Better string handling

The text and pattern are now streamed directly (skipping saving them to
strings).
Knowing exactly what string was matched has now been delegated to
the Rabin_fingerprint_process.

Makefile

@@ -24,8 +24,8 @@ LDLIBS = # library flags
 # -llapacke (LAPACK) NOTE: The order of "-llapacke -llapack -lblas" is very important
 #LINK.o = $(CXX) $(LDFLAGS) # use CXX for linking
 
-simple_string_matching: simple_string_matching.o Rabin_fingerprint.o general_library.o
+simple_string_matching: simple_string_matching.o Rabin_fingerprint.o general_library.o processes.o
-	$(CXX) $(CXXFLAGS) simple_string_matching.o Rabin_fingerprint.o general_library.o -o simple_string_matching
+	$(CXX) $(CXXFLAGS) simple_string_matching.o Rabin_fingerprint.o general_library.o processes.o -o simple_string_matching
 
 simple_string_matching.o: simple_string_matching.cpp
 	$(CXX) $(CXXFLAGS) -c simple_string_matching.cpp
@@ -36,6 +36,9 @@ Rabin_fingerprint.o: Rabin_fingerprint.cpp Rabin_fingerprint.hpp
 general_library.o: general_library.cpp general_library.hpp
 	$(CXX) $(CXXFLAGS) -c general_library.cpp
 
+processes.o: processes.cpp processes.hpp
+	$(CXX) $(CXXFLAGS) -c processes.cpp
+
 porat-porat: porat-porat.cpp
 
 # Tell the compiler that 'clean' isn't referring to a file

README.md

@@ -2,9 +2,20 @@
 
 Code used for my Bachelor's Thesis
 
+# Issues
+When reading the pattern from a file, according to UNIX standards ([read details here](https://unix.stackexchange.com/questions/18743/whats-the-point-in-adding-a-new-line-to-the-end-of-a-file)) a "text file" has to end on a newline.  
+This adds an unintended newline to the end of the pattern, so to avoid adding said newline, write the file using:  
+`echo -n word > pattern.txt`  
+or if you use Vim, remove the newline using  
+```
+:set binary
+:set noeol
+:wq
+```
+
 # Versions
 ## Current Version
-Compared to V1, this current state `generate_initial_irreducible_polynomials.sage`, `generate_random_irreducible_polynomial.sage`, and `multiply_polynomials_modulo_polynomial.sage` have been removed from the main folder, as they have already been used for implementing their corresponding features in `general_library.cpp`. They will be preserved in the V1 folder.  
+Compared to V1, in this current state `generate_initial_irreducible_polynomials.sage`, `generate_random_irreducible_polynomial.sage`, and `multiply_polynomials_modulo_polynomial.sage` have been removed from the main folder, as they have already been used for implementing their corresponding features in `general_library.cpp`. They will be preserved in the V1 folder.  
 
 This version also contains code relating to Porat-Porat and other random code stumps. This code is going to be completely reimplemented, but is preserved for now to avoid redoing work.  
 

Rabin_fingerprint.hpp

@@ -14,7 +14,7 @@ class Rabin_fingerprint {
         void push_bit (bool b);
         void shift_bit (bool b);
         void slide_char (char c_in, char c_out);
-        void slide_bit (bool b1, bool b2);
+        void slide_bit (bool b_in, bool b_out);
 
         uint32_t get_fingerprint();
 

books/pattern.txt

@@ -0,0 +1 @@
+word

processes.cpp

@@ -0,0 +1,59 @@
+#include "processes.hpp"
+
+Rabin_fingerprint_process::Rabin_fingerprint_process(uint32_t irr_poly, size_t window_size_in_bits)
+    : window_size_in_bits(window_size_in_bits),
+      phi(irr_poly, window_size_in_bits)
+{}
+
+void Rabin_fingerprint_process::stream_char(char c) {
+    std::bitset<8> b(c);
+    for (char i = 7; i >= 0; i--) {
+        stream_bit((bool)b[i]);
+    }
+}
+
+void Rabin_fingerprint_process::stream_bit(bool b) {
+    if (window.size() == window_size_in_bits) {
+        window.push(b);
+        bool b_out = window.front();
+        window.pop();
+        phi.slide_bit(b, b_out);
+    } else {
+        window.push(b);
+        phi.push_bit(b);
+    }
+}
+
+uint32_t Rabin_fingerprint_process::get_fingerprint() {
+    return phi.get_fingerprint();
+}
+
+std::string Rabin_fingerprint_process::get_string_in_window() {
+    // check if window contains a whole number of chars
+    if ((window_size_in_bits & 0b111) != 0)
+        throw std::logic_error("The fingerprinting window doesn't contain a whole number of chars (counting the bits), so it doesn't make sense to return it as a string.");
+
+    #ifndef NDEBUG
+    if (window.size() != window_size_in_bits)
+        throw std::logic_error("False match! The sliding window isn't even filled yet, which means you matched the pattern of a substring shorter than the pattern. This case should be handled/avoided elsewhere, so we throw an error.");
+    #endif
+
+    std::ostringstream os;
+    for (size_t i = 0; i < window.size()>>3; i++) {
+        // cycle the char
+        char c = 0;
+        for (size_t j = 0; j < 8; j++) {
+            bool b = window.front();
+            window.pop();
+            window.push(b);
+            c <<= 1;
+            c |= b;
+        }
+
+        os << c;
+    }
+
+    std::string s = os.str();
+
+    return s;
+}

processes.hpp

@@ -0,0 +1,26 @@
+#ifndef PROCESSES_H
+#define PROCESSES_H
+
+#include "Rabin_fingerprint.hpp"
+#include "general_library.hpp"
+
+#include <stdint.h>
+#include <queue>
+#include <string>
+#include <sstream>
+
+class Rabin_fingerprint_process {
+    public:
+        Rabin_fingerprint_process(uint32_t irr_poly, size_t window_size_in_bits);
+        void stream_char(char c);
+        void stream_bit(bool b);
+        uint32_t get_fingerprint();
+        std::string get_string_in_window();
+
+    private:
+        std::queue<bool> window;
+        size_t window_size_in_bits;
+        Rabin_fingerprint phi;
+};
+
+#endif

simple_string_matching.cpp

@@ -1,6 +1,7 @@
 /* #define NDEBUG */
-#include "Rabin_fingerprint.hpp"
+/* #include "Rabin_fingerprint.hpp" */
 #include "general_library.hpp"
+#include "processes.hpp"
 
 #include <iostream>
 #include <stdint.h>
@@ -8,53 +9,42 @@
 #include <string>
 #include <fstream>
 
-void print_match (size_t index, size_t length, std::string &T) {
-    std::cout << "Match found at index " << index << " with the text \"";
-    for (size_t i = 0; i < length; i++)
-        std::cout << T[index + i];
-    std::cout << "\"" << std::endl;
-}
-
 int main() {
-    /* std::ifstream ifs("books/the_complete_works_of_william_shakespeare.txt"); */
+    /* std::ifstream T("books/the_complete_works_of_william_shakespeare.txt"); */
-    std::ifstream ifs("books/genji_monogatari_english.txt");
+    std::ifstream T("books/genji_monogatari_english.txt");
-    std::string T( (std::istreambuf_iterator<char>(ifs) ),
+    std::ifstream P("books/pattern.txt");
-                    (std::istreambuf_iterator<char>() ) );
-
-    /* std::string T = "Hello, this is my test string averylongword is a necessary word to exceed the 32 bit window."; */
-    // Test without the modulo polynomial - and two matches
-    std::string P = "word";
-    // Test with the modulo polynomial
-    /* std::string P = "averylongword"; */
 
+    char c;
+    size_t P_length = 0;
     std::cout << "Searching for pattern:" << std::endl;
-    std::cout << "    " << P << std::endl;
+    std::cout << "    ";
-    /* std::cout << "in text:" << std::endl; */
+    while(P.get(c)) {
-    /* std::cout << "    " << T << std::endl; */
+        std::cout << c;
-    std::cout << std::endl;
+        P_length++;
 
-    /* uint32_t polynomial = pow(2, 30) + pow(2, 2) + 1; // x^31 + x^3 + 1 */
+    }
-    uint32_t polynomial = get_random_irreducible_polynomial_in_Z2(31);
+    std::cout << std::endl << std::endl;
-    /* uint32_t polynomial = 0b11010011100100000111101011110111; */
+
-    // Test without the modulo polynomial
+    P.clear();                 // clear fail and eof bits
-    size_t window_size_in_bits = P.length()*8;
+    P.seekg(0, std::ios::beg); // back to the start!
+
+    uint32_t irreducible_polynomial = get_random_irreducible_polynomial_in_Z2(31);
+    size_t window_size_in_bits = P_length*8;
 
     // Hash the pattern
-    Rabin_fingerprint fP(polynomial, window_size_in_bits);
+    Rabin_fingerprint_process phiP(irreducible_polynomial, (size_t)window_size_in_bits);
-    for (char c : P)
+    while(P.get(c)) {
-        fP.push_char(c);
+        phiP.stream_char(c);
+    }
 
     // Hash the text
-    Rabin_fingerprint fT(polynomial, window_size_in_bits);
+    Rabin_fingerprint_process phiT(irreducible_polynomial, window_size_in_bits);
-    for (size_t i = 0; i < P.length(); i++)
+    size_t index = 0;
-        fT.push_char(T[i]);
+    while(T.get(c)) {
-    if (fT.get_fingerprint() == fP.get_fingerprint())
+        phiT.stream_char(c);
-        print_match(0, P.length(), T);
+        index++;
-
+        if (phiT.get_fingerprint() == phiP.get_fingerprint())
-    for (size_t i = P.length(); i < T.length(); i++) {
+            std::cout << "Match found at index " << index-P_length << " with the text \"" << phiT.get_string_in_window() << "\"" << std::endl;
-        fT.slide_char(T[i], T[i-P.length()]);
-        if (fT.get_fingerprint() == fP.get_fingerprint())
-            print_match(i-P.length()+1, P.length(), T);
     }
 
     std::cout << std::endl;