Word Boundary Edge Cases
You'll learn how \b behaves in surprising situations involving digits, underscores, and string boundaries.
You already know that \b matches word boundaries—positions between "word" and "non-word" characters. But JavaScript defines "word" characters as [A-Za-z0-9_], which leads to unexpected behavior in three key scenarios.
Edge Case 1: Digits Are Word Characters
Since digits count as word characters, \b appears between digits and most punctuation:
/\b\d+\b/.test("Price: 99"); // true - boundaries before 9 and after 99
/\b\d+\b/.test("99degrees"); // false - no boundary between 9 and 'd'
The boundary exists at "9:" but not at "9d" because both are word characters.
Edge Case 2: Underscores Are Word Characters
Underscores belong to the word character set, so no boundary exists between letters and underscores:
/\bmax\b/.test("max_value"); // false - no boundary at 'x_'
/\bmax\b/.test("max-value"); // true - boundary exists at 'x-'
This trips up developers searching for variable names like user in user_id.
Edge Case 3: String Start/End Always Qualify
\b matches at the very beginning or end of a string if that position is adjacent to a word character:
/\btest/.test("test"); // true - start boundary
/test\b/.test("test"); // true - end boundary
/\btest\b/.test(" test "); // true - both boundaries exist
Even without surrounding non-word characters, the string edges count as boundaries.
Key Takeaway: \b treats digits and underscores as word characters, creating boundaries with punctuation but not with letters, and always matches at string edges adjacent to word characters.