File compression and decompression

Compression and decompression of files

u Development Environment : vs2013

u  :

u  : File compression is to compress a large memory file into a small memory file, saving disk space                       &N Bsp      , transmission is also more convenient. Decompression is to restore the compressed file into the original file, read and write more convenient.

√ can compress and unzip small files, can also compress and unzip large files.

√ In the debug, compression and decompression of about 6M files, need about 25s, in the release, need 2s or so.

                                  &N Bsp      √ has profiles, compression vector,

u  problems encountered :√ in the compression process not enough 8 bits will be 0, but in the decompression process will be zero read out, so it is wrong. To resolve                                 &NBSP ; The problem. I defined a total length in the decompression process, it writes a character, the total length is reduced by 1, when the length                         &NB Sp         minus 0, don't go in.

' \r ' but with binary mode it won't have this                                         issues.

√ Sometimes read a large file, you will not read it all. So I swapped EOF for feof. Because EOF is the end flag with-1, but it appears in the file-1 it will not be read anymore. If you switch to feof, it is marked with "End of file". This will not be able to read an endless situation.

In Compress.h

#ifndef  __compress_h_#define __compress_h_#include <string>typedef long long  longtype;struct charinfo{unsigned char _ch;//character longtype _count;//occurrences string _ code;//Huffman Code charinfo (unsigned char ch =  ' 0 '): _ch (CH),  _count (0),  _code ("") {} Charinfo (Longtype count): _count (count),  _ch (0),  _code ("") {}//overloaded operator > "Bool operator  > (const charinfo& com) Const{return _count > com._count;} Overloaded operator "! =" bool operator != (charinfo com) const {return _count != com._count ;} Overloaded operator "+" charinfo operator+ (const charinfo& com) const{return charinfo (_count +  com._count);}}; class filecompress{public://compression filecompress () {for  (size_t i = 0; i <  256; i++) {_info[i]._ch = i;}} Void compress (const char* file) {//Read file file *fout = fopen (file,  "RB"); Unsigned char ch = 0;assert (fout);//Count the occurrences of each character ch = fgetc (fout);while  (!feof (fout)) {_info[(Unsigned char) ch]._count++;ch = fgetc (fout);} Using the number of times to establish Huffman Tree Const charinfo invalid ((unsigned char) 0); Haffmantree<charinfo> tree (_info, 256,invalid);//Generate Huffman Code String tmp;_createhaffmancode ( Tree. Getroot (),  tmp);//Compressed fseek (Fout, 0, seek_set);//compressed/Compressed file name starting from the first character of the document String comfile =  file;comfile +=  ". Haffman"; File *fin = fopen (Comfile.c_str (),  "WB"); assert (Fin); unsigned char value =  0;size_t index = 0;//Mark Current bit ch = fgetc (fout);while  (!feof (Fout)) {string  code = _info[ch]._code;for  (Size_t i = 0; i < code.size ( );  i++) {if  (code[i] ==  ' 1 ') {value <<= 1;value |= 1;} else{value <<= 1;} Over 8Byte, writes it to the compressed file if  (++index == 8) {FPUTC (value, fin);     value =  0;     index = 0;}} CH&NBSP;=&NBSP;FGETC (fout);} If the,value  is not written to 8 bits, the remaining press into the compressed file if  (index != 0) {value <<=  (8 -  index); FPUTC (Value, fin);} Configuration file string configfile = file;configfile +=  ". config"; File* finfo = fopen (Configfile.c_str (),  "WB"), assert (Finfo);//write the total number of characters into the configuration file char  infostr[32];//writes the occurrences of the characters and the number of times into the configuration file string str;for  (size_t j = 0; j < 256 ;  j++) {if  (_info[j]._count > 0) {Str.push_back (Unsigned char) j); Str.push_back (', ') ; _itoa (_info[j]._count, infostr, 10); Str += infostr;fputs (Str.c_str (),  finfo); FPUTC (' \ n ',  finfo); Str.clear ();}} Fclose (Fout); fclose (Fin); fclose (finfo);} Unzip Void uncompress (const char *file) {UNSIGNED&NBSP;CHAR&NBSP;CH&Nbsp;= 0;string fconfig = file;fconfig +=  ". config";//Read Compressed file file* fout =  fopen (Fconfig.c_str (),  "RB"), assert (Fout), string tmp;//character occurrences while  (ReadLine (fout,  TMP)) {if  (!tmp.empty ()) {//That to character ch = tmp[0];//gets the number of characters _info[(Unsigned char) ch]._count =  atoi (TMP.SUBSTR (2). C_STR ()); Tmp.clear (); else{tmp =  ' \ n ';}} Rebuilding Huffman Tree Charinfo invalid ((unsigned char) 0); Haffmantree<charinfo> h (_info, 256, invalid); Haffmantreenode<charinfo>*root = h.getroot ();//Decompression String output = file;output  +=  ". uncom"; File* fin = fopen (Output.c_str (),  "WB"); assert (Fin);//Restore files based on compressed files String haffmanfile  = file; haffmanfile +=  ". Haffman"; File* fcom = fopen (Haffmanfile.c_str (),  "RB"); assert (fcom); HAFFMANTREENODE&LT;CHARINFO&GT;*CUR&NBSP;=&NBSP;ROOT;CH&NBSP;=&NBSP;FGETC (fcom); int pos = 8; longtype len = root->_weight._count;while  (len > 0) {while  (cur->_ Left &&cur->_right) {if  (pos == 0) {ch = fgetc (fcom);p os = 8;} --pos;//with 1, walk right Tree if  (ch& (1 << pos)) {cur = cur->_right;} With 0, walk left tree Else{cur = cur->_left;}} if  (cur) {FPUTC (Cur->_weight._ch, fin);//write in unzip file cur = root;} --len;} Fclose (Fout); fclose (Fin); fclose (fcom);} PROTECTED://Create Huffman Code Void _createhaffmancode (haffmantreenode<charinfo>*root, string tmp) {// Empty if  (Root == null) {return;} if  (root->_left == null&&root->_right == null) {//find subscript, write code to _code int  index =  (root->_weight). _ch;_info[index]._code = tmp;} else{//Zo Shu Write 0if  (root->_left) {tmp.push_back (' 0 '); _createhaffmancode (root->_left, tmp); tmp.pop_ Back ();} Right tree write 1if  (root->_right) {tmp.push_back (' 1 '); _createHaffmancode (root->_right, tmp); Tmp.pop_back ();}} Reads a line bool readline (file* file, string& tmp) {assert (FILE);char ch =  Fgetc (file);if  (feof (file)) {Return false;} while  (ch !=  ' \ n ') {tmp += ch;ch = fgetc (file);} Return true;} protected:charinfo _info[256];}; #endif  //__compress_h_

HaffmanTree.h in

#ifndef  __HaffmanTree_H_#define __HaffmanTree_H_#include  "Heap.h" template<class t> struct  haffmantreenode{typedef haffmantreenode<t> node; t _weight; node* _left; node* _right; Haffmantreenode (const t& weight): _weight (weight),  _left (null),  _right (null) {}}; Template<class t>class haffmantree{typedef haffmantreenode<t> node;public: Haffmantree (): _root (NULL) {}haffmantree (const t* arr, size_t size) {_root = _ Create (arr, size);} Constructor Haffmantree (const t* arr, size_t size, const t& invalid) { _ Root = _create (arr, size, invalid);} ~haffmantree () {if  (_root) {_destroy (_root);}} Node* getroot () {return _root;} protected://Create Huffman Tree node* _create (const t* arr, size_t size, const t&  invalid) {&NBSP;STRUCT&NBSP;COMPARE{BOOL&NBSP;OPERATOR&NBSP; () (Const node *left, const node *right) {return left->_weight >  right->_weight;}}; heap<node*,compare>  _heap;//put the value in the vector for  (size_t i = 0; i  < size; ++i) {if  (arr[i] != invalid) {Node *tmp = new node (arr[i ]); _heap. Push (TMP);}} Construct Huffman Tree while  (_heap. Size ()  > 1) {node *left = _heap. Top (); _heap. Pop (); Node*right = _heap. Top (); _heap. Pop (); Node *parent = new node (left->_weight + right->_weight);p Arent->_left  = left;parent->_right = right;_heap. Push (parent); Return _heap. Top ();} Release node Void _destroy (node* root) { if  (root->_left == null &&  root->_right == null) {delete root;root = null;} Else{_destroy (Root->_left); _destroy (root->_right);}} private:node *_root;}; #Endif //__haffmantree_h_ 

In Heap.h

#ifndef  __Heap_H_#define __Heap_H_#include<vector> #include <assert.h>template<class  t>//Comparator Class less{public:bool operator ()   (const t& left, const  t& right) {return left > right;}}; Template<class t>class greater{public:bool operator ()   (Const t& left,  const t& right) {return left < right;}}; template<class t, class compare = less<t> >class heap{public:// Constructor Heap (): _arr (NULL) {}heap (const t* arr, int size) {assert (arr); _arr.reserve (size); for   (int i = 0; i < size; i++) {_arr.push_back (arr[i]);} Int begin = _arr.size ()  / 2 - 1;while  (begin >= 0) {_ Adjustdown (begin); begin--;         }}//copy construction Heap (vector<T> & s) {_arr = s;int begin = _arr.size ()  / 2 - 1;while  (begin  >= 0) {_adjustdown (BEGIN);begin--;             }}void push (const t& x) {_arr.push_back (x); Int begin = _arr.size ( ); _adjustup (begin);} Void pop ()  {_arr[0] = _arr[_arr.size ()  - 1];_arr.pop_back (); _AdjustDown (0);} Void clear () {_arr.clear ();} Bool empty () {return _arr.empty ();}   t& top ()  {if  (! Empty ()) {return _arr[0];} Exit (1);} Size_t size () {return _arr.size ();} protected://Void _adjustdown (int parent) { //  downward adjustment from the root node int size = _arr.size ();int left = parent * 2 + 1;int right = left + 1; int key = left;while  (left < size) {//compare () functor//if  (right <  size &amP;& array[left] > array[right]) if  (right < size &&  Compare () (_arr[left], _arr[right]))   //right small {key = right;} Else{key = left;} if  ((Array[key] > array[parent])) if  (Compare () (_arr[key], _arr[parent)) {break;} else if  (Array[parent] > array[key]) else if  (Compare () (_arr[parent], _ Arr[key])  //left small {swap (_arr[key], _arr[parent]);} parent = key;left = parent * 2 + 1;right = left +  1;}} Raise Void _adjustup (Int child) {  //to adjust the int size = _arr.size () upward from the leaf node or child node;   (size == 0 | |  size == 1) {return;} int parent =  (child - 2)  / 2;int left = parent * 2  + 1;int right = left + 1;int key = 0;while  (parent >= 0) {//if  (right < size && array[left] >  Array[right]) if  (Right < size && compare () (_arr[left], _arr[right]) { Key = right;} Else{key = left;} if  (Array[parent] > array[key]) if  (Compare () (_arr[parent], _arr[key)) {Swap (_arr [Parent], _arr[key]);} Else{break;} if  (parent == 0) {break;} parent =  (parent - 1)  / 2;left = parent * 2 + 1; right = left + 1;}} private:vector<t> _arr; }; #endif  //__heap_h_

In Test.cpp

#define _crt_secure_no_warnings 1#include<time.h> #include <iostream>using namespace std; #include " HaffmanTree.h "#include" Compress.h "void Test () {double t1;double t2; Filecompress f;//f.compress ("Input.txt"); F.compress ("Mnt.txt"); t1 = Clock ();p rintf ("Time required for compressed files:%f s \ n", T1/clocks_ PER_SEC);//f.uncompress ("Input.txt"); F.uncompress ("Mnt.txt"); t2 = Clock ();p rintf ("Time required for extracting files:%f s \ n", T2/clocks_ PER_SEC);} int main () {Test (); System ("pause"); return 0;}

File compression and decompression