Help in text comparison.

Programming Questions

javiercho..

in Programming Questions • 7 months ago

Hi all,

I'm quite newbie in programming and imagine that this question will problably extremely easy to solve, but I can't get it, so any help will be very welcome.

What i'm trying to do is comparing two text files (a dictionary and a book) and create different arrays with all the matches between both, so to say, i will end up having an array with x "democracy" terms, another with y "state" terms, and so on.

Up to now i've been able to create a counter that counts all the matches between both so i have a counter of total matches, but I can't find the way to split this into different groups?.

Any ideas? Thxs a lot.

Here is the code:

String [] ficherotexto;

String [] libropolitica;

String textolibrojunto;

String textolibrojuntoarray [];

String tokens;

int t, l, m;

boolean c;

int contador, contador2, contador3;

ArrayList coincidencias;

String [] coincidencias3;

void setup () {

size (100, 100);

// Creo la matriz con el fichero a tratar

String [] ficherotexto = loadStrings ("glosario_de_politica.txt");

String [] libropolitica = loadStrings ("aristoteles_politica_parcial.txt");

//Junto todo el texto en una string para que cada elemento no sea un parrafo sino

//que junto todo el texto en una sola String. Asi los podré luego dividir en una array

//de palabras que podré comparar con otra cosa.

String textolibrojuntomay = join (libropolitica, ' ');

//Paso todo el libro a minusculas

String textolibrojunto= textolibrojuntomay.toLowerCase();

//Convierto la cadena en matriz. Split solo se puede usar con cadenas, no con arrays.

//Por eso he usado antes join, para convertir el texto en una unica String de texto,

//y ahora la vuelvo a convertir en una Array de palabras con split.

String [] textolibrojuntoarray = split(textolibrojunto, ' ');

//Elementos a quitar de la cadena

String tokens = " ,.;:-_!?&\"□▫¶~ +%=()[]{}«»#&%§'|";

//Convierto la variable-matriz en variable int

int t= int (ficherotexto.length);// Numero de terminos del glosario de politica.

int l= int (libropolitica.length);//Numero de parrafos del libro.

int m= int (textolibrojuntoarray.length);// Numero de palabras del libro.

println("num terminos del glosario: " + t);

println("num parrafos del libro: " + l);

println("num palabras del libro: " + m);

//println (libropolitica);

//println (textolibrojunto);

//Matriz para las cpincidencias

coincidencias = new ArrayList(); // crea nueva ArrayList vacia

//Comparar los dos textos

int matchedValues = 1;

contador=0;

contador2=0;

contador3=0;

for (int i=0; i < t; i++) {// este es el bucle del glosario

boolean c=true;

contador++;

c= false;

for (int j=0; j < m; j++) {// este es el bucle del libro de politica

// Comparar las dos matrices

if (ficherotexto[i].equals (textolibrojuntoarray[j]) == true) {

if (!c) {

contador3++;

println ("ok"+ contador3 );

}

coincidencias.add (ficherotexto [i]);// incluye la entrada en el Arraylist(coincidencias).

println("Término: " +"'"+ ficherotexto[i]+"'" +" presente en el texto "+ matchedValues);

matchedValues++;

println (coincidencias);

contador2++;//si lo pongo aquí me da las veces que

println (contador2);// encuentra una coincidencia.

}

//Aqui tengo que poner una linea para que mientras no cambie i haga una cadena que reuna todas las

//repeticiones de un termino.

}

//println (contador2);

}

void draw () {

background (255);

}

Replies(3)

PhiLho

Re: Help in text comparison.

7 months ago

First, some general remarks on your code:
- Spanish comments can help you, but doesn't help (most of) us... At least, code is relatively simple, so we can skip them.
- You can omit draw() if your sketch does no drawing at all, and if you don't need events like keyPressed() or mousePressed().
- You declare some variables globally (eg. libropolitica), then re-declare them in setup(). (declare = prefixing them with their type). That means that you create a local variable of same name than the global one. They are distinct, the global one is shadowed, and remains uninitialized. Here, it isn't important, as you have only one function, but it might bite you someday, if you are not careful (that's a common error we see in the forum!).
- Line: int t= int (ficherotexto.length); is redundant (length is already an int). And t isn't a good variable name, unless it is very short lived. Something like textLength is better.
- Better type your array list: ArrayList<String> coincidencias; and coincidencias = new ArrayList<String>(); (here, no shadowing...)
- Your search is inefficient. That's not really important if time isn't a problem, but at some time in your study, you might want to take a look at HashMap (in the reference). Here, HashMap<String, String>.

Split in different groups: you need a glossary for each group. And you repeat your search for each of them.
You might want to do a function taking the content of the book and a glossary, and perhaps an index in a global array of array lists of results, and an array of counters, perhaps. This function would do the search for the given group, avoiding repeating the code, and would store the results in the given index.

javiercho..

Re: Help in text comparison.

7 months ago

Thanks for taking the time of answering to my question PhiLho, all your comments have been very useful.

I will clean the code before posting it next time, I posted them thinking that it might also be useful to any other spanish speaking person, but the truth is that it wouldn`t be useful to anyone if not working :-).

Before posting I had already parsed through 2d arrrays, HashMap and String class in the reference and in the forum, but I haven't succeed in understanding well how it works up to now, I'm stucked at this point.

The problem is that I don't get to know how to split it in different groups as you say, I've tried everything I know and looked in other sketches here and in Openprocessing but haven't found still anything similar to what I'm trying to do.

Thanks again.

javiercho..

Re: Re: Help in text comparison.

7 months ago

Just in case it could be useful to anybody, this code works. It compares two different text files and classifies the coincidences.

Thanks to Rick Companje whose code I started with and PhiLo for your advices. I still don't really understand how it works but i will :-).

/**

hashtable taken from http://www.companje.nl

@author Rick Companje

String list[];

void setup() {

//Texto a analizar

String [] lines = loadStrings ("aristoteles_politica.txt");

String librojunto= join (lines, ',');

String textor= librojunto.toLowerCase();

String[] texto = splitTokens(textor, " ,.?!:;[]-");

//Diccionario

String [] glosario = loadStrings ("glosario_de_politica.txt");

//println (glosario);

String glosar = join (glosario, ',');

String dicc= glosar.toLowerCase();

String[] diccionario = splitTokens(dicc, ",.?!:;[]-");

println ("Hay " +lines.length + " lineas");

Words words = new Words();

for (int j=0; j < texto.length; j++) {

for (int k=0; k<diccionario.length; k++) {

if (texto[j].equals (diccionario[k]) == true) {

words.addWord(texto[j]);

}

words.printWords();

}

class Words {

Hashtable ht=new Hashtable();

void addWord(String txt) {

Word word;

if (ht.containsKey(txt)) {

word = (Word)ht.get(txt);

word.count++;

}

else {

word = new Word(txt);

}

ht.put(txt, word);

}

void printWords() {

for (Enumeration e = ht.keys() ; e.hasMoreElements() ;) {

String name = e.nextElement().toString();

Word word = (Word)ht.get(name);

println (word.txt + ": " + word.count);

}

class Word {

public String txt;

public int

count = 1;

Word(String txt) {

this.txt = txt;

}

Top Reply

Statistics

3 Replies
536 Views
0 Followers

Actions

Permalink

Bible Projects