gu *04-A022, Subject ihe Reogramrsing Languages and Compilers Subject Code- CSE 4Oat Textbook- . 1: Pro meng Lan agp | wit. j 1 By chad 8 epee L. Scott eer Publecation 2 ieiads.2 ‘Srd Edition ' Course. Outttine- -O-Intzoduction +0 Compilation 1. Finite State Machéne (FSM) tant) wastqad aye 2: Lexical, Analy nin (Scanning) Cropore in 3. Syntax fas (Parent on Important) e 4 wy . Syrobol Table Ky 5: Ererzorz Recoverey ; oa 6. Semantic Analysis pe F. Intermediate Code Generation , sid oss -% Code Generation and es Ga ane = = 7 =44. The Arct of Language Deaign: Need of Continuous Language ee © Evolution £2) Special, Purpose % C:Low Level Codzn > Prol Logical Relahonashep > rose bed Manipulation > Per or Ryton Scripting tet Personal. Preference ty) Expreessive Power. Vy Ease es use kore Novice VG Ease of emplernentation NC Standandésation Ney Open Aoutzce Tx) Excellent compilerz %, Economics, Paretionage and Inertia Language Design Criteria- > Reqularzit + Generale > Ortho > Uniforemity > Safe P Extensi bility ee 22-04-2029 (Friday)Y classification of Progreamméng Languages (Ads SU aan 4. Declarative - What computer. Ahould do Example- Functional. Languages: Lips Haskell Scheme Logic - based Languages: Proloqy Data. Flow Languages: td, val a+ Irpercative - How corpurer2 should do sForetens € Example- Van- Neumann Languages Serdiptin Languages: baths OAK. pal Object - 0 fented Languages: Amula 6F 23-04-3022 (Gaturedayy) Language Procossetcs/ Translators: Spurreco larginge af Translators Toxeget Language “sy The translator, translates the High Level Language preogram tnto toreget language quisalent Machtne Learning i ainguage) > Tk provides veniie nyeshagess whenever. the preogreannmer violates the Apeciheation ob High Level Language: Types Ok Tearslarors: sTpore ate three types ok Translators + 4. Complete . Tt translates a. ourece code pro written fo Hegp level Language to equivalent forget language: = iil ag Corsprlarion coral ler P“2 Trfonpretere- + Tt translates a. fOUtee code prLo7 at +p Machéne Leael Languog’s fist} oe ‘Trtereprretotion J ean 3. Asser) saorn ler - Th i 0. special, type a eariat ibe wil +ake only oa erence i aa rare, wariten @n Hégh Le sep by ‘Kep- {, * Trt | eqyereence behean Compiler and Interpreter: pn Fe | we brat Cornpilerz Tntercpreetere : " es | H taxes entire {It takes 0 ae Be J Tnpub progam oF atime. |eode ot aint: Tt produces sae Tt doesn’t preoduce | Oukput -dzate object code- intermediate nga i [object code. aa _ ies | E Compzlation is dene | Compilation and —_| oH before execution. 1 eon ane done Cae a hirmuttaneouny- ; kt is cormparzaltvely | tt ts Ts cormpariatvely | Speed | — kadter- inlower- | Mennorz Frement| Memora ‘owent Mernory-. its aoa ea +o at lens ats a Requircersert, creation of object | wedéate pt t oe laree produces: Display erercowes Dinplouy ercrotzs Ok Errore mmpflation acl one by” all at a He. lone. Bae. (eae Detection fs | Errore Detection is Det i cormparttvely corey anitvely bachest saan deficit. eater. Pertaining | Certs CH, = Python, Perl, PHP, | Rrogrearnnng typescript, ee: Rubyyetes 9 Languages ot Aguasrn nee eee sf Nftxing Comeflason and Interpretation : AoUTRCe prwogrzarn CIAVA Frograrr) CGormpiler /Treanslatere) Gye code) Fienonaéat Freograrn Input Troplemerting Compilation: Pro- Processing Gcannéng) - -yTt rernoyes white Apacer and comments- “Tt groups the characters into tokens: > Tt expends abbreviations: + Ht Identifies high-level ayntactic strcichumes (loops a Addihonal Features, =Macro preocessf > File Incluafory’ : > Rational Fre-Processor + Language Extension > Routines and Linkin P : >The cornpilerz uses a er poyan and a snebithe from kebrarey.forcten Program Compiler nits Mochzne Language Lenker) (Library Routines) Machene language Freogearn Fort Compilation Apsernbly Output: source Language mor rpg king Machene Longuag’ ~ Ir helps aa a and ophimizing: ‘ve C Processor: he Kounce Lace C Pre Processor, Modiffed source proogmarn +g Tooke ae Boom Source we Source Translahon :Pe “Boot AreaPl eng: ary 4 Selk- Homtt "4 Cornptlerts. ieee Salk; Hoating Compileres ame Written fin the AaMNe language they éntend to > The cormpile- |» Exarnple- © compilerz Boo’ Extreapping = yuideculous the ntentionalt STR a tern derived from € ground. by pulléng on noton log letting oneselk Kf +he ones bootstraps - Comptlercs: Jose - In Terme, (IIT) 2 Turk tr-tine (AID) compilation | Lton or Tuuan-tirne corepilattons) % a Way of executive | computerz code that Ton execution | OKO program _ot tun tHime- (Q\so aynamte drzanslat- cles compilation durzin: yather than bekore execution: / 26-0L}-2022 vec (Tuesday) DEcrocode (Firm are) + YThe interpreter te artitien tr low-level enaizuctons Called miterrocodie. Giternriore) sntch i Komied tp Tread Only memory and executed by the handwiane-diate ‘orem, engin. troche oH Compelere > The structure of cornpelere design tp denided ato a bray Categoreles:- Dexign - f | iy Analisis (Front enc) Machine Independent ; Language Mt, fi Synthesis (Back. encl)+Machine Dependent» Language Tring Front end at Lisi Analysis | Ne 2 e Agurece code Phakee Ob Compélatten- Token Cae Atreearm Br ea Farne Rens fav tree - 1 ae hag Abstract Syntax - Tiree or oth ‘ Enteramediate forum —_ Modified interes — Target ——— Modified Target Code 1 1 igh tnizpamedéate Code, ' Bock end Scanner Cexical Analysts) Smee hays) IMachéne Inde lent Target Code Gene reason Coci Biscee 5 Wo cee re lee onl, Semantic Aral ath Oe lintermtedeate Bie Gener r Symbol TableExample- position = tnétfal trate ¥ 60 (Al arze footing vantobest S259 Sten rag] acta | c «tea AswrbitA eae : oe Bae ‘ Xda) K+) oe ik cay reo ton oBribite donutanftitenal < ee Paras ys Aryl)\Intenwedtate code Generation E ; In Inferavediate form, ao try S-Aderess ti tnt tO pleat (60) al +3 Adcheons Cocle, left kozfdgXti | : re: ate Aide muat contain 1 .: I A Operzand) and right side t3= Ido + ta Unt contain 2 fdp=.t3 . Code Optimizaton _. b= Ids ¥60-0 td = ida tO ce Geomatin’ a eens (MULF-Hultiplication| LDF Ra.» tds LDF- Load. ‘Ra, Ras He MULF Ra, R20 #60 / Target Lan @ ADDF-Ada CDF Nata alae ll & x Qe STF-Stone ADDF RisRiy Rae. a Bd Si cipky D3 2F-O4. » (Fata a) m Deerence betreery Phase and Pass: Phase Puss > Compilatvon fsdone — |-»Varctous phases cee | En Varztous No. of Ateps- logicatiy grouped rec ano 3 Oe te hanes of compilation > The process Of cormpila- lenient Analysts, Syntax | —Hon can be carried Aalyatss Semantic 4) out th a angle pass | Ophmizatien , Target code Generation, Machcne ere - =cattan code optimiza ¥ Accorzdéng to the nurnberz passes s.dormpilercs ame of 4Wo types 1 One- Pass Cornpelerz {i Mulis- Pass Cormpilerz one Fos Compilerz - +A one-pass cormpilere isa compilere that passes through the pounce code of each compilation unit only once: Tk ts albo Known ap Naterzon Cornpilercs- Target Code Generacer bi rpecific code 4 Ol w mizahionSia Mult-Pass, Compelere- . + my >A roultpars Comnpeletz 15 O ‘ype the poureee code or2 abstract Y peverzal twes- : ok corapzlete +hat preoceny tax tee Of prrograny Sai Pans "4 Semasnbie ean Seo ode Generator veal Apecific Code a ee [ Machine Code 2 Ophnktin ¢ woe rts — VvChatter 2: Lexical Anabysi gr WAN DAD RAAADRARAMA RARAAAATAAAA Tt th the firent phase of cormpelahon: Lexical Toke 24 sourte J) Anabuizere a Syntare | as program ee | Anodyzere > fa | | getNext Tokenc 2 Pe ee . | Syrrbat Table > Scanning is a Afmnple process that dont require toKenization of the input: S Proconser leKe Aeleton, o cormments, and compaction ok conpeciitive! waite apace, choractere into one are done” t' ™ Within this scanning. > Th Lexical Apallyris fokens are produced from the ourpur oly canner: : ; 50-04-2022 Sosureday Loxeme, Potterun, and Token: _Lexcime- > Lexornes are O Mreany ok characters Which can be Grouped ce based on a specific patteren- Ration: 2 ; ; Patterns are the descriptions that Lexeme can mepresent om can take.t (vor tmp ¥6) . VVUVVY Ve | Reece Ae ; Rawdo d wn | ne e 7tn Barn | avec ; or tc t+ nqt | NY caer ae et | dakt% exo | Aca) er eee. | onze ea | ty on a ae e — ye 0 TW Td + From Lexweme, we construct tokens — Token ih a tuple of 2 elements, Where the and element is optional Ctoken_name’, atietbute-value> -> Optional = = noctune Aymibolic. rep Tm epreacirtaayon Thee ee Ka xpecific lexeme addres of the tokens. Lexcewme token poster, Ka Keyword a $- : > When the Lexical Analyzer encounters a lexennnne sft _— generates a token name, and bls UP Of the mbot - fable with attribute name, type, ete. ——| The aitreibute value of a token fry the xymbot dolble: Fino the Lexemes and Tokens ode. r int geal into, int BS) : Uantie (a@=b) - a te (Arb) 040-5 ; else b=b-; z ‘ return OL; t Hns- Lenemer = 32 and Tokens= 3. WMI pot’ 40 the entra " | Lexemeh Token Pattern for, I while Keyword Roewsitt-; Wen LAYS NUM MO] Idenrfrers —lethert Ort — Clloniedd b oe me em Ok le onyaiqit (ie Sule \a* | 254 B16, OUT nun bere A numenic- conmant — kault” Wrerra Anything bernieen “” cue Ey pun tae Specfal Chonactetz | Example- POLOT = 20-68 > Identifrerz po 101 = 20268 “Tp keynored - Example- korn the pollonting a‘Th y ae 05-05-2099 dnput Butering : Thurs Th input butjertings to avofa the Mon input from the isk the bukkers ane used: The lexical analypere uses the look: ahead fron those Input buygens to determine the tokens based Upon the Patterens watch: ~Bawicatty, a double bugyring ANOId the ‘over: heads axsoctat access. [] fe |= [rahe clx |x [a]é | ; Exponential 4 Ba ) Fem= mxcxec® Begin Lexeme The input buffering Acheme uses & pointers - O) Begin fr Begin Lexeme) by Forward technique fs Used to ed with the Memon Bukbert ends z W a Ino m, >The siriing of characters between the. two patintercs ib nrecognizedt on current lexeme- ahead und! a pattern match lexeme 15 identify, the frward | characterz on tts ruight: of the current > Porinand pointer scans found and whenevere a __ poimterz is set to the next ~ Begin pointerz wiorekhs the begining lexeme-Sorrel The Sentinels are used to ophimize de Mtnten of { oe ten kor the End ok the bile: >This a xpeciat chanactere that cannot occur tha pro (basically EOF) ured O4 Sentinel: “ 06:05: 2022 aw . t Frias Speciffeation af ToKens + md +Reguian Dekinitions ame she nequian expression W specification ok tokens Sireing and banguagen: Syenbot ST ts an alontreact entity which ts nat dekined- tho mechanisms based on hich ame very popular kore Fxample- \etteres 5 digits etc. Alphabet - >Fintte set of mgrobols T= [Olt Strzingp- > Finite Aequence ok mgroibols drann from an alphabet: Example “Hé”» “oloys : > Denoted bye: Length ok the Airting \w\- Po i +h ning o length 0= Null Atraing = ae ‘ Sn het Ok Airings igrorn forre. alphabets:B : EEE ——————— Er > tring te called Emp, language} eieaelel ie with no S" ( (@). } Exar le- Ly = $00, 0151051) ' ? [Fen R= $0, 10MOF L3= $051,00511,010; 101 ++ U } | Tnfinite = $.0300510 1000; 0105100510) «+f ) Nartioun Operrations 09 Steing: : Requlare a Context-freee context- Recutatee : Lanquage ea Sennsdve - Enuumerable Language Language “Type 3 Type S Types Typeo nop Us 2 fabio" n20F i 5 Linear Bounded cae futonnacta- Macite : La= Sab" | m9 OF L2=farb Automata Rurh- Down - Regulan Expressions: Regular Eepriessions ane Tt ih uned {eT descnibé used ‘p.apecity the tokens: all the ait that can be e opercatyorhs fuunfon , concate- built from the pation MO) appiced spthe Ngenbols of the sore alphabets: \ . - > yanious Tules to describe , the > Trere are baxically nequian exepreaion® oyere alphobet Aets overe niga:Rule?! oe a it eXPreANTON- % {a L@)=4 Rule-a? 6% O ‘ ] Rule-a* € th a TE errs expnesnion - 4 whenesL(6)= 16%: whene ‘ath a regroboal of | Rule-3: Fort each ‘EZ, ‘a Requlay. Expremon, Alphabet et 2s ‘a Lo) = §at- mequior expressions denoting Rule-4: If r and 4 are S mreppactnely the languages R a Concatenadoo > ma oft wh Ts a TLEQU operration 1 Known OF d tis also O. THe lon expreesNion L(ms)= R:S> The concatenation Of TWO Tee alarz exprzenicson an quia express ‘ cao] Union :- srs ts an ee expression» L (+8) = RUS: The | aon th known ah Uninn: Of TWO reequiont exxprtenoon and tt th also Regular encpreenion tity (een Closure: - n% th a Regquian expreNson » lanquage ts cloned under Kleen biel aoe L(n®) =R PA The meg lor Stan operation): : -3) = OF ty Powitive Closune= Lcdstaks apiisd (nt)=R'= BRE sit tha mequiare Expression yt | ee pals ad : | gat bab gape (30101 | bs eek Mat oo hccacmnl adk- OF-05-202a3 ee Precedencos Ok Operzatores {fn Requiare Evpresmon « Kleen Closune Concatenatior) » Union x 5, ; to1yU Examples- Ly= {Set ok all Mrciing & Os and ts ith atleark 2 conrery. -Ave Zemoes) = (041) co(oHt), | = {00; 000; 001, 100)='- =? 7 La= §we $017" Kuch that wi has 2 07 3 OccuTeNncerychad And the firext and second. Of which are not connecutive + = O*10%o10* (10% +€). L3- ={Set of all ning Of Os and 1s ening + With out - = (o+1)*on ; Ly = Sifelse; then, dor hele? — : 2 iftelke+ then # dot whele. : ie {atrcings Maxcting with @ Of Zero on more 2 b's followed a‘c’ io aye : \ wilted Erne i > Any lan aq e that can be defrred ie hurt th called ae. Lan “4s re. pat set of oe present in language Hicaets a — teegulare fet Z# Res Gor. Dekinction : See > Regular Definitions Exprrenion and ho! ane ured to give nares to the Regular Ke Names ane Used as thefrz Ayrnbols >A neguiatr defini form di) dana fory ‘TA A AEQMENCE ok definition ok the On > Tn' ch dé th O distinct name. for) ovetZ TU fdvd2 Where, 20 each rz ha requir expres Ato dei? Queahon- : Write a nequian definetvon kort the — identifiers Whech ane Mrdingh oK letters » aegis and —: Ba \ettercs 7 O|\Cl- ar - hes digit > Ol! jajalals sfelt|el9 zd ete (eters + 40g) 5 yab- pan AyBo* ot Oe Quemion: note a unsigned . lar Exprediion toating point) winich ane the _btreings eee \89E- fie : Write dewn the Re nurnbere Gnteget ort Kuch Ob- (5ag0 ,0-01284 5 Aras a eget ay eprint EC anes number =cigits- optyonal. - al Hr gnaRO Extenions ok Requian Fx prresnion: oy : L.One ore Mone Tnxtance - | >The unan pomt fx operator, t+) mepresents the por, Clomune ok ‘Oo. neguian expreension and cts sa uage. - hetfknhwsa mequian ex preesinon then 1’ denote, (L(n)). Srtennter't. 2% Zeno or One Inatance- | | >The unary porthx operator, “2” means 0 or 1 Occuns de, 2? % equivalent. to (Elr2). ney L(n2) = LIDIé 3. Chanacter. Classes - >Arequian expromion ailaz|-+-|an vihene Ae ane the Ambo ok the alphabet can be replaced by @ Khon © hand Qz [0102 +++ An}. “3 >More importantly When 4 42 «-.- An forrrm a logical Kequence (fort exarople>, consecutive uppercare | lawerccay digits) 1 we can Teplace them by [a:— Qn] - | for, efamnple- [a bc] Atonds for. abl and (a-7] fs the Ahorethand kor alblc]---|z- New, lettere —> [A-Za-z-] , digit —> [0-9] id= letere — (letters - [digit . ; Tom aigit 10-4] ‘ digits > digit” a tacsaa ‘ number —> digits ( digits)? (EG-<12dmis)?Recogntsatton oh TeKenh Bernt > ff exprt then Att al lif expre then nem elne Atent ' d <: sia tee ee “9 expr > term nelop Torr od \tern ee a term > Td op |numbere “ a a dégit > {o-9] ie aigrts > digit". wad “numberz > dtgitscaligits) 2(€H-1? digits) Ege letter; > [A-Za-z-] ; ‘s td = leter- (leHere- [aligit)” ban aero~ 10°05-2029 “4 Tuesda Recognixat Yon of ToKens = 4 Tranistion Diagrarn fore Relational, Operator: _Oreore Not equol > ¢ netuny) S Te BCLE) Lexb than eqpiod to {U— otherz es reesturin RELOP (LT) Leas than Sy ecinnt 3S RELOP (ER) equal to wa) @) FELSPLEE) qreeaten than equal ty E | return RELOP(GT) ee ore Traixttion Déagram kor deniers: Zq = letter - (leter- |aigit) ae Aigtt; - soit omen sey! retunn (qetToKenO s wtatdG) Trainston Diagnarn for Nurpbera ii) 5” dagits > ali ate )2 EL] 2 Aeqits) 2 number = aigits Caegitssotens kore a calculator Language : -: 12-05-2029 Th was (pace + tab + newléne) Bees cornment > Lparen > ce rpanen >) plus >+ minus > ~ Her > *& ania Sis , % eq > (letter ae a num > digit-dégit™ stab, |aitgi Gaigit|aigit) dQ”a. ei 13-05-2022. Finite Automata: fine > Finite Automata ane mainig of tno types ~Determintatte Finite Automata (DFA) * Non- Deleremintntic. Finite Automata. (NFA/NDFA) Detétumininbye Finite Automata : A DFA ts a. S-tuple (8,26,405F) Whene, Qk a finite Mote called the Mates. Tir afinite Atate called the alphabet: SAX ZA ththe transition purcten- A & & is the Kromet ware - F CA th the Act ok accept Mater?” | % contaéns no € transition: a) exactly one enitiol Mate. Zéz) contains multiple number Ok Binal Mates . aya Micing ‘w" gets accepted by a DFA by ae mingle 6 pe path tines fatbal Atate 2 the final Nos- Deteremintatic Finite Automata : SANFA th a S-tuple (As) 6) 4oF) where, @ is. finite stote called the Mater, = ts a finite state called the alphabet 6 O% Te —> PUB) fh the trannition funct¥on is go ts the Mart dote ECA th the notch accept mateseo a S contains € tranmbon. ‘nt. to dzfferent mates fpllovitng the the zy A ntate can Hanns, ame input AYmMbdL- 2B a Nin ‘wi gots accepted b ynulple poctins bron initial Mate a. NFA by folloning 4p final prate- ’ Zyp contains 1 tnival state: 4, wulhple Binal Mattes. Ny contain’ _ ar Expresnion and NFA: Equrétence of Regu! 4. RE.= 0 wh 2 =3@ g. RE-= & ©, * ) 3. RE=$ 6 O- Letty ard Te ome. dno trequlare expresxions and 1% 15 represented by NFA N, and 172 ib nepnesented by a NFA Na: 4. RE: =S.RE.= Ttta or Tra fo ae OR : ola 2) 16: RE-=TF em = Oredlere Of a E bracket a-% 2-Concatenahion —— Geuunien nen @)ccna a > atund Ol ind the FA lpr the follonsfng megulan expremion A wx@b)” (olble). a: > Cry b: = -a)~ 4 ANY f Byte rata, NPA Yor es a a ap the expriemvion d* (dla) a die «p:aO—-0 a é NE <€ to DFA: €-closure- |e The tare: Mate of the i | hee oa DFA th the €-clomune ob depea: Find the Modes thot can be freow proent Atose fh each enput OG. ae6 (Ra) = €-clowune (U6 (%2)) d Brample a | er i { 7 gst.) (G6,10))1)33 E 14-05-2022 Aimnizoton ok DFA: Tuesday sep |: 4 Divide the notes o the DFA into two equivalence clawed f-2; nal Atodcs and nen: final ntateh. | : ee Chee OA ale Repeatedly earch kor an equivalence class {x2 and a a mibol. fa’ Nuch that when qinen “Q? ah inputs the woke $x? wake Aranistion to ptates K>1 equivalence dows.m | Ee A LL Step-3 : a nuch a Way that all Then partition ‘y? thto K classes in to a member 4 tates in a ee nen class Would move the same dd clan on tnput ‘a’. eae When We ane unable to find done. then the DFA i4 minimized: 78 (rion- Firat) | a clams paretynon: We ane. A <6 (Non- Final) da B (on Frat). > D (Final) Ay E (Final) pera —t> (Frat).Exarnple- iis. Minimize the follonsing DFA: nner 2 (NF) \ << a) 5 thr) = 2 Sa 2 (NF) > 6 (NF) cde 5 (NF) 5 (NF) set 3(F) ° se (NF).Some ennonr ane detected through Lexical Analysts ANA toy, cannot be detected: : J STF the xicing ‘fis encounterrad tn a ¢ prograrn lereay Onolyser cannot tell wna “A atti ok Keyword “2h” or an undéclaned kunction ident fer. yen Gd and some SThe lexical analyser return the ® the exrcore- Other phase of cormpilatyorn to handle SWhen a lexical anabyserr fails ty detect an @ICIZOTes then thot care home Tecovery nireaegy ta ubed: Papié Mode ok recovery : ; +The ximpleat Tecovery rircateqy % the panic mode ck recovery: > Tr 4his Mizategu, We delete MALCCEMANE characters remaining enput Until the lexical analyser binds a well forawed token. (rom the 19-05-2022 Thursday (Gubmihution Class) * 20-05-2022 gl Errore: : ae > | | j Friday . (Gabmitutton Claas)Irn the Grarnenare grrecntely connplicates the panaere: >SHence Ih some Progranninin languages pragras ane ao | at centatn well Gonbed places. > Frageas Wat aenve ag aAfrrechves turn varelous Kiel ok reunhine checks on ott off (Crarnple- pointers ore nubaereipt checking) Sh dura cerctain ophnizaton on| bb: Ted enable ofr Afaable the perroremance progileng- SPragman that rerve an hints provide 4he cornpilere with Enformation about the source Program that may allow tt todo a bettere job- 2) Voniable ty? te very heavily used}: (x may be Kept fh a tater) . fi) Gubnoute fh “f? Th a pure function: Tt only abkects Hhe reat of the programm by the value tt returns: iit) Gubroute TH &? t not RecuNnANe- fy) 3a bits Auphie fore Kloating point varzionle 2