File Handling

BENEFITS  OF  FILE  HANDLING
  • give permanent storage facility
  • can access huge chunk of data in instant (bcz RAM is of small size -> can't store huge data)
  • easy to transfer data as files
Text files - easy to read file (like simple text editor)

Binary files - hard to read file => sth like compiled form of text files - > (encrypted in binary form)
Can have image, audio, video,etc. {TO check EOF here => FILE size value needs to be checked}


Data - bits of storage
Information - when we make that data meaningful after processing


 We use FILE as data type (eg.- FILE* ptr =NULL)

MODE - 
  • read (r) , write(w), read/write in binary (rb/wb), append(a)
  • r+ => open for both reading and writing
  • w+ => 
  • w , a ,w+, a+ =>If file don't exist -> create new file then do operation on it

Opening of file => ptr = fopen("test.txt", "r") 
        ->     if file don't exist -> ptr = = NULL
file first get copied in RAM then only we can do further operations => ptr store address of file so that it can be accessed easily in RAM

Closing a file => fclose(ptr)

Reading a file => ptr = fscanf(ptr, "%d", &a)  => FILE should be in READ mode   
    {  check EOF to know the end }   

Writing to a file => ptr = fprintf(ptr, "%d", a)   => FILE should be in WRITE mode
        -> always written from scratch (remove earlier written thing and overwrite new stuff)

Appending a file => ptr = fprintf(ptr, "%d", a)   => FILE should be in Append mode
        -> add new thing in earlier file (DON'T overwrite new stuff)

Last byte of file contain EOF (end of file character)
  • fgets(name, 5, ptr) => 5 denotes maximum number of characters to be read
    • It returns NULL if there is an error (such as EOF).

Three special file streams are defined in the <stdio.h> :
        a) stdin => reads input from the keyboard
        b) stdout => send output to the screen
        c) stderr => prints errors to the screen

Comments

Post a Comment

Popular posts from this blog

Graphs

Image
GRAPH  Modelling - Multi-Source BFS --> Use normal BFS but insert all the sources in queue and make their distance = 0 --> Update the distance for others as usual for normal BFS ( Link ) Transform existing Graph and use BFS (non-weighted) & Dijkstra (weighted) Hot (red)  -  Cold (blue)  problem => Update graph s.t. red & blue are alternate  make 2*n nodes as A => A' and A'' connect A' -> B'' (for hot) and A'' -> B' (for cold) Directed - weighted G + can remove k edges make G => k+1 times as A', A'', A''', ..... Add edge of weight 0 in between the k+1 graphs n city + m roads + max storage (c litre) => weighted road (consume w litre) + cost[i] (cost per litre to refuel at i-th point) => min. cost from s to t We have to move from A to B  capacity = 3. From A -> A' -- 1 litre needed at value in cost array. From A''' -> B' because loss of 2 litres (as shown in the below d...
Read more

Basics

Comment -  for In-line comment (means comment in same line) --> Use //                          for multi-line comment --> Use /* ........ */ Array -  int arr[10] = {4, 5, 8} --> creates array of size 10 with first 3 elements known  Access - 10[arr] or arr[10]  --> both correct bcz *(10+arr) and *(arr+10) are same Pair -  pair<int,string> p  = {2 , "fsf"} How to access ? p.first and p.second Vector -  array with dynamic memory size =>  continuous memory allocation How to  initialize vector ? vector <double> v(10) => declares vector of size 10 with each value as 0 vector <double> v(7 , 3) => declares vector of size 7 with each value as 3 vector <double> v = {1, 3, 5, 10} => declares vector of with given values vector <double> v(st.begin(), st.end()) => declares vector from given set vector <int> v1 =v  ...
Read more

Hashing

When we cannot use hashing -  when  order is important =>  Find key which is just smaller than the given key. COMPRESSION  MAP  => H(key) = key % M (M is prime number close to table size)  POLYNOMIAL  ACCUMULATION  =>  H(string) = [(s[0]*x^0) +  (s[1]*x^1) +  ....] % M   Requirements for hash function => H(key) = value No matter how many times you pass key => value should always be same If M is size of hash table => value must lie between 0 to M-1 Value should be uniformly distributed --> less collision UNIVERSAL  HASHING -  choose any hash function uniformly at random
Read more