BOM (Byte Order Mark)

One invisible character at the start of a file, U+FEFF, and the very specific ways it breaks JSON, scripts, and string comparisons.

What it is

A Byte Order Mark is the Unicode character U+FEFF, placed at the very start of a file or text stream. It has no visible glyph — in a normal editor it renders as nothing at all — but it's still a real character occupying a real position in the string.

Why it exists

Some Unicode encodings — UTF-16 and UTF-32 — store each character using more than one byte, and those bytes can be ordered two different ways ("big-endian" or "little-endian") depending on the system that wrote the file. The BOM's original job was to sit at the front of the file and signal which byte order was used, so a reading program could decode the rest correctly.

Why it still shows up in UTF-8 files

UTF-8 doesn't have a byte-order ambiguity to resolve — each character is encoded the same way regardless of system. A BOM is technically meaningless in UTF-8. But some tools, most notably Windows Notepad and a number of older editors, still prepend a BOM to UTF-8 files anyway, using it as an informal "this file is UTF-8, not some legacy encoding" signal. Most modern software ignores it safely. Some doesn't.

What it actually breaks

• JSON parsing — a BOM before the opening { is not valid JSON syntax, and some parsers will throw an error rather than silently skip it.
• Shebang lines — if a BOM precedes #!/usr/bin/env python at the top of a script, some interpreters fail to recognize the shebang at all.
• String comparison — two strings that look identical on screen won't be === equal if one has a leading BOM and the other doesn't. This is a common source of "why doesn't this if-statement work" bugs.
• Rendering — misinterpreted BOM bytes sometimes appear as the mangled character sequence  at the start of a displayed page or file.

How to find and remove one

Because it's invisible, a BOM is easy to carry around for a long time without noticing. Paste suspect text into the invisible character finder — it flags a leading BOM explicitly with its code point, and the cleaned output removes it.

Source

BOM is defined as part of the Unicode Standard, maintained by the Unicode Consortium.

Back to the full glossary — 200 terms covering case conversion, style guides, and text tools.