Regex search
Press / to open search and find text using the same regex syntax as vim's magic mode. Korean, Japanese, and Chinese text search just as well as English.
Basics
| / / Cmd+F | Open search |
| n / N | Next / previous match |
| :nohlsearch / :noh | Turn off search highlighting (turns back on with the next search or n/N) |
Suffixes on the query
A suffix on the query always wins over the search bar's case/regex toggle buttons.
| ...\c / ...\C | Force case-insensitive / case-sensitive |
| ...\r / ...\R | Force regex / force plain text |
Without a suffix, the query is always treated as plain text — the regex toggle just reflects what actually happened.
Supported regex syntax (vim magic mode)
| . * ^ $ | Any character / 0-or-more / line start / line end |
| [...] [^...] | Character class / negated character class |
| \+ \? \= | 1-or-more / 0-or-1 |
| \{n,m} \{-...} | Repeat count (leading - for a minimal match) |
| \( \) \| | Group / OR |
| \< \> | Word boundary (start / end) |
| \zs \ze | Set the start / end of the highlighted match |
| \d \D \w \W \s \S | Digit / non-digit / word / non-word / whitespace / non-whitespace |
| \a \l \u \x \o \h | Alphabetic / lowercase / uppercase / hex digit / octal digit / identifier char (capital = negated) |
Not supported: lookaround (\@= \@! \@<= \@<!), backreferences (\1–\9), \%(...\)
Good to know
Some PDFs — usually old scans or documents produced by legacy toolchains — have text embedded in the file that doesn't actually match what's rendered on screen (a broken ToUnicode mapping). These documents render fine but can fail to search, and that's a defect in the file itself, not something any viewer can fix.