Why does perl use \g1 for backreferences and everyone else uses \1?

Question

I am curious about the historical reason for the divergence in syntax between Perl regex backreferences and everyone else's (C++, grep, emacs, literally every other usage I've seen).

Perl uses \g1 for a group backreference. Everyone else uses a syntax which seems much cleaner, just \1 .

Answer 1

Actually, Perl does accept \1 .

/^(.)\1\z/s    # Equiv* to length($_) == 2 && substr($_, 0, 1) eq substr($_, 1, 1)

\g is a far more recent and far more flexible addition.

\1             # References the text captured by the nth capture
\g{1}   \g1    # References the text captured by the nth capture
\g{-1}  \g-1   # References the text captured by the nth capture before the \g
\g{name}       # References the text captured by (?<name>...)