Offline English to English Dictionary

[cosh]

Hello,
Please enjoy this offline English to English dictionary that I wrote recently.
I just grabbed a raw text format dictionary file from Github and I altered format of this text file as dictionary input.
And I handled the raw file and used a Trie and an AA binary search tree to index all worlds.
The result was now we got a console-based dictionary with maximized searching speed and low memory usage.

$ ./dict
Dict file: ./dict.txt
36447 words loaded.
? Apple
Searching: "Apple"...
        1316 Apple      n. 1 roundish firm fruit with crisp flesh. 2 tree bearing this. apple of one's eye cherished person or thing. [old english]
? .?
Searching: ".?"...
Type [WORD] or [NUMBER] to search.
        For example ? Apple ? 10536
Type .h to show history.
Type .l[A] to show alphabet.
        For example ? .l Z.
Type .? to show this notice.
? 1
Searching: "1"...
        1 A-    prefix (also an- before a vowel sound) not, without (amoral). [greek]
?

The above texts are the console output when you use this dictionary.
I'll print the mainly text file that this dictionary concludes.

Code Select Expand


#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "StoneValley/src/svset.h"
#include "StoneValley/src/svtree.h"

typedef struct st_WORD
{
ptrdiff_t id;
char name[64];
size_t times;
long ltip;
} WORD, * P_WORD;

int cbfcmpid(const void * px, const void * py)
{
return (*(P_WORD)px).id - (*(P_WORD)py).id;
}

int cbfcmpchar(const void * px, const void * py)
{
return *(char *)px - *(char *)py;
}

int cbftvs_history(void * pitem, size_t param)
{
if (((P_WORD)P2P_TNODE_B(pitem)->pdata)->times)
printf("%s\t%lld\n", ((P_WORD)P2P_TNODE_B(pitem)->pdata)->name, ((P_WORD)P2P_TNODE_B(pitem)->pdata)->times);
return CBF_CONTINUE;
}

int cbftvs_alphabet(void * pitem, size_t param)
{
if (toupper(((P_WORD)P2P_TNODE_B(pitem)->pdata)->name[0]) == param)
printf("%s\n", ((P_WORD)P2P_TNODE_B(pitem)->pdata)->name);
return CBF_CONTINUE;
}

static char sWord[BUFSIZ * 2] = { 0 };
static char * p = sWord;

static char sFileName[BUFSIZ] = { 0 };
static char sPattern[BUFSIZ] = { 0 };

int main(int argc, char ** argv)
{
size_t i = 1, j;
WORD w = { 0 };
FILE * fp;
P_SET_T set = setCreateT();
P_TRIE_A trie = treCreateTrieA();
size_t * result = NULL;

strcat(sFileName, ".\\dict.txt");
fp = fopen(sFileName, "rb");

printf("Dict file: %s\n", sFileName);
if (NULL != fp)
{
while (!feof(fp))
{
*p = fgetc(fp);
++p;
if ('#' == *(p - 1))
{
P_BSTNODE pnode = NULL;
*(p - 1) = '\0';
p = sWord;

strcpy(w.name, p);
w.id = i;
w.times = 0;
w.ltip = ftell(fp) + 1;

setInsertT(set, &w, sizeof w, cbfcmpid);
pnode = treBSTFindData_A(*set, &i, cbfcmpid);
treInsertTrieA(trie, p, strlen(p), sizeof(char), (size_t)(pnode->knot.pdata), cbfcmpchar);

++i;
}
if ('\n' == *(p - 1))
{
p = sWord;
}
}
printf("%lld words loaded.\n", i);
do
{
printf("? ");
fgets(sPattern, 100, stdin);
sPattern[strlen(sPattern) - 1] = '\0';
if ('\0' == *sPattern)
break;
printf("Searching: \"%s\"...\n", sPattern);
result = treSearchTrieA(trie, sPattern, strlen(sPattern), sizeof(char), cbfcmpchar);
if (result)
{
printf("\t%lld %s  ", ((P_WORD)*result)->id, ((P_WORD)*result)->name);
++((P_WORD)*result)->times;

/* Redirect to the word on the disk. */
fseek(fp, ((P_WORD)*result)->ltip, SEEK_SET);
/* Read explanations. */
p = sWord;
while ('\n' != (*p = fgetc(fp)))
{
++p;
}
*p = '\0';
p = sWord;
printf("\t%s\n", p);
}
else if ('.' == sPattern[0] && '?' == sPattern[1])
{
printf("Type [WORD] or [NUMBER] to search.\n");
printf("\tFor example ? Apple ? 10536\n");
printf("Type .h to show history.\n");
printf("Type .l[A] to show alphabet.\n");
printf("\tFor example ? .l Z.\n");
printf("Type .? to show this notice.\n");
}
else if ('.' == sPattern[0] && 'h' == sPattern[1])
{
printf("History:\n");
treTraverseBIn(*set, cbftvs_history, 0);
}
else if ('.' == sPattern[0] && 'l' == sPattern[1] && ' ' == sPattern[2])
{
sPattern[3] = toupper(sPattern[3]);
printf("Alphabet:\n");
treTraverseBIn(*set, cbftvs_alphabet, toupper(sPattern[3]));
}
else if (0 != (j = atoi(sPattern)))
{
P_BSTNODE pnode = treBSTFindData_A(*set, &j, cbfcmpid);
if (pnode)
{
printf("\t%lld %s  ", ((P_WORD)(pnode->knot.pdata))->id, ((P_WORD)(pnode->knot.pdata))->name);
++((P_WORD)(pnode->knot.pdata))->times;

/* Redirect to the word on the disk. */
fseek(fp, ((P_WORD)(pnode->knot.pdata))->ltip, SEEK_SET);
/* Read explanations. */
p = sWord;
while ('\n' != (*p = fgetc(fp)))
{
++p;
}
*p = '\0';
p = sWord;
printf("\t%s\n", p);
}
}
else
{
printf("Can not find \"%s\".\n", sPattern);
}
} while ('\0' != sPattern[0]);
fclose(fp);
}
else
printf("Can not open file.\n");

printf("History:\n");
treTraverseBIn(*set, cbftvs_history, 0);
setDeleteT(set);
treDeleteTrieA(trie, sizeof(char));
return 0;
}



And finally hope there are no license and privilege concerns of using the raw dictionary textual file.
You guys may download dictionary from here: https://github.com/coshcage/dict
Please notice that this project needs StoneValley library.

[HellOfMice]

A very clean code.
I downloaded all the files and will try them.
No Pelle's project file.
Why have you made it in a console?
Adding clipboard support would be a good idea.


Thank You to share it.

[cosh]

A very clean code.
I downloaded all the files and will try them.
No Pelle's project file.
Why have you made it in a console?
Adding clipboard support would be a good idea.


Thank You to share it.

HI, HellOfMice, hi, everyone
Happy new year!
The reason I made dict in a console is that it is easy to cross platform and esay for me to achieve.
Now, Windows Terminal supports clipboards.

Recently, I advance the code and let dict support regular expressions.
If you get interested, you could download the source code from github and compile it.
Again, thank you to use and share.