Handling unicode ranges and glyph lookup about opentype.js HOT 8 CLOSED

You can do the subsetting serverside with pyftsubset from https://github.com/behdad/fonttools/

The unicode/glyph lookup is being discussed in #49

from opentype.js.

Not interested in such serverside tools at all. I noticed Glyphr is losing their TTX dependancy in the next version too. I simply want to access specific ranges of glyphs, that is all. A JSON filter/map is the way to do it in this new age of clientside apps.

from opentype.js.

In glyphr-studio/Glyphr-Studio-1#25 I suggested using fontTools python converted to js with emscripten, which means pyftsubset would be available in js too, but, I'm not sure who is going to actively work on that....

from opentype.js.

Thread crashing - @davelab6 I commented in the thread you mentioned above, but i'll give a shortened version here. If something like fontTools.js is made to work, then great (I don't think I can help here). I think TTX is cool, but I agree certainly JSON is a better format than XML for javascripty stuff. The uber scenario is just client-side font file reading / writing.

from opentype.js.

@zoomclub Is sounds like an interesting use case. Any idea/example how that JSON format could look? Is there a reason why you wouldn't want to parse the whole font? (performance? memory?)

from opentype.js.

@fdb I'm working on a possible schema, give me a bit and I'll post it. Yes, I do have concerns about in memory size after parsing, also lookup performance. Most likely because I'm not sure how Opentype manages this for 2000+ glyphs in the font? Although, it did look like it parsed and rendered super fast on the site!

from opentype.js.

@fdb Bravura represents a collection of nearly 3000 glyphs and uses metadata to provide developers with more information about each glyph. This metadata is nice to have but extremely verbose and redundant, being spread out across four JSON files. The metadata is available here: http://www.smufl.org/download

The first thing I did when I saw the Bravura metadata is completely revamp it all into much more logical data models, which extend the basic classes of glyphs found in the Bravura ranges.json file and work better with score data structures. This is a data modeling task which is required for my use of the Bravura font as a whole.

This data modeling aspect is done and thinking it through from there it seems that Opentype should only require a small addition in its Glyph Inspector to make it easier to work with fonts like Bravura. The default way for Opentype to display a large font is in blocks of 100 glyphs, which can be toggled between.

Bravura is laid out by classes of glyphs with each class assigned to a given unicode range. The following generic JSON format could be processed by the Glyph Inspector to show specific glyph ranges with a title per range:

{ xyz-filter : {

{
start-unicode-codepoint : xyz,
end-unicode-codepoint : xyz,
display-picker-title : xyz
}
...
}}

The Glyph Inspector could use this file as an arg and optionally the font loader could as well. The display-picker-title could be used to list the defined ranges so they could be easily picked. It could then also be used to show a title for the picked range in Glyph Inspector. The unicode start and end codepoints could be appended to the title too.

These filters can then be defined by Opentype users but the default display option could still be the generic numbered lots of 100. There can just be an extra button in the Glyph Inspector allowing a defined filter to be loaded. If this seems like a reasonable option to you I can prepare a version of such a filter for the Bravura font, to test with.

from opentype.js.

Closing this old issue. I don't think we need to implement some higher-level functionality for this. You can already filter the glyphset in order to create a subset.

from opentype.js.