Character mappings in Pillow

[ThingamabobsCoding]

Hey there, as far as I understand pillow uses freetype.

There is an utility called character mapping. It can be handy to handle errors or for GUI applications to map the chars.

I don't see a hint that this is somewhat obtainable in documentation of pillow. Nor does font.font.glyphs returns something like that, I just know how many characters are available.
Is there any chance this will be implemented or already a hacky way to achieve this?

Consider something like this: https://www.1001fonts.com/hazard-signs-font.html
If you click on characters you can see how it is mapped.

[radarhere]

From my understanding, you would like to be able to map FreeType glyph indices to characters.

Considering that a proposed feature to check if a character is present at all in a font is stalled over debate about whether it is within our focus, no, I don't think that retrieving the underlying glyph index is in Pillow's scope.

I expect that freetype-py would offer you a solution instead.

[ThingamabobsCoding]

@radarhere I understand that in context of a image-lib there is only rudimentary support for text and fonts.
But I think symbols play a role in image-text and I believe having a character map is missing for it.

From my perspective checking if greek glyphs are in the font isn't in the scope but mapping symbols to the typed character is.

Thank you for your suggestion about freetype-py, but I rather don't have an extra dependency.
For my case it will be enough to refer to the documentation of the font and map them myself with a dictionary.
But I think it would be a cool feature to pillow in the future, since it is more or less already there and just not exported.

EDIT:
To make a clearer case pro implementation.
Working with pillow means working with images, just like Flyer, Handouts, Business Cards or Logos.
And for all examples there is often text needed, that's why imho pillow supports text in the first place.
In all the examples you almost always see symbols be used, such as phone, location, mail and so on.
So I would understand if you would reject an enhancement for Leading/Tracking/Kerning , but rejecting support for symbols is a little bit like rejecting punctuation or unicode, which is already supported by pillow.

So I think it's a complementary feature and not a new branch.

[radarhere]

rejecting support for symbols is a little bit like rejecting punctuation or unicode, which is already supported by pillow.

My understanding so far has been that you are asking for additional information about font characters.

When you say 'support for symbols' though, that sounds like you think that you're unable to draw a particular character in a font with Pillow. If that is the case, please provide an example.

[ThingamabobsCoding]

After thinking my argument above through again, I didn't made a good job with it.

I wanted to point out that you can use '.' and it will show '.' or I could use u"\u250c" and it shows '┌' which are all non-characters and that's why symbols should be accessible as well.
However the important distinction is that a font with symbols, does not necessarily come with a documentation (especially free ones) and therefore it would be appropriated to know which keys correspond with symbols.

Consider this: https://www.1001fonts.com/symbolix-font.html
It would take forever finding the right mapping without a programmatic inspection of the font or a documentation.
So again, it's just about image-font usage and not about font-magic/development.

Consider this font: https://www.1001fonts.com/waveform-font.html
I guess one could do cool stuff with a mask and this font, but without knowing what characters are mapped it's a pain.