Glyph accessibility system

1. In a digital medium glyph environment, a method implemented by a computing device, the method comprising:

receiving, by the computing device, an input selecting a glyph displayed in a user interface using a query font;

identifying, by the computing device, a glyph cluster from a plurality of glyph clusters as corresponding to the glyph, the plurality of glyph clusters formed using a base font;

determining, by the computing device, Unicode identifiers assigned to the identified glyph cluster from the base font;

obtaining, by the computing device, glyphs from the query font that correspond to the Unicode identifiers;

generating, by the computing device, a representative glyph from the obtained glyphs from the query font;

responsive to a result of verifying by the computing device that the representative glyph corresponds to the glyph, determining an amount of similarity, respectively, of the glyph to the obtained glyphs from the query font; and

displaying, by the computing device, representations of at least a portion of the obtained glyphs in the user interface based on the determined amount of similarity.

2. The method as described in claim 1 , wherein the identifying the glyph cluster includes locating the glyph cluster from the plurality of glyph clusters based on a Unicode identifier associated with the glyph.

3. The method as described in claim 1 , wherein the plurality of glyph clusters are maintained in a cache after preprocessing of the base font.

4. The method as described in claim 1 , wherein the verifying includes determining the glyph is similar to the representative glyph based on visual similarity.

5. The method as described in claim 4 , wherein the visual similarity is defined using a similarity threshold.

6. The method as described in claim 1 , wherein the generating the representative glyph from the obtained glyphs from the query font includes generating the representative glyph as an average representation of the obtained glyphs.

7. The method as described in claim 1 , wherein the query font is different than the base font.

8. The method as described in claim 1 , further comprising ordering the obtained glyphs based on the amount of similarity to the glyph, respectively, and wherein the displaying is performed based at least in part on the ordering.

9. The method as described in claim 1 , further comprising:

responsive to a result of verifying that the representative glyph does not correspond to the glyph, determining an amount of similarity, respectively, of the glyph to other glyphs from the query font;

identifying, by the computing device, another portion of the other glyphs by comparing the amount of similarity to a similarity threshold, respectively; and

displaying, by the computing device, representations of the other portion of the other glyphs in the user interface.

10. In a digital medium glyph preprocessing environment, a system comprising:

a font identification module implemented at least partially in hardware of a computing device to select a base font from a plurality of fonts that correspond to a script;

a cluster initiation module implemented at least partially in hardware of the computing device to initiate generation of a plurality of glyph clusters, each glyph cluster of the plurality of glyph clusters corresponding to a respective Unicode identifier of a plurality of Unicode identifiers employed by the base font;

a representative glyph generation module implemented at least partially in hardware of the computing device to generate representative glyphs for the respective glyph clusters; and

a similarity determination module implemented at least partially in hardware of the computing device to assign glyphs of the base font to respective glyph clusters of the plurality of glyph clusters based on the representative glyphs.

11. The system as described in claim 10 , wherein the representative glyphs are generated as an average representation of the glyphs assigned to a respective said cluster.

12. The system as described in claim 10 , wherein the font identification module is configured to select the base font from the plurality of fonts that correspond to the script based on a number of distinct Unicode identifiers included in the base font.

13. The system as described in claim 10 , wherein the similarity determination module is configured to combine glyph clusters based on comparison of respective said representative glyphs using a similarity threshold.

14. The system as described in claim 13 , wherein the similarity threshold is defined based on cosine similarity, earth mover's distance similarity, or kernel density estimation.

15. In a digital medium glyph preprocessing environment, a method implemented by a computing device, the method comprising:

initiating, by the computing device, generation of a plurality of glyph clusters, each glyph cluster of the plurality of glyph clusters corresponding to a respective Unicode identifier of a plurality of Unicode identifiers employed by a base font;

forming, by the computing device, a third said glyph cluster by combining a first said glyph cluster that is determined as visually similar to a second said glyph cluster;

generating, by the computing device, a representative glyph of the third said glyph cluster based on glyphs from the first and second said glyph clusters;

determining, by the computing device, a fourth said glyph cluster is visually similar to the third said glyph cluster based on the representative glyph;

forming, by the computing device, a fifth said glyph cluster by combining the fourth said glyph cluster with the third said glyph cluster; and

generating, by the computing device, glyph metadata and caching the glyph metadata in a storage device, the glyph metadata identifying the fifth said glyph cluster and Unicode identifiers of glyphs from the first, second, and fourth glyph clusters.

16. The method as described in claim 15 , wherein the representative glyph is generated by averaging glyphs from the first and second said glyph clusters.

17. The method as described in claim 15 , wherein the first said glyph cluster is determined as visually similar to the second said glyph cluster based on a similarity threshold.

18. The method as described in claim 17 , wherein the similarity threshold is defined based on cosine similarity, earth mover's distance similarity, or kernel density estimation.

19. The method as described in claim 15 , wherein the initiating is performed responsive to identifying the base font from a plurality of fonts corresponding to a single script.

20. The method as described in claim 19 , wherein the identifying is performed automatically and without user intervention by the computing device based on a number of Unicode identifiers assigned to the plurality of fonts, respectively, and the base font has a maximum said number of Unicode identifiers.