the XML should be tagged with an @xml:lang attribute we can use for this

[wlpotter]

An example of a contents note (once srophe/britishLibrary-data#1973 gets pulled in and re-run) that should have mixed content is record 1015.xm (Add. 17,232)l.

Here's the XML:

<note type="contents-note">It is entitled a <foreign xml:lang="syr">ܦܪܕܝܣܐ</foreign>,
                        "Paradise," or <foreign xml:lang="syr"> ܒܝܬ ܓܐܙܐ</foreign>,"Treasury" (fol. 496 b), and
                        contains a large collection of hymns, chants, and prayers, for various
                        occasions.<num type="item-count" value="33"/>
            </note>

Another example is ms 1037 (which is part 2 of 1035.xml (Add. 17,246, foll. 75–84).:

<note type="contents-note">; 
                  <foreign xml:lang="syr">ܡ̈ܥܢܝܬܐ</foreign> on George
                ; 
                  <foreign xml:lang="syr">ܡ̈ܥܢܝܬܐ</foreign> on Theodore
                ; A Calendar<num type="item-count" value="4"/>
              </note>

[wsalesky]

@emgeier The browse page on the test site does not match the browse page on GitHub, it is missing quite a lit of javascript files, which could account for the broken font style switching. What happens between Github and AWS? I thought it would just copy the static files over?