void build_dict()
{
int edges_count = 0;
for (auto& l : words)
for (auto& r : words)
{
if (l == r)
continue;
auto& lw = dict[l->word_id];
auto& rw = dict[r->word_id];
if (is_similar(lw, rw)) {
edges_count++;
l->similar.insert(r);
}
}
std::cout << "edges count: " << edges_count << std::endl;
//for (auto& w : words)
//{
// std::cout << "item: " << w->word << std::endl;
// for (auto& s : w->similar)
// std::cout << "\t -> " << s->word << std::endl;
//}
}
// A value of type T1 can be converted to a type T2 if they are
// similar and... something about the cv-signature.
//
// Note that the top-level cv-qualifiers can be removed by this
// conversion since it applies to values (i.e., copies).
Expr&
convert_qualifier(Expr& e, Type& t)
{
if (is_similar(e.type(), t)) {
Qualifier_list sa = get_qualification_signature(e.type());
Qualifier_list sb = get_qualification_signature(t);
if (can_convert_signature(sa, sb))
return *new Qualification_conv(t, e);
}
return e;
}
void
mysql__add_row_to_model(MYSQL_ROW row, unsigned long* lengths)
{
GdkPixbuf* iconbuf = NULL;
char length[64];
char keywords[256];
char name[256];
float samplerate; char samplerate_s[32];
unsigned channels, colour;
gboolean online = FALSE;
GtkTreeIter iter;
//db__iter_to_result(result);
//deserialise the pixbuf field:
GdkPixdata pixdata;
GdkPixbuf* pixbuf = NULL;
if(row[MYSQL_PIXBUF]){
if(gdk_pixdata_deserialize(&pixdata, lengths[MYSQL_PIXBUF], (guint8*)row[MYSQL_PIXBUF], NULL)){
pixbuf = gdk_pixbuf_from_pixdata(&pixdata, TRUE, NULL);
}
}
format_time(length, row[MYSQL_LENGTH]);
if(row[MYSQL_KEYWORDS]) snprintf(keywords, 256, "%s", row[MYSQL_KEYWORDS]); else keywords[0] = 0;
if(!row[MYSQL_SAMPLERATE]) samplerate = 0; else samplerate = atoi(row[MYSQL_SAMPLERATE]); samplerate_format(samplerate_s, samplerate);
if(row[7]==NULL) channels = 0; else channels = atoi(row[7]);
if(row[MYSQL_ONLINE]==NULL) online = 0; else online = atoi(row[MYSQL_ONLINE]);
if(row[MYSQL_COLOUR]==NULL) colour = 0; else colour = atoi(row[MYSQL_COLOUR]);
strncpy(name, row[MYSQL_NAME], 255);
//TODO markup should be set in cellrenderer, not model!
#if 0
if(GTK_WIDGET_REALIZED(app.view)){
//check colours dont clash:
long c_num = strtol(app.config.colour[colour], NULL, 16);
dbg(2, "rowcolour=%s", app.config.colour[colour]);
GdkColor row_colour; color_rgba_to_gdk(&row_colour, c_num << 8);
if(is_similar(&row_colour, &app.fg_colour, 0x60)){
snprintf(name, 255, "%s%s%s", "<span foreground=\"blue\">", row[MYSQL_NAME], "</span>");
}
}
#endif
#ifdef USE_AYYI
//is the file loaded in the current Ayyi song?
if(ayyi.got_song){
gchar* fullpath = g_build_filename(row[MYSQL_DIR], name, NULL);
if(pool__file_exists(fullpath)) dbg(0, "exists"); else dbg(0, "doesnt exist");
g_free(fullpath);
}
#endif
//icon (only shown if the sound file is currently available):
if(online){
MIME_type* mime_type = mime_type_lookup(row[MYSQL_MIMETYPE]);
type_to_icon(mime_type);
if ( ! mime_type->image ) dbg(0, "no icon.");
iconbuf = mime_type->image->sm_pixbuf;
} else iconbuf = NULL;
#if 0
//strip the homedir from the dir string:
char* path = strstr(row[MYSQL_DIR], g_get_home_dir());
path = path ? path + strlen(g_get_home_dir()) + 1 : row[MYSQL_DIR];
#endif
gtk_list_store_append(app.store, &iter);
gtk_list_store_set(app.store, &iter, COL_ICON, iconbuf,
#ifdef USE_AYYI
COL_AYYI_ICON, ayyi_icon,
#endif
COL_IDX, atoi(row[MYSQL_ID]), COL_NAME, name, COL_FNAME, row[MYSQL_DIR], COL_KEYWORDS, keywords,
COL_OVERVIEW, pixbuf, COL_LENGTH, length, COL_SAMPLERATE, samplerate_s, COL_CHANNELS, channels,
COL_MIMETYPE, row[MYSQL_MIMETYPE], COL_NOTES, row[MYSQL_NOTES], COL_COLOUR, colour, -1);
if(pixbuf) g_object_unref(pixbuf);
#if 0
if(app.no_gui){
printf(" %s\n", name);
}
#endif
}
/** Function make_cluster()
*
* Given a list of fragments denoted by seeds, make pairwise comparison
* and clustering conforming max_mismatch criteria
*
* Output: clusters in 2d vector format, where each row of the vector
* stores the clustered fragment IDs.
*/
void make_cluster (iivec_t& clusters, const ii64vec_t& list_seeds,
const ivec_t& init_cluster, int max_mismatch, const ivec_t& uf_clst) {
if (list_seeds.size() == 0) {
abording ("DuplRm.cpp -- make_cluster(): SC failed");
}
//--------- union find: (1) initialize the cluster ---------
int sz = init_cluster.size();
bvec_t visited (sz, false);
ivec_t clst (sz);
for (int i = 0; i < sz; ++ i) clst[i] = i;
//--------- pairwise comparison ---------
for (int i = 0; i < sz - 1; ++ i) {
if (visited[i]) continue; // to speed up
int idx_i = init_cluster[i];
for (int j = i + 1; j < sz; ++ j) {
if (visited[j]) continue; // to speed up, avoid of comparison
// if this is already clustered
int idx_j = init_cluster[j];
// check global uf structure according to fragID
int root_uf_i = uf_clsfind ((int) list_seeds[idx_i].back(), uf_clst),
root_uf_j = uf_clsfind ((int) list_seeds[idx_j].back(), uf_clst);
if (root_uf_i != root_uf_j) {
int root_i = uf_find (i, clst),
root_j = uf_find (j, clst);
if (root_i != root_j) {
if (is_similar (list_seeds[idx_i], list_seeds[idx_j],
max_mismatch)) {
clst[root_j] = root_i;
visited[j] = true;
}
}
} // if
} // for (int j = i + 1
} // for (int i = 0
//----- generate final cluster { clusterID --> fragment IDs } ------
std::map<int, ivec_t> clstID_fragIDs;
std::map<int, ivec_t>::iterator it;
for (int i = 0; i < sz; ++ i) {
int idx_i = init_cluster[i];
int fragID = list_seeds[idx_i].back();
int root_i = uf_clsfind (i, clst);
it = clstID_fragIDs.find (root_i);
if (it != clstID_fragIDs.end()) it->second.push_back(fragID);
else clstID_fragIDs[root_i] = ivec_t (1, fragID);
} // for (int i = 0
// go through the map and produce clusters in sorted vector format
for (it = clstID_fragIDs.begin(); it != clstID_fragIDs.end(); ++ it) {
if (it->second.size() > 1) {
std::sort (it->second.begin(), it->second.end());
clusters.push_back(it->second);
}
} // for (it
} // make_cluster
int main()
{
FILE *fp, *fpd, *fpo;
char c;
char buffer[MAXS];
char outbuffer[MAXS2] = "";
char dictionary[DICS];
char* token;
double flag;
int isreplaced=0;
int numcount=1;
/* Open file for both reading and writing */
fprintf(stderr, "Opening files ...\n");
fp = fopen("input.txt", "r+");
fpd = fopen("dictionary.txt", "r+");
fpo = fopen("output.txt", "w+");
/* Write data to the file */
//fwrite(c, strlen(c) + 1, 1, fp);
/* Seek to the beginning of the file */
//fseek(fp, SEEK_SET, 0);
/* Read and display data */
//fread(buffer, sizeof(buffer)+1, 1, fp);
fgets(buffer, sizeof(buffer), fp);
remove_newline_ch(buffer);
printf("Input paragraph;\n%s\n", buffer);
/* while(!feof(fpd))
{
fgets(dictionary, sizeof(dictionary), fpd);
printf("%s\n", dictionary);
}*/
token = strtok(buffer, " ");
while (token) {
remove_punct_and_make_lower_case(token);
// printf("\n\ntoken: %s, lenght: %d\n\n\n", token, strlen(token));
// checking for numbers
if(is_valid_int(token)){
printf("\nNumber %d: %s",numcount, token);
numcount++;
convert_to_words(token);
strcat(outbuffer, token);
strcat(outbuffer, " ");
} else {
if(strlen(token) >= 4){
// checking for spelling
fseek(fpd, SEEK_SET, 0);
isreplaced = 0;
while(!feof(fpd))
{
flag = 0.0;
fgets(dictionary, sizeof(dictionary), fpd);
remove_punct_and_make_lower_case(dictionary);
remove_newline_ch(dictionary);
// printf("%s\n", dictionary);
if(strlen(token) >= 4){
flag = is_similar(token, dictionary);
/*{
strcat(outbuffer, token);
strcat(outbuffer, " ");
}if(flag == 0.0){
strcat(outbuffer, token);
strcat(outbuffer, " ");
} else */
if(isreplaced != 1){
if(flag > 0.6){
strcat(outbuffer, dictionary);
strcat(outbuffer, " ");
isreplaced = 1;
} else if(flag == 1.0){
strcat(outbuffer, token);
strcat(outbuffer, " ");
isreplaced = 1;
}
}
}
}
if(isreplaced == 0){
strcat(outbuffer, token);
strcat(outbuffer, " ");
}
} else {
strcat(outbuffer, token);
strcat(outbuffer, " ");
}
}
token = strtok(NULL, " ");
}
fwrite(outbuffer, strlen(outbuffer) + 1, 1, fpo);
fclose(fp);
fclose(fpd);
fclose(fpo);
return(0);
}
bool operator()(Array_type const& a) { return is_similar(a, cast<Array_type>(b)); }
bool operator()(Sequence_type const& a) { return is_similar(a, cast<Sequence_type>(b)); }
bool operator()(Pointer_type const& a) { return is_similar(a, cast<Pointer_type>(b)); }
bool
is_similar(Sequence_type const& a, Sequence_type const& b)
{
return is_similar(a.type(), b.type());
}
bool
is_similar(Pointer_type const& a, Pointer_type const& b)
{
return is_similar(a.type(), b.type());
}
bool operator()(Tuple_type const& a) { return is_similar(a, cast<Tuple_type>(b)); }
// FIXME: Arrays are only similar when they have the same extent.
bool
is_similar(Array_type const& a, Array_type const& b)
{
return is_similar(a.element_type(), b.element_type());
}
