reference_perl_operators.md

reference_perl_operators.md

Regular Expression Reference

Perl specific operators

m/regex/mods - simple match

echo "Russia" | perl -ne 'print "yes\n" if(/russ/i);'

---

s/regex/replacement/mods - substitute (or count)
substitute and (if needed) gives back how often the match has been substituted

echo "superman" | perl -pe 's/man/woman/;' # output: superwoman
echo "abc" | perl -pe 'print s/ab/X/;'     # output: 1Xc
echo "abcdabdeab" | perl -pe '$a= s/ab/X/g; print "---$a---\n";'
# output:
# ---3---
# XcdXdeX

---

tr/before/after/ - transliteration operator
the tr operator does not accept regexes, everything is treated as a character

echo "paperino" | perl -pe 'tr/pao/PAO/;' # output: PAPerinO
echo "p.*" | perl -pe 'tr/p.*/P?+/;'      # output: P?+

---

** =~ !~ - positive and negative match operators**

---

regex delimiter - how can I delimit my regex /.../
/ works, but you can use almost every non alphanumeric simple visible ascii char
the reason can be: save as many backslashes as possible

echo "a&b&c" | perl -pe 's/&/X/g;'             # output: aXbXc
echo "a/b/c" | perl -pe 's&/&X&g;'             # output: aXbXc
echo "a/b/c" | perl -pe 's+/+X+g;'             # output: aXbXc
echo "a*b+c/d&e" | perl -pe 's#[*+/&]#X#g;'    #output: aXbXcXdXe

---

built in variable
$` text before match
$& text matched
$' text after match
$1 text matched with first parentheses
$2 text matched with second parentheses
$+ text matched with highest numbered parentheses
$^N text matched with most recently closed parentheses
@- array of match-start indices into target text
@+ array of match-end indices into target text
for better efficiency avoid $& $' $` and backref where possible

---

pos() Used to obtain the actual position of the match. But you can also use the function to set the start of the match. NEVER FORGET the g modifier when doing it!

perl -e '$a="a1a2a3"; pos($a)=3; if($a =~ /(a.)/g){print "$1\n";}'    # output: a3

---

(X|Y) - about alternation
Some Implementation study the length of X and Y in order to support greedy alternation.
This may take a longer time.
Perl always takes the left first match of an alternation.

echo "ab" | perl -pe 's/(ab|a)/X/;'   # output: X
echo "ab" | perl -pe 's/(a|ab)/X/;'   # output: Xb

---

study()
When the string to analyze is very long and you need to match many regexex on it, you can first study it.

echo 'dudududadada' | perl -pe 'study($_); s/(u|a)/X/g;'     # output: dXdXdXdXdXdX

---