PerlOne-Liners130ProgramsThatGetThingsDone

PeterisKrumins

PublishedbyNoStarchPress

AbouttheAuthor

PeterisKruminsisaprogrammer,systemsadministrator,blogger,andall-aroundhacker.Heiscurrentlyrunninghisowncompany,Browserling,whichfocusesoncross-browsertesting.Hehasself-publishedthreebooksonessentialUNIXtools,andheenjoysopen-sourcinghundredsofsmallprojectsonGitHub.

Findhiswebsiteandblogathttp://www.catonmat.net/,follow@pkruminson

Twitter,andseehisopensourceprojectsathttp://github.com/pkrumins/.

AbouttheTechnicalReviewer

AlastairMcGowan-DouglaslivesinRugbyintheUK.HehasbeenaPerldevelopersince2008andisnowstuckwritingPHPforaliving.HisfavoritepastimeatworkiswritingPerlscriptsforinternalusetoencourageotherstoembracethelanguage.AlsoaJavaScriptdeveloperandGitaficionado,hisrantingsandmusingsonthesevarioussubjectscanbefoundathttp://altreus.blogspot.com/.

Acknowledgments

I’dliketothankEricPementforinspiringmetowritethisbook;BillPollockforgivingmetheopportunitytopublishitatNoStarchPress;RileyHoffmanandLaurelChunforworkingwithmetomakeitperfect;AlastairMcGowan-Douglasforhistechnicalreview;andDavidPrecious,GaborSzabo,JimDavis,ChrisFedde,AndyLester,JohnD.Cook,JonathanScottDuff,andJeremyMatesforreviewingthebookandmakinggreatsuggestionsforimprovements.I’dalsoliketothankeveryonewhohelpedmeonthe#perlIRCchannelonfreenode.IfIforgotanyone,I’msorry,butthanksforhelpingmetogetthisbookwritten!

Chapter1.IntroductiontoPerlOne-Liners

Perlone-linersaresmallandawesomePerlprogramsthatfitinasinglelineofcode.Theydoonethingreallywell—likechanginglinespacing,numberinglines,performingcalculations,convertingandsubstitutingtext,deletingandprintingspecificlines,parsinglogs,editingfilesin-place,calculatingstatistics,carryingoutsystemadministrationtasks,orupdatingabunchoffilesatonce.Perlone-linerswillmakeyouashellwarrior:whattookyouminutes(orevenhours)tosolvewillnowtakeyouonlyseconds!

Inthisintroductorychapter,I’llshowyouwhatone-linerslooklikeandgiveyouatasteofwhat’sintherestofthebook.ThisbookrequiressomePerlknowledge,butmostoftheone-linerscanbetweakedandmodifiedwithoutknowingthelanguageindepth.

Let’slookatsomeexamples.Here’sone:

perl -pi -e 's/you/me/g' file

Thisone-linerreplacesalloccurrencesofthetextyouwithmeinthefilefile.Veryusefulifyouaskme.Imagineyou’reonaremoteserverandyouneedtoreplacetextinafile.Youcaneitheropenthefileinatexteditorandexecutefind-replaceorsimplyperformthereplacementthroughthecommandlineand,bam,bedonewithit.

Thisone-linerandothersinthisbookworkwellinUNIX.I’musingPerl5.8torunthem,buttheyalsoworkinnewerPerlversions,suchasPerl5.10andlater.Ifyou’reonaWindowscomputer,you’llneedtochangethemalittle.Tomakethisone-linerworkonWindows,swapthesinglequotesfordoublequotes.TolearnmoreaboutusingPerlone-linersonWindows,seeAppendixB.

I’llbeusingPerl’s-ecommand-lineargumentthroughoutthebook.Itallows

youtousethecommandlinetospecifythePerlcodetobeexecuted.Inthepreviousone-liner,thecodesays“dothesubstitution(s/you/me/gcommand)andreplaceyouwithmeglobally(/gflag).”The-pargumentensuresthatthecodeisexecutedoneverylineofinputandthatthelineisprintedafterexecution.The-iargumentensuresthatfileiseditedin-place.Editingin-placemeansthatPerlperformsallthesubstitutionsrightinthefile,overwritingthecontentyouwanttoreplace.Irecommendthatyoualwaysmakeabackupofthefileyou’reworkingwithbyspecifyingthebackupextensiontothe-iargument,likethis:

perl -pi.bak -e 's/you/me/g' file

NowPerlcreatesafile.bakbackupfilefirstandonlythenchangesthecontentsoffile.

Howaboutdoingthissamereplacementinmultiplefiles?Justspecifythefilesonthecommandline:

perl -pi -e 's/you/me/g' file1 file2 file3

Here,Perlfirstreplacesyouwithmeinfile1andthendoesthesameinfile2andfile3.

Youcanalsoperformthesamereplacementonlyonlinesthatmatchwe,assimplyasthis:

perl -pi -e 's/you/me/g if /we/' file

Here,youusetheconditionalif /we/toensurethats/you/me/gisexecutedonlyonlinesthatmatchtheregularexpression/we/.

Theregularexpressioncanbeanything.Sayyouwanttoexecutethesubstitutiononlyonlineswithdigitsinthem.Youcouldusethe/\d/regularexpressiontomatchnumbers:

perl -pi -e 's/you/me/g if /\d/' file

Howaboutfindingalllinesinafilethatappearmorethanonce?

perl -ne 'print if $a{$_}++' file

Thisone-linerrecordsthelinesyou’veseensofarinthe%ahashandcountsthenumberoftimesitseesthelines.Ifithasalreadyseentheline,thecondition$a{$_}++istrue,soitprintstheline.Otherwiseit“automagically”createsanelementthatcontainsthecurrentlineinthe%ahashandincrementsitsvalue.The$_specialvariablecontainsthecurrentline.Thisone-lineralsousesthe-ncommand-lineargumenttoloopovertheinput,butunlike-p,itdoesn’tprintthelinesautomatically.(Don’tworryaboutallthecommand-lineargumentsrightnow;you’lllearnaboutthemasyouworkthroughthisbook!)

Howaboutnumberinglines?Supersimple!Perl’s$.specialvariablemaintainsthecurrentlinenumber.Justprintittogetherwiththeline:

perl -ne 'print "$. $_"' file

Youcandothesamethingbyusingthe-pargumentandmodifyingthe$_variable:

perl -pe '$_ = "$. $_"' file

Here,eachlineisreplacedbythestring"$. $_",whichisequaltothecurrentlinenumberfollowedbythelineitself.(Seeone-liner3.1onpage17forafullexplanation.)

Ifyouomitthefilenameattheendoftheone-liner,Perlreadsdatafromstandardinput.Fromnowon,I’llassumethedatacomesfromthestandardinputanddropthefilenameattheend.Youcanalwaysputitbackifyouwanttorunone-linersonwholefiles.

Youcanalsocombinetheprevioustwoone-linerstocreateonethatnumbersonlytherepeatedlines:

perl -ne 'print "$. $_" if $a{$_}++'

AnotherthingyoucandoissumthenumbersineachlineusingthesumfunctionfromtheList::UtilCPANmodule.CPAN(ComprehensivePerlArchiveNetwork;http://www.cpan.org/)isanarchiveofover100,000reusablePerl

modules.List::UtilisoneofthemodulesonCPAN,anditcontainsvariouslistutilityfunctions.Youdon’tneedtoinstallthismodulebecauseitcomeswithPerl.(It’sinPerlcore.)

perl -MList::Util=sum -alne 'print sum @F'

The-MList::Utilcommand-lineargumentimportstheList::Utilmodule.The=sumpartofthisone-linerimportsthesumfunctionfromtheList::Utilmodulesothattheprogramcanusethefunction.Next,-aenablestheautomaticsplittingofthecurrentlineintofieldsinthe@Farray.Thesplittinghappensonthewhitespacecharacterbydefault.The-largumentensuresthatprintoutputsanewlineattheendofeachline.Finally,sum @Fsumsalltheelementsinthe@Flist,andprintprintstheresultfollowedbyanewline(whichIaddedwiththe-largument).(Seeone-liner4.2onpage30foramoredetailedexplanation.)

Howaboutfindingthedate1299daysago?Trythis:

perl -MPOSIX -le

'@t = localtime; $t[3] -= 1299; print scalar localtime mktime @t'

Iexplainthisexampleindetailinone-liner4.19(page41),butbasicallyyoumodifythefourthelementofthestructurereturnedbylocaltime,whichhappenstobedays.Yousimplysubtract1299daysfromthecurrentdayandthenreassembletheresultintoanewtimewithlocaltime mktime @tandprinttheresultinthescalarcontexttodisplayhuman-readabletime.

Howaboutgeneratinganeight-letterpassword?Hereyougo:

perl -le 'print map { ("a".."z")[rand 26] } 1..8'

The"a".."z"generatesalistoflettersfromatoz(foratotalof26letters).Thenyourandomlychoosealettereighttimes!(Thisexampleisexplainedindetailinone-liner5.4onpage51.)

OrsupposeyouwanttofindthedecimalnumberthatcorrespondstoanIPaddress.Youcanuseunpacktofinditreallyquickly:

perl -le 'print unpack("N", 127.0.0.1)'

Thisone-linerusesav-string,whichisaversionliteral.V-stringsofferawaytocomposeastringwiththespecifiedordinals.TheIPaddress127.0.0.1istreatedasav-string,meaningthenumbers127,0,0,1areconcatenatedtogetherintoastringoffourcharacters,wherethefirstcharacterhasordinalvalue127,thesecondandthirdcharactershaveordinalvalues0,andthelastcharacterhasordinalvalue1.Next,unpackunpacksthemtoasingledecimalnumberin“network”(big-endian)order.(Seeone-liner4.27onpage45formore.)

Whataboutcalculations?Let’sfindthesumofthenumbersinthefirstcolumninatable:

perl -lane '$sum += $F[0]; END { print $sum }'

Thelinesareautomaticallysplitintofieldswiththe-aargument,whichcanbeaccessedthroughthe@Farray.Thefirstelementofthearray,$F[0],isthefirstcolumn,soyousimplysumallthecolumnswith$sum += $F[0].WhenthePerlprogramfinishes,itexecutesanycodeintheENDblock,which,inthiscase,printsthetotalsum.Easy!

Nowlet’sfindouthowmanypacketshavepassedthroughiptablesrules:

iptables -L -nvx | perl -lane '$pkts += $F[0]; END { print $pkts }'

Theiptablesprogramoutputsthepacketsinthefirstcolumn.Allyouhavetodotofindouthowmanypacketshavepassedthroughthefirewallrulesissumthenumbersinthefirstcolumn.Althoughiptableswilloutputtableheadersaswell,youcansafelyignorethesebecausePerlconvertsthemtozeroforthe+=operation.

Howaboutgettingalistofallusersonthesystem?

perl -a -F: -lne 'print $F[4]' /etc/passwd

Combining-awiththe-Fargumentletsyouspecifythecharacterwherelinesshouldbesplit,which,bydefault,iswhitespace.Here,yousplitlinesonthecoloncharacter,therecordseparatorof/etc/passwd.Next,youprintthefifthfield,$F[4],whichcontainstheuser’srealname.

Ifyouevergetlostwithcommand-linearguments,rememberthatPerlcomeswithafantasticdocumentationsystemcalledperldoc.Typeperldoc perlrunatthecommandline.ThiswilldisplaythedocumentationabouthowtorunPerlandallthecommand-linearguments.It’sveryusefulwhenyousuddenlyforgetwhichcommand-lineargumentdoeswhatandneedtolookitupquickly.Youmayalsowanttoreadperldocperlvar,whichexplainsvariables;perldocperlop,whichexplainsoperators;andperldocperlfunc,whichexplainsfunctions.

Perlone-linersletyouaccomplishmanytasksquickly.You’llfindover130one-linersinthisbook.Readthem,trythem,andsoonenoughyou’llbethelocalshellwizard.(Justdon’ttellyourfriends—unlessyouwantcompetition.)

Enjoy!

Chapter2.Spacing

Inthischapter,welookatvariousone-linersthatchangelineandwordspacing,performingsuchtasksasdouble-andtriple-spacinglinesinafile,removingblanklines,anddouble-spacingwords.You’llalsolearnaboutvariouscommand-linearguments,suchas-p,-e,-n,andspecialvariables,suchas$_and$\.

2.1Double-spaceafile

perl -pe '$\ = "\n"' file

Thisone-linerdouble-spacesafile.Ineedtoexplainthreethingshere:the-pand-ecommand-lineoptionsandtheshort$\ = "\n"Perlprogram.

Usethe-eoptiontoenteraPerlprogramdirectlyatthecommandline.Typicallyyouwon’twanttocreatesourcefilesforeverysmallprogram;with-eyoucaneasilywriteaprogramdirectlyatthecommandlineasaone-liner.Inthiscase,theentirePerlprogramcontainedinthisone-lineris$\ = "\n".Besuretousesinglequotes(')aroundtheprogram,oryourshellwillinterpretthingslike$\asshellvariables,whichhavenovalue,effectivelyremovingthem!

Nowlet’slookatthe-pswitch.Specifying-ptellsPerltoassumethefollowinglooparoundyourprogram:

while (<>) {

# your program goes here (specified by -e)

} continue {

print or die "-p failed: $!\n";

}

Broadlyspeaking,thisconstructloopsoveralltheinput,executesyourcode,and

printsthevalueof$_(theprintstatementprintsthevalueof$_),whichallowsyoutomodifyallorsomelinesoftheinputquickly.The$_variableisaspecialvariablethatgetsreplacedwiththecurrentlineoftext.Itcanbereplacedwithotherstuffaswell.You’lllearnallabout$_asyouworkthroughthebook.(SeeAppendixAforasummaryofitsusecases.)

Butunderstandingwhatisgoingoninthisloopinmoredetailisimportant.First,thewhile (<>)looptakeseachlinefromthestandardinputandputsitinthe$_variable.Next,thecodespecifiedby-eisexecuted,followedbytheprint ordieportion.

Thecontinuestatementexecutestheprint or diestatementaftereachline,whichattemptstoprintthecontentsofthe$_variable.Iftheattemptfails(forexample,theterminalisnotwritableorstandardoutputhasbeenredirectedtosomethingthatisn’twritable),diemakesPerlexit(die)withanerrormessage.

Inthisone-liner,thecodespecifiedby-eis$\ = "\n",sotheprogramthatPerlexecuteslookslikethis:

while (<>) {

$\ = "\n";

} continue {

print or die "-p failed: $!\n";

}

ThisPerlprogramreadseachlineintothe$_variableandthensets$\toanewlineandcallsprint.Anotherspecialvariableis$\.ItissimilartotheORS(OutputRecordSeparator)variableinAwkinthatitisappendedaftereveryprintoperation.Aprintstatementwithnoargumentsprintsthecontentsof$_andappends$\attheendoftheoutput.Asaresult,eachlineisprintedunmodified,followedbythe$\,whichwassettonewline.Theinputisnowdouble-spaced.

Infact,youactuallydonotneedtoset$\tonewlineforeveryline;youcansetitjustonceatthebeginningoftheprogram:

perl -pe 'BEGIN { $\ = "\n" }' file

Thisone-linersets$\tothenewlinecharacterjustoncebeforePerldoesanythingintheBEGINcodeblock.TheBEGINblockisaspecialcodeblockthatisexecutedbeforeeverythingelseinaPerlprogram.Here’swhattheexpandedPerlprogramlookslike,anditworksexactlythesamewayasthepreviousone-liner:

BEGIN { $\ = "\n" }

while (<>) {

} continue {

print or die "-p failed: $!\n";

}

Hereisanotherwaytodouble-spaceafile.Thisone-linerappendsanothernewlinecharacterattheendofeachlineandthenprintstheline:

perl -pe '$_ .= "\n"' file

Thisone-linerisequivalentto

while (<>) {

$_ = $_ . "\n"

} continue {

print or die "-p failed: $!\n";

}

Writing$_ = $_ . "\n"isthesameaswriting$_ .= "\n".Thisexpressionsimplyconcatenates$_with"\n".(Theperiod(.)isthestringconcatenationoperator.)

Butprobablythecleanestwaytodouble-spaceafileistousethesubstitutionoperators:

perl -pe 's/$/\n/' file

Thisone-linerreplacestheregularexpression$thatmatchestheendofalinewithanewline,effectivelyaddinganewlineattheend.

Ifyou’rerunningPerl5.10orlater,youcanusethesayoperator.Thesayoperatoractsmuchlikeprint,butitalwaysaddsanewlineattheendofthe

line.InPerl5.10,thissameone-linercanbewrittenlikethis:

perl -nE 'say' file

The-Ecommand-lineargumentworksexactlythesamewayasthe-ecommand-lineargument,butitalsoenablesPerl5.10features,includingthesayoperator.The-nargumentissimilarto-p,butyouhavetoprintthelineyourself.(Iexplainthe-nargumentinmoredetailinone-liner2.6.)Thisone-linerprintstheline,followedbyanothernewlinethat’sappendedbythesayoperator.

Forexample,ifafilecontainsfourlines:

line1

line2

line3

line4

runninganyoftheseone-linersoutputsthefollowing:

line1

line2

line3

line4

Inthesefirstfewexamples,Ipassedafilenameasthelastargumenttotheone-liners.WhenIdothat,theone-linersoperateonthecontentsofthatfile.IfIdidn’tpassafilenametotheone-liners,they’doperateonthedatafromthestandardinput.FromnowonIwon’tspecifyafileattheendoftheone-liners,butyoucanalwaysadditbackifyouwanttoruntheone-linersonfiles.Whenwritingone-liners,it’sagoodideatoquicklytestifthey’recorrectbytypingsomethingdirectlytothestandardinput.Thenwhenyou’resuretheone-linerworks,youcanpassoneormorefilenamesattheend.

Again,don’tforgetaboutPerl’shandydocumentationsystem,perldoc.Justtypeperldoc perlrunatthecommandlinetodisplayinformationabouthowtorunPerlandallthecommand-linearguments.

2.2Double-spaceafile,excludingtheblanklines

2.2Double-spaceafile,excludingtheblanklines

perl -pe '$_ .= "\n" unless /^$/'

Thisone-linerdouble-spacesalllinesthatarenotcompletelyemptybyappendinganewlinecharacterattheendofeachnonblankline.Theunlessmeans“ifnot,”andunless /^$/means“ifnot‘beginningthenendofline.’”Thecondition“beginningthenendofline”istrueonlyforblanklines.

Here’showthisone-linerlookswhenexpanded:

while (<>) {

unless (/^$/) {

$_ .= "\n"

}

} continue {

print or die "-p failed: $!\n";

}

Hereisabettertest,whichtakesintoaccountspacesandtabsonaline:

perl -pe '$_ .= "\n" if /\S/'

Here,thelineismatchedagainst\S—aregularexpressionsequencethatistheinverseof\s,whichmatchesanywhitespacecharacter(includingtab,verticaltab,space,newline,andcarriagereturn).Theinverseof\sisanynonwhitespacecharacter.Theresultisthateverylinewithatleastonenonwhitespacecharacterisdouble-spaced.

2.3Triple-spaceafileYoucanalsotriple-spaceafilesimplybyoutputtingmorenewlinesattheendofeachline:

perl -pe '$\ = "\n\n"'

Or

perl -pe '$_ .= "\n\n"'

perl -pe '$_ .= "\n\n"'

Or

perl -pe 's/$/\n\n/'

Theseone-linersaresimilartothefirstone-linerinthischapter,exceptthattwonewlinesareappendedaftereachline.

2.4N-spaceafile

perl -pe '$_ .= "\n"x7'

Thisone-linerinsertssevennewlinesaftereachline.NoticehowI’veused"\n"x 7torepeatthenewlinecharacterseventimes.ThexoperatorrepeatsthevalueontheleftNtimes.

Forexample,theline

perl -e 'print "foo"x5'

printsfoofoofoofoofoo.

Asasidenote,sometimeswhenyouneedtogenerateacertainamountofdata,thexoperatorcomesinreallyhandy.Forexample,togenerate1KBofdata,youcoulddothis:

perl -e 'print "a"x1024'

Thisone-linerprintstheacharacter1024times.

2.5Addablanklinebeforeeveryline

perl -pe 's/^/\n/'

Thisone-linerusesthes/regex/replace/operator.Itsubstitutesthegivenregularexpressionwiththereplacement.Inthisone-liner,theoperatoris

s/^/\n/,theregularexpressionis^,andthereplacementis\n.The^patternmatchesthebeginningpositionofthetext,andthesoperatorreplacesitwith\n,anewline.Asaresult,thenewlinecharacterisinsertedbeforetheline.Toinsertsomethingelsebeforetheline,simplyreplace\nwiththebittobeinserted.

2.6Removeallblanklines

perl -ne 'print unless /^$/'

Thisone-linerusesthe-nflag,whichtellsPerltoassumeadifferentloop,otherthan-p,aroundtheprogram:

while (<>) {

# your program goes here

}

ComparethislooptotheonethatPerlassumeswhenyouspecify-p,andyou’llseethatthisloophasnocontinue { print or die }part.Inthisloop,eachlineisreadbythediamondoperator<>andisplacedinthespecialvariable$_,butit’snotprinted!Youhavetoprintthelineyourself—ausefulfeatureifyouwanttoprint,modify,ordeletelinesselectively.

Inthisone-liner,thecodeisprint unless /^$/,sotheentirePerlprogrambecomes

while (<>) {

print unless /^$/

}

Unravelingthisabitfurther,yougetthis:

while (<>) {

print $_ unless $_ =~ /^$/

}

Thisone-linerprintsallnonblanklines.(Yousawthe/^$/regularexpressioninone-liner2.2onpage11.)

Thisone-lineralsoremovesallblanklines:

Thisone-lineralsoremovesallblanklines:

perl -lne 'print if length'

Thisone-linerusesthe-lcommand-lineargument,whichautomaticallychompstheinputline(basicallyremovesthenewlineattheend)andappendsitbackattheendofthelinewhenprinting.Thecodespecifiedtothe-eargumentis'print if length',whichsays“printthelineifithassomelength.”Emptylineshavealengthof0,sothey’renotprinted(0isafalsevalueinPerl,sotheif lengthconditionevaluatestofalse).Allotherlineshavelengthandareprinted.Without-l,thestringwouldstillhavethenewlineattheendandthusbe1or2characterslong![1]

Here’sanotherone-linertoremoveallblanklines:

perl -ne 'print if /\S/'

Thisone-linerbehavesslightlydifferentlyfromtheprevioustwo.Bothprintunless /^$/andprint if lengthalsoprintthelinesthatconsistofonlyspacesand/ortabs.Suchlinesappeartobeemptytotheeye,andyoumaywanttofilterthem.Thisone-lineruses\S(explainedinone-liner2.2onpage11),aregularexpressionsequencethatmatchesnonblankcharacters.Linescontainingonlyspacesand/ortabsdon’tmatch\Sandthereforearenotprinted.

Asyoucansee,youcanwritethesameprograminmanydifferentways.Infact,Perl’smottoisThere’sMoreThanOneWayToDoIt,whichisabbreviatedasTIMTOWTDIandpronounced“TimToady.”(Funtrivia:LarryWall,theinventorofPerl,usesthehandle@TimToadyonTwitterandIRC.)

2.7Removeallconsecutiveblanklines,leavingonlyone

perl -00 -pe ''

Thisone-linerisreallytricky,isn’tit?First,itdoesn’thaveanycode!The-eisempty.Next,ithasasilly-00command-lineoptionthatturnsparagraphslurp

modeon,meaningPerlreadstextparagraphbyparagraph,ratherthanlinebyline.(Aparagraphistextbetweentwoormorenewlines.)Theparagraphisputinto$_,andthe-poptionprintsitout.

Youcanevenwritethisinashorterfashion:

perl -00pe0

Here,thecodespecifiedto-eis0,whichdoesnothing.

Thisisoneofmyfavoriteone-linersbecauseifyouhaven’tseenitbefore,itcanbetrickytofigureout,andIlovemindgames.(There’snocodespecifiedto-e!Howcanitpossiblydosomething?)

2.8Compress/expandallblanklinesintoNconsecutivelinesSayyouhaveafilewithtwoblanklinesaftereachparagraph,andyouwishtoexpandthelinespacingbetweentheparagraphstothreelines.Todoso,youcansimplycombineone-liners2.4and2.7likethis:

perl -00 -pe '$_ .= "\n"x2'

Thisone-linerslurpslinesparagraph-wiseviathe-00optionandthenappendsthreenewlinecharactersaftereachparagraph.Thecode"\n"x2printstwonewlines,whichareaddedtotheblanklinealreadyattheendoftheparagraph.

Inasimilarvein,youcanalsoreducethespacingbetweenparagraphs.Sayyou’vegotafilethatforsomecrazyreasonhastenblanklinesbetweenparagraphs,andyouwanttocompresstheseblanklinestojustthree.Youcanusethesameone-lineragain!

2.9Double-spacebetweenallwords

perl -pe 's/ / /g'

Here,youusethesubstitutionoperatorstoreplaceonespace“”withtwospaces“”globallyoneachline(the/gflagmakesthereplacementglobal),andyou’redone.It’sthatsimple!

Here’sanexample.Let’ssayyouhavethislineoftext:

this line doesn't have enough whitespace!

Runningthisone-linerincreasesthespacingbetweenwords:

this line doesn't have enough whitespace!

2.10Removeallspacingbetweenwords

perl -pe 's/ +//g'

Thisone-linerusesthe“+”regularexpressiontomatchoneormorespaces.Whenitfindsamatch,itsubstitutesitwithnothing,globally,whichdeletesallspacesbetweenwords.

Ifyoualsowanttogetridoftabsandotherspecialcharactersthatmightaddspacing,usethe\s+regularexpression,whichmeans“matchaspace,atab,averticaltab,anewline,oracarriagereturn”:

perl -pe 's/\s+//g'

Here’sanexample.Let’ssayyouhavethislineoftext:

this line has too much whitespace said cowboy neal

Runningthisone-linerremovesallspaces:

thislinehastoomuchwhitespacesaidcowboyneal

2.11Changeallspacingbetweenwordstoonespace

perl -pe 's/ +/ /g'

Thisone-linerissimilartothepreviousone,exceptthatitreplacesoneormorespaceswithjustonespace.

Forexample,ifyouhavethisline:

this line has really messed-up spacing

runningthisone-linernormalizesthespacingbetweenwordstoonespace:

this line has really messed-up spacing

2.12Insertaspacebetweenallcharacters

perl -lpe 's// /g'

Hereyoumatchseeminglynothingandreplaceitwithasinglespace.Thenothingnessactuallymeans“matchbetweencharacters,”withtheresultthatyouinsertaspacebetweenallcharacters.(Thematchingincludesthebeginningandendofthetext.)

Forexample,giventhisline:

today was a great day

runningthisone-linerproducesthisresult:

t o d a y w a s a g r e a t d a y

Itmightbedifficulttoseewhereallthespacesareadded,solet’sillustratethatbymodifyingthisone-linertoinsertacolonbetweenallcharacters:

perl -lpe 's//:/g'

Thiswilloutput:

:t:o:d:a:y: :w:a:s: :a: :g:r:e:a:t: :d:a:y:

Asyoucansee,spaces(orcolons)arealsoinsertedatthebeginningandendofthetext.Alsonotethattheexistingspacescountascharacters,sothey’retriple-spaced.

[1]Windowsusestwocharactersforthenewline.

Chapter3.Numbering

Inthischapter,we’lllookatvariousone-linersfornumberinglinesandwords,andyou’llgettoknowthe$.specialvariable.You’llalsolearnaboutPerlgolfing,a“sport”thatinvolveswritingtheshortestPerlprogramtogetataskdone.

3.1Numberalllinesinafile

perl -pe '$_ = "$. $_"'

AsIexplainedinone-liner2.1(page7),-ptellsPerltoassumealooparoundtheprogram(specifiedby-e)thatreadseachlineofinputintothe$_variable,executestheprogram,andthenprintsthecontentsofthe$_variable.

Thisone-linersimplymodifies$_byprependingthe$.variabletoit.Thespecialvariable$.containsthecurrentlinenumberoftheinput.Theresultisthateachlinehasitslinenumberprepended.

Similarly,youcanalsousethe-nargumentandprintthestring"$. $_",whichisthecurrentlinenumberfollowedbytheline:

perl -ne 'print "$. $_"'

Sayafilecontainsthreelines:

foo

bar

baz

Runningthisone-linernumbersthem:

1 foo

2 bar

2 bar

3 baz

3.2Numberonlynon-emptylinesinafile

perl -pe '$_ = ++$x." $_" if /./'

Hereyouemployan“actionifcondition”statementthatexecutesanactiononlyiftheconditionistrue.Inthiscase,theconditionistheregularexpression/./,whichmatchesallcharactersotherthannewline(thatis,itmatchesanon-emptyline).Theaction$_ = ++$x." $_"prependsthevariable$xincrementedbyonetothecurrentline.Becauseyou’renotusingthestrictpragma,$xiscreatedautomaticallythefirsttimeit’sincremented.

Theresultisthatateachnon-emptyline$xisincrementedbyoneandprependedtothatline.Nothingismodifiedattheemptylines,andtheyareprintedasis.

One-liner2.2(page11)showsanotherwaytomatchnon-emptylinesthroughthe\Sregularexpression:

perl -pe '$_ = ++$x." $_" if /\S/'

Sayafilecontainsfourlines,twoofwhichareempty:

line1

line4

Runningthisone-linernumbersonlylinesoneandfour:

1 line1

2 line4

3.3Numberandprintonlynon-emptylinesinafile(dropemptylines)

perl -ne 'print ++$x." $_" if /./'

Thisone-linerusesthe-nprogramargument,whichplacesthelineinthe$_variableandthenexecutestheprogramspecifiedby-e.Unlike-p,-ndoesnotprintthelineafterexecutingthecodein-e,soyouhavetocallprintexplicitlytoprintthecontentsofthe$_variable.

Theone-linercallsprintonlyonlineswithatleastonecharacter,andasinthepreviousone-liner,itincrementsthelinenumberinthevariable$xbyoneforeachnon-emptyline.Theemptylinesareignoredandneverprinted.

Sayafilecontainsthesamefourlinesasinone-liner3.2:

line1

line4

Runningthisone-linerdropstheemptylinesandnumberslinesoneandfour:

1 line1

2 line4

3.4Numberalllinesbutprintlinenumbersonlyfornon-emptylines

perl -pe '$_ = "$. $_" if /./'

Thisone-linerissimilartoone-liner3.2.Here,youmodifythe$_variablethatholdstheentirelineonlyifthelinehasatleastonecharacter.Allotheremptylinesareprintedasis,withoutlinenumbers.

Sayafilecontainsfourlines:

line1

line4

Runningthisone-linernumbersallthelinesbutprintsthelinenumbersonlyforlinesoneandfour:

1 line1

4 line4

3.5Numberonlylinesthatmatchapattern;printothersunmodified

perl -pe '$_ = ++$x." $_" if /regex/'

Here,again,youusean“actionifcondition”statement,andagaintheconditionisapattern(aregularexpression):/regex/.Theactionisthesameasinone-liner3.2.

Sayafilecontainstheselines:

record foo

bar baz

record qux

andyouwanttonumberthelinesthatcontainthewordrecord.Youcanreplace/regex/intheone-linerwith/record/:

perl -pe '$_ = ++$x." $_" if /record/'

Whenyourunthisone-liner,itgivesyouthefollowingoutput:

1 record foo

bar baz

2 record qux

3.6Numberandprintonlylinesthatmatchapattern

perl -ne 'print ++$x." $_" if /regex/'

Thisone-linerisalmostexactlylikeone-liner3.3,exceptitonlynumbersandprintslinesthatmatch/regex/.Itdoesn’tprintnonmatchinglines.

Forexample,afilecontainsthesamelinesasinone-liner3.5:

record foo

bar baz

bar baz

record qux

Andlet’ssayyouwanttonumberandprintonlythelinesthatcontainthewordrecord.Inthiscase,changing/regex/to/record/andrunningtheone-linergivesyouthisresult:

1 record foo

2 record qux

3.7Numberalllinesbutprintlinenumbersonlyforlinesthatmatchapattern

perl -pe '$_ = "$. $_" if /regex/'

Thisone-linerissimilartoone-liners3.4and3.6.Here,thelinenumberisprependedtothelineifthelinematchesa/regex/;otherwise,it’ssimplyprintedwithoutalinenumber.

Replacing/regex/with/record/andrunningthisone-lineronthesameexamplefileshowninone-liner3.6givesyouthisoutput:

1 record foo

bar baz

3 record qux

3.8Numberalllinesinafileusingacustomformat

perl -ne 'printf "%-5d %s", $., $_'

Thisone-linerusesprintftoprintthelinenumbertogetherwiththelinecontents.printfdoesformattedprinting.Youspecifytheformatandsenditthedatatoprint,anditformatsandprintsthedataaccordingly.Here,theformatforthelinenumbersis%-5d,whichalignsthelinenumbersfivepositionstotheleft.

Here’sanexample.Saytheinputtothisone-lineris

hello world

bye world

Thentheoutputis

1 hello world

2 bye world

Otherformatstringsinclude%5d,whichalignsthelinenumbersfivepositionstotheright,and%05d,whichzero-fillsandright-justifiesthelinenumbers.Here’stheoutputyougetwiththe%5dformatstringforlinenumbers:

1 hello world

2 bye world

Andthisiswhatyougetwiththe%05dformatstring:

00001 hello world

00002 bye world

Tolearnmoreaboutthevariousformatsthatareavailable,runperldoc -fsprintfatthecommandline.

3.9Printthetotalnumberoflinesinafile(emulatewc-l)

perl -lne 'END { print $. }'

Thisone-linerusestheENDblockthatPerltookasafeaturefromtheAwklanguage.TheENDblockisexecutedoncethePerlprogramhasexecuted.Here,thePerlprogramisthehiddenloopovertheinputcreatedbythe-nargument.Onceitloopsovertheinput,thespecialvariable$.containsthenumberoflinesintheinput,andtheENDblockprintsthisvariable.The-lparametersetstheoutputrecordseparatorforprinttoanewline,soyoudon’thavetoprintthenewlineyourself,likethis:print "$.\n".

Youcandothesamethingwiththisone-liner:

perl -le 'print $n = () = <>'

Thisone-lineriseasytograspifyouunderstandPerlcontextswell.The() = <>parttellsPerltoevaluatethe<>operator(thediamondoperator)inthelistcontext,whichmakesthediamondoperatorreadthewholefileasalistoflines.Next,youassignthislistto$n.Because$nisscalar,thislistassignmentisevaluatedinthescalarcontext.

What’sreallyhappeninghereisthe=operatorisright-associative,meaningthe=ontherightisdonefirstandthe=ontheleftisdonesecond:

perl -le 'print $n = (() = <>)'

Evaluatingalistassignmentinthescalarcontextreturnsthenumberofelementsinthelist;thus,the$n = () = <>constructionisequaltothenumberoflinesintheinput,thatis,thenumberoflinesinthefile.Theprintstatementprintsthisnumber.The-largumentensuresanewlineisaddedafterprintingthisnumber.

Youcanalsodropthevariable$nfromthisone-linerandforcethescalarcontextthroughthescalaroperator:

perl -le 'print scalar(() = <>)'

Here,insteadofevaluatingalistassignmentinthescalarcontextbyassigningitagaintoanotherscalar,yousimplyevaluatethelistassignmentinthescalarcontextusingthescalaroperator.

Andnowforamoreobviousversion:

perl -le 'print scalar(@foo = <>)'

Here,insteadofusinganemptylist()toforcethelistcontexton<>,youusethevariable@footoachievethesameeffect.

Andhere’sanotherwaytodoit:

perl -ne '}{print $.'

perl -ne '}{print $.'

Thisone-linerusestheso-calledeskimooperator}{(actuallyacleverconstruct)inconjunctionwiththe-ncommand-lineargument.AsIexplainedearlier,the-nargumentforcesPerltoassumeawhile(<>) { }looparoundtheprogram.TheeskimooperatorforcesPerltoescapetheloop,andtheone-linerexpandsto

while (<>) {

}{ # eskimo operator here

print $.;

}

Asyoucansee,thisprogramjustloopsoveralltheinput,andonceit’sfinished,itprintsthe$.,whichisthenumberoflinesintheinput.Itbecomesevenmoreobviousifyouchangetheformattingalittle:

while (<>) {}

{

print $.;

}

Asyoucansee,thisisjustanemptyloopthatloopsoveralltheinput,followedbytheprintstatementwrappedincurlybrackets.

3.10Printthenumberofnon-emptylinesinafile

perl -le 'print scalar(grep { /./ } <>)'

Thisone-linerusesPerl’sgrepfunction,whichissimilartothegrepUNIXcommand.Givenalistofvalues,grep { condition } listreturnsonlythosevaluesinthelistthatmaketheconditiontrue.Inthiscase,theconditionisaregularexpressionthatmatchesatleastonecharacter,sotheinputisfilteredandgrep{ /./ }returnsallnon-emptylines.Togetthenumberoflines,youevaluategrepinthescalarcontextandprinttheresult.

SomePerlprogrammersliketocreatetheshortestPerlprogramthatdoessomeparticulartask—anexercisecalledPerlgolfing.Agolfer’sversionofthisone-

linerwouldreplacescalar()with~~(doublebitwisenegate)anddropthespaces,shorteningitlikethis:

perl -le 'print ~~grep{/./}<>'

Thisdoublebitwisenegatetrickiseffectivelyasynonymforscalarbecausethebitwisenegationworksonscalarvalues,sogrepisevaluatedinthescalarcontext.

Youcanmakethisevenshorterbydroppingthespaceafterprintandgettingridofthecurlybrackets:

perl -le 'print~~grep/./,<>'

IfyouhavePerl5.10orlater,youcanalsousethe-Ecommand-lineswitchandthesayoperator:

perl -lE 'say~~grep/./,<>'

Atruegolfer’smasterpiece!

3.11Printthenumberofemptylinesinafile

perl -lne '$x++ if /^$/; END { print $x+0 }'

Here,youusethevariable$xtocountthenumberofemptylinesencountered.Onceyou’veloopedoverallthelines,youprintthevalueof$xintheENDblock.Youusethe$x+0constructiontoensurethat0isoutputifnolinesareempty.(Otherwise$xisnevercreatedandisundefined.Adding+0totheundefinedvalueproduces0.)Analternativeto$x+0istheintoperator:

perl -lne '$x++ if /^$/; END { print int $x }'

Youcouldalsomodifythepreviousone-linerbydoingthis:

perl -le 'print scalar(grep { /^$/ } <>)'

Orwriteitwith~~:

perl -le 'print ~~grep{ /^$/ } <>'

The~~doesbitwisenegationtwice,whichmakesgrepexecuteinthescalarcontextandreturnthenumberofemptylines.

Theselasttwoversionsarenotaseffectiveastheone-linerwiththeENDblockbecausetheyreadthewholefileintomemory,whereastheone-linerwiththeENDblockdoesitlinebylineand,therefore,keepsonlyonelineofinputinmemory.

3.12Printthenumberoflinesinafilethatmatchapattern(emulategrep-c)

perl -lne '$x++ if /regex/; END { print $x+0 }'

Thisone-linerisbasicallythesameas3.11,exceptitincrementsthelinecounter$xbyonewhenalinematchestheregularexpression/regex/.The$x+0trickmakessure0isprintedwhennolinesmatch/regex/.(Seeone-liner3.11foramoredetailedexplanationofthe$x+0trick.)

3.13Numberwordsacrossalllines

perl -pe 's/(\w+)/++$i.".$1"/ge'

Thisone-linerusesthe/eflag,whichmakesPerlevaluatethereplacepartofthes/regex/replace/expressionascode!

Thecodehereis++$i.".$1",whichmeans“incrementvariable$ibyoneandthenadditinfrontofthestring".$1"(thatis,adotfollowedbythecontentsofthematchedgroup$1).”Thematchedgrouphereiseachword:(\w+).

Inonesentence,thisone-linermatchesaword(\w+),putsitin$1,andthenexecutesthe++$i.".$1"codethatnumbersthewordglobally(/gflag).There

—allwordsarenumbered.

Forexample,ifyouhaveafilewiththefollowingthreelines:

just another

perl hacker

hacking perl code

runningthisone-linernumberseachwordinthefileandproducesthefollowingoutput:

1.just 2.another

3.perl 4.hacker

5.hacking 6.perl 7.code

3.14Numberwordsoneachindividualline

perl -pe '$i=0; s/(\w+)/++$i.".$1"/ge'

Thisissimilartoone-liner3.13,exceptthatyouresetthe$ivariableto0oneachline.Here’stheresultofrunningthisone-linerontheexamplefromone-liner3.13:

1.just 2.another

1.perl 2.hacker

1.hacking 2.perl 3.code

Asyoucansee,ineachlinethewordsarenumberedindependentlyofotherlines.

3.15Replaceallwordswiththeirnumericpositions

perl -pe 's/(\w+)/++$i/ge'

Thisone-linerisalmostthesameasone-liner3.13.Here,yousimplyreplaceeachwordwithitsnumericposition,whichiskeptinthevariable$i.For

example,ifyourunthisone-lineronthefilefromone-liners3.13and3.14,itreplacesthewordsinthefilewiththeirnumericpositionstoproducethisoutput:

1 2

3 4

5 6 7

Fun!

Chapter4.Calculations

Inthischapter,we’lllookatvariousone-linersforperformingcalculations,suchasfindingminimumandmaximumelements,counting,shufflingandpermutingwords,andcalculatingdatesandnumbers.You’llalsolearnaboutthe-a,-M,and-Fcommand-linearguments,the$,specialvariable,andthe@{[ ... ]}constructionthatletsyouruncodeinsidedoublequotes.

4.1Checkifanumberisaprime

perl -lne '(1x$_) !~ /^1?$|^(11+?)\1+$/ && print "$_ is prime"'

Thisone-linerusesaningeniousregularexpressionbyAbigailtodetectwhetheragivennumberisaprime.(Don’ttakethisregularexpressiontooseriously;I’veincludeditforitsartisticvalue.Forseriouspurposes,usetheMath::PrimalitymodulefromCPANtoseewhetheranumberisprime.)

Here’showthisingeniousone-linerworks:First,thenumberisconvertedintoitsunaryrepresentationby(1x$_).Forexample,5isconvertedinto1x5,whichis11111(1repeated5times).Next,theunarynumberistestedagainsttheregularexpression.Ifitdoesn’tmatch,thenumberisaprime;otherwiseit’sacomposite.The!~operatoristheoppositeofthe=~operatorandistrueiftheregularexpressiondoesn’tmatch.

Theregularexpressionconsistsoftwoparts:Thefirstpart,^1?$,matches1andtheemptystring.Theemptystringand1areclearlynotprimenumbers,sothispartoftheregularexpressiondiscardsthem.

Thesecondpart,^(11+?)\1+$,determineswhethertwoormore1srepeatedlymakeupthewholenumber.Ifso,theregularexpressionmatches,whichmeansthenumberisacomposite.Ifnot,it’saprime.

Nowconsiderhowthesecondpartoftheregularexpressionwouldactonthe

number5.Thenumber5inunaryis11111,sothe(11+?)matchesthefirsttwo1s,theback-reference\1becomes11,andthewholeregularexpressionnowbecomes^11(11)+$.Becauseitcan’tmatchfive1s,itfails.Next,itattemptstomatchthefirstthree1s.Theback-referencebecomes111,andthewholeregularexpressionbecomes^111(111)+$,whichdoesn’tmatch.Theprocessrepeatsfor1111and11111,whichalsodon’tmatch,andasaresulttheentireregularexpressiondoesn’tmatchandthenumberisaprime.

Whataboutthenumber4?Thenumber4is1111inunary.The(11+?)matchesthefirsttwo1s.Theback-reference\1becomes11,andtheregularexpressionbecomes^11(11)+$,whichmatchestheoriginalstringandconfirmsthatthenumberisnotprime.

4.2Printthesumofallfieldsoneachline

perl -MList::Util=sum -alne 'print sum @F'

Thisone-linerturnsonfieldauto-splittingwiththe-acommand-lineoptionandimportsthesumfunctionfromtheList::Utilmodulewith-Mlist::Util=sum.(List::UtilispartofthePerlcore,soyoudon’tneedinstallit.)Auto-splittinghappensonwhitespacecharactersbydefault,andtheresultingfieldsareputinthe@Fvariable.Forexample,theline1 4 8wouldbesplitoneachspacesothat@Fwouldbecome(1, 4, 8).Thesum @Fstatementsumstheelementsinthe@Farray,givingyou13.

The-Mmodule=argoptionimportsargfrommodule.It’sthesameaswriting

use module qw(arg);

Thisone-linerisequivalentto

use List::Util qw(sum);

while (<>) {

@F = split(' ');

print sum @F, "\n";

}

Youcanchangeauto-splitting’sdefaultbehaviorbyspecifyinganargumenttothe-Fcommand-lineswitch.Sayyouhavethefollowingline:

1:2:3:4:5:6:7:8:9:10

Andyouwishtofindthesumofallthesenumbers.Youcansimplyspecify:asanargumenttothe-Fswitch,likethis:

perl -MList::Util=sum -F: -alne 'print sum @F'

Thissplitsthelineonthecoloncharacterandsumsallthenumbers.Theoutputis55becausethat’sthesumofthenumbers1through10.

4.3Printthesumofallfieldsonalllines

perl -MList::Util=sum -alne 'push @S,@F; END { print sum @S }'

Thisone-linerkeepspushingthesplitfieldsin@Ftothe@Sarray.OncetheinputstopsandPerlisabouttoquit,theEND { }blockisexecutedanditoutputsthesumofallitemsin@F.Thissumsallfieldsoveralllines.

Noticehowpushingthe@Farraytothe@Sarrayactuallyappendselementstoit.Thisdiffersfrommanyotherlanguages,wherepushingarray1toarray2wouldputarray1intoarray2,ratherthanappendingtheelementsofarray1ontoarray2.Perlperformslistflatteningbydesign.

Unfortunately,summingallfieldsonalllinesusingthissolutioncreatesamassive@Sarray.Abettersolutionistokeeponlytherunningsum,likethis:

perl -MList::Util=sum -alne '$s += sum @F; END { print $s }'

Here,eachlineissplitinto@Fandthevaluesaresummedandstoredintherunningsumvariable$s.Onceallinputhasbeenprocessed,theone-linerprintsthevalueof$s.

4.4Shuffleallfieldsoneachline

4.4Shuffleallfieldsoneachline

perl -MList::Util=shuffle -alne 'print "@{[shuffle @F]}"'

Thetrickiestpartofthisone-lineristhe@{[shuffle @F]}construction.Thisconstructionallowsyoutoexecutethecodeinsidethequotationmarks.Normallytextandvariablesareplacedinsidequotationmarks,butwiththe@{[... ]}constructionyoucanruncode,too.

Inthisone-liner,thecodetoexecuteinsideofthequotationmarksisshuffle@F,whichshufflesthefieldsandreturnstheshuffledlist.The[shuffle @F]createsanarrayreferencecontainingtheshuffledfields,andthe@{ ... }dereferencesit.Yousimplycreateareferenceandimmediatelydereferenceit.Thisallowsyoutorunthecodeinsidethequotationmarks.

Let’slookatseveralexamplestounderstandwhyIchosetorunthecodeinsidethequotationmarks.IfIhadwrittenprint shuffle @F,thefieldsonthelinewouldbeconcatenated.Comparetheoutputofthisone-liner:

$ echo a b c d | perl -MList::Util=shuffle -alne 'print "@{[shuffle @F]}"'

b c d a

tothis:

$ echo a b c d | perl -MList::Util=shuffle -alne 'print shuffle @F'

bcda

Inthefirstexample,thearrayofshuffledfields(insidethedoublequotationmarks)isinterpolated,andthearray’selementsareseparatedbyaspace,sotheoutputisb c d a.Inthesecondexample,interpolationdoesn’thappen,andPerlsimplydumpsoutelementbyelementwithoutseparatingthem,andtheoutputisbcda.

Youcanusethe$,specialvariabletochangetheseparatorbetweenarrayelementswhenthey’reprinted.Forexample,here’swhathappenswhenIchangetheseparatortoacolon:

$ echo a b c d | perl -MList::Util=shuffle -alne '$,=":"; print shuffle @F'

b:c:d:a

Youcanalsousethejoinfunctiontojointheelementsof@Fwithaspace:

perl -MList::Util=shuffle -alne 'print join " ", shuffle @F'

Here,thejoinfunctionjoinstheelementsofanarrayusingthegivenseparator,butthe@{[ ... ]}constructionisthecleanestwaytodoit.

4.5Findthenumericallysmallestelement(minimumelement)oneachline

perl -MList::Util=min -alne 'print min @F'

Thisone-linerissomewhatsimilartothepreviousones.ItusestheminfunctionfromList::Util.Oncethelinehasbeenautomaticallysplitby-aandtheelementsareinthe@Farray,theminfunctionfindsthenumericallysmallestelement,whichitprints.

Forexample,ifyouhaveafilethatcontainstheselines:

-8 9 10 5

7 0 9 3

5 -25 9 999

Runningthisone-linerproducesthefollowingoutput:

-8

0

-25

Thesmallestnumberonthefirstlineis-8;onthesecondline,thesmallestnumberis0;andonthethirdline,-25.

4.6Findthenumericallysmallestelement(minimumelement)overalllines

perl -MList::Util=min -alne '@M = (@M, @F); END { print min @M }'

perl -MList::Util=min -alne '@M = (@M, @F); END { print min @M }'

Thisone-linercombinesone-liners4.3and4.5.The@M = (@M, @F)constructisthesameaspush @M, @F.Itappendsthecontentsof@Ftothe@Marray.

Thisone-linerstoresalldatainmemory,andifyourunitonareallyhugefile,Perlwillrunoutofmemory.Yourbestbetistofindthesmallestelementoneverylineandcomparethatelementwiththesmallestelementonthepreviousline.Iftheelementonthecurrentlineislessthanthepreviousone,it’sthesmallestelementsofar.Oncealllineshavebeenprocessed,youcanjustprintthesmallestelementfoundthroughtheENDblock:

perl -MList::Util=min -alne '

$min = min @F;

$rmin = $min unless defined $rmin && $min > $rmin;

END { print $rmin }

'

Here,youfirstfindtheminimumelementonthecurrentlineandstoreitin$min.Thenyouchecktoseeifthesmallestelementonthecurrentlineisthesmallestelementsofar.Ifso,assignitto$rmin.Onceyou’veloopedoverthewholeinput,theENDblockexecutesandyouprintthe$rmin.

Sayyourfilecontainstheselines:

-8 9 10 5

7 0 9 3

5 -25 9 999

Runningthisone-lineroutputs-25becausethat’sthesmallestnumberinthefile.

Ifyou’reusingPerl5.10orlater,youcandothesamethingwiththisone-liner:

perl -MList::Util=min -alne '$min = min($min // (), @F); END { print $min }'

Thisone-linerusesthe//operator,whichisnewtoPerl5.10.ThisoperatorissimilartothelogicalORoperator(||),exceptthatitteststheleftside’sdefinednessratherthanthetruth.Whatthatmeansisittestswhethertheleftsideisdefinedratherthanwhetheritistrueorfalse.Inthisone-liner,theexpression$min // ()returns$minif$minhasbeendefined,orelseitreturnsanempty

list().The//operatorsavesyoufromhavingtousedefinedtotestdefinedness.

Considerwhathappenswhenthisone-linerisrunonthepreviousfile.First,Perlreadstheline-8 9 10 5,splitsit,andputsthenumbersinthe@Farray.The@Farrayisnow(-8, 9, 10, 5).Next,itexecutes$min = min ($min // (),@F).Because$minhasn’tbeendefined,$min // ()evaluatesto(),sothewholeexpressionbecomes$min = min ((), (-8, 9, 10, 5)).

Perldoeslistflatteningbydesign,soafterflatteningtheargumentstotheminfunction,theexpressionbecomes$min = min(-8, 9, 10, 5).Thisdefines$min,settingitto-8.Perlproceedstothenextline,whereitsets@Fto(7, 0,9, 3)andagainevaluates$min = min($min // (), @F).Because$minhasnowbeendefined,$min // ()evaluatesto$minandtheexpressionbecomes$min = min(-8, 7, 0, 9, 3).Atthispoint,-8isstillthesmallestelement,so$minremains-8.Finally,Perlreadsinthelastline,andafterevaluating$min =min(-8, 5, -25, 9, 999),itfindsthat-25isthesmallestelementinthefile.

4.7Findthenumericallylargestelement(maximumelement)oneachline

perl -MList::Util=max -alne 'print max @F'

Thisworksthesameasone-liner4.5,exceptthatyoureplaceminwithmax.

4.8Findthenumericallylargestelement(maximumelement)overalllines

perl -MList::Util=max -alne '@M = (@M, @F); END { print max @M }'

Thisone-linerissimilartoone-liners4.6and4.7.Inthisone-liner,eachlineisauto-splitandputinthe@Farray,andthenthisarrayismergedwiththe@Marray.Whentheinputhasbeenprocessed,theENDblockexecutesandthemaximumelementisprinted.

Here’sanotherwaytofindthemaximumelement,keepingjusttherunningmaximumelementinsteadofallelementsinmemory:

perl -MList::Util=max -alne '

$max = max @F;

$rmax = $max unless defined $rmax && $max < $rmax;

END { print $rmax }

'

Ifyou’reusingPerl5.10orlater,youcanusethe//operatortoshortenthisone-liner:

perl -MList::Util=max -alne '$max = max($max // (), @F); END { print $max }'

Thisisthesameasone-liner4.6,exceptyoureplaceminwithmax.

4.9Replaceeachfieldwithitsabsolutevalue

perl -alne 'print "@{[map { abs } @F]}"'

Thisone-linerfirstauto-splitsthelineusingthe-aoption.Thesplitfieldsendupinthe@Fvariable.Next,itcallstheabsolutevaluefunctionabsoneachfieldwiththehelpofthemapfunction.Essentially,themapfunctionappliesagivenfunctiontoeachelementofthelistandreturnsanewlistthatcontainstheresultsofapplyingthefunction.Forexample,ifthelist@Fis(-4, 2, 0),mappingabsoveritproducesthelist(4, 2, 0).Finally,thisone-linerprintsthenewlistofpositivevalues.

The@{[ ... ]}construct,introducedinone-liner4.4,allowsyoutoexecutethecodeinsidethequotationmarks.

4.10Printthetotalnumberoffieldsoneachline

perl -alne 'print scalar @F'

Thisone-linerforcestheevaluationof@Finthescalarcontext,whichinPerl

means“thenumberofelementsin@F.”Asaresult,itprintsthenumberofelementsoneachline.

Forexample,ifyourfilecontainsthefollowinglines:

foo bar baz

foo bar

baz

Runningthisone-linerproducesthefollowingoutput:

3

2

1

Thefirstlinehasthreefields,thesecondlinehastwofields,andthelastlinehasonefield.

4.11Printthetotalnumberoffieldsoneachline,followedbytheline

perl -alne 'print scalar @F, " $_"'

Thisone-lineristhesameasone-liner4.10,withtheadditionof$_attheend,whichprintsthewholeline.(Rememberthat-nputseachlineinthe$_variable.)

Let’srunthisone-lineronthesameexamplefilethatIusedinone-liner4.10:

foo bar baz

foo bar

baz

Runningtheone-linerproducesthefollowingoutput:

3 foo bar baz

2 foo bar

1 baz

4.12Printthetotalnumberoffieldsonalllines

perl -alne '$t += @F; END { print $t }'

Here,theone-linerkeepsaddingthenumberoffieldsoneachlinetovariable$tuntilalllineshavebeenprocessed.Next,itprintstheresult,whichcontainsthenumberofwordsonalllines.Noticethatyouaddthe@Farraytothescalarvariable$t.Because$tisscalar,the@Farrayisevaluatedinthescalarcontextandreturnsthenumberofelementsitcontains.

Runningthisone-lineronthefollowingfile:

foo bar baz

foo bar

baz

producesthenumber6asoutputbecausethefilecontainsatotalofsixwords.

4.13Printthetotalnumberoffieldsthatmatchapattern

perl -alne 'map { /regex/ && $t++ } @F; END { print $t || 0 }'

Thisone-linerusesmaptoapplyanoperationtoeachelementinthe@Farray.Inthisexample,theoperationcheckstoseeifeachelementmatches/regex/,andifitdoes,itincrementsthe$tvariable.Itthenprintsthe$tvariable,whichcontainsthenumberoffieldsthatmatchthe/regex/pattern.The$t || 0constructisnecessarybecauseifnofieldsmatch,$twouldn’texist,soyoumustprovideadefaultvalue.Insteadof0,youcanprovideanyotherdefaultvalue,evenastring!

Loopingwouldbeabetterapproach:

perl -alne '$t += /regex/ for @F; END { print $t }'

Here,eachelementin@Fistestedagainst/regex/.Ifitmatches,/regex/

returnstrue;otherwiseitreturnsfalse.Whenusednumerically,trueconvertsto1andfalseconvertsto0,so$t += /regex/addseither1or0tothe$tvariable.Asaresult,thenumberofmatchesiscountedin$t.YoudonotneedadefaultvaluewhenprintingtheresultintheENDblockbecausethe+=operatorisrunregardlessofwhetherthefieldmatches.Youwillalwaysgetavalue,andsometimesthatvaluewillbe0.

Anotherwaytodothisistousegrepinthescalarcontext:

perl -alne '$t += grep /regex/, @F; END { print $t }'

Here,grepreturnsthenumberofmatchesbecauseit’sevaluatedinthescalarcontext.Inthelistcontext,grepreturnsallmatchingelements,butinthescalarcontext,itreturnsthenumberofmatchingelements.Thisnumberisaccumulatedin$tandprintedintheENDblock.Inthiscase,youdon’tneedtoprovideadefaultvaluefor$tbecausegrepreturns0inthosesituations.

4.14Printthetotalnumberoflinesthatmatchapattern

perl -lne '/regex/ && $t++; END { print $t || 0 }'

Here,/regex/evaluatestotrueifthecurrentlineofinputmatchesthisregularexpression.Writing/regex/ && $t++isthesameaswritingif ($_ =~/regex/) { $t++ },whichincrementsthe$tvariableifthelinematchesthespecifiedpattern.IntheENDblock,the$tvariablecontainsthetotalnumberofpatternmatchesandisprinted;butifnolinesmatch,$tisonceagainundefined,soyoumustprintadefaultvalue.

4.15Printthenumberπ

perl -Mbignum=bpi -le 'print bpi(21)'

Thebignumpackageexportsthebpifunctionthatcalculatestheπconstanttothedesiredaccuracy.Thisone-linerprintsπto20decimalplaces.(Noticethatyouneedtospecifyn+1toprintittoanaccuracyofn.)

Thebignumlibraryalsoexportstheconstantπ,precomputedto39decimalplaces:

perl -Mbignum=PI -le 'print PI'

4.16Printthenumbere

perl -Mbignum=bexp -le 'print bexp(1,21)'

Thebignumlibraryexportsthebexpfunction,whichtakestwoarguments:thepowertoraiseeto,andthedesiredaccuracy.Thisone-linerprintstheconstanteto20decimalplaces.

Forexample,youcouldprintthevalueofe2to30decimalplaces:

perl -Mbignum=bexp -le 'print bexp(2,31)'

Aswithπ,bignumalsoexportstheconstanteprecomputedto39decimalplaces:

perl -Mbignum=e -le 'print e'

4.17PrintUNIXtime(secondssinceJanuary1,1970,00:00:00UTC)

perl -le 'print time'

Thebuilt-intimefunctionreturnssecondssincetheepoch.Thisone-linersimplyprintsthetime.

4.18PrintGreenwichMeanTimeandlocalcomputertime

computertime

perl -le 'print scalar gmtime'

Thegmtimefunctionisabuilt-inPerlfunction.Whenusedinthescalarcontext,itreturnsthetimelocalizedtoGreenwichMeanTime(GMT).

Thebuilt-inlocaltimefunctionactslikegmtime,exceptitreturnsthecomputer’slocaltimewhenit’susedinthescalarcontext:

perl -le 'print scalar localtime'

Inthelistcontext,bothgmtimeandlocaltimereturnanine-elementlist(knownasstruct tmtoUNIXprogrammers)withthefollowingelements:

($second, [0]

$minute, [1]

$hour, [2]

$month_day, [3]

$month, [4]

$year, [5]

$week_day, [6]

$year_day, [7]

$is_daylight_saving [8]

)

Youcanslicethislist(thatis,extractelementsfromit)orprintindividualelementsifyouneedjustsomepartoftheinformationitcontains.Forexample,toprintH:M:S,slicetheelements2,1,and0fromlocaltime,likethis:

perl -le 'print join ":", (localtime)[2,1,0]'

Tosliceelementsindividually,specifyalistofelementstoextract,forinstance[2,1,0].Orslicethemasarange:

perl -le 'print join ":", (localtime)[2..6]'

Thisone-linerprintsthehour,date,month,year,anddayoftheweek.

Youcanalsousenegativeindexestoselectelementsfromtheoppositeendofalist:

list:

perl -le 'print join ":", (localtime)[-2, -3]'

Thisone-linerprintselements7and6,whicharethedayoftheyear(forexample,the200thday)andoftheweek(forexample,the4thday),respectively.

4.19Printyesterday’sdate

perl -MPOSIX -le '

@now = localtime;

$now[3] -= 1;

print scalar localtime mktime @now

'

Rememberthatlocaltimereturnsanine-itemlist(seeone-liner4.18)ofvariousdateelements.Thefourthelementinthelististhecurrentmonth’sday.Ifyousubtract1fromthiselement,yougetyesterday.

ThemktimefunctionconstructstheUNIXepochtimefromthismodifiednine-elementlist,andthescalar localtimeconstructprintsthenewdate,whichisyesterday.Thisone-lineralsoworksinedgecases,suchaswhenthecurrentdayisthefirstdayofthemonth.YouneedthePOSIXpackagebecauseitexportsthemktimefunction.

Forexample,ifit’sMonMay2005:49:55rightnow,runningthisone-linerprintsSunMay1905:49:55.

4.20Printthedate14months,9days,and7secondsago

perl -MPOSIX -le '

@now = localtime;

$now[0] -= 7;

$now[3] -= 9;

$now[4] -= 14;

print scalar localtime mktime @now

'

'

Thisone-linermodifiesthefirst,fourth,andfifthelementsofthe@nowlist.Thefirstelementisseconds,thefourthisdays,andthefifthismonths.ThemktimecommandgeneratestheUNIXtimefromthisnewstructure,andlocaltime,whichisevaluatedinthescalarcontext,printsthedate14months,9days,and7secondsago.

4.21Calculatethefactorial

perl -MMath::BigInt -le 'print Math::BigInt->new(5)->bfac()'

Thisone-linerusesthebfac()functionfromtheMath::BigIntmoduleinthePerlcore.(Inotherwords,youdon’tneedtoinstallit.)TheMath::BigInt->new(5)constructioncreatesanewMath::BigIntobjectwithavalueof5,afterwhichthebfac()methodiscalledonthenewlycreatedobjecttocalculatethefactorialof5.Change5toanynumberyouwishtofinditsfactorial.

Anotherwaytocalculateafactorialistomultiplythenumbersfrom1tontogether:

perl -le '$f = 1; $f *= $_ for 1..5; print $f'

Here,Iset$fto1andthenloopfrom1to5andmultiply$fbyeachvalue.Theresultis120(1*2*3*4*5),thefactorialof5.

4.22Calculatethegreatestcommondivisor

perl -MMath::BigInt=bgcd -le 'print bgcd(@list_of_numbers)'

Math::BigInthasseveralotherusefulmathfunctionsincludingbgcd,whichcalculatesthegreatestcommondivisor(gcd)ofalistofnumbers.Forexample,tofindthegreatestcommondivisorof(20, 60, 30),executetheone-linerlikethis:

perl -MMath::BigInt=bgcd -le 'print bgcd(20,60,30)'

perl -MMath::BigInt=bgcd -le 'print bgcd(20,60,30)'

Tocalculatethegcdfromafileoruser’sinput,usethe-acommand-lineargumentandpassthe@Farraytothebgcdfunction:

perl -MMath::BigInt=bgcd -anle 'print bgcd(@F)'

(Iexplainedthe-aargumentandthe@Farrayinone-liner4.2onpage30.)

YoucouldalsouseEuclid’salgorithmtofindthegcdof$nand$m.Thisone-linerdoesjustthatandstorestheresultin$m:

perl -le '

$n = 20; $m = 35;

($m,$n) = ($n,$m%$n) while $n;

print $m

'

Euclid’salgorithmisoneoftheoldestalgorithmsforfindingthegcd.

4.23CalculatetheleastcommonmultipleTheleastcommonmultiple(lcm)function,blcm,isincludedinMath::BigInt.Usethisone-linertofindtheleastcommonmultipleof(35, 20, 8):

perl -MMath::BigInt=blcm -le 'print blcm(35,20,8)'

Tofindthelcmfromafilewithnumbers,usethe-acommand-lineswitchandthe@Farray:

perl -MMath::BigInt=blcm -anle 'print blcm(@F)'

Ifyouknowabitofnumbertheory,youmayrecallthatthereisaconnectionbetweenthegcdandthelcm.Giventwonumbers$nand$m,youknowthattheirlcmis$n*$m/gcd($n,$m).Thisone-liner,therefore,follows:

perl -le '

$a = $n = 20;

$b = $m = 35;

$b = $m = 35;

($m,$n) = ($n,$m%$n) while $n;

print $a*$b/$m

'

4.24Generate10randomnumbersbetween5and15(excluding15)

perl -le 'print join ",", map { int(rand(15-5))+5 } 1..10'

Thisone-linerprints10randomnumbersbetween5and15.Itmaylookcomplicated,butit’sactuallysimple.int(rand(15-5))isjustint(rand(10)),whichreturnsarandomintegerfrom0to9.Adding5toitmakesitreturnarandomintegerfrom5to14.Therange1..10makesitdraw10randomintegers.

Youcanalsowritethisone-linermoreverbosely:

perl -le '

$n=10;

$min=5;

$max=15;

$, = " ";

print map { int(rand($max-$min))+$min } 1..$n;

'

Here,allvariablesaremoreexplicit.Tomodifythisone-liner,changethevariables$n,$min,and$max.The$nvariablerepresentshowmanyrandomnumberstogenerate,and$min-$maxistherangeofnumbersforuseinthatgeneration.

The$,variableissettoaspacebecauseit’stheoutputfieldseparatorforprintandit’sundefbydefault.Ifyoudidn’tset$,toaspace,thenumberswouldbeprintedconcatenated.(Seeone-liner4.4onpage32foradiscussionof$,.)

4.25Generateallpermutationsofalist

perl -MAlgorithm::Permute -le '

$l = [1,2,3,4,5];

$l = [1,2,3,4,5];

$p = Algorithm::Permute->new($l);

print "@r" while @r = $p->next

'

Thisone-linerusestheobject-orientedinterfaceofthemoduleAlgorithm::Permutetofindallpermutationsofalist,thatis,allwaystorearrangeitems.TheconstructorofAlgorithm::Permutetakesanarrayreferenceofelementstopermute.Inthisparticularone-liner,theelementsarethenumbers1, 2, 3, 4, 5.

Thenextmethodreturnsthenextpermutation.Callingitrepeatedlyiteratesoverallpermutations,andeachpermutationisputinthe@rarrayandthenprinted.(Beware:Theoutputlistgetslargereallyquickly.Therearen!(nfactorial)permutationsforalistofnelements.)

Anotherwaytoprintallpermutationsiswiththepermutesubroutine:

perl -MAlgorithm::Permute -le '

@l = (1,2,3,4,5);

Algorithm::Permute::permute { print "@l" } @l

'

Here’swhatyougetifyouchange@ltojustthreeelements(1, 2, 3)andrunit:

1 2 3

1 3 2

3 1 2

2 1 3

2 3 1

3 2 1

4.26Generatethepowerset

perl -MList::PowerSet=powerset -le '

@l = (1,2,3,4,5);

print "@$_" for @{powerset(@l)}

'

Thisone-linerusestheList::PowerSetmodulefromCPAN.Themoduleexportsthepowersetfunction,whichtakesalistofelementsandreturnsareferencetoanarraycontainingreferencestosubsetarrays.Youcaninstallthismodulebyrunningcpan List::PowerSetatthecommandline.

Intheforloop,youcallthepowersetfunctionandpassitthelistofelementsof@l.Next,youdereferencethereturnvalueofpowerset,whichisareferencetoanarrayofsubsets,andthendereferenceeachindividualsubset@$_andprintit.

Thepowersetisthesetofallsubsets.Forasetofnelements,thereareexactly2n

subsetsinthepowerset.Here’sanexampleofthepowersetof(1, 2, 3):

1 2 3

2 3

1 3

3

1 2

2

1

4.27ConvertanIPaddresstoanunsignedinteger

perl -le '

$i=3;

$u += ($_<<8*$i--) for "127.0.0.1" =~ /(\d+)/g;

print $u

'

Thisone-linerconvertstheIPaddress127.0.0.1intoanunsignedintegerbyfirstdoingaglobalmatchof(\d+)ontheIPaddress.Performingaforloopoveraglobalmatchiteratesoverallthematches,whicharethefourpartsoftheIPaddress:127,0,0,1.

Next,thematchesaresummedinthe$uvariable.Thefirstbitisshifted8×3=24places,thesecondisshifted8×2=16places,andthethirdisshifted8places.Thelastissimplyaddedto$u.Theresultingintegerhappenstobe2130706433(averygeekynumber).

Herearesomemoreone-liners:

Herearesomemoreone-liners:

perl -le '

$ip="127.0.0.1";

$ip =~ s/(\d+)\.?/sprintf("%02x", $1)/ge;

print hex($ip)

'

Thisone-linerutilizesthefactthat127.0.0.1canbeeasilyconvertedtohex.Here,the$ipismatchedagainst(\d+),andeachIPpartistransformedintoahexnumberwithsprintf("%02x", $1)insidethesoperator.The/eflagofthesoperatormakesitevaluatethesubstitutionpartasaPerlexpression.Asaresult,127.0.0.1istransformedinto7f000001andtheninterpretedasahexadecimalnumberbyPerl’shexoperator,whichconvertsittoadecimalnumber.

Youcanalsouseunpack:

perl -le 'print unpack("N", 127.0.0.1)'

Thisone-linerisprobablyasshortaspossible.Itusesvstringliterals(versionstrings)toexpresstheIPaddress.Avstringformsastringliteralcomposedofcharacterswiththespecifiedordinalvalues.Thenewlyformedstringliteralisunpackedintoanumberfromastringinnetworkbyteorder(big-endianorder)andthenprinted.

IfyouhaveastringwithanIP(ratherthanavstring),youfirsthavetoconvertittobyteformwiththefunctioninet_aton:

perl -MSocket -le 'print unpack("N", inet_aton("127.0.0.1"))'

Here,inet_atonconvertsthestring127.0.0.1tothebyteform(equivalenttothepurevstring127.0.0.1)andthenunpackunpacksit,asinthepreviousone-liner.

4.28ConvertanunsignedintegertoanIPaddress

perl -MSocket -le 'print inet_ntoa(pack("N", 2130706433))'

Here,theinteger2130706433ispackedintoanumberinbig-endianbyteorderandthenpassedtotheinet_ntoafunctionthatconvertsanumberbacktoanIPaddress.(Noticethatinet_ntoaistheoppositeofinet_aton.)

Youcandothesamethinglikethis:

perl -le '

$ip = 2130706433;

print join ".", map { (($ip>>8*($_))&0xFF) } reverse 0..3

'

Here,the$ipisshifted24bitstotherightandthenbitwiseANDedwith0xFFtoproducethefirstpartoftheIP,whichis127.Next,it’sshifted16bitsandbitwiseANDedwith0xFF,producing0,andthenshifted8bitsandbitwiseANDedwith0xFF,producinganother0.Finally,thewholenumberisbitwiseANDedwith0xFF,producing1.

Theresultfrommap { ... }isalist(127, 0, 0, 1).Thatlistisnowjoinedbyadot"."toproducetheIPaddress127.0.0.1.

Youcanreplacejoinwiththespecialvariable$,,whichactsasavalueseparatorfortheprintstatement:

perl -le '

$ip = 2130706433;

$, = ".";

print map { (($ip>>8*($_))&0xFF) } reverse 0..3

'

Becausereverse 0..3isthesameas3,2,1,0,youcouldalsowrite:

perl -le '

$ip = 2130706433;

$, = ".";

print map { (($ip>>8*($_))&0xFF) } 3,2,1,0

'

Chapter5.WorkingWithArraysandStrings

Inthischapter,we’lllookatvariousone-linersforcreatingstringsandarrays,fordoingthingslikegeneratingpasswords,creatingstringsofcertainlength,findingthenumericvaluesofcharacters,andcreatingarraysofnumbers.You’llalsolearnabouttherangeoperator..,thexoperator,the$,specialvariable,andthe@ARGVarray.

5.1Generateandprintthealphabet

perl -le 'print a..z'

Thisone-linerprintsalllettersfromatozintheEnglishalphabetasabcdefghijklmnopqrstuvwxyz.Thelettersaregeneratedbytherangeoperator..,which,whenusedonstringsinthelistcontext(providedherebyprint),appliesthemagicalauto-incrementalgorithmthatadvancesthestringtothenextcharacter.Therefore,inthisone-liner,theauto-incrementalgorithmontherangea..zproducesalllettersfromatoz.

Ireallygolfedthisone-liner.IfIhadusedstrict,itwouldn’thaveworkedbecauseofthebarewordsaandz.Thisversionismorecorrectsemantically:

perl -le 'print ("a".."z")'

Rememberthattherangeoperator..producesalistofvalues.Ifyouwish,youcouldprintthevalueswithcommaseparationsbysettingthe$,specialvariable:

perl -le '$, = ","; print ("a".."z")'

The$,isthefieldseparator.It’soutputbyprintbetweeneachfield.Semantically,though,usingjointoseparatethelistofletterswithacommaismoreappealingbecauseitworksevenwhennotusingprintdirectly:

perl -le '$alphabet = join ",", ("a".."z"); print $alphabet'

Here,thelista..zisjoinedbyacommabeforeprinting,andtheoutputis

a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z

5.2Generateandprintallthestringsfrom“a”to“zz”

perl -le 'print join ",", ("a".."zz")'

Thisone-linerusestherangeoperator..again,butthistime,itdoesn’tstopatzasinthepreviousone-liner.Instead,itadvanceszbyonecharacter,producingaa.Thenitkeepsgoing,producingab,ac,andsoon,untilithitsaz.Atthatpoint,itadvancesthestringtoba,continueswithbb,bc,andsoon,untileventuallyitreacheszz.

Youcouldalsogenerateallstringsfromaatozzbydoingthis:

perl -le 'print join ",", "aa".."zz"'

Theoutputfromthisone-lineris

aa, ab, ..., az, ba, bb, ..., bz, ca, ..., zz

5.3Createahexlookuptable

@hex = (0..9, "a".."f")

Inthisone-liner,the@hexarrayisfilledwiththenumbers0,1,2,3,4,5,6,7,8,9andthelettersa,b,c,d,e,f.Youcouldusethisarraytoconvertanumber(in

variable$num)fromdecimaltohexwiththefollowingbaseconversionformula.(Thisisn’taone-liner;Iincludeittoillustratehowtousethe@hexlookuparray.)

perl -le '

$num = 255;

@hex = (0..9, "a".."f");

while ($num) {

$s = $hex[($num % 16)].$s;

$num = int $num/16;

}

print $s

'

Butsurely,convertinganumbertohexismucheasierifIuseprintf(orsprintf)withthe%xformatspecifier.

perl -le 'printf("%x", 255)'

Toconvertthenumberbackfromhextodec,usethehexoperator:

perl -le '$num = "ff"; print hex $num'

Thehexoperatortakesahexstring(beginningwithorwithout0x)andconvertsittodecimal.

5.4Generatearandomeight-characterpassword

perl -le 'print map { ("a".."z")[rand 26] } 1..8'

Here,themapoperatorexecutesthecode("a".."z")[rand 26]eighttimesbecauseititeratesovertherange1..8.Ineachiteration,thecodechoosesarandomletterfromthealphabet.Whenmaphasfinishediterating,itreturnsthegeneratedlistofcharacters,andprintprintsit,therebyconcatenatingallthecharacters.

Toalsoincludenumbersinthepassword,add0..9tothelistofcharactersto

choosefromandchange26to36becauseyounowhave36possiblecharacters:

perl -le 'print map { ("a".."z", 0..9)[rand 36] } 1..8'

Ifyouneedalongerpassword,change1..8to1..20togenerateonethat’s20characterslong.

5.5Createastringofspecificlength

perl -le 'print "a"x50'

Thisone-linercreatesastringof50lettersaandprintsit.Operatorxistherepetitionoperator.Here,theletteraisrepeated50timesbyx50.Thisone-linerishandywhenyouneedtogenerateaspecificamountofdatafordebuggingorothertasks.Forexample,ifyouneed1KBofdata,justdothis:

perl -e 'print "a"x1024'

Iremovedthe-largumentbecauseitwouldhaveoutputanadditionalnewlinesymbol,producing1025bytesofdata.

Whenyouusetherepetitionoperatorinthelistcontext,withalistasitsfirstoperand,youcreatealistwiththegivenelementsrepeated,likethis:

perl -le '@list = (1,2)x20; print "@list"'

Thisone-linercreatesalistof20repetitionsof(1, 2)thatlookslike(1, 2,1, 2, 1, 2, ...).(Theparenthesestotheleftofxmakealist.)

5.6Createanarrayfromastring

@months = split ' ', "Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec"

Here,@monthsisfilledwithvaluesfromthestringcontainingmonthnames.Becauseallmonthnamesareseparatedbyaspace,thesplitoperatorsplits

themandputsthemin@months.Asaresult,$months[0]containsJan,$months[1]containsFeb,…,and$months[11]containsDec.

Youcoulddothesamethingwiththeqw/.../operator:

@months = qw/Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec/

Theqw/.../operatortakesaspace-separatedstringandcreatesanarrayinwhicheachwordisanarrayelement.

Althoughnotaone-linerperse,thisisauseful,idiomaticwaytocreatearraysthatcancomeinhandywhenwritingone-liners.

5.7Createastringfromthecommand-linearguments

perl -le 'print "(", (join ",", @ARGV), ")"' val1 val2 val3

Thisone-linerusesthe@ARGVarray,whichcontainsalltheargumentsthathavebeenpassedtoPerl.Inthisone-liner,thevaluespassedtoPerlareval1,val2,andval3,so@ARGVcontainsthestringsval1,val2,andval3.Thisone-linerprintsthestring(val1,val2,val3)andwouldbeuseful,forexample,togenerateaSQLquery.

Ifyou’refamiliarwiththeINSERTqueryinSQL,youknowitsmostbasicformlookslikeINSERT INTO table VALUES (val1, val2, val3, ...).Asyoucansee,thisone-linergeneratestheVALUESpartoftheSQLquery.

Youcaneasilymodifythisone-linertoprintthewholeINSERTquery:

perl -le '

print "INSERT INTO table VALUES (", (join ",", @ARGV), ")"

' val1 val2 val3

Here’swhattheone-linerprints:

INSERT INTO table VALUES (val1,val2,val3)

5.8Findthenumericvaluesforcharactersinastring

perl -le 'print join ", ", map { ord } split //, "hello world"'

Thisone-linertakesthestring"hello world"andsplitsitintoalistofcharacterswithsplit //, "hello world".Itthenmapstheordoperatorontoeachofthecharacters,whichreturnseachcharacter’snumericvalue.Finally,allofthenumericvaluesarejoinedtogetherbyacommaandprinted.Here’stheoutput:

104, 101, 108, 108, 111, 32, 119, 111, 114, 108, 100

YoucouldalsodothiswiththeunpackoperatorbyspecifyingC*astheunpackingtemplate:

perl -le 'print join ", ", unpack("C*", "hello world")'

TheCinthetemplatemeans“unsignedcharacter”and*means“allcharacters.”

Tofindthehexadecimalvaluesofthecharacters,youcoulddothis:

perl -le '

print join ", ", map { sprintf "0x%x", ord $_ } split //, "hello world"

'

Here,themapoperatorexecutessprintf "0x%x", ord $_foreverycharacter,whichreturnsthecharacter’shexadecimalvalueprependedwith'0x'.Here’stheoutput:

0x68, 0x65, 0x6c, 0x6c, 0x6f, 0x20, 0x77, 0x6f, 0x72, 0x6c, 0x64

Similarly,togettheoctalvaluesofcharacters,youdothis:

perl -le '

print join ", ", map { sprintf "%o", ord $_ } split //, "hello world"

'

Here’stheoutput:

150, 145, 154, 154, 157, 40, 167, 157, 162, 154, 144

Andfinally,togenerateproperoctalvaluesthatbeginwith0,youcanspecifythe%#oformattothesprintffunction:

perl -le '

print join ", ", map { sprintf "%#o", ord $_ } split //, "hello world"

'

Andhere’stheoutput:

0150, 0145, 0154, 0154, 0157, 040, 0167, 0157, 0162, 0154, 0144

5.9ConvertalistofnumericASCIIvaluesintoastring

perl -le '

@ascii = (99, 111, 100, 105, 110, 103);

print pack("C*", @ascii)

'

JustasIunpackedastringintoalistofvalueswiththeC*templateinthepreviousone-liner,Icanpackthemintoastringbyusingthesametemplate.Here’stheoutputfromtheone-liner:

coding

AnotherwaytoconvertalistofnumericASCIIvaluesintoastringistousethechroperator,whichtakesthecodepointvalueandreturnsthecorrespondingcharacter:

perl -le '

@ascii = (99, 111, 100, 105, 110, 103);

$str = join "", map chr, @ascii;

print $str

'

Here,yousimplymapthechroperatorontoeachnumericvalueinthe@asciiarray,whichproducesalistofcharactersthatcorrespondtothenumericvalues.Next,youjointhecharacterstogetherandproducethe$str,andthenyouprintitout.

Youcanalsogolfthisone-linerandcomeupwiththefollowing:

perl -le 'print map chr, 99, 111, 100, 105, 110, 103'

Youcanalsousethe@ARGVarrayandpasstheASCIIvaluesasargumentstotheone-liner:

perl -le 'print map chr, @ARGV' 99 111 100 105 110 103

5.10Generateanarraywithoddnumbersfrom1to100

perl -le '@odd = grep {$_ % 2 == 1} 1..100; print "@odd"'

Thisone-linergeneratesanarrayofoddnumbersfrom1to99(thatis,1,3,5,7,9,11,…,99).Itusesgreptoevaluatethecode$_ % 2 == 1foreachelementinthelist1..100andreturnsonlythoseelementsforwhichthegivencodeevaluatestotrue.Inthiscase,thecodeteststoseeiftheremainderwhendividingby2is1.Ifitis,thenumberisodd,andit’sputinthe@oddarray.

Youcouldalsowritethisusingthefactthatoddnumbershavetheleastsignificantbitsetandtestfortheleastsignificantbit:

perl -le '@odd = grep { $_ & 1 } 1..100; print "@odd"'

Theexpression$_ & 1isolatestheleastsignificantbit,andgrepselectsonlythosenumberswiththeleastsignificantbitset—thatis,alloddnumbers.

5.11Generateanarraywithevennumbersfrom1to100

perl -le '@even = grep {$_ % 2 == 0} 1..100; print "@even"'

Thisone-linerisalmostthesameastheonein5.10,exceptthatgreptestsforthecondition“isthenumbereven(remainderafterdividingbytwois0)?”

5.12Findthelengthofastring

perl -le 'print length "one-liners are great"'

Thelengthsubroutinefindsthelengthofastring.

5.13Findthenumberofelementsinanarray

perl -le '@array = ("a".."z"); print scalar @array'

Evaluatinganarrayinthescalarcontextreturnsitsnumberofelements.

Youcouldalsodothisbyadding1tothelastindexofanarray:

perl -le '@array = ("a".."z"); print $#array + 1'

Here,$#arrayreturnsthelastindexin@array.Becausethatnumberisonelessthanthenumberofelementsinthearray,youadd1totheresulttofindthetotalnumberofelementsinthearray.

Forexample,sayyouwanttofindouthowmanytextfilesareinthecurrentdirectory.Youcanuse@ARGVandpassthe*.txtwildcardtoPerl.Theshellexpandsthe*.txtwildcardtoalistoffilenamesthatmatch*.txt,andPerlputsthemintothe@ARGVarrayandprintsthearrayinthescalarcontext.Theoutputwillbethenumberoftextfilesinthecurrentdirectory:

perl -le 'print scalar @ARGV' *.txt

Ifyourshelldoesn’tsupportfilenameexpansion(alsoknownasglobbing)orifyou’reonWindows,youcanusethediamondoperatorwiththe*.txtargument:

perl -le 'print scalar (@ARGV=<*.txt>)'

Inthiscase,thediamondoperatordoestheglobbingandreturnsalistoffilenamesthatmatch*.txt.Evaluatingthislistinthescalarcontextreturnsthenumberoffilesthatmatched.

Chapter6.TextConversionandSubstitution

Inthischapter,we’lllookatvariousone-linersthatchange,convert,andsubstitutetext,includingbase64encodinganddecoding,URLescapingandunescaping,HTMLescapingandunescaping,convertingtextcase,andreversinglines.You’llalsogettoknowthey,tr,uc,lc,andreverseoperatorsandstring-escapesequences.

6.1ROT13astring

perl -le '$string = "bananas"; $string =~ y/A-Za-z/N-ZA-Mn-za-m/; print $string'

Thisone-linerusestheyoperator(alsoknownasthetroperator)todoROT13.Theoperatorsyandtrperformstringtransliteration.Giveny/search/replace/,theyoperatortransliteratesalloccurrencesofthecharactersfoundinthesearchlistwiththecharactersinthecorrespondingpositionsinthereplacelist.Theyandtroperatorsareoftenmistakenfortakingaregularexpression,buttheydon’t.Theytransliteratethingsandtakealistofcharactersinboththesearchandreplaceparts.

Inthisone-liner,A-Za-zcreatesthefollowinglistofcharacters:

ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz

AndN-ZA-Mn-za-mcreatesthislist:

NOPQRSTUVWXYZABCDEFGHIJKLMnopqrstuvwxyzabcdefghijklm

Noticethatinthesecondlisttheuppercaseandlowercasealphabetsareoffsetby

13characters.Nowtheyoperatortranslateseachcharacterinthefirstlisttoacharacterinthesecondlist,thusperformingtheROT13operation.(OnefunfactaboutROT13isthatapplyingittwiceproducesthesamestring;thatis,ROT13(ROT13(string))equalsstring.)

ToROT13thewholefilebananas.txtandprintittothescreen,justdothis:

perl -lpe 'y/A-Za-z/N-ZA-Mn-za-m/' bananas.txt

YoucanalsousePerl’s-iargumenttodoin-placereplacementofthefile.Forexample,toROT13oranges.txtin-place,writethis:

perl -pi.bak -e 'y/A-Za-z/N-ZA-Mn-za-m/' oranges.txt

Thisone-linerfirstcreatesabackupfilecalledoranges.txt.bakandthenreplacesthecontentsoforanges.txtwithROT13-edtext.The.bakpartofthe-icommandcreatesthebackupfile.Youcanomitthe.bakpartofthecommandifyou’resureabouttheresult,butIrecommendalwaysusing-i.bakbecauseonedayyoumightmakeamistakeandmessupanimportantfile.(Ispeakfromexperience.)

6.2Base64-encodeastring

perl -MMIME::Base64 -e 'print encode_base64("string")'

Thisone-linerusestheMIME::Base64module.Itexportstheencode_base64function,whichtakesastringandreturnsabase64-encodedversionofit.

Tobase64-encodethewholefile,usethis:

perl -MMIME::Base64 -0777 -ne 'print encode_base64($_)' file

Here,the-0777argumenttogetherwith-ncausesPerltoslurpthewholefileintothe$_variable.Next,thefileisbase64-encodedandprinted.(IfPerldidn’tslurptheentirefile,itwouldbeencodedlinebyline,andyou’dendupwithamess.)

6.3Base64-decodeastring

perl -MMIME::Base64 -le 'print decode_base64("base64string")'

TheMIME::Base64modulealsoexportsthedecode_base64function,whichtakesabase64-encodedstringanddecodesit.

Theentirefilecanbedecodedsimilarlywith

perl -MMIME::Base64 -0777 -ne 'print decode_base64($_)' file

6.4URL-escapeastring

perl -MURI::Escape -le 'print uri_escape("http://example.com")'

Tousethisone-liner,youneedtoinstalltheURI::Escapemodulefirstbyenteringcpan URI::Escapeintheshell.Themoduleexportstwofunctions:uri_escapeanduri_unescape.ThefirstfunctionperformsURLescaping(sometimesreferredtoasURLencoding),andtheotherdoesURLunescaping(orURLdecoding).Now,toURL-escapeastring,justcalluri_escape($string)andyou’redone!

Theoutputofthisone-linerishttp%3A%2F%2Fexample.com.

6.5URL-unescapeastring

perl -MURI::Escape -le 'print uri_unescape("http%3A%2F%2Fexample.com")'

Thisone-linerusestheuri_unescapefunctionfromtheURI::EscapemoduletoperformURLunescaping.Itunescapestheoutputofthepreviousone-linertoreversetheoperation.

Theoutputofthisone-linerishttp://example.com.

6.6HTML-encodeastring

perl -MHTML::Entities -le 'print encode_entities("<html>")'

Thisone-linerusestheencode_entitiesfunctionfromtheHTML::EntitiesmoduletoencodeHTMLentities.Forexample,youcanturn<and>into&lt;and&gt;.

6.7HTML-decodeastring

perl -MHTML::Entities -le 'print decode_entities("&lt;html&gt;")'

Thisone-linerusesthedecode_entitiesfunctionfromtheHTML::Entitiesmodule.Forexample,youcanturn&lt;and&gt;backinto<and>.

6.8Convertalltexttouppercase

perl -nle 'print uc'

Thisone-linerusestheucfunction,which,bydefault,operatesonthe$_variableandreturnsanuppercaseversionofthetextitcontains.

Youcoulddothesamethingwiththe-pcommand-lineoption,whichenablestheautomaticprintingofthe$_variableandmodifiesitin-place:

perl -ple '$_ = uc'

Oryoucanapplythe\Uescapesequencetostringinterpolation:

perl -nle 'print "\U$_"'

Thisone-lineruppercaseseverythingthatfollowsit(oreverythinguptothefirstoccurrenceof\E).

6.9Convertalltexttolowercase

perl -nle 'print lc'

perl -nle 'print lc'

Thisone-linerissimilartothepreviousone.Thelcfunctionconvertsthecontentsof$_tolowercase.

Youcouldalsousetheescapesequence\Landstringinterpolation:

perl -nle 'print "\L$_"'

Here,\Llowercaseseverythingafterit(oruntilthefirstinstanceof\E).

6.10Uppercaseonlythefirstletterofeachline

perl -nle 'print ucfirst lc'

Thisone-linerfirstlowercasestheinputwiththelcfunctionandthenusesucfirsttouppercaseonlythefirstcharacter.Forexample,ifyoupassitalinethatsaysfoobarbaz,itproducesthetextFoobarbaz.Similarly,ifyoupassitalineFOOBARBAZ,itlowercasesthelinefirstandthenuppercasesthefirstletter,producingFoobarbazagain.

Youcandothesamethingusingescapecodesandstringinterpolation:

perl -nle 'print "\u\L$_"'

First\Llowercasesthewholelineandthen\uuppercasesthefirstcharacter.

6.11Invertthelettercase

perl -ple 'y/A-Za-z/a-zA-Z/'

Thisone-linerchangesthecaseoftheletters:Thecapitallettersbecomelowercaseletters,andthelowercaselettersbecomecapitalletters.Forexample,thetextCowsareCOOLbecomescOWSAREcool.Thetransliterationoperatory(explainedinone-liner6.1onpage59)createsamappingfromcapitallettersA-Ztolowercaselettersa-zandamappingfromlowercaselettersa-ztocapital

lettersA-Z.

6.12Title-caseeachline

perl -ple 's/(\w+)/\u$1/g'

Thisone-linerattemptstotitle-caseastring,meaningthefirstletterofeachwordisuppercased;forexample,ThisTextIsWrittenInTitleCase.Thisone-linerworksbymatchingeverywordwith\w+andreplacingthematchedwordwith\u$1,whichuppercasesthefirstletteroftheword.

6.13Stripleadingwhitespace(spaces,tabs)fromthebeginningofeachline

perl -ple 's/^[ \t]+//'

Thisone-linerdeletesallwhitespacefromthebeginningofeverylinewiththehelpofthesubstitutionoperators.Givens/regex/replace/,itreplacesthematchedregexwiththereplacestring.Inthiscase,theregexis^[ \t]+,whichmeans“matchoneormorespacesortabsatthebeginningofthestring,”andreplaceisempty,meaning“replacethematchedpartwithanemptystring.”

Theregularexpressionclass[ \t]canalsobereplacedby\s+tomatchanywhitespace(includingtabsandspaces):

perl -ple 's/^\s+//'

6.14Striptrailingwhitespace(spaces,tabs)fromtheendofeachline

perl -ple 's/[ \t]+$//'

Thisone-linerdeletesallwhitespacefromtheendofeachline.Theregexofthe

soperatorsays“matchoneormorespacesortabsattheendofthestring.”Thereplacepartisemptyagain,whichmeans“erasethematchedwhitespace.”

Youcanalsoachievethesamebywriting:

perl -ple 's/\s+$//'

Here,youreplacewith[ \t]+$with\s+,asinone-liner6.13.

6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

perl -ple 's/^[ \t]+|[ \t]+$//g'

Thisone-linercombinesone-liners6.13and6.14.Itspecifiestheglobal/gflagtothesoperatorbecauseyouwantittodeletewhitespaceatthebeginningandtheendofthestring.Ifyoudon’tspecifythis,itdeleteswhitespaceonlyatthebeginning(ifthereiswhitespace)oronlyattheend(iftherewasnowhitespaceatthebeginning).

Youcanalsoreplace[ \t]+$with\s+andgetthesameresults:

perl -ple 's/^\s+|\s+$//g'

Writing\s+isshorterthanwriting[ \t]+.Andsstandsforspace,whichmakesiteasiertoremember.

6.16ConvertUNIXnewlinestoDOS/Windowsnewlines

perl -pe 's|\012|\015\012|'

Thisone-linersubstitutestheUNIXnewlinecharacter\012(LF)fortheWindows/DOSnewlinecharacter\015\012(CRLF)oneachline.Onenicefeatureofs/regex/replace/isthatitcantakecharactersotherthanforward

slashesasdelimiters.Here,itusesverticalpipestodelimitregexfromreplacetoimprovereadability.

Newlinesareusuallyrepresentedas\nandcarriagereturnsas\r,butacrossplatforms,themeaningsofthe\nand\rsequencescanvary.TheUNIXnewlinecharacter,however,isalwaysavailableas\012(LF),andthecarriage-returncharacterrepresentedby\risalwaysavailableas\015(CR).That’swhyyouusethosenumericcodes:Sometimesusingtheflexiblesequenceispreferable,butnothere.

6.17ConvertDOS/WindowsnewlinestoUNIXnewlines

perl -pe 's|\015\012|\012|'

Thisone-linerworksintheoppositedirectionfromone-liner6.16.IttakesWindowsnewlines(CRLF)andconvertsthemtoUNIXnewlines(LF).

6.18ConvertUNIXnewlinestoMacnewlines

perl -pe 's|\012|\015|'

MacOSpreviouslyused\015(CR)asnewlines.Thisone-linerconvertsUNIX’s\012(LF)toMacOS’s\015(CR).

6.19Substitute(findandreplace)“foo”with“bar”oneachline

perl -pe 's/foo/bar/'

Thisone-linerusesthes/regex/replace/commandtosubstitutethefirstoccurrenceoffoowithbaroneachline.

Toreplaceallfooswithbars,addtheglobal/gflag:

perl -pe 's/foo/bar/g'

6.20Substitute(findandreplace)“foo”with“bar”onlinesthatmatch“baz”

perl -pe '/baz/ && s/foo/bar/'

Thisone-linerisroughlyequivalentto

while (defined($line = <>)) {

if ($line =~ /baz/) {

$line =~ s/foo/bar/

}

}

Thisexpandedcodeputseachlineintothevariable$lineandthencheckstoseeifalineinthatvariablematchesbaz.Ifso,itreplacesfoowithbarinthatline.

Youcouldalsowrite

perl -pe 's/foo/bar/ if /baz/'

6.21Printparagraphsinreverseorder

perl -00 -e 'print reverse <>' file

Thisone-linerusesthe-00argumentdiscussedinone-liner2.7(page14)toturnparagraphslurpmodeon,meaningthatPerlreadstextparagraphbyparagraph,ratherthanlinebyline.Next,itusesthe<>operatortomakePerlreadtheinputfromeitherstandardinputorfilesspecifiedasarguments.Here,I’vespecifiedfileastheargumentsoPerlwillreadfileparagraphbyparagraph(thanksto-00).OncePerlfinishesreadingthefile,itreturnsallparagraphsasalistandcallsreversetoreversetheorderoftheparagraphlist.Finally,printprintsthelistofreversedparagraphs.

6.22Printalllinesinreverseorder

perl -lne 'print scalar reverse $_'

Thisone-linerevaluatesthereverseoperatorinthescalarcontext.Inthepreviousone-liner,yousawthatevaluatingreverseinthelistcontextreversesthewholelist,thatis,theorderoftheelements.Todothesameforscalarvaluessuchas$_thatcontainthewholeline,youhavetocallreverseinthescalarcontext.Otherwise,itsimplyreversesalistwithonlyoneelement,whichisthesamelist!Onceyou’vedonethat,yousimplyprintthereversedline.

Oftenyoucandropthe$_variablewhenusingoperatorsandPerlwillstillapplythefunctiononthe$_variable.Inotherwords,youcanrewritethesameone-lineras

perl -lne 'print scalar reverse'

Oryoucansubstitute-nfor-p,modifythe$_variable,andsetitsvaluetoreverse:

perl -lpe '$_ = reverse $_'

Youcanalsowritethisas

perl -lpe '$_ = reverse'

Here,$_isdroppedbecausemostPerloperatorsdefaultto$_whennotgivenanargument.

6.23Printcolumnsinreverseorder

perl -alne 'print "@{[reverse @F]}"'

Thisone-linerreversestheorderofcolumnsinafile.The-acommand-lineargumentsplitseachlineintocolumnsatspacesandputstheminthe@Farray,

whichisthenreversedandprinted.Thisone-linerissimilartoone-liner4.4onpage32;Iexplainedthe@{[ ... ]}constructthere.Itsimplyletsyouruncodeinsideofdoublequotes.Forexample,giventhefollowinginputfile:

one two three four

five six seven eight

theone-linerreversestheorderofthecolumns,andtheoutputisthefollowing:

four three two one

eight seven six five

Ifthecolumnsinyourinputareseparatedbyanycharacterotherthanaspace,youcanusethe-Fcommand-lineargumenttosetadifferentdelimiter.Forexample,giventhefollowinginputfile:

one:two:three:four

five:six:seven:eight

youcanaddthe-F:command-lineargumenttotheone-linerlikethis:

perl -F: -alne 'print "@{[reverse @F]}"'

anditproducesthisoutput:

four three two one

eight seven six five

Notice,however,thatthe:charactersaremissinginthisoutput.Togetthemback,youneedtomodifytheone-linerabitandsetthe$"variableto":",asshownhere:

perl -F: -alne '$" = ":"; print "@{[reverse @F]}"'

Thisproducestheexpectedoutput:

four:three:two:one

eight:seven:six:five

The$"variablechangesthecharacterthat’sprintedbetweenarrayelementswhenanarrayisinterpolatedwithinadouble-quotedstring.

Chapter7.SelectivelyPrintingandDeletingLines

Inthischapter,we’llexaminevariousone-linersthatprintanddeletecertainlines.Theseone-linerswill,forexample,printrepeatedlines,printtheshortestlineinafile,andprintlinesthatmatchcertainpatterns.

Buteveryone-linerthatprintscertainlinescanalsobeviewedasonethatdeletesthelinesthataren’tprinted.Forexample,aone-linerthatprintsalluniquelinesdeletesallrepeatedlines.I’lldiscussonlytheone-linersthatprintsomething,ratherthandeletesomething,becauseoneisalwaystheinverseoftheother.

7.1Printthefirstlineofafile(emulatehead-1)

perl -ne 'print; exit' file

Thisone-linerisquitesimple.Perlreadsthefirstlineintothe$_variable,thankstothe-noption,andthencallsprinttoprintthecontentsofthe$_variable.Thenitjustexits.That’sit.Thefirstlineisprintedandthat’swhatyouwant.

Youmightalsosaythatthisone-linerdeletesalllinesexceptthefirstone.Butdon’tworry.Thisparticularone-linerwon’tdeletethecontentsofthefileunlessyoualsospecifythe-icommand-lineargument,likethis:

perl -i -ne 'print; exit' file

AsIexplainedinChapter1andinone-liner6.1onpage59,the-iargumenteditsthefilein-place.Inthiscase,allthelinesinthefilewouldbedeletedexceptforthefirst.Whenusing-i,alwaysspecifyabackupextensiontoit,likethis:

perl -i.bak -ne 'print; exit' file

Thiswillcreateabackupfilefile.bakbeforethecontentsareoverwritten.

Youcanaddthe-icommand-lineargumenttoanyoftheone-linerstochangethefilecontent.Ifyoudon’tusethe-iargument,theone-linerssimplyprintthenewcontentofthefiletoscreenratherthanmodifyingthefile.

7.2Printthefirst10linesofafile(emulatehead-10)

perl -ne 'print if $. <= 10' file

Thisone-linerusesthe$.specialvariable,whichstandsfor“thecurrentlinenumber.”EachtimePerlreadsinaline,itincrements$.by1,soclearlythisone-linersimplyprintsthefirst10lines.

Thisone-linercanalsobewrittenwithouttheifstatement:

perl -ne '$. <= 10 && print' file

Here,printiscalledonlyiftheBooleanexpression$. <= 10istrue,andthisexpressionistrueonlyifthecurrentlinenumberislessthanorequalto10.

Another,thoughsomewhattrickier,waytodothisiswiththerangeoperator(..)inthescalarcontext:

perl -ne 'print if 1..10' file

TherangeoperatorinthescalarcontextreturnsaBooleanvalue.Theoperatorisbistable,likeaflip-flop,andemulatestheline-range(comma)operatorinsed,awk,andvarioustexteditors.Itsvalueisfalseaslongasitsleftoperandisfalse.Oncetheleftoperandistrue,therangeoperatoristrueuntiltherightoperandistrue,afterwhichtherangeoperatorbecomesfalseagain.Asaresult,thisbistableoperatorbecomestrueatthefirstline,staystrueuntilthetenthline,andthenbecomesandremainsfalse.

Afourthoptionistofollowthefirstexampleinthischapter:

Afourthoptionistofollowthefirstexampleinthischapter:

perl -ne 'print; exit if $. == 10' file

Here,Iputaconditiononexit,whichisthatthecurrentline(whichIjustprinted)isnumber10.

7.3Printthelastlineofafile(emulatetail-1)

perl -ne '$last = $_; END { print $last }' file

Printingthelastlineofafileistrickierthanprintingthefirstline,becauseyouneverknowwhichisthelastline.Asaresult,youalwayshavetokeepthelineyoujustreadinmemory.Inthisone-liner,youalwayssavethecurrentlineheldin$_intothe$lastvariable.WhenthePerlprogramends,itexecutesthecodeintheENDblock,whichprintsthelastlineread.

Here’sanotherwaytodothis:

perl -ne 'print if eof' file

Thisone-linerusestheeof(orend-of-file)function,whichreturns1ifthenextreadreturnstheend-of-file.Becausethenextreadafterthelastlineinthefilewillreturntheend-of-file,thisone-linerdoesthejob.ThenextreadmeansthatPerlwillattempttoreadacharacterfromthecurrentfile,andifreadingthecharacterfails,itwillsignalthattheend-of-filehasbeenreached,meaningthewholefilehasbeenread.Ifthereadsucceeds,Perlsecretlyputsthecharacterbackintheinputstreamasifnothinghadhappened.

7.4Printthelast10linesofafile(emulatetail-10)

perl -ne 'push @a, $_; @a = @a[@a-10..$#a] if @a>10; END { print @a }' file

Thisone-linerisabittricky.Here,youpusheachlinetothe@aarrayandthen

replace@awithasliceofitselfifthelistcontainsmorethan10elements.Thephrase@a = @a[@a-10..$#a]means“replace@awiththelast10elementsof@a.”Thebit@a-10causes@atobeevaluatedinthescalarcontext,andthusitreturnsthenumberofelementsinthearrayminus10.Theexpression$#aisthelastindexinthe@aarray.Andfinally,@a[@a-10..$#a]slices(returns)thelast10elementsofthearray,withwhichitoverwrites@aitselfsoitalwayscontainsonlythelast10elements.

Forexample,suppose@acontains(line1, line2, line3, line4),andyouwanttoprintthelastfourlinesofthefile.Whenyoureadthefifthline,thearraybecomes(line1, line2, line3, line4, line5),andthevalueof@a-4is1because@ainthescalarcontextis5.Butthevalueof$#ais4becauseit’sthelastindexinthearray.Thereforewhenyoutaketheslice@a[@a-4..$#a],itbecomes@a[1..4],whichdropsthefrontelementfromthearray,andthe@aarraybecomes(line2, line3, line4, line5).

Asimplerwaytowritethisiswithshift:

perl -ne 'push @a, $_; shift @a if @a>10; END { print @a }' file

Thisone-linerdoesn’tneedtoslice@abecauseyoucanguaranteethatif@a >10,then@a == 11.shiftisanoperatorthatremovesthefirstelementofanarray.Sointhisloop,youcansimplyshiftoffthefirststoredlinewhenyouhavemorethan10lines.

7.5Printonlylinesthatmatcharegularexpression

perl -ne '/regex/ && print'

Thisone-linerteststoseewhetherthecurrentlinematches/regex/.Ifso,the/regex/matchsucceedsandprintiscalled.

Insteadofusing&&,youcanalsouseiftoreversethe/regex/andprintstatements:

perl -ne 'print if /regex/'

7.6Printonlylinesthatdonotmatcharegularexpression

perl -ne '!/regex/ && print'

Thisone-linerinvertsthepreviousone-liner.Here,Itesttoseethatthelinedoesn’tmatch/regex/byinvertingthematchviathe!operator.Ifthelinedoesn’tmatch,Icallprinttoprinttheline.

Youcanalsowritethistheotherwayaround:

perl -ne 'print if !/regex/'

Andyoucanalsouseunlessinsteadofif !:

perl -ne 'print unless /regex/'

AnotherwaytowritethisistoapplyDeMorgan’slawto!/regex/ && print:

perl -ne '/regex/ || print'

7.7Printeverylineprecedingalinethatmatchesaregularexpression

perl -ne '/regex/ && $last && print $last; $last = $_'

Thisone-linerprintsalinewhenitisabovealinethatmatches/regex/.Let’swalkthroughit,beginningatthelaststatement,$last = $_,whichsaveseachlineinthe$lastvariable.Supposethenextlineisreadanditmatches/regex/.Becausethepreviouslineissavedin$last,theone-linersimplyprintsit.Theseriesof&&meansfirstthattheregularexpressionmustmatchandsecondthat$lastmustbeatruevalue.(Blanklinesarestillprintedbecausetheycontain

thenewlinecharacter.)

Sayyouhaveafilewithfourlines:

hello world

magic line

bye world

magic line

andyouwanttoprintalllinesabovethosethatmatchmagic.Youcandothis:

perl -ne '/magic/ && $last && print $last; $last = $_'

andtheone-linerwillprint:

hello world

bye world

7.8Printeverylinefollowingalinethatmatchesaregularexpression

perl -ne 'if ($p) { print; $p = 0 } $p++ if /regex/'

Here,Isetthevariable$pto1ifthecurrentlinematchestheregularexpression.Thefactthatthevariable$pis1indicatesthatthenextlineshouldbeprinted.Now,whenthenextlineisreadinand$pisset,thatlineisprintedand$pisresetto0.Quitesimple.

Let’ssayyouhavethisfour-linefile:

science

physics

science

math

andyouwanttoprintalllinesbelowthosethatmatchscience.Dothis:

perl -ne 'if ($p) { print; $p = 0 } $p++ if /science/'

Theone-linerwillprint:

physics

math

Ifyouwanttowritethiswith&&andavoidusingifandcurlybrackets,dothis:

perl -ne '$p && print && ($p = 0); $p++ if /science/'

Youcanalsobeverysmartaboutthisandsimplifythisone-linertothefollowing:

perl -ne '$p && print; $p = /science/'

Ifthecurrentlinematchesscience,thenvariable$pissettoatruevalueandthenextlinegetsprinted.Ifthecurrentlinedoesn’tmatchscience,then$pbecomesundefinedandthenextlinedoesn’tgetprinted.

7.9PrintlinesthatmatchregularexpressionsAAAandBBBinanyorder

perl -ne '/AAA/ && /BBB/ && print'

Thisone-linerteststoseewhetheralinematchestworegularexpressions.Ifalinematches/AAA/and/BBB/,it’sprinted.Specifically,thisone-linerprintsthelinefooAAAbarBBBbazbecauseitcontainsbothAAAandBBB,butitwon’tprintthelinefooAAAbarAAAbecauseitdoesn’tcontainBBB.

7.10Printlinesthatdon’tmatchregularexpressionsAAAandBBB

perl -ne '!/AAA/ && !/BBB/ && print'

Thisone-linerisalmostthesameasthepreviousone.Here,Itesttoseeifaline

doesn’tmatchbothregularexpressions.Ifitdoesn’tmatch/AAA/or/BBB/,itprints.

7.11PrintlinesthatmatchregularexpressionAAAfollowedbyBBBfollowedbyCCC

perl -ne '/AAA.*BBB.*CCC/ && print'

Here,IsimplychaintheregularexpressionsAAA,BBB,andCCCwith.*,whichmeans“matchanythingornothingatall.”IfAAAisfollowedbyBBB,whichisfollowedbyCCC,thelineprints.Forexample,thisone-linermatchesandprintsstringslike123AAA880BBB222CCC,xAAAyBBBzCCC,andAAABBBCCC.

7.12Printlinesthatareatleast80characterslong

perl -ne 'print if length >= 80'

Thisone-linerprintsalllinesthatareatleast80characterslong.InPerl,youcansometimesomittheparentheses()forfunctioncalls,sohereI’veomittedthemforthelengthfunctioncall.Infact,theinvocationslength,length(),andlength($_)areallthesameasfarasPerlisconcerned.

Ifyoudon’twanttocountlineendings,youcanturnonautomatichandlingoflineendingswith-l:

perl -lne 'print if length >= 80'

Thisswitchensuresthatablanklinehaszerolength,whereasitusuallyhaslength1or2,dependingonthefileformat.(UNIXnewlineshavelength1;Windowsnewlineshavelength2.)

7.13Printlinesthatarefewerthan80characterslong

long

perl -ne 'print if length() < 80'

Thisone-linerreversesthepreviousone.Itcheckstoseewhetherthelengthofalineislessthan80characters.Again,youuse-lifyoudon’twantthelineendingstobecounted.

7.14Printonlyline13

perl -ne '$. == 13 && print && exit'

AsIexplainedinone-liner7.2onpage70,the$.specialvariablestandsfor“thecurrentlinenumber.”Therefore,if$.hasavalueof13,thisone-linerprintsthelineandexits.

7.15Printalllinesexceptline27

perl -ne '$. != 27 && print'

Asinthepreviousone-liner,thisonecheckstoseewhetherthelinenumberofthecurrentlineis27.Ifalineisnot27,itprints;ifitis,itdoesn’tprint.

Youcanaccomplishthesamethingbyreversingprintand$. != 27andusingtheifstatementmodifier—justlikethis:

perl -ne 'print if $. != 27'

Oryoucanuseunless:

perl -ne 'print unless $. == 27'

7.16Printonlylines13,19,and67

perl -ne 'print if $. == 13 || $. == 19 || $. == 67'

perl -ne 'print if $. == 13 || $. == 19 || $. == 67'

Thisone-linerprintsonlylines13,19,and67.Itdoesn’tprintanyotherlines.Here’showitworks:Itcallsprintifthecurrentlinenumber,storedinthe$.variable,is13,19,or67.Youcanuseanylinenumberstoprintspecificlines.Forexample,toprintthelines13,19,88,290,and999,youdothis:

perl -ne 'print if $. == 13 || $. == 19 || $. == 88 || $. == 290 || $. == 999'

Ifyouwanttoprintmorelines,youcanputtheminaseparatearrayandthentestwhether$.isinthisarray:

perl -ne '

@lines = (13, 19, 88, 290, 999, 1400, 2000);

print if grep { $_ == $. } @lines

'

Thisone-linerusesgreptotestifthecurrentline$.isinthe@linesarray.Ifthecurrentlinenumberisfoundinthe@linesarray,thegrepfunctionreturnsalistofoneelementthatcontainsthecurrentlinenumberandthislistevaluatestotrue.Ifthecurrentlinenumberisnotfoundinthe@linesarray,thegrepfunctionreturnsanemptylistthatevaluatestofalse.

7.17Printalllinesfrom17to30

perl -ne 'print if $. >= 17 && $. <= 30'

Inthisone-liner,the$.variablestandsforthecurrentlinenumber.Asaresult,theone-linercheckstoseewhetherthecurrentlinenumberisgreaterthanorequalto17andlessthanorequalto30.

Youcandothesamethingusingtheflip-flopoperator,whichisexplainedinone-liner7.2onpage70.Theflip-flopoperatoroperateson$.whenusedwithintegers:

perl -ne 'print if 17..30'

7.18Printalllinesbetweentworegular

7.18Printalllinesbetweentworegularexpressions(includingthelinesthatmatch)

perl -ne 'print if /regex1/../regex2/'

Thisone-linerusestheflip-flopoperator(explainedinone-liner7.2onpage70).Whenusedwithintegers,theoperandsaretestedagainstthe$.variable.Whenusedwithregularexpressions,theoperandsaretestedagainstthecurrentline,storedinthe$_variable.Initiallytheoperatorreturnsfalse.Whenalinematchesregex1,theoperatorflipsandstartsreturningtrueuntilanotherlinematchesregex2.Atthatpoint,theoperatorreturnstrueforthelasttimeandthenflipstothefalsestate.Fromnowontheoperatorreturnsfalse.Thisone-liner,therefore,printsalllinesbetween(andincluding)thelinesthatmatchregex1andregex2.

7.19Printthelongestline

perl -ne '

$l = $_ if length($_) > length($l);

END { print $l }

'

Thisone-linerkeepsthelongestlineseensofarinthe$lvariable.Ifthelengthofthecurrentline$_exceedsthelengthofthelongestline,thevaluein$lisreplacedwiththevalueofthecurrentline.Beforeexiting,theENDblockisexecuted,anditprintsthelongestlinevaluethat’sheldin$l.

Remembertouse-lifyouwanttopreventthenewlinecharactersfromcountingtowardthelinelength.

7.20Printtheshortestline

perl -ne '

$s = $_ if $. == 1;

$s = $_ if length($_) < length($s);

END { print $s }

'

Thisone-lineristheoppositeofthepreviousone.Becauseit’sfindingtheshortestlineand$sisnotdefinedforthefirstline,youhavetosetitsvaluetothefirstlineexplicitlythrough$s = $_ if $. == 1.Thenitsimplydoestheoppositeofthepreviousone-liner.Thatis,itcheckstoseewhetherthecurrentlineistheshortestlinesofarand,ifso,assignsitto$s.

7.21Printalllinescontainingdigits

perl -ne 'print if /\d/'

Thisone-linerusestheregularexpression\d(whichstandsfor“adigit”)toseewhetheralinecontainsadigit.Ifso,thechecksucceeds,andthelineisprinted.Forexample,thislinewouldbeprintedbecauseitcontainsdigits:

coding is as easy as 123

However,thislinewouldn’tbeprintedbecauseitdoesn’tcontaindigits:

coding is as easy as pie

7.22Printalllinescontainingonlydigits

perl -ne 'print if /^\d+$/'

Inthisone-liner,theregularexpression^\d+$means“matchalineifitcontainsonlydigitsfromthebeginninguntiltheend.”Forexample,thislinewouldbeprintedbecauseitcontainsonlydigits:

3883737189170238912377

However,thislinewouldn’tbeprintedbecauseitalsocontainssomecharacters:

8388338 foo bar random data 999

Youcanalsoinvertthe^\d$regularexpressionanduse\D:

perl -lne 'print unless /\D/'

Thisone-linerisgreatfordevelopingyourlogicalreasoningbecauseituseslogicalnegationtwice.Here,thelineprintsonlyifitdoesnotcontainanon-numericcharacter.Inotherwords,itprintsonlyifallthecharactersarenumeric.(NoticethatIusedthe-lcommand-lineargumentforthisone-linerbecauseofthenewlinecharacterattheendoftheline.IfIdidn’tuse-l,thelinewouldcontainthenewlinecharacter—anon-numericcharacter—anditwouldn’tbeprinted.)

7.23Printalllinescontainingonlyalphabeticcharacters

perl -ne 'print if /^[[:alpha:]]+$/

Thisone-linercheckstoseewhetheralinecontainsonlyalphabeticcharacters.Ifso,itprintstheline.The[[:alpha:]]standsfor“anyalphabeticcharacter.”And[[:alpha:]]+standsfor“allalphabeticcharacters.”

7.24Printeverysecondline

perl -ne 'print if $. % 2'

Thisone-linerprintsthefirst,third,fifth,andseventhlines(andsoon).Itdoessobecause$. % 2istruewhenthecurrentlinenumberisoddandfalsewhenthecurrentlinenumberiseven.

7.25Printeverysecondline,beginningwiththesecondline

perl -ne 'print if $. % 2 == 0'

Thisone-linerislikepreviousone,exceptitprintsthesecond,fourth,sixth,andeighthlines(andsoon)because$. % 2 == 0istruewhenthecurrentlinenumberiseven.

Alternatively,youcansimplyinvertthetestfromthepreviousexample:

perl -ne 'print unless $. % 2'

7.26Printallrepeatedlinesonlyonce

perl -ne 'print if ++$a{$_} == 2'

Thisone-linertracksthelinesithasseensofarandcountsthenumberoftimesithasseenthelinespreviously.Ifitseesalineasecondtime,itprintsthelinebecause++$a{$_} == 2istrue.Ifitseesalinemorethantwotimes,itdoesnothingbecausethecountforthislineisgreaterthan2.

7.27Printalluniquelines

perl -ne 'print unless $a{$_}++'

Thisone-linerprintsalineonlyifthehashvalue$a{$_}forthatlineisfalse.EverytimePerlreadsinaline,itincrements$a{$_},whichensuresthatthisone-linerprintsonlynever-before-seenlines.

Chapter8.UsefulRegularExpressions

Inthischapter,we’lllookatvariousregularexpressionsandhowtousetheminsomehandyone-liners.TheregularexpressionsincludematchingIPaddresses,HTTPheaders,andemailaddresses;matchingnumbersandnumberranges;andextractingandchangingmatches.I’llalsosharesomeregularexpressionpuzzlesandbestpractices.ThischapterwillbeabitdifferentfrompreviousonesbecauseI’llstartwitharegularexpressionandthenwriteaone-linerthatusesit.

8.1MatchsomethingthatlookslikeanIPaddress

/^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}$/

Thisregularexpressiondoesn’tactuallyguaranteethatthethingthatmatchedis,infact,avalidIP;itsimplymatchessomethingthatlookslikeanIPaddress.Forexample,itmatchesavalidIPsuchas81.198.240.140aswellasaninvalidIPsuchas936.345.643.21.

Here’showitworks.The^atthebeginningoftheregularexpressionisananchorthatmatchesthebeginningofthestring.Next,\d{1,3}matchesone,two,orthreeconsecutivedigits.The\.matchesadot.The$attheendisananchorthatmatchestheendofthestring.(Youuseboth^and$anchorstopreventstringslikefoo213.3.1.2barfrommatching.)

Youcansimplifythisregularexpressionbygroupingthefirstthreerepeated\d{1,3}\.expressions:

/^(\d{1,3}\.){3}\d{1,3}$/

Sayyouhaveafilewiththefollowingcontentandyouwanttoextractonlythe

SayyouhaveafilewiththefollowingcontentandyouwanttoextractonlythelinesthatlooklikeIPaddresses:

81.198.240.140

1.2.3.4

5.5

444.444.444.444

90.9000.90000.90000

127.0.0.1

Toextractonlythematchinglines,youcanwritethis:

perl -ne 'print if /^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}$/'

whichshouldprint

81.198.240.140

1.2.3.4

444.444.444.444

127.0.0.1

One-liner8.3explainshowtomatchanIPprecisely,notjustsomethingthatlookslikeanIP.

8.2Testwhetheranumberisintherange0to255

/^([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])$/

Iliketochallengepeoplewithpuzzles.Oneofmyfavoritesistoasksomeonetocomeupwitharegularexpressionthatmatchesanumberrange.Writingoneisactuallyquitetrickyifyou’veneverdonesobefore.

Here’showitworks.Anumbercanhaveone,two,orthreedigits.Ifthenumberhasonedigit,youallowittobeanything[0-9].Ifithastwodigits,youalsoletitbeanycombinationof[0-9][0-9].Butifthenumberhasthreedigits,ithastobeeitheronehundred–somethingortwohundred–something.Ifthenumberisonehundred–something,1[0-9][0-9]matchesit.Ifthenumberistwo

hundred–something,thenumberiseither200to249(whichismatchedby2[0-4][0-9])orit’s250to255(whichismatchedby25[0-5]).

Let’sconfirmthisregularexpressionreallymatchesallnumbersintherange0to255andwriteaone-linertodoit:

perl -le '

map { $n++ if /^([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])$/ } 0..255;

END { print $n }

'

Thisone-lineroutputs256,thetotalnumbersintherange0to255.Ititeratesovertherange0to255andincrementsthe$nvariableforeverynumberthatmatches.Iftheoutputvaluewaslessthan256,you’dknowthatsomenumbersdidn’tmatch.

Let’salsomakesurethisone-linerdoesn’tmatchnumbersabove255:

perl -le '

map { $n++ if /^([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])$/ } 0..1000;

END { print $n }

'

Althoughthereare1001iterations,from0to1000,thefinalvalueof$nandtheoutputshouldstillbe256becausenumbersgreaterthan255shouldnotmatch.Ifthevaluewasgreaterthan256,you’dknowthattoomanynumbersmatchedandtheregularexpressionwasincorrect.

8.3MatchanIPaddress

my $ip_part = qr/[0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5]/;

if ($ip =~ /^$ip_part\.$ip_part\.$ip_part\.$ip_part$/) {

print "valid ip\n";

}

Thisregularexpressioncombinestheideasfromtheprevioustworegularexpressions(8.1and8.2)andintroducestheqr/.../operator,whichletsyouconstructaregularexpressionandsaveitinavariable.Here,I’msavingthe

regularexpressionthatmatchesallnumbersintherange0to255inthe$ip_partvariable.Next,the$ip_partmatchesthefourpartsoftheIPaddress.

YoucansimplifythisbygroupingthefirstthreeIPparts:

if ($ip =~ /^($ip_part\.){3}$ip_part$/) {

print "valid ip\n";

}

Let’srunthisonthesamefilefromone-liner8.1.Ifyouhavethisfileasinput:

81.198.240.140

1.2.3.4

5.5

444.444.444.444

90.9000.90000.90000

127.0.0.1

andyourone-lineris

perl -ne '

$ip_part = qr|([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])|;

print if /^($ip_part\.){3}$ip_part$/

'

thentheoutputis

81.198.240.140

1.2.3.4

127.0.0.1

Asyoucansee,onlythevalidIPaddressesareprinted.

8.4Checkwhetherastringlookslikeanemailaddress

/\S+@\S+\.\S+/

Thisregularexpressionmakessurethestringlookslikeanemailaddress;it

doesn’tguaranteethestringisanemailaddress,however.First,itmatchessomethingthat’snotwhitespace(\S+)uptothe@symbol;thenitmatchesasmuchaspossibleuntilitfindsadot;thenitmatchessomemore.

Ifthematchessucceed,youknowthestringatleastlookslikeanemailaddresswiththe@symbolandadotinit.Forexample,cats@catonmat.netmatches,butcats@catonmatdoesn’tbecausetheregularexpressioncan’tfindthedotthat’srequiredinafullyqualifieddomainname.

Here’samuchmorerobustwaytoseewhetherastringisavalidemailaddress,usingtheEmail::Validmodule:

use Email::Valid;

print Email::Valid->address('cats@catonmat.net') ? 'valid email' : 'invalid email';

Here,youusetheternaryoperatorcond ? true : false.Ifthecondistrue,thetruepartexecutes;otherwisethefalsepartexecutes.Thisprintsvalidemailiftheemailisvalid;ifnot,itprintsinvalid email.

Soaone-linerwouldlooklikethis:

perl -MEmail::Valid -ne 'print if Email::Valid->address($_)'

Here,iftheemailaddressisvalid,yousimplyprintit.

8.5CheckwhetherastringisanumberDeterminingwhetherastringisanumberisdifficultwitharegularexpression.Thisisaderivationofaregularexpressionthatmatchesdecimalnumbers.

IstartwithPerl’s\dregularexpression,whichmatchesthedigits0through9:

/^\d+$/

Thisregularexpressionmatchesoneormoredigits\dfromthebeginningofthestring^totheendat$.Butitdoesn’tmatchnumberssuchas+3and-3.Let’smodifytheregularexpressiontomatchthem:

/^[+-]?\d+$/

/^[+-]?\d+$/

Here,the[+-]?means“matchanoptionalplusoraminusbeforethedigits.”Thisregularexpressionnowmatches+3and-3butnot-0.3.Let’saddthat:

/^[+-]?\d+\.?\d*$/

I’veexpandedthepreviousregularexpressionbyadding\.?\d*,whichmatchesanoptionaldotfollowedbyzeroormorenumbers.Nowwe’reinbusiness.Thisregularexpressionalsomatchesnumberslike-0.3and0.3,thoughitwouldnotmatchnumberssuchas123,456or.5.

AmuchbetterwaytomatchadecimalnumberistousetheRegexp::Commonmodule.Forexample,tomatchadecimalnumber,youcanuse$RE{num}{real}fromRegexp::Common.Here’saone-linerthatfilterstheinputandprintsonlythedecimalnumbers:

perl -MRegexp::Common -ne 'print if /$RE{num}{real}/'

Thisone-lineralsomatchesandprintsnumberssuchas123,456and.5.

Howaboutmatchingpositivehexadecimalnumbers?Here’show:

/^0x[0-9a-f]+$/i

Thisone-linermatchesthehexprefix0xfollowedbythehexnumberitself.The/iflagattheendensuresthematchiscaseinsensitive.Forexample,0x5afmatches,0X5Famatches,but97doesn’tbecause97hasnohexprefix.

Betterstill,use$RE{num}{hex}becauseitsupportsnegativenumbers,decimalplaces,andnumbergrouping.

Howaboutmatchingoctals?

/^0[0-7]+$/

Octalnumbersareprefixedby0,whichisfollowedbytheoctaldigits0-7.Forexample,013matchesbut09doesn’tbecauseit’snotavalidoctalnumber.Using$RE{num}{oct}isbetterbecauseitsupportsnegativeoctalnumbers,octal

numberswithdecimalplaces,andnumbergrouping.

Finally,wecometobinarymatching:

/^[01]+$/

Binarybaseconsistsofonly0sand1s,so010101matchesbut210101doesn’tbecause2isnotavalidbinarydigit.

Regexp::Commonalsooffersabetterregularexpressionformatchingbinarynumbers:$RE{num}{bin}.

8.6Checkwhetherawordappearsinastringtwice

/(word).*\1/

Thisregularexpressionmatchesawordfollowedbysomethingornothingatall,followedbythesameword.Here,(word)capturesthewordingroup1,and\1referstothecontentsofgroup1,whichisthesameaswriting/(word).*word/.Forexample,silly things are sillymatches/(silly).*\1/,butsillythings are boringdoesn’tbecausesillyisnotrepeatedinthestring.

8.7Increaseallintegersinastringbyone

$str =~ s/(\d+)/$1+1/ge

Here,youusethesubstitutionoperatorstomatchallintegers(\d+),putthemincapturegroup1,andthenreplacethemwiththeirvalueincrementedbyone:$1+1.Thegflagfindsallnumbersinthestring,andtheeflagevaluates$1+1asaPerlexpression.Forexample,this 1234 is awesome 444isturnedintothis 1235 is awesome 445.

Notethatthisregularexpressiondoesn’tincrementfloating-pointnumbersbecauseituses\d+tomatchintegers.Toincrementfloating-pointnumbers,use

the$RE{num}{real}regularexpressionfromone-liner8.5.Here’sasampleone-linerthatuses$RE{num}{real}:

perl -MRegexp::Common -pe 's/($RE{num}{real})/$1+1/ge'

Ifyoupassthisone-linertheinputweird 44.5 line -1.25,itprintsweird45.5 line -0.25.

8.8ExtracttheHTTPUser-AgentstringfromHTTPheaders

/^User-Agent: (.+)$/

HTTPheadersareformattedasKey: Valuepairs.YoucaneasilyparsesuchstringsbyinstructingtheregularexpressionenginetosavetheValuepartinthe$1groupvariable.Forexample,iftheHTTPheaderscontainthefollowing:

Host: www.catonmat.net

Connection: keep-alive

User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_0_0; en-US)

Accept: application/xml,application/xhtml+xml,text/html

Accept-Encoding: gzip,deflate,sdch

Accept-Language: en-US,en;q=0.8

Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.3

thentheregularexpressionwillextractthestringMozilla/5.0 (Macintosh;U; Intel Mac OS X 10_0_0; en-US).

8.9MatchprintableASCIIcharacters

/[ -~]/

Thisregularexpressionistrickyandsmart.Tounderstandit,takealookatmanascii,andyou’llseethatspacestartsatvalue0x20andthe~characteris0x7e.Theexpression[ -~]definesarangeofcharactersfromthespaceuntil~.

Becauseallcharactersbetweenthespaceand~areprintable,thisregularexpressionmatchesallprintablecharacters.Thisismyfavoriteregularexpressionofalltimebecauseit’squitepuzzlingwhenyoufirstseeit.Whatdoesitmatch?Aspace,adash,andatilde?No,itmatchesarangeofcharactersfromthespaceuntilthetilde!

Toinvertthematch,place^asthefirstcharacterinthegroup:

/[^ -~]/

Thismatchestheoppositeof[ -~],thatis,allnonprintablecharacters.

8.10ExtracttextbetweentwoHTMLtags

m|<strong>([^<]*)</strong>|

BeforeIexplainthisregularexpression,letmesaythatit’sokaytomatchHTMLwithregularexpressionsonlyforquickhackswhenyouneedtogetthingsdoneandmoveon.YoushouldneveruseregularexpressionstomatchandparseHTMLinseriousapplicationsbecauseHTMLisactuallyacomplicatedlanguage,and,ingeneral,itcan’tbeparsedbyaregularexpression.Instead,usemoduleslikeHTML::TreeBuildertoaccomplishthetaskmorecleanly!

Thisregularexpressionsavestextbetweenthe<strong>...</strong>HTMLtagsinthe$1specialvariable.Thetrickiestpartofthisone-lineris([^<]*),whichmatcheseverythinguptothe<character.It’saregularexpressionidiom.

Forexample,iftheHTMLyou’retryingtomatchis<strong>hello</strong>,thenthisregularexpressioncaptureshellointhe$1variable.However,iftheHTMLyou’retryingtomatchis<strong><em>hello</em> </strong>,thenthisregularexpressiondoesn’tmatchatallbecausethereisanotherHTMLtagbetween<strong>and</strong>.

ToextracteverythingbetweentwoHTMLtags,includingotherHTMLtags,youcanwrite:

m|<strong>(.*?)</strong>|

m|<strong>(.*?)</strong>|

Thisregularexpressionsaveseverythingbetween<strong>...</strong>inthe$1variable.Forexample,iftheHTMLis<strong><em>hello</em></strong>,thisregularexpressionsets$1to<em>hello</em>.The(.*?)partoftheregularexpressionmatcheseverythingbetweenthetwonearest<strong>and</strong>tags.Thequestionmark?inthisregularexpressioncontrolsitsgreediness.

IfyouwanttobeagoodcitizenanduseHTML::TreeBuilder,thenaPerlprogramthatdoesthesamethingwouldlooklikethis:

use warnings;

use strict;

use HTML::TreeBuilder;

my $tree = HTML::TreeBuilder->new_from_content(

"<strong><em>hello</em></strong>"

);

my $strong = $tree->look_down(_tag => 'strong');

if ($strong) {

print $_->as_HTML for $strong->content_list;

}

$tree->delete;

Here,IcreatedanewHTML::TreeBuilderinstancefromthegivenstring;thenIfoundthe<strong>taganddumpedallthechildelementsofthe<strong>tagasHTML.Asyoucansee,althoughwritingaprogramlikethisisn’tsuitableasaone-liner,it’samuchmorerobustsolution.

8.11Replaceall<b>tagswith<strong>

$html =~ s|<(/)?b>|<$1strong>|g

Here,IassumethattheHTMLisinthevariable$html.Theexpression<(/)?b>matchestheopeningandclosing<b>tags,capturestheoptionalclosingtagslashinthegroup$1,andthenreplacesthematchedtagwitheither<strong>or</strong>,dependingonwhetheritfindsanopeningorclosingtag.

RememberthatthecorrectwaytodothisistouseHTML::TreeBuilderandwriteaproperprogram.Youshouldonlyusethisregularexpressionforaquickhack.Here’swhataprogramthatusesHTML::TreeBuilderlookslike:

use warnings;

use strict;

use HTML::TreeBuilder;

my $tree = HTML::TreeBuilder->new_from_content("

<div><p><b>section 1</b></p><p><b>section 2</b></p></div>

");

my @bs = $tree->look_down(_tag => 'b');

$_->tag('strong') for @bs;

print $tree->as_HTML;

$tree->delete;

Here,I’vecreatedtheHTML::TreeBuilderobjectfromthegivenstring;next,Ifoundallthe<b>tags,storedtheminthe@bsarray,andthenloopedoverall@bsandchangedtheirtagnameto<strong>.

8.12Extractallmatchesfromaregularexpression

my @matches = $text =~ /regex/g;

Here,theregularexpressionmatchisevaluatedinthelistcontext,whichmakesitreturnallmatches.Thematchesareputinthe@matchesvariable.

Forexample,thefollowingregularexpressionextractsallintegersfromastring:

my $t = "10 hello 25 moo 30 foo";

my @nums = $text =~ /\d+/g;

Afterexecutingthiscode,@numscontains(10, 25, 30).Youcanalsouseparenthesestocaptureonlypartofthestring.Forexample,here’showtocaptureonlythevaluesfromalinecontaininglotsofkey-valuepairs(suchaskey=value),separatedbysemicolons:

my @vals = $text =~ /[^=]+=([^;]+)/g;

Thisregularexpressionfirstmatchesthekeysthrough[^=]+,thenitmatchesthe=characterthatseparatesthekeysandvalues,andthenitmatchesthevalues([^;]+).Asyoucansee,thevaluepartoftheregularexpressioniswrappedinparenthesessothevaluesarecaptured.

Here’sanexample.Sayyouhaveafilewiththefollowingcontents:

access=all; users=peter,alastair,bill; languages=awk,sed,perl

andyouwritethisone-liner:

perl -nle 'my @vals = $_ =~ /[^=]+=([^;]+)/g; print "@vals"'

Runningitoutputsthefollowing:

all peter,alastair,bill awk,sed,perl

Thesearethevaluesfortheaccess,users,andlanguageskeys!

AppendixA.Perl’sSpecialVariables

Inthisappendix,IsummarizePerl’smostcommonlyusedspecial(predefined)variables,suchas$_,$.,$/,$\,$1,$2,$3(andsoon),$,,@F,and@ARGV,amongothers.

A.1Variable$_The$_variable,calledthedefaultvariable,isthemostcommonlyusedvariableinPerl.Oftenthisvariableispronounced“it”(whennotpronounced“dollar-underscore”);asyoureadon,you’llunderstandwhy.

Whenusingthe-nand-pcommand-linearguments,it’s(see?)wheretheinputisstored.Also,manyoperatorsandfunctionsactonitimplicitly.Here’sanexample:

perl -le '$_ = "foo"; print'

Here,Iplacethestring"foo"inthe$_variableandthencallprint.Whengivennoarguments,printprintsthecontentsofthe$_variable,whichis"foo".

Similarly,$_isusedbythes/regex/replace/and/regex/operatorswhenusedwithoutthe=~operator.Considerthisexample:

perl -ne '/foo/ && print'

Thisone-linerprintsonlylinesthatmatch/foo/.The/foo/operatorimplicitlyoperatesonthe$_variablethatcontainsthecurrentline.Youcouldrewritethisasfollows,butdoingsowouldrequiretoomuchtyping:

perl -ne 'if ($_ =~ /foo/) { print $_ }'

“Ifitmatches/foo/,printit”—yougettheidea.Youcouldalsoreplacetextinallthelinessimplybycallings/foo/bar/:

perl -pe 's/foo/bar/'

Interestingly,Perlborrowsthe$_variablefromsed.Rememberthatsedhasapatternspace?The$_variablecanalsobecalledPerl’spatternspace.Ifyouwrotethepreviousone-liner(perl -pe 's/foo/bar/')insed,itwouldlooklikesed 's/foo/bar/'becausesedputseachlineinthepatternspaceandthescommandactsonitimplicitly.Perlborrowsmanyconceptsandcommandsfromsed.

Using$_withthe-nargumentWhenusingthe-nargument,Perlputsthefollowinglooparoundyourprogram:

while (<>) {

# your program goes here (specified by -e)

}

Thewhile (<>)loopreadslinesfromstandardinputorfilesnamedonthecommandlineandputseachlineintothe$_variable.Youcanthenmodifythelinesandprintthem.Forexample,youcanreversethelines:

perl -lne 'print scalar reverse'

BecauseI’musingthe-nargumenthere,thisprogrambecomes

while (<>) {

print scalar reverse

}

whichisequivalentto

while (<>) {

print scalar reverse $_

}

}

ThetwoprogramsareequivalentbecausemanyPerlfunctionsacton$_implicitly,whichmakeswritingreverseandreverse $_functionallythesamething.Youneedscalartoputthereversefunctioninthescalarcontext.Otherwiseit’sinthelistcontext(printforcesthelistcontext)andwon’treversestrings.(Iexplainthe-nflagingreatdetailinone-liner2.6onpage12andlinereversinginone-liner6.22onpage67.)

Using$_withthe-pargumentWhenyouusethe-pargument,Perlputsthefollowinglooparoundyourprogram:

while (<>) {

# your program goes here (specified by -e)

} continue {

print or die "-p failed: $!\n";

}

Theresultisalmostthesameasforthe-nargument,exceptthataftereachiterationthecontentof$_isprinted(throughprintinthecontinueblock).

ToreversethelinesasIdidwith-n,Icandothis:

perl -pe '$_ = reverse $_'

Theprogramnowbecomes:

while (<>) {

$_ = reverse $_;

} continue {

print or die "-p failed: $!\n";

}

I’vemodifiedthe$_variableandsetittoreverse $_,whichreversestheline.Thecontinueblockmakessurethatit’sprinted.(One-liner2.1onpage7explainsthe-pargumentinmoredetail.)

Using$_explicitly

Using$_explicitlyThe$_variableisalsooftenusedexplicitly.Herearesomeexamplesofusingthe$_variableexplicitly:

perl -le '@vals = map { $_ * 2 } 1..10; print "@vals"'

Theoutputofthisone-lineris2 4 6 8 10 12 14 16 18 20.Here,Iusethemapfunctiontomapanexpressionovereachelementinthegivenlistandreturnanewlist,whereeachelementistheresultoftheexpression.Inthiscase,thelistis1..10(1 2 3 4 5 6 7 8 9 10)andtheexpressionis$_ * 2,whichmeansmultiplyeachelement(“it”)by2.Asyoucansee,I’musing$_explicitly.Whenthemapfunctioniteratesoverthelist,eachelementisputinto$_formyconvenience.

Nowlet’susemapinahandyone-liner.Howaboutonethatmultiplieseachelementonalineby2?

perl -alne 'print "@{[map { $_ * 2 } @F]}"'

Thisone-linermapstheexpression$_ * 2ontoeachelementin@F.Thecrazy-looking"@{[...]}"isjustawaytoexecutecodeinsidequotes.(One-liner4.2onpage30explains@F,andone-liner4.4onpage32explains"@{[...]}".)

Anotherfunctionthatexplicitlyuses$_isgrep,whichletsyoufiltertheelementsfromalist.Here’sanexample:

perl -le '@vals = grep { $_ > 5 } 1..10; print "@vals"'

Theoutputofthisone-lineris6 7 8 9 10.Asyoucansee,grepfilteredelementsgreaterthan5fromthelist.Thecondition$_ > 5asks,“Isthecurrentelementgreaterthan5?”—or,moresuccinctly,“Isitgreaterthan5?”

Let’susegrepinaone-liner.Howaboutonethatfindsandprintsallelementsonthecurrentlinethatarepalindromes?

perl -alne 'print "@{[grep { $_ eq reverse $_ } @F]}"'

Theconditionspecifiedtothegrepfunctionhereis$_ eq reverse $_,whichasks,“Isthecurrentelementthesameasitsreverse?”Thisconditionistrueonlyforpalindromes.Forexample,giventhefollowinginput:

civic foo mom dad

bar baz 1234321 x

theone-lineroutputsthis:

civic mom dad

1234321 x

Asyoucansee,alloftheseelementsarepalindromes.

Youcanlearnevenmoreaboutthe$_variablebytypingperldoc perlvaratthecommandline.TheperlvardocumentationexplainsallthepredefinedvariablesinPerl.

A.2Variable$.Whenreadingafile,the$.variablealwayscontainsthelinenumberofthelinecurrentlybeingread.Forexample,thisone-linernumbersthelinesinfile:

perl -lne 'print "$. $_"' file

Youcandothesamethingwiththisone-liner,whichreplacesthecurrentlinewiththelinenumberfollowedbythesameline:

perl -pe '$_ = "$. $_"' file

The$.variableisn’tresetacrossfiles,sotonumbermultiplefilessimultaneously,youwrite

perl -pe '$_ = "$. $_"' file1 file2

Thisone-linercontinuesnumberinglinesinfile2wherefile1leftoff.(Iffile1contains10lines,thefirstlineoffile2isnumbered11.)

Toresetthe$.variable,youuseanexplicitcloseonthecurrentfilehandleARGV:

perl -pe '$_ = "$. $_"; close ARGV if eof' file1 file2

ARGVisaspecialfilehandlethatcontainsthecurrentlyopenfile.Bycallingeof,I’mcheckingtoseeifit’stheendofthecurrentfile.Ifso,closeclosesit,whichresetsthe$.variable.

YoucanchangewhatPerlconsiderstobealinebymodifyingthe$/variable.Thenextsectiondiscussesthisvariable.

A.3Variable$/The$/variableistheinputrecordseparator,whichisanewlinebydefault.ThisvariabletellsPerlwhattoconsideraline.Sayyouhavethissimpleprogramthatnumberslines:

perl -lne 'print "$. $_"' file

Because$/isanewlinebydefault,Perlreadseverythinguptothefirstnewline,putsitinthe$_variable,andincrementsthe$.variable.Next,itcallsprint"$. $_",whichprintsthecurrentlinenumberandtheline.Butifyouchangethevalueof$/totwonewlines,like$/ = "\n\n",Perlreadseverythinguptothefirsttwonewlines;thatis,itreadstextparagraphbyparagraphratherthanlinebyline.

Here’sanotherexample.Ifyouhaveafilelikethefollowing,youcanset$/to:,andPerlwillreadthefiledigitbydigit.

3:9:0:7:1:2:4:3:8:4:1:0:0:1:... (goes on and on)

Orifyouset$/toundef,Perlreadstheentirefileinasingleread(calledslurping):

perl -le '$/ = undef; open $f, "<", "file"; $contents = <$f>"

Thisone-linerslurpstheentirefilefileinvariable$contents.

Youcanalsoset$/toreferenceaninteger:

$/ = \1024

Inthiscase,Perlreadsthefile1024bytesatatime.(Thisisalsocalledrecord-by-recordreading.)

Youcanalsousethe-0command-lineswitchtoprovidethisvariablewithavalue,butnotethatyoucan’tdotherecord-by-recordversionlikethis.Forexample,toset$/to:,specify-0072because072istheoctalvalueofthe:character.

Torememberwhatthisvariabledoes,recallthatwhenquotingpoetry,linesareseparatedby/.

A.4Variable$\Thedollar-backslashvariableisappendedaftereveryprintoperation.Forexample,youcouldappendadotfollowedbyaspace". "aftereachprint:

perl -e '$\ = ". "; print "hello"; print "world"'

Thisone-linerproducesthefollowingoutput:

hello. world.

Modifyingthisvariableisespeciallyhelpfulwhenyouwanttoseparateprintoutsbydoublenewlines.

Torememberthisvariable,justrecallthatyouprobablywanttoprint\naftereveryline.NotethatforPerl5.10andlater,thefunctionsayisavailable,whichislikeprint,exceptthatitalwaysaddsanewlineattheendanddoesn’tusethe$\variable.

A.5Variables$1,$2,$3,andsoon

Variables$1,$2,$3,andsooncontainthematchesfromthecorrespondingsetofcapturingparenthesesinthelastpatternmatch.Here’sanexample:

perl -nle 'if (/She said: (.*)/) { print $1 }'

Thisone-linermatcheslinesthatcontainthestringShe said:andthencaptureseverythingafterthestringinvariable$1andprintsit.

Whenyouuseanotherpairofparentheses,thetextiscapturedinvariable$2,andsoon:

perl -nle 'if (/(She|He) said: (.*)/) { print "$1: $2" }'

Inthisone-liner,firsteither"She"or"He"iscapturedinvariable$1andthenanythingsheorhesaidiscapturedinvariable$2andprintedas"$1: $2".You’llgetthesamenumberofcapturevariablesasyouhavepairsofparentheses.

Toavoidcapturingtextinavariable,usethe?:symbolsinsidetheopeningparenthesis.Forexample,changing(She|He)to(?:She|He):

perl -nle 'if (/(?:She|He) said: (.*)/) { print "Someone said: $1" }'

willnotcapture"She"or"He"invariable$1.Instead,thesecondpairofparenthesescaptureswhatsheorhesaidinvariable$1.

BeginningwithPerl5.10,youcanusenamedcapturegroupsasin(?<name>...).Whenyoudo,insteadofusingvariables$1,$2,andsoon,youcanuse$+{name}torefertothegroup.Forexample,thiscaptures"She"or"He"inthenamedgroupgenderandthesaidtextinthenamedgrouptext:

perl -nle 'if (/(?<gender>She|He) said: (?<text>.*)/) {

print "$+{gender}: $+{text}"

}'

A.6Variable$,The$,variableistheoutputfieldseparatorforprintwhenprintingmultiple

values.It’sundefinedbydefault,whichmeansthatallitemsprintedareconcatenatedtogether.Indeed,ifyoudothis:

perl -le 'print 1, 2, 3'

youget123printedout.Ifyouset$,toacolon,however:

perl -le '$,=":"; print 1, 2, 3'

youget1:2:3.

Now,supposeyouwanttoprintanarrayofvalues.Ifyoudothis:

perl -le '@data=(1,2,3); print @data'

theoutputis123.Butifyouquotethevariable,thevaluesarespaceseparated:

perl -le '@data=(1,2,3); print "@data"'

Sotheoutputis1 2 3becausethearrayisinterpolatedinadouble-quotedstring.

A.7Variable$”Thisbringsustothe$"variable:asinglewhitespace(bydefault)that’sinsertedbetweeneveryarrayvaluewhenit’sinterpolated.Whenyouwritethingslikeprint "@data",the@dataarraygetsinterpolated,andthevalueof$"isinsertedbetweeneveryarrayelement.Forexample,thisprints1 2 3:

perl -le '@data=(1,2,3); print "@data"'

Butifyouchange$"to,say,adash-,theoutputbecomes1-2-3:

perl -le '@data=(1,2,3); $" = "-"; print "@data"'

Recallthe@{[...]}trickhere.Ifyouprint "@{[...]}",youcanexecutecodeplacedbetweenthesquarebrackets.Forexamplesandmoredetails,see

sectionA.1Variable$_’sdiscussionofthe$_variableonpage95andone-liner4.4onpage32.

A.8Variable@FThe@FvariableiscreatedinyourPerlprogramwhenyouusethe-aargument,whichstandsforauto-splitfields.Whenyouuse-a,theinputissplitonwhitespacecharactersandtheresultingfieldsareputin@F.Forexample,iftheinputlineisfoo bar baz,then@Fisanarray("foo", "bar", "baz").

Thistechniqueallowsyoutooperateonindividualfields.Forinstance,youcanaccess$F[2]toprintthethirdfieldasfollows(rememberingthatarraysstartfromindex0):

perl -ane 'print $F[2]'

Youcanalsoperformvariouscalculations,likemultiplyingthefifthfieldby2:

perl -ane '$F[4] *= 2; print "@F"'

Here,thefifthfield$F[4]ismultipliedby2,andprint "@F"printsallthefields,separatedbyaspace.

Youcanalsousethe-aargumentwiththe-Fargument,whichspecifiesthecharactertospliton.Forexample,toprocessthecolon-separatedentriesin/etc/passwdentries,youwrite

perl -a -F: -ne 'print $F[0]' /etc/passwd

whichprintstheusernamesfrom/etc/passwd.

A.9Variable@ARGVThe@ARGVvariablecontainstheargumentsthatyoupasstoyourPerlprogram.Forexample,thisprintsfoo bar baz:

perl -le 'print "@ARGV"' foo bar baz

Whenyouuse-nor-pflags,theargumentsthatyoupasstoyourPerlprogramareopenedonebyoneasfilesandremovedfrom@ARGV.Toaccessthefilenamespassedtoyourprogram,savetheminanewvariableintheBEGINblock:

perl -nle 'BEGIN { @A = @ARGV }; ...' file1 file2

Nowyoucanuse@Ainyourprogram,whichcontains("file1", "file2").Ifyoudidn’tdothisandyouused@ARGV,itwouldcontain("file2")atfirst,butwhenfile1wasprocessed,itwouldbeempty().Becarefulhere!

Asimilar-lookingvariable,$ARGV,containsthefilenameofthefilecurrentlybeingread,whichis"-"iftheprogramiscurrentlyreadingfromthestandardinput.

A.10Variable%ENVThe%ENVhashcontainsenvironmentvariablesfromyourshell.ThisvariablecomesinhandywhenyouwishtopredefinesomevaluesinyourscriptandthenusethesevaluesinyourPerlprogramorone-liner.

Sayyouwanttousethesystemfunctiontoexecuteaprogramthat’snotinthepath.Youcouldmodifythe$ENV{PATH}variableandappendtheneededpath:

perl -nle '

BEGIN { $ENV{PATH} .= ":/usr/local/yourprog/bin" }

...

system("yourprog ...");

'

Thisone-linerprintsallenvironmentvariablesfromPerl:

perl -le 'print "$_: $ENV{$_}" for keys %ENV'

Itloopsoverthekeys(environmentvariablenames)ofthe%ENVhash,putseachkeyintothe$_variable,andthenprintsthenamefollowedby$ENV{$_},whichisthevalueoftheenvironmentvariable.

AppendixB.UsingPerlOne-LinersOnWindows

Inthisappendix,I’llshowyouhowtorunPerlonWindows,installabashportonWindows,andusePerlone-linersinthreedifferentways:fromtheWindowsbashport,theWindowscommandprompt(cmd.exe),andPowerShell.

B.1PerlonWindowsBeforeyoucanrunPerlonWindows,youneedtoinstallPerlforWindows.MyfavoriteWindowsPerlportisStrawberryPerl(http://strawberryperl.com/),aPerlenvironmentwitheverythingyouneedtorunanddevelopPerlapplicationsonWindows.StrawberryPerlisdesignedtofunctionasmuchaspossiblelikethePerlenvironmentonUNIXsystems.ItincludesPerlbinaries,thegcccompilerandrelatedbuildtools,andmanyexternallibraries.

ToinstallStrawberryPerl,downloadandruntheinstaller,clickthroughabunchofmenusafewtimes,andyou’redone.Mychoicefortheinstallationdirectoryisc:\strawberryperl.(InstallinganyUNIXsoftwareforWindowsinadirectorywithnospacesinitisalwaysagoodidea.)Oncetheinstallationisdone,theinstallershouldputtheinstallationdirectoryinyourpathenvironmentvariablesoyoucanrunPerlfromthecommandlinerightaway.

Unfortunately,theWindowscommandlineisreallybasiccomparedtothecommandlineonUNIXsystems.AUNIXsystemrunsarealshellwithwell-definedcommand-lineparsingrules,butWindowsdoesn’treallyhaveanythinglikethat.TheWindowscommandlinehasweirdrulesabouthowittreatscertainsymbols,thequotingrulesaren’twelldefined,andtheescapingrulesarestrange,allofwhichmakesitdifficulttorunPerlone-liners.Therefore,thepreferredwaytorunone-linersonWindowsistouseaUNIXshell(suchasbash)forWindows,asyou’lllearninthenextsection.

B.2BashonWindowsGettingabashshelltorunonWindowsissimple.Irecommendwin-bash(http://win-bash.sourceforge.net/),astand-alonebashportforWindowsthatdoesn’tneedaspecialenvironmentoradditionalDLLs.Thedownloadisasinglezipfilethatcontainsthebashshell(bash.exe)andabunchofUNIXutilities(suchasawk,cat,cp,diff,find,grep,sed,vi,wc,andabout100others).

Toinstallbashandalltheutilities,simplyunzipthefileandyou’redone.Mychoicefortheinstallationdirectoryisc:\winbash,againwithnospacesinthedirectory.Runbash.exefromc:\winbashtostartthebashshell.

Ifyoustartbash.exeafteryouinstallStrawberryPerl,PerlshouldbeavailableforuserightawaybecausetheStrawberryPerlinstallershouldhaveupdatedthepathwiththeinstallationdirectory.Toconfirm,runperl --version.ItshouldoutputtheversionoftheinstalledPerl.Ifyougetanerrorsayingthatperlwasnotfound,manuallyappendtheC:\strawberryperl\perl\bindirectorytothePATHenvironmentvariablebyenteringthisinthecommandline:

PATH=$PATH:C:\\strawberryperl\\perl\\bin

BashusesthePATHvariabletofindexecutablestorun.ByappendingStrawberryPerl’sbinarydirectorytothePATHvariable,youtellbashwheretolookfortheperlexecutable.

B.3PerlOne-LinersinWindowsBashTherearesomeimportantdifferencesbetweenbashonWindowsandUNIX.Thefirstdifferencepertainstofilepaths.Win-bashsupportsbothUNIX-styleandWindows-stylepaths.

Sayyouinstallwin-bashinC:\winbash.Whenyoustartbash.exe,itshouldmaptherootdirectory/tothecurrentC:drive.Tochangetherootdirectorytoanotherdrive,suchasD:,entercd d:inthebashshell.TochangebacktoC:,entercd c:intheshell.NowyoucanaccessafilesuchasC:\work\report.txtvia/work/report.txt,c:/work/report.txt,orc:\\work\\report.txt.

Thebestthingaboutusingwin-bashisthatalloftheone-linersinthisbook

shouldworkbecauseyou’rerunningarealshelljustlikeinaUNIXenvironment!Forexample,tonumberthelinesintheC:\work\report.txtfile(one-liner3.1onpage17),youcanrun:

perl -pe '$_ = "$. $_"' C:/work/report.txt

OryoucanrefertothefileasifyouwereinUNIX:

perl -pe '$_ = "$. $_"' /work/report.txt

OryoucanuseWindows-stylepaths:

perl -pe '$_ = "$. $_"' C:\\work\\report.txt

Toavoidusingdoublebackslashes,youcansingle-quotethefilepath:

perl -pe '$_ = "$. $_"' 'C:\work\report.txt'

Ifthefilenamehasspacesinit,thenyoualwayshavetoquoteit.Forexample,toworkwithC:\DocumentsandSettings\Peter\MyDocuments\report.txt,quotetheentirepathwhenpassingittoaone-liner:

perl -pe '$_ = "$. $_"' 'C:\Documents and Settings\Peter\My Documents\report.txt'

OrusetheUNIX-stylepathtothefile:

perl -pe '$_ = "$. $_"' '/Documents and Settings/Peter/My Documents/report.txt'

QuotingthefilenameisnecessaryherebecausewithoutitPerlthinksyou’repassingitabunchoffilesratherthanasinglefilewithspacesinit.

B.4PerlOne-LinersintheWindowsCommandPromptIf,forsomereason,youcan’tusewin-bashasrecommended,youcanrunone-linersthroughtheWindowscommandprompt(cmd.exe).Youwillneedto

changetheone-linersinthisbookabitifyou’rerunningthemintheWindowscommandpromptbecauseofthewayWindowsparsesandtreatsthecommand-linearguments.Here’swhattodo.

First,verifythatPerlisavailablefromthecommandprompt.Startcmd.exeandenterperl --versioninthecommandline.IfyoudothisafterhavinginstalledStrawberryPerl,thecommandshouldoutputthePerlversion,andyou’regoodtogo.Otherwise,you’llhavetoupdatethePATHenvironmentvariablewiththepathtoStrawberryPerl’sbinarydirectory:

set PATH=%PATH%;C:\strawberryperl\perl\bin

AsinUNIX,thePATHvariabletellsthecommandpromptwheretolookforexecutables.

ConvertingOne-LinersintheWindowsCommandPromptNowlet’sseehowtoconvertone-linersforthecommandprompt,beginningwithone-liner2.1(page7),whichdouble-spacesafile.InUNIX,yousimplyrun:

perl -pe '$\ = "\n"' file

Ifyou’rerunningthisone-linerintheWindowscommandprompt,however,youhavetomakesurethatit’salwayswrappedindoublequotesfromtheoutsideandthatyou’veescapedanydoublequotesandspecialcharactersinsideit.Withthosechanges,theone-linerlookslikethisonWindows:

perl -pe "$\ = \"\n\"" file

Thisone-linerisgettinguglyquickly,butyoucanemployacoupleofPerltrickstomakeitlookalittlenicer.First,replacedoublequotesinsidetheone-linerwiththeqq/.../operator,whichdouble-quotesanythingbetweentheslashes.Writingqq/text/inPerlisthesameaswriting"text".Nowyourewritethisone-linerlikethis:

perl -pe "$\ = qq/\n/" file

That’salittlenicer.Youcanalsochangethecharacterthattheqqoperatorusestoseparateitscontent.Forexample,thesyntaxqq|...|double-quotesanythingbetweenthepipes:

perl -pe "$\ = qq|\n|" file

Youcouldevenusematchingparenthesesorcurlybrackets,likethis:

perl -pe "$\ = qq(\n)" file

orthis:

perl -pe "$\ = qq{\n}" file

Let’sseehowtoconvertseveralmoreone-linerstoWindows.HowaboutconvertinganIPaddresstoaninteger(one-liner4.27onpage45)?InUNIXyourun:

perl -MSocket -le 'print unpack("N", inet_aton("127.0.0.1"))'

OnWindows,youneedtochangethequotesoutsidetheone-linertodoublequotesandescapethedoublequotesinsidetheone-liner:

perl -MSocket -le "print unpack(\"N\", inet_aton(\"127.0.0.1\"))"

Oryoucanusetheqq|...|operatortoavoidescapingdoublequotesinsidetheone-liner:

perl -MSocket -le "print unpack(qq|N|, inet_aton(qq|127.0.0.1|))"

Forthingsthatdon’tneedinterpolation,suchastheformatstringNandtheIPaddress127.0.0.1,youcanalsousesinglequotesinsteadofdoublequotes:

perl -MSocket -le "print unpack('N', inet_aton('127.0.0.1'))"

Anothertrickistousetheq/.../operator,whichsingle-quotesanytextbetweentheslashes:

perl -MSocket -le "print unpack(q/N/, inet_aton(q/127.0.0.1/))"

Writingq/N/andq/127.0.0.1/isthesameaswriting'N'and'127.0.0.1'.

Let’sconvertanotherone-linerfromUNIXtoWindows.I’veexpandedittomultiplelinesforclarity:

perl -le '

$ip="127.0.0.1";

$ip =~ s/(\d+)\.?/sprintf("%02x", $1)/ge;

print hex($ip)

'

Unfortunately,toconvertthistoWindows,youhavetojoinallofthelinestogether(makingtheresultlessreadable),andapplythenewquotingrules:

perl -le "$ip=\"127.0.0.1\"; $ip =~ s/(\d+)\.?/sprintf(\"%02x\", $1)/ge; print

hex($ip)"

Youcanincreasethereadabilityalittlebyusingtheqqoperator:

perl -le "$ip=qq|127.0.0.1|; $ip =~ s/(\d+)\.?/sprintf(qq|%02x|, $1)/ge; print

hex($ip)"

orbyusingsinglequotes:

perl -le "$ip='127.0.0.1'; $ip =~ s/(\d+)\.?/sprintf('%02x', $1)/ge; print hex($ip)"

SymbolChallengesYoumightalsorunintoissueswiththecaret(^)symbolinone-linersbecausetheWindowscommandpromptusesthecaretastheescapesymbol.TotellWindowstotreatthecaretsymbolliterally,youusuallyhavetoreplaceeachcaretwithtwocarets:^^.

Let’slookatseveralexamplesthatsimplytrytoprintthe^character.Here’smyfirstattempt:

perl -e "print \"^\""

Nooutput!The^symboldisappeared.Let’stry^twice:

perl -e "print \"^^\""

Thisworked!Itprintedthe^symbol.Nowlet’stryusingsinglequotes:

perl -e "print '^'"

Thisalsoworkedandprinted^,andIdidn’tneedtoenter^twice.Usingqq/^/alsoworks:

perl -e "print qq/^/"

Asyoucansee,runningone-linersonWindowscanbetrickybecausetherearenouniformparsingrulesforthecommand-linearguments.Youmayrunintosimilarissueswhenwritingone-linerswiththe%,&,<,>,and|symbols.Ifso,tryprecedingthemwiththe^escapecharactersothat%becomes^%,&becomes^&,<becomes^<,>becomes^>,and|becomes^|.Ortrywrappingthemintheqqoperator,asIdiscussedpreviously.(Betteryet,installwin-bashandusetheone-linersthroughittoavoidalltheseissues.)

WindowsFilePathsWhenusingtheWindowscommandprompt,youcanpassfilenamestoone-linersinseveraldifferentways.Forexample,toaccessthefileC:\work\wrong-spacing.txt,youcanenter:

perl -pe "$\ = qq{\n}" C:\work\wrong-spacing.txt

Oryoucanreversetheslashes:

perl -pe "$\ = qq{\n}" C:/work/wrong-spacing.txt

Ifthefilenamecontainsspaces,youhavetoquotethepath:

perl -pe "$\ = qq{\n}" "C:\Documents and Settings\wrong-spacing.txt"

FormoreWindowsPerlusagehints,seetheWin32Perldocumentationathttp://perldoc.perl.org/perlwin32.html.

B.5PerlOne-LinersinPowerShellRunningone-linersinPowerShellisabitdifferentthanrunningtheminthecommandprompt(cmd.exe).ThemaindifferenceisthatPowerShellisamodernshellimplementationwithdifferentparsingrulesthanthecommandprompt.Inthissection,I’llshowyouhowtorunPerlone-linersinPowerShell.

First,youneedtoverifythatPerlworksinthePowerShellenvironment.Runperl --versioninthePowerShell.IfthecommandoutputsthePerlversion,thenPerlisavailable,andyoushouldbeabletoruntheone-liners.Otherwise,updatethePathenvironmentvariableandappendStrawberryPerl’sbinarydirectorytoitbyrunningthefollowingcommandinPowerShell:

$env:Path += ";C:\strawberryperl\perl\bin"

ThePathvariabletellsPowerShellwheretolookforexecutables,sowhenyourunperl,itsearchesallthedirectories(separatedbythe;character)forperl.exe.

ConvertingOne-LinersinPowerShellConsiderone-liner2.1(page7),whichdouble-spacesafile.InUNIX,theone-linerlookslikethis:

perl -pe '$\ = "\n"' file

Tomakethisone-linerworkinPowerShell,youhavetochangethreethings:

Escapethe$symbol,whichPowerShellusesforvariables,byaddingthe`character(backtick)beforeit:`$.

Aswiththecmd.execommandprompt,makesuredoublequotesareusedontheoutsideoftheone-liner.

Usetheqq/.../operatorforthedoublequotesinsidetheone-liner,asexplainedin“ConvertingOne-LinersintheWindowsCommandPrompt”onpage108.Youcan’tjustescapethedoublequoteswithabackslashaswiththecommandprompt,however;youmustusetheqq/.../operator.

Whenyouputitalltogether,thePowerShellversionofthisone-linerbecomes:

perl -pe "`$\ = qq/\n/" file

Tospecifyfullpathstofiles,useWindows-stylepaths.Forexample,toreferenceafileatC:\work\wrong-spacing.txt,enterthatpathdirectlyaftertheone-liner:

perl -pe "`$\ = qq/\n/" C:\work\wrong-spacing.txt

Ifthefilenameorfilepathcontainsspaces,enteritlikethis,withdoublequotesaroundit:

perl -pe "`$\ = qq/\n/" "C:\Documents and Settings\wrong-spacing.txt"

Nowforanotherversionofthesameone-liner.InUNIXtheone-linerlookslikethis:

perl -pe '$_ .= "\n" unless /^$/' file

ButinPowerShellyouhavetochangetheoutersinglequotestodoublequotes,escapethe$symbol,andchangedoublequotestoqq/.../insidetheone-liner:

perl -pe "`$_ .= qq/\n/ unless /^`$/" file

Nowlet’slookattheone-linerthatnumbersthenon-emptylinesinafile(one-liner3.2onpage18):

perl -pe '$_ = ++$x." $_" if /./'

WhenconvertedtoPowerShell,theone-linerlookslikethis:

perl -pe "`$_ = ++`$a.qq/ `$_/ if /./"

perl -pe "`$_ = ++`$a.qq/ `$_/ if /./"

Howabouttheartisticone-linerthatchecksifanumberisprime(one-liner4.1onpage29)?

perl -lne '(1x$_) !~ /^1?$|^(11+?)\1+$/ && print "$_ is prime"'

InPowerShell,theone-linerlookslikethis:

perl -lne "(1x`$_) !~ /^1?`$|^(11+?)\1+`$/ && print qq/`$_ is prime/"

Remembertheone-lineronpage46thatconvertsIPstointegers?Here’showitlooksinUNIX:

perl -le '

$ip="127.0.0.1";

$ip =~ s/(\d+)\.?/sprintf("%02x", $1)/ge;

print hex($ip)

'

Andhere’sthesameone-linerforPowerShell:

perl -le "

`$ip=qq|127.0.0.1|;

`$ip =~ s/(\d+)\.?/sprintf(qq|%02x|, `$1)/ge;

print hex(`$ip)

"

One-LinersinPowerShell3.0+Ifyou’rerunningPowerShellversion3.0orlater,youcanusethe--%escapesequencetopreventPowerShellfromdoinganyadditionalparsing.

TofindoutwhichPowerShellversionyou’rerunning,enter$PSVersionTable.PSVersionintheshell.Itshouldoutputatablelikethis:

PS C:\Users\Administrator> $PSVersionTable.PSVersion

Major Minor Build Revision

----- ----- ----- --------

3 0 -1 -1

Thistableshowsthatyou’rerunningPowerShellversion3.0,whichhasthe--%escapesequence.(OlderversionsofPowerShelldon’thavethissequence,inwhichcaseyouhavetousethetricksIdescribedearlier.)

Whenusingthe--%escapesequence,youdon’thavetoescapethe$symbol.Italsoletsyouescapethedoublequoteswithbackslashesinsidetheone-liner.Forexample,here’showtheone-linerthatdouble-spaceslineslookswiththe--%escapesequence:

perl --% -pe "$\ = \"\n\""

Youcanalsousetheqq/.../operatortoavoidescapingdoublequotesinsidetheone-liner:

perl --% -pe "$\ = qq/\n/"

Here’showyoucanwritetheotherversionofthesameone-linerinPowerShell3.0andlater:

perl --% -pe "$_ .= \"\n\" unless /^$/" file

Andthisishowtheone-linerthatnumbersthelineslooks:

perl --% -pe "$_ = ++$a.qq/ $_/ if /./"

Here’stheone-linerthatusesaregularexpressiontoseeifanumberisprime:

perl --% -lne "(1x$_) !~ /^1?$|^(11+?)\1+$/ && print \"$_ is prime\""

Andhere’stheone-linerthatconvertsIPstointegers:

perl --% -le "

$ip=\"127.0.0.1\";

$ip =~ s/(\d+)\.?/sprintf(\"%02x\", $1)/ge;

print hex($ip)

"

Asyoucansee,runningone-linersinPowerShellisquitetrickyandrequiresseveralworkarounds.Again,Irecommendthatyouinstallwin-bashasdescribed

severalworkarounds.Again,Irecommendthatyouinstallwin-bashasdescribedin“BashonWindows”onpage106toavoidhavingtoimplementtheseworkarounds.

AppendixC.Perl1Line.Txt

AsIwaswritingthisbook,Icompiledalltheone-linersinafilecalledperl1line.txt.Thisappendixisthatfile.Itcomesinveryhandywhenyouneedtoquicklylookupaone-liner.Youcanjustopenperl1line.txtinatexteditorandsearchfortheactionyouwanttoperform.Thelatestversionofthisfilecanalwaysbefoundathttp://www.catonmat.net/download/perl1line.txt.

C.1Spacing

Double-spaceafile

perl -pe '$\ = "\n"'

perl -pe 'BEGIN { $\ = "\n" }'

perl -pe '$_ .= "\n"'

perl -pe 's/$/\n/'

perl -nE 'say'

Double-spaceafile,excludingtheblanklines

perl -pe '$_ .= "\n" unless /^$/'

perl -pe '$_ .= "\n" if /\S/'

Triple-spaceafile

perl -pe '$\ = "\n\n"'

perl -pe '$_ .= "\n\n"'

perl -pe 's/$/\n\n/'

N-spaceafile

perl -pe '$_ .= "\n"x7'

Addablanklinebeforeeveryline

Addablanklinebeforeeveryline

perl -pe 's/^/\n/'

Removeallblanklines

perl -ne 'print unless /^$/'

perl -lne 'print if length'

perl -ne 'print if /\S/'

Removeallconsecutiveblanklines,leavingonlyone

perl -00 -pe ''

perl -00pe0

Compress/expandallblanklinesintoNconsecutivelines

perl -00 -pe '$_ .= "\n"x2'

Double-spacebetweenallwords

perl -pe 's/ / /g'

Removeallspacingbetweenwords

perl -pe 's/ +//g'

perl -pe 's/\s+//g'

Changeallspacingbetweenwordstoonespace

perl -pe 's/ +/ /g'

Insertaspacebetweenallcharacters

perl -lpe 's// /g'

C.2Numbering

Numberalllinesinafile

Numberalllinesinafile

perl -pe '$_ = "$. $_"'

perl -ne 'print "$. $_"'

Numberonlynon-emptylinesinafile

perl -pe '$_ = ++$x." $_" if /./'

perl -pe '$_ = ++$x." $_" if /\S/'

Numberandprintonlynon-emptylinesinafile(dropemptylines)

perl -ne 'print ++$x." $_" if /./'

Numberalllinesbutprintlinenumbersonlyfornon-emptylines

perl -pe '$_ = "$. $_" if /./'

Numberonlylinesthatmatchapattern;printothersunmodified

perl -pe '$_ = ++$x." $_" if /regex/'

Numberandprintonlylinesthatmatchapattern

perl -ne 'print ++$x." $_" if /regex/'

Numberalllinesbutprintlinenumbersonlyforlinesthatmatchapattern

perl -pe '$_ = "$. $_" if /regex/'

Numberalllinesinafileusingacustomformat

perl -ne 'printf "%-5d %s", $., $_'

Printthetotalnumberoflinesinafile(emulatewc-l)

perl -lne 'END { print $. }'

perl -le 'print $n = () = <>'

perl -le 'print $n = (() = <>)'

perl -le 'print scalar(() = <>)'

perl -le 'print scalar(() = <>)'

perl -le 'print scalar(@foo = <>)'

perl -ne '}{print $.'

Printthenumberofnon-emptylinesinafile

perl -le 'print scalar(grep { /./ } <>)'

perl -le 'print ~~grep{/./}<>'

perl -le 'print~~grep/./,<>'

perl -lE 'say~~grep/./,<>'

Printthenumberofemptylinesinafile

perl -lne '$x++ if /^$/; END { print $x+0 }'

perl -lne '$x++ if /^$/; END { print int $x }'

perl -le 'print scalar(grep { /^$/ } <>)'

perl -le 'print ~~grep{ /^$/ } <>'

Printthenumberoflinesinafilethatmatchapattern(emulategrep-c)

perl -lne '$x++ if /regex/; END { print $x+0 }'

Numberwordsacrossalllines

perl -pe 's/(\w+)/++$i.".$1"/ge'

Numberwordsoneachindividualline

perl -pe '$i=0; s/(\w+)/++$i.".$1"/ge'

Replaceallwordswiththeirnumericpositions

perl -pe 's/(\w+)/++$i/ge'

C.3Calculations

Checkifanumberisaprime

perl -lne '(1x$_) !~ /^1?$|^(11+?)\1+$/ && print "$_ is prime"'

perl -lne '(1x$_) !~ /^1?$|^(11+?)\1+$/ && print "$_ is prime"'

Printthesumofallfieldsoneachline

perl -MList::Util=sum -alne 'print sum @F'

perl -MList::Util=sum -F: -alne 'print sum @F'

Printthesumofallfieldsonalllines

perl -MList::Util=sum -alne 'push @S,@F; END { print sum @S }'

perl -MList::Util=sum -alne '$s += sum @F; END { print $s }'

Shuffleallfieldsoneachline

perl -MList::Util=shuffle -alne 'print "@{[shuffle @F]}"'

perl -MList::Util=shuffle -alne 'print join " ", shuffle @F'

Findthenumericallysmallestelement(minimumelement)oneachline

perl -MList::Util=min -alne 'print min @F'

Findthenumericallysmallestelement(minimumelement)overalllines

perl -MList::Util=min -alne '@M = (@M, @F); END { print min @M }'

perl -MList::Util=min -alne '

$min = min @F;

$rmin = $min unless defined $rmin && $min > $rmin;

END { print $rmin }

'

perl -MList::Util=min -alne '$min = min($min // (), @F); END { print $min }'

Findthenumericallylargestelement(maximumelement)oneachline

perl -MList::Util=max -alne 'print max @F'

Findthenumericallylargestelement(maximumelement)overalllines

perl -MList::Util=max -alne '@M = (@M, @F); END { print max @M }'

perl -MList::Util=max -alne '

$max = max @F;

$rmax = $max unless defined $rmax && $max < $rmax;

END { print $rmax }

'

perl -MList::Util=max -alne '$max = max($max // (), @F); END { print $max }'

Replaceeachfieldwithitsabsolutevalue

perl -alne 'print "@{[map { abs } @F]}"'

Printthetotalnumberoffieldsoneachline

perl -alne 'print scalar @F'

Printthetotalnumberoffieldsoneachline,followedbytheline

perl -alne 'print scalar @F, " $_"'

Printthetotalnumberoffieldsonalllines

perl -alne '$t += @F; END { print $t }'

Printthetotalnumberoffieldsthatmatchapattern

perl -alne 'map { /regex/ && $t++ } @F; END { print $t || 0 }'

perl -alne '$t += /regex/ for @F; END { print $t }'

perl -alne '$t += grep /regex/, @F; END { print $t }'

Printthetotalnumberoflinesthatmatchapattern

perl -lne '/regex/ && $t++; END { print $t || 0 }'

Printthenumberπ

perl -Mbignum=bpi -le 'print bpi(21)'

perl -Mbignum=PI -le 'print PI'

Printthenumbere

perl -Mbignum=bexp -le 'print bexp(1,21)'

perl -Mbignum=e -le 'print e'

PrintUNIXtime(secondssinceJanuary1,1970,00:00:00UTC)

perl -le 'print time'

PrintGreenwichMeanTimeandlocalcomputertime

perl -le 'print scalar gmtime'

perl -le 'print scalar localtime'

Printyesterday’sdate

perl -MPOSIX -le '

@now = localtime;

$now[3] -= 1;

print scalar localtime mktime @now

'

Printthedate14months,9days,and7secondsago

perl -MPOSIX -le '

@now = localtime;

$now[0] -= 7;

$now[3] -= 9;

$now[4] -= 14;

print scalar localtime mktime @now

'

Calculatethefactorial

perl -MMath::BigInt -le 'print Math::BigInt->new(5)->bfac()'

perl -le '$f = 1; $f *= $_ for 1..5; print $f'

Calculatethegreatestcommondivisor

perl -MMath::BigInt=bgcd -le 'print bgcd(@list_of_numbers)'

perl -MMath::BigInt=bgcd -le 'print bgcd(20,60,30)'

perl -MMath::BigInt=bgcd -anle 'print bgcd(@F)'

perl -le '

$n = 20; $m = 35;

($m,$n) = ($n,$m%$n) while $n;

($m,$n) = ($n,$m%$n) while $n;

print $m

'

Calculatetheleastcommonmultiple

perl -MMath::BigInt=blcm -le 'print blcm(35,20,8)'

perl -MMath::BigInt=blcm -anle 'print blcm(@F)'

perl -le '

$a = $n = 20;

$b = $m = 35;

($m,$n) = ($n,$m%$n) while $n;

print $a*$b/$m

'

Generate10randomnumbersbetween5and15(excluding15)

perl -le 'print join ",", map { int(rand(15-5))+5 } 1..10'

perl -le '

$n=10;

$min=5;

$max=15;

$, = " ";

print map { int(rand($max-$min))+$min } 1..$n;

'

Generateallpermutationsofalist

perl -MAlgorithm::Permute -le '

$l = [1,2,3,4,5];

$p = Algorithm::Permute->new($l);

print "@r" while @r = $p->next

'

perl -MAlgorithm::Permute -le '

@l = (1,2,3,4,5);

Algorithm::Permute::permute { print "@l" } @l

'

Generatethepowerset

perl -MList::PowerSet=powerset -le '

@l = (1,2,3,4,5);

print "@$_" for @{powerset(@l)}

'

'

ConvertanIPaddresstoanunsignedinteger

perl -le '

$i=3;

$u += ($_<<8*$i--) for "127.0.0.1" =~ /(\d+)/g;

print $u

'

perl -le '

$ip="127.0.0.1";

$ip =~ s/(\d+)\.?/sprintf("%02x", $1)/ge;

print hex($ip)

'

perl -le 'print unpack("N", 127.0.0.1)'

perl -MSocket -le 'print unpack("N", inet_aton("127.0.0.1"))'

ConvertanunsignedintegertoanIPaddress

perl -MSocket -le 'print inet_ntoa(pack("N", 2130706433))'

perl -le '

$ip = 2130706433;

print join ".", map { (($ip>>8*($_))&0xFF) } reverse 0..3

'

perl -le '

$ip = 2130706433;

$, = ".";

print map { (($ip>>8*($_))&0xFF) } reverse 0..3

'

perl -le '

$ip = 2130706433;

$, = ".";

print map { (($ip>>8*($_))&0xFF) } 3,2,1,0

'

C.4WorkingwithArraysandStrings

Generateandprintthealphabet

perl -le 'print a..z'

perl -le 'print ("a".."z")'

perl -le '$, = ","; print ("a".."z")'

perl -le '$alphabet = join ",", ("a".."z"); print $alphabet'

Generateandprintallthestringsfrom“a”to“zz”

perl -le 'print join ",", ("a".."zz")'

perl -le 'print join ",", "aa".."zz"'

Createahexlookuptable

@hex = (0..9, "a".."f")

perl -le '

$num = 255;

@hex = (0..9, "a".."f");

while ($num) {

$s = $hex[($num % 16)].$s;

$num = int $num/16;

}

print $s

'

perl -le 'printf("%x", 255)'

perl -le '$num = "ff"; print hex $num'

Generatearandomeight-characterpassword

perl -le 'print map { ("a".."z")[rand 26] } 1..8'

perl -le 'print map { ("a".."z", 0..9)[rand 36] } 1..8'

Createastringofspecificlength

perl -le 'print "a"x50'

perl -e 'print "a"x1024'

perl -le '@list = (1,2)x20; print "@list"'

Createanarrayfromastring

@months = split ' ', "Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec"

@months = qw/Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec/

Createastringfromthecommand-linearguments

perl -le 'print "(", (join ",", @ARGV), ")"' val1 val2 val3

perl -le '

print "INSERT INTO table VALUES (", (join ",", @ARGV), ")"

' val1 val2 val3

Findthenumericvaluesforcharactersinastring

perl -le 'print join ", ", map { ord } split //, "hello world"'

perl -le 'print join ", ", unpack("C*", "hello world")'

perl -le '

print join ", ", map { sprintf "0x%x", ord $_ } split //, "hello world"

'

perl -le '

print join ", ", map { sprintf "%o", ord $_ } split //, "hello world"

'

perl -le '

print join ", ", map { sprintf "%#o", ord $_ } split //, "hello world"

'

ConvertalistofnumericASCIIvaluesintoastring

perl -le '

@ascii = (99, 111, 100, 105, 110, 103);

print pack("C*", @ascii)

'

perl -le '

@ascii = (99, 111, 100, 105, 110, 103);

$str = join "", map chr, @ascii;

print $str

'

perl -le 'print map chr, 99, 111, 100, 105, 110, 103'

perl -le 'print map chr, @ARGV' 99 111 100 105 110 103

Generateanarraywithoddnumbersfrom1to100

perl -le '@odd = grep {$_ % 2 == 1} 1..100; print "@odd"'

perl -le '@odd = grep { $_ & 1 } 1..100; print "@odd"'

Generateanarraywithevennumbersfrom1to100

perl -le '@even = grep {$_ % 2 == 0} 1..100; print "@even"'

Findthelengthofastring

perl -le 'print length "one-liners are great"'

Findthenumberofelementsinanarray

perl -le '@array = ("a".."z"); print scalar @array'

perl -le '@array = ("a".."z"); print $#array + 1'

perl -le 'print scalar @ARGV' *.txt

perl -le 'print scalar (@ARGV=<*.txt>)'

C.5TextConversionandSubstitution

ROT13astring

perl -le '$str = "bananas"; $str =~ y/A-Za-z/N-ZA-Mn-za-m/; print $str'

perl -lpe 'y/A-Za-z/N-ZA-Mn-za-m/' file

perl -pi.bak -e 'y/A-Za-z/N-ZA-Mn-za-m/' file

Base64-encodeastring

perl -MMIME::Base64 -e 'print encode_base64("string")'

perl -MMIME::Base64 -0777 -ne 'print encode_base64($_)' file

Base64-decodeastring

perl -MMIME::Base64 -le 'print decode_base64("base64string")'

perl -MMIME::Base64 -0777 -ne 'print decode_base64($_)' file

URL-escapeastring

perl -MURI::Escape -le 'print uri_escape("http://example.com")'

URL-unescapeastring

perl -MURI::Escape -le 'print uri_unescape("http%3A%2F%2Fexample.com")'

HTML-encodeastring

perl -MHTML::Entities -le 'print encode_entities("<html>")'

HTML-decodeastring

perl -MHTML::Entities -le 'print decode_entities("&lt;html&gt;")'

Convertalltexttouppercase

perl -nle 'print uc'

perl -ple '$_ = uc'

perl -nle 'print "\U$_"'

Convertalltexttolowercase

perl -nle 'print lc'

perl -nle 'print "\L$_"'

Uppercaseonlythefirstletterofeachline

perl -nle 'print ucfirst lc'

perl -nle 'print "\u\L$_"'

Invertthelettercase

perl -ple 'y/A-Za-z/a-zA-Z/'

Title-caseeachline

perl -ple 's/(\w+)/\u$1/g'

Stripleadingwhitespace(spaces,tabs)fromthebeginningofeachline

perl -ple 's/^[ \t]+//'

perl -ple 's/^\s+//'

Striptrailingwhitespace(spaces,tabs)fromtheendofeachline

perl -ple 's/[ \t]+$//'

perl -ple 's/\s+$//'

Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

perl -ple 's/^[ \t]+|[ \t]+$//g'

perl -ple 's/^\s+|\s+$//g'

ConvertUNIXnewlinestoDOS/Windowsnewlines

perl -pe 's|\012|\015\012|'

ConvertDOS/WindowsnewlinestoUNIXnewlines

perl -pe 's|\015\012|\012|'

ConvertUNIXnewlinestoMacnewlines

perl -pe 's|\012|\015|'

Substitute(findandreplace)“foo”with“bar”oneachline

perl -pe 's/foo/bar/'

perl -pe 's/foo/bar/g'

Substitute(findandreplace)“foo”with“bar”onlinesthatmatch“baz”

perl -pe '/baz/ && s/foo/bar/'

perl -pe 's/foo/bar/ if /baz/'

Printparagraphsinreverseorder

perl -00 -e 'print reverse <>' file

Printalllinesinreverseorder

perl -lne 'print scalar reverse $_'

perl -lne 'print scalar reverse'

perl -lpe '$_ = reverse $_'

perl -lpe '$_ = reverse'

Printcolumnsinreverseorder

perl -alne 'print "@{[reverse @F]}"'

perl -F: -alne 'print "@{[reverse @F]}"'

perl -F: -alne '$" = ":"; print "@{[reverse @F]}"'

perl -F: -alne '$" = ":"; print "@{[reverse @F]}"'

C.6SelectivelyPrintingandDeletingLines

Printthefirstlineofafile(emulatehead-1)

perl -ne 'print; exit' file

perl -i -ne 'print; exit' file

perl -i.bak -ne 'print; exit' file

Printthefirst10linesofafile(emulatehead-10)

perl -ne 'print if $. <= 10' file

perl -ne '$. <= 10 && print' file

perl -ne 'print if 1..10' file

perl -ne 'print; exit if $. == 10' file

Printthelastlineofafile(emulatetail-1)

perl -ne '$last = $_; END { print $last }' file

perl -ne 'print if eof' file

Printthelast10linesofafile(emulatetail-10)

perl -ne 'push @a, $_; @a = @a[@a-10..$#a] if @a>10; END { print @a }' file

perl -ne 'push @a, $_; shift @a if @a>10; END { print @a }' file

Printonlylinesthatmatcharegularexpression

perl -ne '/regex/ && print'

perl -ne 'print if /regex/'

Printonlylinesthatdonotmatcharegularexpression

perl -ne '!/regex/ && print'

perl -ne 'print if !/regex/'

perl -ne 'print unless /regex/'

perl -ne '/regex/ || print'

Printeverylineprecedingalinethatmatchesaregularexpression

perl -ne '/regex/ && $last && print $last; $last = $_'

Printeverylinefollowingalinethatmatchesaregularexpression

perl -ne 'if ($p) { print; $p = 0 } $p++ if /regex/'

perl -ne '$p && print && ($p = 0); $p++ if /regex/'

perl -ne '$p && print; $p = /regex/'

PrintlinesthatmatchregularexpressionsAAAandBBBinanyorder

perl -ne '/AAA/ && /BBB/ && print'

Printlinesthatdon’tmatchregularexpressionsAAAandBBB

perl -ne '!/AAA/ && !/BBB/ && print'

PrintlinesthatmatchregularexpressionAAAfollowedbyBBBfollowedbyCCC

perl -ne '/AAA.*BBB.*CCC/ && print'

Printlinesthatareatleast80characterslong

perl -ne 'print if length >= 80'

perl -lne 'print if length >= 80'

Printlinesthatarefewerthan80characterslong

perl -ne 'print if length() < 80'

Printonlyline13

perl -ne '$. == 13 && print && exit'

Printalllinesexceptline27

perl -ne '$. != 27 && print'

perl -ne 'print if $. != 27'

perl -ne 'print unless $. == 27'

perl -ne 'print unless $. == 27'

Printonlylines13,19,and67

perl -ne 'print if $. == 13 || $. == 19 || $. == 67'

perl -ne '

@lines = (13, 19, 88, 290, 999, 1400, 2000);

print if grep { $_ == $. } @lines

'

Printalllinesfrom17to30

perl -ne 'print if $. >= 17 && $. <= 30'

perl -ne 'print if 17..30'

Printalllinesbetweentworegularexpressions(includingthelinesthatmatch)

perl -ne 'print if /regex1/../regex2/'

Printthelongestline

perl -ne '

$l = $_ if length($_) > length($l);

END { print $l }

'

Printtheshortestline

perl -ne '

$s = $_ if $. == 1;

$s = $_ if length($_) < length($s);

END { print $s }

'

Printalllinescontainingdigits

perl -ne 'print if /\d/'

Printalllinescontainingonlydigits

perl -ne 'print if /^\d+$/'

perl -ne 'print if /^\d+$/'

perl -lne 'print unless /\D/'

Printalllinescontainingonlyalphabeticcharacters

perl -ne 'print if /^[[:alpha:]]+$/

Printeverysecondline

perl -ne 'print if $. % 2'

Printeverysecondline,beginningwiththesecondline

perl -ne 'print if $. % 2 == 0'

perl -ne 'print unless $. % 2'

Printallrepeatedlinesonlyonce

perl -ne 'print if ++$a{$_} == 2'

Printalluniquelines

perl -ne 'print unless $a{$_}++'

C.7UsefulRegularExpressions

MatchsomethingthatlookslikeanIPaddress

/^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}$/

/^(\d{1,3}\.){3}\d{1,3}$/

perl -ne 'print if /^\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}$/'

Testwhetheranumberisintherange0to255

/^([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])$/

perl -le '

map { $n++ if /^([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])$/ } 0..255;

END { print $n }

'

perl -le '

perl -le '

map { $n++ if /^([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])$/ } 0..1000;

END { print $n }

'

MatchanIPaddress

my $ip_part = qr/[0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5]/;

if ($ip =~ /^$ip_part\.$ip_part\.$ip_part\.$ip_part$/) {

print "valid ip\n";

}

if ($ip =~ /^($ip_part\.){3}$ip_part$/) {

print "valid ip\n";

}

perl -ne '

$ip_part = qr|([0-9]|[0-9][0-9]|1[0-9][0-9]|2[0-4][0-9]|25[0-5])|;

print if /^($ip_part\.){3}$ip_part$/

'

Checkwhetherastringlookslikeanemailaddress

/\S+@\S+\.\S+/

use Email::Valid;

print Email::Valid->address('cats@catonmat.net') ? 'valid email' : 'invalid email';

perl -MEmail::Valid -ne 'print if Email::Valid->address($_)'

Checkwhetherastringisanumber

/^\d+$/

/^[+-]?\d+$/

/^[+-]?\d+\.?\d*$/

perl -MRegexp::Common -ne 'print if /$RE{num}{real}/'

perl -MRegexp::Common -ne 'print if /$RE{num}{hex}/'

perl -MRegexp::Common -ne 'print if /$RE{num}{oct}/'

perl -MRegexp::Common -ne 'print if /$RE{num}{bin}/'

Checkwhetherawordappearsinastringtwice

/(word).*\1/

Increaseallintegersinastringbyone

$str =~ s/(\d+)/$1+1/ge

perl -MRegexp::Common -pe 's/($RE{num}{real})/$1+1/ge'

perl -MRegexp::Common -pe 's/($RE{num}{real})/$1+1/ge'

ExtracttheHTTPUser-AgentstringfromHTTPheaders

/^User-Agent: (.+)$/

MatchprintableASCIIcharacters

/[ -~]/

ExtracttextbetweentwoHTMLtags

m|<strong>([^<]*)</strong>|

m|<strong>(.*?)</strong>|

Replaceall<b>tagswith<strong>

$html =~ s|<(/)?b>|<$1strong>|g

Extractallmatchesfromaregularexpression

my @matches = $text =~ /regex/g;

Index

ANOTEONTHEDIGITALINDEX

Alinkinanindexentryisdisplayedasthesectiontitleinwhichthatentryappears.Becausesomesectionshavemultipleindexmarkers,itisnotunusualforanentrytohaveseverallinkstothesamesection.Clickingonanylinkwilltakeyoudirectlytotheplaceinthetextinwhichthemarkerappears.

Symbols

!operator,7.6Printonlylinesthatdonotmatcharegularexpression

!~operator,4.1Checkifanumberisaprime

$,(fieldseparator)variable,4.3Printthesumofallfieldsonalllines,A.5Variables$1,$2,$3,andsoon

$.variable,3.1Numberalllinesinafile,Using$_explicitly

$/variable,A.2Variable$.

$1,$2,$3,…variables,3.13Numberwordsacrossalllines,A.4Variable$\

$variable+0construct,3.10Printthenumberofnon-emptylinesinafile

$variable=()=<>construct,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

$variable||0construct,4.13Printthetotalnumberoffieldsthatmatchapattern

$\variable,2.1Double-spaceafile,A.4Variable$\

$_variable,2.1Double-spaceafile,6.22Printalllinesinreverseorder,Perl’sSpecialVariables

$”variable,6.23Printcolumnsinreverseorder,A.5Variables$1,$2,$3,andsoon

%ENVvariable,A.9Variable@ARGV

-0command-lineargument,2.6Removeallblanklines,A.2Variable$.

-00command-lineargument,2.6Removeallblanklines,6.19Substitute(findandreplace)“foo”with“bar”oneachline

-0777command-lineargument,6.1ROT13astring

-acommand-lineargument,4.1Checkifanumberisaprime,A.7Variable$”

-ecommand-lineargument,IntroductiontoPerlOne-Liners,2.1Double-spaceafile

-Ecommand-lineargument,2.1Double-spaceafile

-Fcommand-lineargument,4.2Printthesumofallfieldsoneachline,6.23Printcolumnsinreverseorder

-icommand-lineargument,IntroductiontoPerlOne-Liners,6.1ROT13astring,7.1Printthefirstlineofafile(emulatehead-1)

-i.bakcommand-lineargument,IntroductiontoPerlOne-Liners,6.1ROT13astring,7.1Printthefirstlineofafile(emulatehead-1)

-lcommand-lineargument,IntroductiontoPerlOne-Liners,2.6Removeallblanklines

-Mcommand-lineargument,4.1Checkifanumberisaprime

-Mmodule=arg,4.2Printthesumofallfieldsoneachline

-ncommand-lineargument,IntroductiontoPerlOne-Liners,2.4N-spaceafile,A.1Variable$_

-pcommand-lineargument,IntroductiontoPerlOne-Liners,2.1Double-spaceafile,Using$_withthe-nargument

.operator,2.1Double-spaceafile

..(range)operator,WorkingWithArraysandStrings,7.2Printthefirst10linesofafile(emulatehead-10),7.17Printalllinesfrom17to30

//operator,4.6Findthenumericallysmallestelement(minimumelement)overalllines

/etc/passwd,IntroductiontoPerlOne-Liners,A.7Variable$”

<>(diamond)operator,2.6Removeallblanklines,3.9Printthetotalnumberoflinesinafile(emulatewc-l),5.13Findthenumberofelementsinanarray,6.19Substitute(findandreplace)“foo”with“bar”oneachline

=~operator,4.1Checkifanumberisaprime

@ARGVarray,5.6Createanarrayfromastring,5.6Createanarrayfromastring,A.7Variable$”,A.7Variable$”

@array1=(@array1,@array2)construct,4.4Shuffleallfieldsoneachline

@Farray,4.1Checkifanumberisaprime,4.1Checkifanumberisaprime,4.9Replaceeachfieldwithitsabsolutevalue,4.9Replaceeachfieldwithitsabsolutevalue,A.7Variable$”,A.7Variable$”

@TimToady(LarryWall),2.6Removeallblanklines

@{[…]}construct,4.3Printthesumofallfieldsonalllines

\Eescapesequence,6.6HTML-encodeastring,6.9Convertalltexttolowercase

\Lescapesequence,6.9Convertalltexttolowercase

\Uescapesequence,6.6HTML-encodeastring

\uescapesequence,6.9Convertalltexttolowercase

||(logicalOR)operator,4.6Findthenumericallysmallestelement(minimumelement)overalllines

}{(eskimooperator)construct,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

~~(doublebitwisenegate)construct,3.9Printthetotalnumberoflinesinafile(emulatewc-l),3.10Printthenumberofnon-emptylinesinafile

πconstant,4.15Printthenumberπ,4.15Printthenumberπ

“actionifcondition”statement,3.1Numberalllinesinafile,3.4Numberalllinesbutprintlinenumbersonlyfornon-emptylines

“actionunlesscondition”statement,2.2Double-spaceafile,excludingtheblanklines

A

Abigail,Calculations

abs(absolutevalue)function,4.9Replaceeachfieldwithitsabsolutevalue

adding

blanklines,Spacing

fields(calculatingsumof),IntroductiontoPerlOne-Liners,4.1Checkifanumberisaprime

spaces,betweencharacters,2.11Changeallspacingbetweenwordstoonespace

addresses

email,matchinglookalikeswithregularexpressions,8.3MatchanIPaddress

IP

convertingunsignedintegers,toandfrom,4.26Generatethepowerset

lookalikes,matching,UsefulRegularExpressions

valid,matching,8.2Testwhetheranumberisintherange0to255

Algorithm::Permutemodule,4.24Generate10randomnumbersbetween5and15(excluding15)

alphabet,generating,WorkingWithArraysandStrings

arguments.(seecommand-linearguments)

ARGVfilehandle,Using$_explicitly

arrays.(seelists)

ASCII

characters,printable,findingwithregularexpressions,8.9MatchprintableASCIIcharacters

numericvalues,convertingintostrings,5.9ConvertalistofnumericASCIIvaluesintoastring

auto-splitting,4.1Checkifanumberisaprime,A.7Variable$”

B

backupfiles,IntroductiontoPerlOne-Liners

base64encoding,6.1ROT13astring

bash,onWindows,UsingPerlOne-LinersOnWindows

one-linersin,B.3PerlOne-LinersinWindowsBash

bash.exe(win-bash),UsingPerlOne-LinersOnWindows,B.3PerlOne-LinersinWindowsBash

BEGINblock,2.1Double-spaceafile

bignummodule,4.15Printthenumberπ

blanklines,addingorremoving,Spacing

blocks

BEGIN,2.1Double-spaceafile

END,3.8Numberalllinesinafileusingacustomformat

C

calculating

absolutevalue,4.9Replaceeachfieldwithitsabsolutevalue

array,numberofelementsin,5.10Generateanarraywithoddnumbersfrom1to100

econstant,4.15Printthenumberπ

factorials,4.19Printyesterday’sdate

greatestcommondivisor(gcd),4.21Calculatethefactorial

increasing,integersinastring,8.7Increaseallintegersinastringbyone

leastcommonmultiple(lcm),4.23Calculatetheleastcommonmultiple

maximumelement,4.6Findthenumericallysmallestelement(minimumelement)overalllines

minimumelement,4.4Shuffleallfieldsoneachline

number

ofelements,inanarray,5.10Generateanarraywithoddnumbersfrom1to100

offields,4.9Replaceeachfieldwithitsabsolutevalue

oflines,3.8Numberalllinesinafileusingacustomformat,4.13Printthetotalnumberoffieldsthatmatchapattern

numericpositions,ofwords,3.15Replaceallwordswiththeirnumericpositions

permutations,oflists,4.24Generate10randomnumbersbetween5and15(excluding15)

powerset,4.26Generatethepowerset

range,whetheranumberisin,8.1MatchsomethingthatlookslikeanIPaddress

sum,offields,IntroductiontoPerlOne-Liners,4.1Checkifanumberisaprime

πconstant,4.15Printthenumberπ

case,converting,6.6HTML-encodeastring

characters

ASCII,printable,8.9MatchprintableASCIIcharacters

caseof,6.6HTML-encodeastring

numericvaluesof,5.6Createanarrayfromastring

spacingof,2.11Changeallspacingbetweenwordstoonespace

chroperator,5.9ConvertalistofnumericASCIIvaluesintoastring

closefunction,Using$_explicitly

cmd.exe(Windowscommandprompt),B.4PerlOne-LinersintheWindowsCommandPrompt

columns

orderof,reversing,6.22Printalllinesinreverseorder

sumofelementsin,IntroductiontoPerlOne-Liners

commandprompt,Windows,B.4PerlOne-LinersintheWindowsCommandPrompt

command-linearguments

-0,2.6Removeallblanklines,A.2Variable$.

-00,2.6Removeallblanklines,6.19Substitute(findandreplace)“foo”with“bar”oneachline

-0777,6.1ROT13astring

-a,4.1Checkifanumberisaprime,A.7Variable$”

-e,IntroductiontoPerlOne-Liners,2.1Double-spaceafile

-E,2.1Double-spaceafile

-F,4.2Printthesumofallfieldsoneachline,6.23Printcolumnsinreverseorder

-i,IntroductiontoPerlOne-Liners,6.1ROT13astring,7.1Printthefirstlineofafile(emulatehead-1)

-i.bak,IntroductiontoPerlOne-Liners,6.1ROT13astring,7.1Printthefirstlineofafile(emulatehead-1)

-l,IntroductiontoPerlOne-Liners,2.6Removeallblanklines

-M,4.1Checkifanumberisaprime

-Mmodule=arg,4.2Printthesumofallfieldsoneachline

-n,IntroductiontoPerlOne-Liners,2.4N-spaceafile,A.1Variable$_

-p,IntroductiontoPerlOne-Liners,2.1Double-spaceafile,Using$_withthe-nargument

ComprehensivePerlArchiveNetwork(CPAN),IntroductiontoPerlOne-Liners

conditionalstatements(if,unless),2.2Double-spaceafile,excludingtheblanklines,3.1Numberalllinesinafile,3.4Numberalllinesbutprintlinenumbersonlyfornon-emptylines

constants

e,4.15Printthenumberπ

π,4.15Printthenumberπ

constructs

$variable+0construct,3.10Printthenumberofnon-emptylinesinafile

$variable=()=<>construct,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

$variable||0construct,4.13Printthetotalnumberoffieldsthatmatchapattern

@array1=(@array1,@array2)construct,4.4Shuffleallfieldsoneachline

@{[…]}construct,4.3Printthesumofallfieldsonalllines

}{(eskimooperator),3.9Printthetotalnumberoflinesinafile(emulatewc-l)

~~(doublebitwisenegate),3.9Printthetotalnumberoflinesinafile(emulatewc-l),3.10Printthenumberofnon-emptylinesinafile

converting

case,6.6HTML-encodeastring

IPaddresses,toandfromunsignedintegers,4.26Generatethepowerset

newlines,fordifferentoperatingsystems,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

numericASCIIvalues,intostrings,5.9ConvertalistofnumericASCIIvaluesintoastring

CPAN(ComprehensivePerlArchiveNetwork),IntroductiontoPerlOne-Liners

cpan(utility),6.9Convertalltexttolowercase,7.20Printtheshortestline

D

date,IntroductiontoPerlOne-Liners,4.18PrintGreenwichMeanTimeandlocalcomputertime,4.19Printyesterday’sdate

DeMorgan’slaw,7.6Printonlylinesthatdonotmatcharegularexpression

decoding

frombase64,6.2Base64-encodeastring

fromHTML,6.6HTML-encodeastring

URLs,6.2Base64-encodeastring

definedfunction,4.6Findthenumericallysmallestelement(minimumelement)overalllines

deleting

blanklines,2.4N-spaceafile

lines,byprintingothers,SelectivelyPrintingandDeletingLines

spacingbetweenwords,2.9Double-spacebetweenallwords

whitespace,6.13Stripleadingwhitespace(spaces,tabs)fromthebeginningofeachline

diamondoperator(<>),2.6Removeallblanklines,3.9Printthetotalnumberoflinesinafile(emulatewc-l),5.13Findthenumberofelementsinanarray,6.19Substitute(findandreplace)“foo”with“bar”oneachline

documentation

perldoc,IntroductiontoPerlOne-Liners

Win32Perl,SymbolChallenges

DOS/Windowsnewlines,converting,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

doublebitwisenegate~~construct,3.9Printthetotalnumberoflinesinafile(emulatewc-l),3.10Printthenumberofnon-emptylinesinafile

E

econstant,4.15Printthenumberπ,4.15Printthenumberπ

elements.,4.9Replaceeachfieldwithitsabsolutevalue

(seealsofields)

maximum,4.6Findthenumericallysmallestelement(minimumelement)overalllines

minimum,4.4Shuffleallfieldsoneachline

numberof,inanarray,5.10Generateanarraywithoddnumbersfrom1to100

emailaddresses,matchinglookalikeswithregularexpressions,8.3MatchanIPaddress

Email::Validmodule,8.4Checkwhetherastringlookslikeanemailaddress

emptylines,numberof,3.10Printthenumberofnon-emptylinesinafile

encoding

asbase64,6.1ROT13astring

asHTML,6.6HTML-encodeastring

URLs,6.2Base64-encodeastring

ENDblock,3.8Numberalllinesinafileusingacustomformat

eoffunction,7.2Printthefirst10linesofafile(emulatehead-10)

epoch,timesince,4.15Printthenumberπ

escapesequences

\E,6.6HTML-encodeastring,6.9Convertalltexttolowercase

\L,6.9Convertalltexttolowercase

\U,6.6HTML-encodeastring

\u,6.9Convertalltexttolowercase

escaping

characters,touseone-liners

inPowerShell,B.5PerlOne-LinersinPowerShell

inWindows,B.4PerlOne-LinersintheWindowsCommandPrompt

loops,withtheeskimooperator}{construct,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

URLs,6.2Base64-encodeastring

eskimooperator}{construct,3.9Printthetotalnumberoflinesinafile

(emulatewc-l)

Euclid’salgorithm,4.21Calculatethefactorial

exitfunction,7.1Printthefirstlineofafile(emulatehead-1)

F

factorials,calculating,4.19Printyesterday’sdate

fields.,3.13Numberwordsacrossalllines

(seealsowords)

absolutevalue,replacingwith,4.9Replaceeachfieldwithitsabsolutevalue

auto-splitting,4.1Checkifanumberisaprime

numberof,4.9Replaceeachfieldwithitsabsolutevalue

separator($,variable),4.3Printthesumofallfieldsonalllines,A.5Variables$1,$2,$3,andsoon

shuffling,4.3Printthesumofallfieldsonalllines

sumof,calculating,IntroductiontoPerlOne-Liners,4.1Checkifanumberisaprime

filepaths,inWindows,B.3PerlOne-LinersinWindowsBash,SymbolChallenges

filenameexpansion(globbing),5.13Findthenumberofelementsinanarray

files

backup,IntroductiontoPerlOne-Liners

numberoflinesin,3.8Numberalllinesinafileusingacustomformat

numberinglinesin,Numbering

printing,firstandlastlinesin,7.1Printthefirstlineofafile(emulatehead-1)

slurping,A.2Variable$.

spacing,Spacing

find-and-replace

HTMLtags,8.10ExtracttextbetweentwoHTMLtags

text,IntroductiontoPerlOne-Liners,6.19Substitute(findandreplace)“foo”with“bar”oneachline

flip-flopoperator,7.2Printthefirst10linesofafile(emulatehead-10),7.17Printalllinesfrom17to30

functions.,4.9Replaceeachfieldwithitsabsolutevalue

(seealsooperators)

abs,4.9Replaceeachfieldwithitsabsolutevalue

close,Using$_explicitly

defined,4.6Findthenumericallysmallestelement(minimumelement)overalllines

eof,7.2Printthefirst10linesofafile(emulatehead-10)

exit,7.1Printthefirstlineofafile(emulatehead-1)

gmtime,4.18PrintGreenwichMeanTimeandlocalcomputertime

grep,3.9Printthetotalnumberoflinesinafile(emulatewc-l),Using$_withthe-pargument

inet_aton,4.27ConvertanIPaddresstoanunsignedinteger

inet_ntoa,4.28ConvertanunsignedintegertoanIPaddress

join,4.4Shuffleallfieldsoneachline,5.1Generateandprintthealphabet

lc,6.6HTML-encodeastring

length,2.6Removeallblanklines,5.10Generateanarraywithoddnumbersfrom1to100

localtime,4.18PrintGreenwichMeanTimeandlocalcomputertime,4.19Printyesterday’sdate

map,4.9Replaceeachfieldwithitsabsolutevalue,Using$_withthe-pargument

mktime,4.19Printyesterday’sdate

pack,4.28ConvertanunsignedintegertoanIPaddress,5.9ConvertalistofnumericASCIIvaluesintoastring

printf,3.8Numberalllinesinafileusingacustomformat

push,4.2Printthesumofallfieldsoneachline

rand,4.23Calculatetheleastcommonmultiple,5.3Createahexlookuptable

sprintf,5.8Findthenumericvaluesforcharactersinastring

system,A.9Variable@ARGV

time,4.15Printthenumberπ

uc,6.6HTML-encodeastring

ucfirst,6.9Convertalltexttolowercase

unpack,4.27ConvertanIPaddresstoanunsignedinteger,5.8Findthenumericvaluesforcharactersinastring

G

gcd(greatestcommondivisor),4.21Calculatethefactorial

generating

alphabet,WorkingWithArraysandStrings

numbers

constants,4.15Printthenumberπ

even,5.10Generateanarraywithoddnumbersfrom1to100

odd,5.9ConvertalistofnumericASCIIvaluesintoastring

random,4.23Calculatetheleastcommonmultiple

passwords,5.3Createahexlookuptable

permutations,4.24Generate10randomnumbersbetween5and15(excluding15)

powerset,4.26Generatethepowerset

strings,from“a”to“zz”,5.1Generateandprintthealphabet

globbing(filenameexpansion),5.13Findthenumberofelementsinanarray

GMT(GreenwichMeanTime),4.18PrintGreenwichMeanTimeandlocalcomputertime

gmtimefunction,4.18PrintGreenwichMeanTimeandlocalcomputertime

golfing(Perl),3.9Printthetotalnumberoflinesinafile(emulatewc-l)

greatestcommondivisor(gcd),4.21Calculatethefactorial

GreenwichMeanTime(GMT),4.18PrintGreenwichMeanTimeandlocalcomputertime

grepfunction,3.9Printthetotalnumberoflinesinafile(emulatewc-l),Using$_withthe-pargument

H

hexlookuptable,5.3Createahexlookuptable

hexoperator,4.27ConvertanIPaddresstoanunsignedinteger,5.3Createahexlookuptable

HTML

encoding,6.6HTML-encodeastring

tags

extractingtextbetween,8.9MatchprintableASCIIcharacters

replacing,8.10ExtracttextbetweentwoHTMLtags

HTML::Entitiesmodule,6.6HTML-encodeastring

HTML::TreeBuildermodule,8.9MatchprintableASCIIcharacters

HTTPheaders,extractingUser-Agentstringfrom,8.7Increaseallintegersinastringbyone

I

I

ifstatement,3.1Numberalllinesinafile,3.4Numberalllinesbutprintlinenumbersonlyfornon-emptylines

inet_atonfunction,4.27ConvertanIPaddresstoanunsignedinteger

inet_ntoafunction,4.28ConvertanunsignedintegertoanIPaddress

initialcapitalization,6.9Convertalltexttolowercase

intoperator,3.10Printthenumberofnon-emptylinesinafile

integers

convertingIPaddresses,toandfrom,4.26Generatethepowerset

increasing,8.7Increaseallintegersinastringbyone

invertingcase,ofletters,6.9Convertalltexttolowercase

IPaddresses

converting

tounsignedintegers,4.26Generatethepowerset

unsignedintegersto,4.28ConvertanunsignedintegertoanIPaddress

matching

lookalikes,UsefulRegularExpressions

valid,8.2Testwhetheranumberisintherange0to255

iptables,IntroductiontoPerlOne-Liners

J

joinfunction,4.4Shuffleallfieldsoneachline,5.1Generateandprintthealphabet

L

lcfunction,6.6HTML-encodeastring

lcm(leastcommonmultiple),4.23Calculatetheleastcommonmultiple

leastcommonmultiple(lcm),4.23Calculatetheleastcommonmultiple

length

creatingstringsby,5.4Generatearandomeight-characterpassword

none,2.6Removeallblanklines

ofstrings,finding,5.10Generateanarraywithoddnumbersfrom1to100

printinglinesby,7.12Printlinesthatareatleast80characterslong,7.17Printalllinesfrom17to30

lengthfunction,2.6Removeallblanklines,5.10Generateanarraywithoddnumbersfrom1to100

letters.(seecharacters)

lines

blank,addingorremoving,2.4N-spaceafile

case,converting,6.6HTML-encodeastring

numberof

emptyornon-empty,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

inafile,total,3.8Numberalllinesinafileusingacustomformat

matchingapattern,4.13Printthetotalnumberoffieldsthatmatchapattern

numbering,Numbering

withtotalnumberofwordson,4.11Printthetotalnumberoffieldsoneachline,followedbytheline

wordson,3.13Numberwordsacrossalllines

order,reversing,6.22Printalllinesinreverseorder

printing

bycontents,7.20Printtheshortestline

bylength,7.12Printlinesthatareatleast80characterslong,7.17Printalllinesfrom17to30

bymatchingregularexpressions,3.10Printthenumberofnon-

emptylinesinafile,7.4Printthelast10linesofafile(emulatetail-10),7.17Printalllinesfrom17to30

bynumber,7.12Printlinesthatareatleast80characterslong

everysecond,7.22Printalllinescontainingonlydigits

firstandlast,7.1Printthefirstlineofafile(emulatehead-1)

repeatedorunique,7.25Printeverysecondline,beginningwiththesecondline

replacingtexton,6.19Substitute(findandreplace)“foo”with“bar”oneachline

spacing,Spacing

List::PowerSetmodule,4.26Generatethepowerset

List::Utilmodule,4.1Checkifanumberisaprime,4.3Printthesumofallfieldsonalllines

lists

convertingintostrings,5.9ConvertalistofnumericASCIIvaluesintoastring

creatingfromstrings,5.4Generatearandomeight-characterpassword

generating

alphabet,WorkingWithArraysandStrings

evennumbers,5.10Generateanarraywithoddnumbersfrom1to100

fromregularexpressionmatches,8.11Replaceall<b>tagswith<strong>

oddnumbers,5.9ConvertalistofnumericASCIIvaluesintoastring

numberofelementsin,5.10Generateanarraywithoddnumbersfrom1to100

permutationsof,4.24Generate10randomnumbersbetween5and15(excluding15)

powerset,4.26Generatethepowerset

slicing,4.18PrintGreenwichMeanTimeandlocalcomputertime

localtime,4.18PrintGreenwichMeanTimeandlocalcomputertime

localtimefunction,4.18PrintGreenwichMeanTimeandlocalcomputertime,4.19Printyesterday’sdate

logicalORoperator,4.6Findthenumericallysmallestelement(minimumelement)overalllines

lookuptable,hex,5.3Createahexlookuptable

lowercase,convertingto,6.6HTML-encodeastring

M

MacOSnewlines,convertingto,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

mapfunction,4.9Replaceeachfieldwithitsabsolutevalue,Using$_withthe-pargument

matching.(seeregularexpressions)

Math::BigIntmodule,4.19Printyesterday’sdate

maximumelement,calculating,4.6Findthenumericallysmallestelement(minimumelement)overalllines

MIME::Base64module,6.1ROT13astring

minimumelement,calculating,4.4Shuffleallfieldsoneachline

mktimefunction,4.19Printyesterday’sdate

modules

Algorithm::Permute,4.24Generate10randomnumbersbetween5and15(excluding15)

bignum,4.15Printthenumberπ

Email::Valid,8.4Checkwhetherastringlookslikeanemailaddress

HTML::Entities,6.6HTML-encodeastring

HTML::TreeBuilder,8.9MatchprintableASCIIcharacters

List::PowerSet,4.26Generatethepowerset

List::Util,4.1Checkifanumberisaprime,4.3Printthesumofallfieldsonalllines

Math::BigInt,4.19Printyesterday’sdate

MIME::Base64,6.1ROT13astring

POSIX,4.19Printyesterday’sdate

Regexp::Common,8.4Checkwhetherastringlookslikeanemailaddress

Socket,4.27ConvertanIPaddresstoanunsignedinteger

URI::Escape,6.2Base64-encodeastring

months,days,andsecondsago,4.19Printyesterday’sdate

N

newlines

adding,Spacing

convertingfordifferentoperatingsystems,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

non-emptylines

numberof,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

numbering,3.1Numberalllinesinafile

number

ofelements,inanarray,5.10Generateanarraywithoddnumbersfrom1to100

offields,4.9Replaceeachfieldwithitsabsolutevalue

oflines,3.8Numberalllinesinafileusingacustomformat,4.13Printthetotalnumberoffieldsthatmatchapattern

numbering

lines,Numbering,4.11Printthetotalnumberoffieldsoneachline,followedbytheline

words,3.13Numberwordsacrossalllines

numbers

constants,4.15Printthenumberπ

even,5.10Generateanarraywithoddnumbersfrom1to100

odd,5.9ConvertalistofnumericASCIIvaluesintoastring

random,4.23Calculatetheleastcommonmultiple

numeric

positions,ofwords,3.15Replaceallwordswiththeirnumericpositions

values,ofcharacters,5.6Createanarrayfromastring

O

one-linersinWindows

inbash,B.3PerlOne-LinersinWindowsBash

inPowerShell,SymbolChallenges

inWindowscommandprompt,B.4PerlOne-LinersintheWindowsCommandPrompt

operatingsystems.,B.4PerlOne-LinersintheWindowsCommandPrompt

(seealsoWindows)

newlines,convertingfor,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

operators.

!operator,7.6Printonlylinesthatdonotmatcharegularexpression

!~operator,4.1Checkifanumberisaprime

.operator,2.1Double-spaceafile

..(range),WorkingWithArraysandStrings,7.2Printthefirst10linesofafile(emulatehead-10),7.17Printalllinesfrom17to30

//operator,4.6Findthenumericallysmallestelement(minimumelement)overalllines

<>(diamond),2.6Removeallblanklines,3.9Printthetotalnumberoflinesinafile(emulatewc-l),5.13Findthenumberofelementsinanarray,6.19Substitute(findandreplace)“foo”with“bar”oneachline

=~operator,4.1Checkifanumberisaprime

chr,5.9ConvertalistofnumericASCIIvaluesintoastring

flip-flop,7.2Printthefirst10linesofafile(emulatehead-10),7.17Printalllinesfrom17to30

hex,4.27ConvertanIPaddresstoanunsignedinteger,5.3Createahexlookuptable

int,3.10Printthenumberofnon-emptylinesinafile

ord,5.6Createanarrayfromastring

q/…/,ConvertingOne-LinersintheWindowsCommandPrompt

qq/…/,B.4PerlOne-LinersintheWindowsCommandPrompt

qr/…/,8.2Testwhetheranumberisintherange0to255

qw/…/,5.6Createanarrayfromastring

reverse,6.22Printalllinesinreverseorder

s/regex/replace/,2.4N-spaceafile

say,2.1Double-spaceafile

scalar,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

shift,7.4Printthelast10linesofafile(emulatetail-10)

split,5.4Generatearandomeight-characterpassword

ternary,8.4Checkwhetherastringlookslikeanemailaddress

tr,TextConversionandSubstitution

x,2.4N-spaceafile,5.4Generatearandomeight-characterpassword

y,TextConversionandSubstitution

||(logicalOR),4.6Findthenumericallysmallestelement(minimumelement)overalllines

}{(eskimo),3.9Printthetotalnumberoflinesinafile(emulatewc-l)

ORoperator,logical,4.6Findthenumericallysmallestelement(minimumelement)overalllines

ordoperator,5.6Createanarrayfromastring

order,reversing,6.19Substitute(findandreplace)“foo”with“bar”oneachline

P

packfunction,4.28ConvertanunsignedintegertoanIPaddress,5.9ConvertalistofnumericASCIIvaluesintoastring

paragraphs

orderof,reversing,6.19Substitute(findandreplace)“foo”with“bar”oneachline

slurping,2.6Removeallblanklines

passwords,generatingrandom,5.3Createahexlookuptable

patterns.(seeregularexpressions)

Perl

golfing,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

onWindows.,UsingPerlOne-LinersOnWindows

(seealsoWindows)

Strawberry,UsingPerlOne-LinersOnWindows,B.4PerlOne-LinersintheWindowsCommandPrompt

versions,IntroductiontoPerlOne-Liners

perldoc,IntroductiontoPerlOne-Liners

permutations,oflists,4.24Generate10randomnumbersbetween5and15(excluding15)

POSIXmodule,4.19Printyesterday’sdate

powerset,generating,4.26Generatethepowerset

PowerShell,SymbolChallenges

prime,checkingifanumberis,Calculations

printffunction,3.8Numberalllinesinafileusingacustomformat

printing

fields,numberof,4.9Replaceeachfieldwithitsabsolutevalue

lines

bycontents,7.20Printtheshortestline

bylength,7.12Printlinesthatareatleast80characterslong,7.17Printalllinesfrom17to30

bymatchingregularexpressions,3.10Printthenumberofnon-emptylinesinafile,7.4Printthelast10linesofafile(emulatetail-10),7.17Printalllinesfrom17to30

bynumber,7.12Printlinesthatareatleast80characterslong

everysecond,7.22Printalllinescontainingonlydigits

firstandlast,7.1Printthefirstlineofafile(emulatehead-1)

non-empty,3.1Numberalllinesinafile

numberof,3.8Numberalllinesinafileusingacustomformat,4.13Printthetotalnumberoffieldsthatmatchapattern

repeatedorunique,7.25Printeverysecondline,beginningwiththesecondline

pushfunction,4.2Printthesumofallfieldsoneachline

Q

q/…/operator,ConvertingOne-LinersintheWindowsCommandPrompt

qq/…/operator,B.4PerlOne-LinersintheWindowsCommandPrompt

qr/…/operator,8.2Testwhetheranumberisintherange0to255

qw/…/operator,5.6Createanarrayfromastring

R

randfunction,4.23Calculatetheleastcommonmultiple,5.3Createahexlookuptable

random

numbers,generating,4.23Calculatetheleastcommonmultiple

passwords,generating,5.3Createahexlookuptable

rangeoperator,WorkingWithArraysandStrings,7.2Printthefirst10linesofafile(emulatehead-10)

range,calculatingwhetheranumberisin,8.1MatchsomethingthatlookslikeanIPaddress

Regexp::Commonmodule,8.4Checkwhetherastringlookslikeanemailaddress

regularexpressions

ASCIIcharacters,tomatchprintable,8.9MatchprintableASCIIcharacters

emailaddresses,tomatchlookalikes,8.3MatchanIPaddress

HTMLtags

toextracttextbetween,8.9MatchprintableASCIIcharacters

toreplace,8.10ExtracttextbetweentwoHTMLtags

IPaddresses,andlookalikes,tomatch,UsefulRegularExpressions

linesmatching

numberof,4.13Printthetotalnumberoffieldsthatmatchapattern

tonumberandprint,3.5Numberonlylinesthatmatchapattern;printothersunmodified

toprint,3.10Printthenumberofnon-emptylinesinafile,7.4Printthelast10linesofafile(emulatetail-10),7.17Printalllinesfrom17to30

listingallmatches,8.11Replaceall<b>tagswith<strong>

range,totestwhetheranumberisin,8.1MatchsomethingthatlookslikeanIPaddress

removing.(seedeleting)

repeated

lines,7.25Printeverysecondline,beginningwiththesecondline

words,8.5Checkwhetherastringisanumber

replacing

HTMLtags,8.10ExtracttextbetweentwoHTMLtags

text,IntroductiontoPerlOne-Liners,6.19Substitute(findandreplace)“foo”with“bar”oneachline

reverseoperator,6.22Printalllinesinreverseorder

ROT13,astring,TextConversionandSubstitution

S

s/regex/replace/operator,2.4N-spaceafile

sayoperator,2.1Double-spaceafile

scalaroperator,3.9Printthetotalnumberoflinesinafile(emulatewc-l)

shiftoperator,7.4Printthelast10linesofafile(emulatetail-10)

shufflingfields(words),4.3Printthesumofallfieldsonalllines

slicing,lists,4.18PrintGreenwichMeanTimeandlocalcomputertime

slurping,2.6Removeallblanklines,A.2Variable$.

Socketmodule,4.27ConvertanIPaddresstoanunsignedinteger

spacing

ofcharacters,2.11Changeallspacingbetweenwordstoonespace

oflines,Spacing

ofwords,2.9Double-spacebetweenallwords

splitoperator,5.4Generatearandomeight-characterpassword

splitting,auto-,4.1Checkifanumberisaprime,A.7Variable$”

sprintffunction,5.8Findthenumericvaluesforcharactersinastring

StrawberryPerl,UsingPerlOne-LinersOnWindows,B.4PerlOne-LinersintheWindowsCommandPrompt

strings

arrays,creatingfrom,5.4Generatearandomeight-characterpassword

base64encoding,6.1ROT13astring

case,converting,6.6HTML-encodeastring

characters,numericvaluesof,5.6Createanarrayfromastring

emailaddresses,andlookalikes,8.3MatchanIPaddress

generating

alphabet,WorkingWithArraysandStrings

bylength,5.4Generatearandomeight-characterpassword

fromASCIIvalues,5.9ConvertalistofnumericASCIIvaluesintoastring

fromcommand-linearguments,5.6Createanarrayfromastring

from“a”to“zz”,5.1Generateandprintthealphabet

passwords,random,5.3Createahexlookuptable

HTMLencoding,6.6HTML-encodeastring

integers,increasing,8.7Increaseallintegersinastringbyone

IPaddresses,andlookalikes,UsefulRegularExpressions

length,finding,5.10Generateanarraywithoddnumbersfrom1to100

numbers,checkingif,8.4Checkwhetherastringlookslikeanemailaddress

numericvalues,ofcharacters,5.6Createanarrayfromastring

order,reversing,6.19Substitute(findandreplace)“foo”with“bar”oneachline

ROT13,TextConversionandSubstitution

shufflingfields(words),4.3Printthesumofallfieldsonalllines

spacing,2.9Double-spacebetweenallwords

URLescaping,6.2Base64-encodeastring

withwordsappearingtwice,8.5Checkwhetherastringisanumber

strippingwhitespace,6.13Stripleadingwhitespace(spaces,tabs)fromthebeginningofeachline

substituting

HTMLtags,8.10ExtracttextbetweentwoHTMLtags

text,IntroductiontoPerlOne-Liners,6.19Substitute(findandreplace)“foo”with“bar”oneachline

sum(offields),calculating,IntroductiontoPerlOne-Liners,4.1Checkifanumberisaprime

systemfunction,A.9Variable@ARGV

T

ternaryoperator,8.4Checkwhetherastringlookslikeanemailaddress

text.(seefiles;lines;words)

There’sMoreThanOneWayToDoIt(TIMTOWTDI),2.6Removeallblanklines

time

date,IntroductiontoPerlOne-Liners,4.18PrintGreenwichMeanTimeandlocalcomputertime,4.19Printyesterday’sdate

GreenwichMeanTime(GMT),4.18PrintGreenwichMeanTimeandlocalcomputertime

local,4.18PrintGreenwichMeanTimeandlocalcomputertime

UNIX,4.15Printthenumberπ

timefunction,4.15Printthenumberπ

TIMTOWTDI(There’sMoreThanOneWayToDoIt),2.6Removeallblanklines

titlecase,convertingto,6.9Convertalltexttolowercase

troperator,TextConversionandSubstitution

U

ucfunction,6.6HTML-encodeastring

ucfirstfunction,6.9Convertalltexttolowercase

unescaping,URLs,6.2Base64-encodeastring

uniquelines,7.25Printeverysecondline,beginningwiththesecondline

UNIX

newlines,converting,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

time,4.15Printthenumberπ

unlessstatement,2.2Double-spaceafile,excludingtheblanklines

unpackfunction,4.27ConvertanIPaddresstoanunsignedinteger,5.8Findthenumericvaluesforcharactersinastring

unsignedintegers,convertingIPaddressestoandfrom,4.26Generatethepowerset

uppercase,convertingto,6.6HTML-encodeastring

URI::Escapemodule,6.2Base64-encodeastring

URL-escaping,6.2Base64-encodeastring

usernames,listingfrom/etc/passwd,IntroductiontoPerlOne-Liners,A.7Variable$”

V

v-string(versionstring)literal,6.13Stripleadingwhitespace(spaces,tabs)fromthebeginningofeachline

validIPaddresses,matching,8.2Testwhetheranumberisintherange0to255

values,ofcharacters,5.6Createanarrayfromastring

variables

$,(fieldseparator),4.3Printthesumofallfieldsonalllines,A.5Variables$1,$2,$3,andsoon

$.variable,3.1Numberalllinesinafile,Using$_explicitly

$/variable,A.2Variable$.

$1,$2,$3,…variables,3.13Numberwordsacrossalllines,A.4Variable$\

$\variable,2.1Double-spaceafile,A.4Variable$\

$_variable,2.1Double-spaceafile,6.22Printalllinesinreverseorder,Perl’sSpecialVariables

$”variable,6.23Printcolumnsinreverseorder,A.5Variables$1,$2,$3,andsoon

%ENV,A.9Variable@ARGV

@ARGV,5.6Createanarrayfromastring,A.7Variable$”

@F,4.1Checkifanumberisaprime,4.9Replaceeachfieldwithitsabsolutevalue,A.7Variable$”

W

Wall,Larry,2.6Removeallblanklines

whitespace,stripping,6.13Stripleadingwhitespace(spaces,tabs)fromthebeginningofeachline

win-bash(bash.exe),UsingPerlOne-LinersOnWindows,B.3PerlOne-LinersinWindowsBash

Windows

bashon,UsingPerlOne-LinersOnWindows

bash.exe(win-bash),UsingPerlOne-LinersOnWindows,B.3PerlOne-LinersinWindowsBash

cmd.exe,B.4PerlOne-LinersintheWindowsCommandPrompt

commandprompt,B.4PerlOne-LinersintheWindowsCommandPrompt

filepaths

bash,B.3PerlOne-LinersinWindowsBash

commandprompt,SymbolChallenges

PowerShell,B.5PerlOne-LinersinPowerShell

newlines,6.15Stripwhitespace(spaces,tabs)fromthebeginningandendofeachline

one-liners

inbash,B.3PerlOne-LinersinWindowsBash

inPowerShell,SymbolChallenges

inWindowscommandprompt,B.4PerlOne-LinersintheWindowsCommandPrompt

Perlon,UsingPerlOne-LinersOnWindows

StrawberryPerl,UsingPerlOne-LinersOnWindows

win-bash,UsingPerlOne-LinersOnWindows,B.3PerlOne-LinersinWindowsBash

words.,4.9Replaceeachfieldwithitsabsolutevalue

(seealsofields)

numberof,onaline,4.9Replaceeachfieldwithitsabsolutevalue

numbering,onaline,3.13Numberwordsacrossalllines

numericpositionsof,3.15Replaceallwordswiththeirnumericpositions

passwords,generatingrandom,5.3Createahexlookuptable

repeated,8.5Checkwhetherastringisanumber

shuffling,4.3Printthesumofallfieldsonalllines

spacingof,2.9Double-spacebetweenallwords

X

xoperator,2.4N-spaceafile,5.4Generatearandomeight-characterpassword

Y

yoperator,TextConversionandSubstitution

yesterday’sdate,4.19Printyesterday’sdate

AbouttheAuthorPeterisKruminsisaprogrammer,systemsadministrator,start-upmanager,andall-aroundhacker.HeisthecofounderofBrowserlingandTestling,andherunsthepopularprogrammingblogcatonmat.net.

PerlOne-Liners130ProgramsThatGetThingsDone

PeterisKruminsCopyright©2013Allrightsreserved.Nopartofthisworkmaybereproducedortransmittedinanyformorbyanymeans,electronicormechanical,includingphotocopying,recording,orbyanyinformationstorageorretrievalsystem,withoutthepriorwrittenpermissionofthecopyrightownerandthepublisher.

Publisher:WilliamPollockProductionEditor:RileyHoffmanCoverIllustration:TinaSalamehInteriorDesign:OctopodStudiosDevelopmentalEditor:WilliamPollockTechnicalReviewer:AlastairMcGowan-DouglasCopyeditor:LeeAnnPickrellCompositor:RileyHoffmanProofreader:ElaineMerrill

Forinformationondistribution,translations,orbulksales,pleasecontactNoStarchPress,Inc.directly:

LibraryofCongressCataloging-in-PublicationData

Krumins, Peteris.

Perl one-liners : 130 programs that get things done / by Peteris Krumins.

pages cm

Summary: "Snappy Perl programs to streamline tasks and sharpen coding skills"--

Provided by publisher.

ISBN 978-1-59327-520-4 (paperback) -- ISBN 1-59327-520-X (paperback)

1. Perl (Computer program language) I. Title.

QA76.73.P22K78 2013

005.13'3--dc23

2013030613

NoStarchPressandtheNoStarchPresslogoareregisteredtrademarksofNoStarchPress,Inc.Otherproductandcompanynamesmentionedhereinmaybethetrademarksoftheirrespectiveowners.Ratherthanuseatrademarksymbolwitheveryoccurrenceofatrademarkedname,weareusingthenamesonlyinaneditorialfashionandtothebenefitofthetrademarkowner,withnointentionofinfringementofthetrademark.

Theinformationinthisbookisdistributedonan“AsIs”basis,withoutwarranty.Whileeveryprecautionhasbeentakeninthepreparationofthiswork,neithertheauthornorNoStarchPress,Inc.shallhaveanyliabilitytoanypersonorentitywithrespecttoanylossordamagecausedorallegedtobecauseddirectlyorindirectlybytheinformationcontainedinit.

NoStarchPress

2013-12-20T11:08:53-08:00