void letterChain(char *inFile, char *outFile, int maxSize)
/* letterChain - Make Markov chain of letters in text. */
{
struct dlList *ll = dlListNew();
int llSize = 0;
int c;
FILE *in = mustOpen(inFile, "r");
FILE *out;
struct dlNode *node;
UBYTE *s;
struct trie *trie;
AllocVar(trie);
while ((c = getc(in)) >= 0)
{
if (llSize < maxSize)
{
s = needMem(1);
*s = c;
dlAddValTail(ll, s);
++llSize;
if (llSize == maxSize)
addToTrie(trie, ll);
}
else
{
node = dlPopHead(ll);
s = node->val;
*s = c;
dlAddTail(ll, node);
addToTrie(trie, ll);
}
}
if (llSize < maxSize)
addToTrie(trie, ll);
while ((node = dlPopHead(ll)) != NULL)
{
addToTrie(trie, ll);
freeMem(node->val);
freeMem(node);
}
dlListFree(&ll);
carefulClose(&in);
out = mustOpen(outFile, "w");
rDumpTrie(0, trie->useCount, trie, out);
carefulClose(&out);
}
void addToTrie(Node *node, int from, int to) {
int ch;
node->degree++;
if (from == to)
return;
ch = string[from] - ALPHA_BASE;
if (node->children[ch] == NULL)
node->children[ch] = createNode();
addToTrie(node->children[ch], from+1, to);
}
int main(int argc, char **argv) {
int i, n;
Node *root;
while (scanf("%s", string) != EOF) {
poolPos = 0;
n = strlen(string);
root = createNode();
for (i = 0; i < n; i++)
addToTrie(root, i, n);
findSuffix(root);
getSuffix(root, 0);
printf("%s\n", suffix);
// destroyNode(root);
}
return EXIT_SUCCESS;
}
// Builds add, rep and rem tries from given content of transformations file
int trieFromFile(char *data) {
char *string1;
char *string2;
wchar_t *wstr1;
wchar_t *wstr2;
int w1;
int w2;
double v;
int datalen;
int i;
int j;
datalen = strlen(data);
string1 = malloc(2);
string2 = malloc(2);
i = 0;
j = 0;
while(i < datalen) {
/*
* Setting weights of default edit operations:
* the line must begin with character '>', followed
* by "add" for addition, "rep" for replacement and
* "del" for deletion operation; after the operation
* marker, ':' is placed and finally comes the new
* weight as double.
*/
if(data[i] == '>') {
i = i +1;
/* Change weight of default add operation */
if(strncmp((char *)(data+i), "add", 3) == 0) {
i +=4;
j = i;
while(data[j] != '\n' && data[j] != '\r' && j < strlen(data))
j++;
add = findValue(data, i, j);
if(data[j] == '\r') /* In case we are under Windows */
j = j + 2;
else j = j+1;
i = j;
}
/* Change weight of default replace operation */
else if(strncmp((char *)(data+i), "rep", 3) == 0) {
i +=4;
j = i;
while(j < strlen(data) && data[j] != '\n' && data[j] != '\r') j++;
rep = findValue(data, i, j);
if(data[j] == '\r') /* In case we are under Windows */
j = j + 2;
else j = j+1;
i = j;
}
/* Change weight of default remove operation */
else if(strncmp((char *)(data+i), "rem", 3) == 0) {
i +=4;
j = i;
while(j < strlen(data) && data[j] != '\n' && data[j] != '\r') j++;
rem = findValue(data, i, j);
if(data[j] == '\r') /* In case we are under Windows */
j = j + 2;
else j = j+1;
i = j;
}
}
/*
* Setting weights of generalized edit distance transformations;
*/
else {
j = i;
/* Locate the left side string of transformation */
while(data[j] != ':') j++;
string1 = (char *)realloc(string1, (j-i+1));
if(string1 == NULL) {
perror("Memory");
exit(1);
}
string1[j-i] ='\0';
strncpy(string1, (data +i), (j-i));
/* Locate the right side string of transformation */
j++;
i = j;
while(data[j] != ':') j++;
string2 = (char *)realloc(NULL, (j-i+1));
if(string2 == NULL) {
perror("Memory");
exit(1);
}
string2[(j-i)]='\0';
strncpy(string2, (data +i), (j-i));
j++;
i = j;
while(data[j] != '\n' && data[j] != '\r' && j < datalen) j++;
/* Find the weight of transformation */
v = findValue(data, i, j);
/* According to the type of transformation, add into suitable trie */
if(strlen(string1) == 0) {
/* Add to add-operations trie */
w2 = mbstowcs(NULL, string2, 0);
wstr2 = (wchar_t *)localeToWchar(string2);
if(caseInsensitiveMode)
wstr2 = makeStringToIgnoreCase(wstr2, w2);
addToARTrie(addT, wstr2, w2, v);
free(wstr2);
free(string2);
} else if(strlen(string2) == 0) {
/* Add to remove-operations trie */
w1 = mbstowcs(NULL, string1, 0);
wstr1 = (wchar_t *)localeToWchar(string1);
if(caseInsensitiveMode)
wstr1 = makeStringToIgnoreCase(wstr1, w1);
addToARTrie(remT, wstr1, w1, v);
} else {
/* Add to replace-operations trie */
wstr2 = (wchar_t *)localeToWchar(string2);
wstr1 = (wchar_t *)localeToWchar(string1);
w1 = mbstowcs(NULL, string1, 0);
w2 = mbstowcs(NULL, string2, 0);
if(caseInsensitiveMode) {
wstr1 = makeStringToIgnoreCase(wstr1, w1);
wstr2 = makeStringToIgnoreCase(wstr2, w2);
}
addToTrie(wstr1, w1, wstr2,v);
free(string2);
free(wstr1);
}
if(data[j] == '\r') /* In case we are under Windows */
j = j + 2;
else
j = j+1;
i = j;
}
}
free(string1);
/* Release the memory under data. */
munmap(data, strlen(data));
return 0;
}