Icon fonts and odd bullets extract as gibberish
A phone icon or a decorative bullet is a character with no Unicode meaning. Parsers receive junk where you see a neat symbol.
Every character in a PDF is a glyph drawn from a font, plus a mapping that says which Unicode character that glyph represents. When the mapping is missing or points into a private range, extraction produces something unreadable — often a box, a question mark, or an invisible control character.
Where this comes from
Icon fonts. Font Awesome, Material Icons, Wingdings, and their relatives assign pictures to codepoints in the Unicode private use area. A phone icon beside your number is, as far as a parser is concerned, character U+F095. It carries no meaning and can break the token immediately after it.
Decorative bullets. Template designers reach for symbols like ▪ ◆ ➤ ✦. Most extract fine. Some, taken from a symbol font, do not — and a broken bullet at the start of a line can swallow the first word of the bullet.
Ligature-heavy display fonts. A font that renders “ti” as a single glyph without a proper mapping produces opera on from operation. Rare, but it happens with free display fonts, and it destroys keyword matching on the exact words you most want matched.
Non-embedded fonts. If a font is referenced but not embedded, the reader substitutes one. Character mapping usually survives; spacing does not, which can merge or split words.
Fonts that are safe
Calibri, Arial, Helvetica, Georgia, Garamond, Cambria, Times New Roman, Verdana, Lato, Source Sans, Open Sans.
The list is boring on purpose. These are widely installed, reliably embedded, and correctly mapped by every PDF generator in common use.
Practical rules
- Use
•or a plain hyphen for bullets. Nothing else is worth the risk. - Delete icons entirely. A label reading
Email:does the same work and parses. - Two font families at most: one for headings, one for body.
- Body text between 10 and 12 points. Below 9pt, some extractors treat text as a graphic annotation.
- Avoid text effects — outlines, shadows, letter-spacing pushed to extremes. Heavy tracking in particular can cause a parser to insert a space between every letter.
How to verify
Copy the text out of your PDF and paste it somewhere plain. Look for boxes, question marks, or missing letters — especially around bullets and any symbol beside your contact details.