FontFamily `default` name mapping missing about imscjs

[mattjuggins]

My test case was using something along the lines of tts:fontFamily="myCustomFont, Arial, Helvetica, default". So default was used as a fallback rather than a single entry.

With the above in mind, I think there might be a problem similar to #245
default doesn't get converted to monospaceSerif correctly with the space character at the start when the array is split. It could be solved using trim() again?

[nigelmegitt]

Hi @palemieux yes I do. The reason is that it is where we have it in the BBC fork, and the reason we put it there is because we apply styling options when HTML mapping, so that they happen whenever we render an ISD, not just on document parsing. The specific example here is at https://github.com/bbc/imscJS/blob/4b194d20e590f255e4168f9d3ddfd28430b35590/src/main/js/html.js#L1470-L1472

                    if (context.options.fontFamily) {
                        attr = context.options.fontFamily.split(",");
                    }

Then we process the split attr. That way if the customisation options change then we only have to generate new HTML for the current ISD, and the syntax of the fontFamily option is the same as the syntax in the TTML document, which can include spaces.

[palemieux]

@mattjuggins It sounds like adding a trim() at time of parsing should work.

[palemieux]

When specified, tts:fontFamily="default" is parsed as "monospaceSerif" per

imscJS/src/main/js/styles.js

Line 191 in bf2e98a

function (str) {

.

The initial value of tts:fontFamily is also aliased to "monospaceSerif" since initial values, as specified in TTML, are subjected to the same parsing.

Do you have a test case that demonstrates the issue you are encoutering?

[palemieux]

@nigelmegitt Do you recall why trim() at #246 was applied during HTML mapping instead of TTML parsing?

https://github.com/nigelmegitt/imscJS/blob/13c449b22d570d8ec722f6ff11934de2f5ef4fa9/src/main/js/html.js#L1403

instead of:

imscJS/src/main/js/styles.js

Line 197 in bf2e98a

if (ffs[i].charAt(0) !== "'" && ffs[i].charAt(0) !== '"') {

[nigelmegitt]

I agree, that should work.

[mattjuggins]

Excellent. Thanks both. I've opened up #251 for this.