Perl Reference Card



字数:0 关键词: Perl开发 Perl

Perl Reference Card This is version 2 of the perl reference card. (cl) 2008 Michael Goerz . Information taken liberally from the perl documentation and various other sources. You may freely distribute this document. 1 Variable Types 1.1 Scalars and Strings chomp($str); discard trailing \n $v = chop($str); $v becomes trailing char eq, ne, lt, gt, le, ge, cmp string comparison $str = “0” x 4; $str is now “0000” $v = index($str, $x); $v = rindex($str, $x); find index of $x in $str, starting from left or right $v = substr($str, $strt, $len); extract substring $cnt = $sky =~ tr/0-9//; count the digits in $sky $str =~ tr/a-zA-Z/ /cs; change non-alphas to space $v = sprintf(“%10s %08d”,$s,$n); format string Format String: %[flags][0][width][.precision][mod]type types: c character d(i) signed decimal int e(E) scientific notation f decimal floating point g, G shorter %e or %f / %E or %f o signed octal s string of chars u, x, X unsigned decimal int / hex int / hex int in caps p address pointer n nothing printed modifiers: h,l,L arg is short int / long int, double/ long double More: chr, crypt, hex, lc, lcfirst, length, oct, ord, pack, q/STRING/, qq/STRING/, reverse, uc, ucfirst 1.2 Arrays and Lists @a = (1..5); array initialization $i = @a; number of elements in @a ($a, $b) = ($b, $a); swap $a and $b $x = $a[1]; access to index 1 $i = $#a; last index in @a push(@a, $s); appends $s to @a $a = pop(@a); removes last element chop(@a); remove last char (per el.) $a = shift(@a); removes first element reverse(@a); reverse @a @a = sort{$ela <=> $elb}(@a); sort numerically @a = split(/-/,$s); split string into @a $s = join(“, ” @c); join @a elements into string @a2 = @a[1,2,6..9]; array slice @a2 = grep(!/^#/, @a); remove comments from @a 1.3 Hashes %h=(k1 => “val1”,k2 => 3); hash initialization $val = $map{k1}; recall value @a = %h; array of keys and values %h = @a; create hash from array foreach $k (keys(%h)){..} iterate over list of keys foreach $v (vals(%h)){..} iterate over list of values while (($k,$v)=each %h){..} iterate over key-value-pairs delete $h{k1}; delete key exists $h{k1} does key exist? defined $h{k1} is key defined? 2 Basic Syntax ($a, $b) = shift(@ARGV); read command line params sub p{my $var = shift; ...} define subroutine p(“bla”); execute subroutine if(expr){} elsif {} else {} conditional unless (expr){} negative conditional while (expr){} while-loop until (expr){} until-loop do {} until (expr) postcheck until-loop for($i=1; $i<=10; $i++){} for-loop foreach $i (@list){} foreach-loop last, next, redo end loop, skip to next, jump to top eval {$a=$a/$b; }; warn $@ if $@; exception handling 3 References and Data Structures $aref = \@a; reference to array $aref = [1,"foo",undef,13]; anonymous array $el = $aref->[0]; $el = @{$aref}[0]; access element of array $aref2 = [@{$aref1}]; copy array $href = \%h; reference to hash $href ={APR => 4,AUG => 8}; anonymous hash $el = $href->{APR}; $el = %{$href}{APR}; access element of hash $href2 = {%{$href1}}; copy hash if (ref($r) eq "HASH") {} checks if $r points to hash @a = ([1, 2],[3, 4]); 2-dim array $i = $a[0][1]; access 2-dim array %HoA=(fs=>["f","b"], sp=>["h","m"]); hash of arrays $name = $HoA{sp}[1]; access to hash of arrays $fh = \*STDIN globref $coderef = \&fnc; code ref (e.g. callback) $coderef =sub{print "bla"}; anon subroutine &$coderef(); calling anon subroutine sub createcnt{ my $c=shift; return sub { print "$c++"; }; } closure, $c persists *foo{THING} foo-syntax for creating refs 4 System Interaction system(“cat $f|sort -u>$f.s”); system call @a = readpipe(“lsmod”); catch output $today = “Today: “.`date`; catch output chroot(“/home/user/”); change root while (<*.c>) {} operate on all c-files unlink(“/tmp/file”); delete file if (-f “file.txt”){...} file test File Tests: -r, -w readable, writeable -x executable -e exists -f, -d, -l is file, directory, symlink -T, -B text file, binary file -M, -A mod/access age in days @stats = stat(“filename”); 13-element list with status More: chmod, chown, chroot, fcntl, glob, ioctl, link, lstat, mkdir, opendir, readlink, rename, rmdir, symlink, umask, utime 5 Input/Output open(INFILE,"in.txt") or die; open file for input open(INFILE,"<:utf8","file"); open file with encoding open(TMP, "+>", undef); open anonymous temp file open(MEMORY,'>', \$var); open in-memory-file open(OUT,">out.txt") or die; open output file open(LOG,">>my.log") or die; open file for append open(PRC,"caesar <$file |"); read from process open(EXTRACT, "|sort >Tmp$$"); write to process $line = ; get next line @lines = ; slurp infile foreach $line (){...} loop of lines from STDIN print STDERR "Warning 1.\n"; print to STDERR close INFILE; close filehandle More: binmode, dbmopen, dbmclose, fileno, flock, format, getc, read, readdir, readline, rewinddir, seek, seekdir, select, syscall, sysreed, sysseek, tell, telldir,truncate, pack, unpack, vec 6 Regular Expressions ($var =~ /re/), ($var !~ /re/) matches / does not match m/pattern/igmsoxc matching pattern qr/pattern/imsox store regex in variable s/pattern/replacement/igmsoxe search and replace Modifiers: i case-insensitive o compile once g global x extended m multiline c don't reset pos (with g) s as single line (. matches \n) e evaluate replacement Syntax: \ escape . any single char ^ start of line $ end of line *, *? 0 or more times (greedy / nongreedy) +, +? 1 or more times (greedy / nongreedy) ?, ?? 0 or 1 times (greedy / nongreedy) \b, \B word boundary ( \w - \W) / match except at w.b. \A string start (with /m) \Z string end (before \n) \z absolute string end \G continue from previous m//g [...] character set (...) group, capture to $1, $2 (?:...) group without capturing {n,m} , {n,m}? at least n times, at most m times {n,} , {n,}? at least n times {n} , {n}? exactly n times | or \1, \2 text from nth group ($1, ...) Escape Sequences: \a alarm (beep) \e escape \f formfeed \n newline \r carriage return \t tab \cx control-x \l lowercase next char \L lowercase until \E \U uppercase until \E \Q diable metachars until \E \E end case modifications Character Classes: [amy] 'a', 'm', or 'y' [f-j.-] range f-j, dot, and dash [^f-j] everything except range f-j \d, \D digit [0-9] / non-digit \w, \W word char [a-zA-Z0-9_] / non-word char \s, \S whitepace [ \t\n\r\f] / non-space \C match a byte \pP, \PP match p-named unicode / non-p-named-unicode \p{...}, \P{...} match long-named unicode / non-named-unicode \X match extended unicode Posix: [:alnum] alphanumeric [:alpha] alphabetic [:ascii:] any ASCII char [:blank:] whitespace [ \t] [:cntrl:] control characters [:digit:] digits [:graph:] alphanum + punctuation [:lower:] lowercase chars [:print:] alphanum, punct, space [:punct:] punctuation [:space:] whitespace [\s\ck] [:upper:] uppercase chars [:word:] alphanum + '_' [:xdigit:] hex digit [:^digit:] non-digit Extended Constructs (?#text) comment (?imxs-imsx:...) enable or disable option (?=...), (?!...) positive / negative look-ahead (?<=..), (?...) prohibit backtracking (?{ code }) embedded code (??{ code }) dynamic regex (?(cond)yes|no) condition corresponding to captured parentheses (?(cond)yes) condition corresponding to look-around Variables $& entire matched string $` everything prior to matched string $' everything after matched string $1, $2 ... n-th captured expression $+ last parenthesis pattern match $^N most recently closed capt. $^R result of last (?{...}) @-, @+ offsets of starts / ends of groups 7 Object-Oriented Perl and Modules Defining a new class: package Person; use strict; sub new { #constructor, any name is fine my $class = shift; my $self = {}; $self->{NAME} = undef; # field $self->{"_CENSUS"} = \$Census; # class data ++ ${ $self->{"_CENSUS"} }; bless ($self, $class); return $self; } sub name { #method my $self = shift; if (@_) { $self->{NAME} = shift } return $self->{NAME}; } sub DESTROY { #destructor my $self = shift; -- ${$self->{"_CENSUS"} };} 1; # so the ‘require’ or ‘use’ succeeds Using the class: use Person; $him = Person->new(); $him->name("Jason"); printf "There's someone named %s.\n", $him->name; use Data::Dumper; print Dumper($him); # debug Installing Modules: perl -MCPAN -e shell; 8 One-Liners -0 (zero) specify the input record separator -a split data into an array named @F -F specify pattern for -a to use when splitting -i edit files in place -n run through all the @ARGV arguments as files, using <> -p same as -n, but will also print the contents of $_ Interactive Mode: perl -de 42 Examples: 1. just lines 15 to 17, efficiently perl -ne 'print if $. >= 15; exit if $. >= 17;' 2. just lines NOT between line 10 and 20 perl -ne 'print unless 10 .. 20' 3. lines between START and END perl -ne 'print if /^START$/ .. /^END$/' 4. in-place edit of *.c files changing all foo to bar perl -pi.bak -e 's/\bfoo\b/bar/g' *.c 5. delete first 10 lines perl -i.old -ne 'print unless 1 .. 10' foo.txt 6. change all the isolated oldvar occurrences to newvar perl -i.old -pe 's{\boldvar\b}{newvar}g' *.[chy] 7. printing each line in reverse order perl -e 'print reverse <>' file1 file2 file3 .... 8. find palindromes in the /usr/dict/words dictionary file perl -lne '$_ = lc $_; print if $_ eq reverse' /usr/dict/words 9. command-line that reverses all the bytes in a file perl -0777e 'print scalar reverse <>' f1 f2 f3 10. word wrap between 50 and 72 chars perl -p000e 'tr/ \t\n\r/ /; s/(.{50,72})\s/$1\n/g;$_.="\n"x2' 11. strip and remove double spaces perl -pe '$_ = " $_ "; tr/ \t/ /s; $_ = substr($_,1,-1)' 12. move '*.txt.out' to '*.out' perl -e '($n = $_) =~ s/\.txt(\.out)$/$1/ and not -e $n and rename $_, $n for @ARGV' *



需要 2 金币 [ 分享文档获得金币 ] 0 人已下载