tinyfont.WriteLine renders wrong glyphs for missing codepoints #58

When trying to render characters that are not included in the font, instead of skipping or showing a placeholder (e.g. □), the renderer draws an incorrect glyph from a different character.
In this case, for "①②③④⑤", only ①–③ are included in the font. However, ④ (U+2463) and ⑤ (U+2464) incorrectly display as "字" (U+5B57).

Environment

• TinyGo × TinyFont × WebAssembly (browser rendering)
• Example font snippet (only ①–③ are registered):

"\x00\x24\x60" + "\x00\x10\x3F" + // ① U+2460
"\x00\x24\x61" + "\x00\x10\x6F" + // ② U+2461
"\x00\x24\x62" + "\x00\x10\x9F" + // ③ U+2462
"\x00\x5B\x57" + "\x00\x10\xCF" + // 字 U+5B57

Steps to Reproduce

1. Prepare a font with only ①②③ and "字" registered.
2. Run:

tinyfont.WriteLine(target, &testfont, 10, 50, "①②③④⑤", color.RGBA{0, 0, 0, 255})

3. Observe the output.

Expected Behavior

• Unregistered characters (④⑤) should not be drawn, or should render as a placeholder (e.g., □ or ?).
• At minimum, they should not render as a completely unrelated glyph.

Actual Behavior

• Output shows: ①②③字字
• Unregistered ④⑤ are replaced with the glyph for "字".

Impact

• Any unregistered character may render as a wrong but valid glyph.
• This can lead to confusing or misleading text output, especially in UI or educational use cases where accuracy of glyphs is critical.
• The issue is not limited to circled numbers; any missing codepoint can incorrectly map to the glyph of a different registered character.