Quick snippets/tutorial on how to condense a font by selecting a subset of characters. First, create a `.txt` file with the characters (as unicode chars) you want included.

1_example-glyphs.txt

U+0039
U+003A
U+002F
U+0023
U+0050
U+0075
U+007A
U+006C
U+0065
U+0041
U+0042
U+0043
U+0044
U+0045
U+0046

2_condense_font.sh

# pip3 install fonttools

pyftsubset ${FONT}.ttf --output-file=${FONT}-Subset.ttf --unicodes-file=1_example-glyphs.txt

3_base64_encode_font.py

import base64

FONT = "Font"

with open(f"{FONT}.txt", "w") as output_file:
    with open(f"{FONT}.ttf", "rb") as input_file:
        output_file.write(
            f"data:font/ttf;utf-8;base64,{base64.b64encode(input_file.read()).decode('utf-8')}"
        )

"""
You can now use it as follows:

<svg {...} >
    <style type="text/css">
        @font-face {
            font-family: Font-Name;
            src: url(...);
        }
    </style>
    <text font-family="Font-Name">
        some text
    </text>
</svg>
"""

4_generate_glyphs.py

CHARACTERS = "This is the string with whatever characters you want to include"

for char in sorted(CHARACTERS):
  print(f"U+{str(hex(ord(char))[2:]).zfill(4).upper()}")

adapted this into a blog post btw: https://fiveoutofnine.com/blog/on-chain-font