4.1 Introduction An array is a very simple and extremely applications such as stacks and queues arrays. However, useful data structure. A large number ¢ an be easily implemented with the help «: there are many problems associated with this data structure Some of the problems are given below : 1 42 Linked I An array is a static data structure and, therefore, its size should be known in advance before its usage. All the locations assigned to an array are retained by it throughout the execution of a program. % Ifa list of data is to be stored then an array of approximate size is used. This is totally a guess work and can result in overflow or wastage of computer memory. A very important issue-is what if the number of elements in the list is no! known in advance ? Insertions and deletions of elements in an array involve a large amount of shifting of components in one directions or the other. This is a very time consming process. The elements of an array must be homogenous. The above given problems can be avoided by the use of linked structures discussed in the following sections. (One-Way Lists) A linked list is a linear collection line’ is ea ction of data elements, call linear saa ean led nodes, where th! [128] a Ulican Ullictiam, Coen eliment (except Or | List dilpot voolt) hor a Wage preslocess yr |petement OF oe : a linked list contains two fields or parts - 4 ‘An information field (INFO) : + pointer field (LINK) : 4 js shown in figure 4.1 It contains the information of the elements, It contains the address of next node. Node structure NODE INFO LINK lp FIGURE 4.1 Here P denotes the address of the node. Figure 4.2 is a diagram of a linked list that contains four elements. INFOLINK INFO LINK INFOLINK INFO LINK rinst— - | . { = x FIGURE 4.2 _ Pointer FIRST store the address of the first node in the linked list. Each node has a link to the next node of the list. The last node has a NULL pointer, denoted by X in the diagram. It is clear from the diagram that we need only the address of the first node (FIRST) to trace through the list. A list with no nodes is called an empty list. So if FIRST = NULL, then list is empty. The linked list do not have to be physically adjacent as shown in figure 4.3. INFO LINK Finst— -— INFOLINK INFO_LINK x INFO LINK FIGURE 4.3 [129]» Representation of Linked Lists in Memory To maintain the linked list in memory, we need two linear arrays denotey and LINK : Since the subscripts of the array INFO and LINK will be positive, therefore E choose NULL = 0. : Ls Example 4.1 Figure 4.4 shows a linked list in memory where each node of the fist conta, single character. ty INFO LINK rc 1 1 10; (= 2 3 U 7 r4 a FIRST [5}—> 5 s 4 6 ro f- Ss 8 “ai 8 1; > 9 > 10 L 0 FIGURE 4.4 in this figure, FIRST = 5, therefore, address of first node is 5, now INFO [5] = S and LINK [5] = 4 is the first node INFO [4] = U and LINK [4] = 7 is the 2nd node INFO [7] = S and LINK [7] = 8 is the 3rd node INFO [8] = H and LINK [8] = 1 is the 4th node INFO [1] =! and LINK [1] = 10 is the Sth node NFO [10] = L and LINK [10] = 0 is the 6th node eaanlshat, LINK [10] = 0, the NULL value, so the fist has ended than one fist May be Maintained in the same arrays INFO and LINK 42 4.6 shows two lists in memory where the nodes of both lists are stored in the same linear arrays INFO and LINK. INFOLINK 4 u 0 Node 3 of Second (End of list) 2 3 N i a Node 2 of Second FIRST[4/+—> 4 | A 7 | Node 4 of FIRST 5 6 7 | ™M 9 | Node 2 of FIRST SECOND a | A 3 | Node 1 of Second 9 1 11. | Node 3 of FIRST 10 au] oT © _| Node 4 of FIRST (END of list) 12 FIGURE 4.5 The Information pant of a node may be recorded with more than one data item. In ‘such a case, the data should be stored in a collection of paralled arrays. Example 4.3 Assume that the STUDENT record have the following data items or fields Rol Number, NAME and Marks. Figure 4.6 shows how the data may be stored in the Using linked list [131]____—_— ~ 1 ONAME waa ROLL NAME MARKS —» 4 | 1020 Mohan [re] pie 2 woe 3 Pr] 4 FIRST[5}—— 5 | 1001 Aman 83 a 6 1) z > 8 1011 Savita 88 1 9 ‘+ 10 | 1025 Anu 85 0 | FIGURE 4.6 It is also clear from the above example that elements of linked list ma | hetrogeneous. apata Structures Bee aa ts indicate that the nodes of a list need not occupy adjacen han one list may be maintained in the sap ie and that more : . f list must have its own pointer variable Biving the Jes of Hinked Tis ¢ following examP a ee in the arrays INFO and | me ae fine rays INFO. and LINK. However. near arrays ation of its first ode loca Example 5.2 st in memory where each node of the list contains a gure 5.4 pictures a linked li w h irae tain the actual list of characters, or, in other words, the single character. We can ol string, as follows: START = 9, so INFO[9] = N is the first character. LINK[9] = 3, so INFO[3] = 0 is the second character. LINK[3 so INFO[6] = OI (blank) is the third character. LINK[6) 1, so INFO[11] = E is the fourth character. LINK[11] = 7, so INFO[7] = X is the fifth character. LINK[7] = 10, so INFO[10] = I is the sixth character. LINK[10] = 4, so INFO[4] = T is the seventh character. LINK[4] = 0, the NULL value, so the list has ended. In other words, NO EXIT is the character string. INFO LINK START e 3]0 a LE oDLinked Lists es) Example 5.3 Figure 5.5 pictures show tw maintained in memo arrays TEST and LIN pointer variables. H © lists of test scores, here A\ b . LG and GEOM, may be He utere the nodes of both lists are stored in the same linear eee that the names of the lists are also used as the list Fontaine 5, the location ce Gotaits 1, the location of is Fist node, and GEOM cnsists of the teak semen tS Fist node. Following the pointers, we see that ALG and GEOM consists of the test scores?” °% ® 84, 62, 74, 100, 74, 78 TEST LINK 1 [eee 2| 74 | | 14 | Nodezot atc 3 aus [1 | 02 © | Node 4 of ALG s| ea | | 12 | NodetotGeom s| 7 ° 7| 7 8 | Nodes of GEOM GEOM 5 8 | t00 | | 13 8 10 WW 88 2 Node 1 of ALG 12 62 7 Node 2 of GEOM 13 74 6 14 93 4 Node 3 of ALG 15 16 fetieai) Fig. 5.5 (The nodes of ALG and some of the nodes of GEOM are explicitly labeled in the gram.) _—_——————— ] |Example 5.4 ers and each broker has his own list of Suppose a brokerage firm has four brovel’ « Fig. 56. That is, all Four Lists ee customers. Such data may be organize” TPT an anray LINK contains the customers appear in the same array c JstOl ei is also an array BROKER which nextpointer Fields of the nodes of the WSIS. TNT rar such that POINTIK] points to contains the list of brokers, and a pointer array the beg e t fF BROKER[K]. hi inning of the list of customers 0 BROKER[K] CUSTOMER —_LINK BROKER POINT [poe] [2 rit ee 2| Kelly 3 a 3| Hal | 0 GI Hunter 4 Nelson 9 4 | Katz 2 5 6 | Evans 0 7 8 | Rogers 15 9 | Teller 10 10 | Jones 19 ani 12 | Grant 17 13 as 14 | McBride 6 15 | Weston 0 16 17 | Scott 1 18 19 Adams | 20; Fig. 5.6 an Accordingly, Bond's {is 5 list of Customers, as indicated by the arroy WS, Consists of Grant, Scott, Vito, Katzsimilarly. Kelly's list consists of and Nelson's list consists of Linked Lists Munter, McBride, Evans T eller, Jones, Adams, Rogers, Weston xall’s list is the null list. since the null Generally speaking, the information p, an. In such @ case, the d such as that illustrated in the Pointer 0 appears in POINT(3]- alu art of a node may be a record with more thar St be stored in some type of record structure or in @ CO! following example. n one data lection of . h suppose the personnel file of a small company contains the following data on its mine employees: | Name, Social Security Number, Sex, Monthly Salary Normally, four parallel arrays, say NAME, SSN, SEX, SALARY, are required to store the list. Observe that 0 is used as the null pointer. data as discussed in Sec. 4.12. Figure 5.7 shows h d c 4.12. 5 ow the dat | sorted (alphabetically) inked list using only an aditional | Textpointer field of the list and the variable START to poin | NAME SSN SEX SALARY LINK | 1 6} 2 | pavis | |192-38-7282| | Female | | 22800 | | 12 (s Kelly 165-64-3351| | Male 19000 | | 7 4 | Green | [175-56-2251| | Male 27200 | | 14 5 6 | Brown | |178-52-1065| | Female | | 14700 9 7 | tows | [te1-se-0039] | Femate | | 16400) | 10 a | +4 S| Cohen | | 17-44-4557] | Male 19000 2 to | Rubin | |135-46-6262| | Female 15.500 0 11} —_+4 Ge | Evans | | 168-56-6113] | Male 34.200 4 | eas | oe 13 | 14 | Harris | |208-56-1654 Fomale | | 22800 3 ae — —_—— Fig. 5.7 2 may be stored as 2 array LINK for the Yt to the first record in the4.4 The basic operations that are generally performed on linked list are CO aa 1. Traversing 2. Searching 3. Insertion 4. Deletion Traversing a Linked List Traversing means accessing each node exactly once in the list successively sta from the beginning and making the desired changes in one or more data fields of node. The operation of making changes is called CHANGE without actuz specifying what changes are to be made. CHANGE will depend on a spec’ application, for example, printing data. Assume that PTR is the pointer var: which points to the node that is currently being processed. Initially, the PTR m: have the value of FIRST and then successively, it must have the address of next node so as to access the next node in the linked list. The algorithm will termin when the current value of PTR is NULL. The algorithm is : [132]algorithm = TRAVERS ( INFO, PTR, FIRST) ; gtep1 «Set PTR := FIRST | step 2 Repeat steps 3 and 4 while PTR < > NULL Rep? Apply CHANGE to INFO (PTR) step 4 Set PTR := LINK (PTR) [End of Step 2 loop] Lt ~--------- Searching An element pose WE have a list of names stored in a linked list. We want to find out whether the name ‘Shivank’, stored in variable FIND, is in the list or not. The PTR is set to ‘the address of the first node in the linked list and the information field is compared “with the FIND. If INFO (PTR) = FIND, the algorithm terminates and LOC is set to ine address of the node containing the FIND else FIND is searched in the next node. This process continues until 'Shivank' is found or PTR = NULL and then Loc is set to NULL. 7 Algorithm SEARCH (INFO, LINK, FIRST, FIND, LOC) is | | | \ Step 1 Set PTR := FIRST Step2 Repeat steps 3 while PTR < > NULL i Step 3 If INFO (PTR) = FIND then t Set LOC | ! | \ \ | | PTR and exit Else Set PTR := LINK (PTR) [End of If Statement] [End of Step 2 Loop] Set LOC := NULL | | | \ | \ | \ | | | | | Step 5 Exit | |4.5.1 Searching in Sorted Linked List If the linked list is in sorted order then the following algorithm will g, element. In sorted list, We can stop once FIND exceeds INFO Prete ) ey pocccee nn nnn | Algorithm : SSEARCH (INFO, LINK, FIRST, FIND, Loc) =~ Step 1 Set PTR := FIRST Repeat Steps 3 to 5 while PTR < > NULL I Step 2 | Step 3 If INFO (PTR) > FIND then Set LOC := NULL and Exit [End of If Statement] Step 4 If INFO (PTR) = FIND then set LOC := PTR and Exit | | | | | I [End of If Statement] Set PTR := LINK (PTR) | Set LOC := NULL Exit e The complexity of both algorithm is the same as that the linear search algori arrays. A binary search algorithm can not be applied to a sorted linked list, : there is no way of indexing the middle element in the list. Hence searching is list and it is the main drawback 0 inkedtist. consuming in linked4.5.1 Searching in Sorted Linked List I 1the tivked list in in sorted order then the following algarithen wt gay, teh element. In warted list, We oan stop once FIND exceeds INFO (Fry My, ' | Algorithm » SSRARCH (INFO, LINK, FIABT, FIND, Loc) ! Step 1 Set PTR ce FIRST : Repent Steps 3 to 6 while PTR < > NULL. 1 Step 2 | Step 3 It INFO (PTR) > FIND then i Set LOC :@ NULL and Exit [End of If Statement) Step 4 If INFO (PTR) = FIND then set LOC := PTR and Exit i 1 i ! | [End of If Statement] 1 Set PTR := LINK (PTR) Step 5 : Step6 Set LOC := NULL ‘ ; Step7 Exit bm ee EEE as algorit, The complexity of both algorithm is the same as that the linear search ked list». errays. A binary search algorithm can not be applied to a sorted lin iddle element in the list. Hence Searching is. main drawback of linkedtist. there is no way of indexing 1 consuming in linked list and Free-Storage List Together with the linked lists in memory, a special list is maintained which conse: of unused memory locations. This list is called the free-storage list or the lst » available space or the free pool or AVAIL List. This list has its own pointer, denot: by AVAIL as shown in figure 4.7 FIGURE 4.7 I we want to insert 4 element or node into the linked list then we have to check ™ free-storage list. If free-storage list is empty, then we can not insert the oloment a this situation is known as overflow. ree-storage list is not empty, then the nf node of thi be used tor the new node. The fields of this node can be filed” with the appropriate Values aid can be inserted into the linked list. In program™ i ; 7134.5, this process is known as creation of a new node. The f 5 ». The free stor ages anon afd after obtaining a node from it is pe’ eee eT Toe ‘age list hown in figure 4.8 ot... - : After | | FIGURE 4.8 | qe Algorithm of creation a NEW node is : {tpt I AVAIL = NULL then 1 \ Write "OVERFLOW" and Exit \ | ep 2 [obtain address of NEW (first) node] | | NEW := AVAIL | | Stap 3 [Remove first node from available list] 1 | AVAIL := LINK [AVAIL] | L-—— We can also use the special assignment statement NEW = NODE in algorithm to create the new node and to ignore the details of the free storage list. When anode N is deleted from linked list, we have to return its memory space to the AVAIL list. Generally, it will be returned to the beginning of the AVAIL list. Free- storage list before and after a free node has been returned, is shown in figure 49 Pointer to list to which it i once belonged (Before) [135]4.8 ae oH avait [o} La _ T+: Fe (after) FIGURE 4.9 : The Algorithm for Restore the node pointed to by N to the free-storage jig is he Algor S 4. LINK (N) := AVAIL 2 AVAIL:=N Garbage Collection i | Suppose some memory space becomes available due 0 deletion of some Node, from the linked list. One method is to insert this Space immediately into the fes, storage list. But this method is very time consuming. So we may choose ¢ alternative method as follows. The operating system may periodically collect all the deleted space onto the free. storage list. This method is known as garbage collection. Garbage collection Usually takes places in two steps. a: First the computer scan all linked lists and tag all the memory locations which are currently in use. 2. Secondly the computer runs through the memory and collects all untagged memory locations onto the free-storage list. Insertion into a linked list Linked lists are most suitable and efficient for insert and delete operations. The insert operation involves the following two steps. : a a > ‘he location where the node is to be inserted. beginning or into an arbitary position in the list.insertion at the beginning : So address of the node to be ines - IEW is the pointer variable and has the acne now node is NX. denelurs at the beginning. Let information contents 1. Obtain a new node. 2. Initialize the information field as NX 3. Change FIRST poi it poi : a pointer so it points to the new node. The algorithm to insert a element at the beginning of a list is aigorithm : INSERT_FIRST ( INFO, LINK, FIRST, AVAIL, NX) 1 | step 1 IF AVAIL = NULL then | write ‘OVERFLOW’ and Exit siep2 [Obtain address of the new node] \ ; NEW := AVAIL | step 3 [Remove the node from available list] ; AVAIL := LINK [AVAIL] \ | step 4 __ [Initialize information and link field] \ | INFO (NEW) := NX I LINK (NEW) = FIRST - \ | Step 5 [Update FIRST] \ | Set FIRST := NEW i Step 6 Exit | | $e Er eee ee eae zs We can write this algorithm equivalently as follows Eee eee as -—- a | Step 1 [create a new node] \ NEW <= NODE | | Step 2 [Initialize information and link fields] | \ INFO (NEW) := NX ! \ LINK (NEW) := FIRST Step3 [Update FIRST] \ Set FIRST := NEW | | Step 4 Exit | et Bet 4 a [137]NEW [+ FIGURE 4.10 (b) __ Insertion at the end of a linked list : Step are 1. 2 Obtain a new node. Initialize the information part of the new node as NX ang Make LINK pointer equal to NULL. 7 'f list is empty then insert the new node as first node. If list is not empty then first locate the la: st node by traversing throug the original list until a node with a LINI : IK field of NULL is found. the non emply list by setting the ks le New node. Insert new node to the end of node LINK pointer to point to th END Figure 4.11 shows the insert Operation at the end of the list. "HOE. ayaigorithm : INSERT LAST (INFO, LINK, FIRST, ‘AVAIL, NX) fend 1 in this algorithm, SAVE is the temporary pointer variables. ‘ [create a new node] NEW = NODE 2 [Initialize new node] INFO (NEW) := NX LINK (NEW) := NULL IF FIRST = NULL then FIRST := NEW and Exit 7 | | ! | | | | | | i | i | \ | | I | | | 3. [Is the original list empty ?] i | | | i [End of If Statement] | \ | \ | \ | \ | | | ! ! | | | ! | | 4 4. [Initalize search for last node] SAVE := FIRST 5. [Search for last node] REPEAT WHILE LINK (SAVE) < > NULL SAVE := LINK (SAVE) 6. [Set Link field of last node] LINK (SAVE) := NEW 7. Exit L = Given Node : Suppose we want to insert a element into (c) Inserting after a i the list after the node pointed to by the pointer P as shown in figure 4.12. Steps are: 1. Obtain the new node. 2. Put INFO into the new node. 3. If list is empty (i.e. P = null or First = null) then insert the new node as first node. 4. If list is not empty, insert the new node after node with location P. End [139]— 14 jt (a) Before insertion | NEW (b) After insertion [Create a new node] NEW = NODE IInitalize new node] INFO (NEW) := NX [Is the original list IF FIRST = NULL [insert after node empty ?] then with Location P] LINK [NEW] LINK [P} ang LINK [P] := NEW Exit | Algorithm : INSERT_LOC (INFO, LINK, FIRST, AVAIL, NX; P)