gint search4string(gint srchtype, GjitenDicfile *dicfile, gchar *srchstrg, guint32 *res_index, gint *hit_pos, gint *res_len, gchar *res_str) {
gint search_result;
gchar *linestart, *lineend;
gint copySize = 1023;
static gchar *linsrchptr;
if (dicfile->status == DICFILE_NOT_INITIALIZED) {
if (dicfile_init(dicfile) == FALSE) return SRCH_FAIL;
}
if (dicfile->status != DICFILE_OK) return SRCH_FAIL;
if ((dicfile != conf.mmaped_dicfile) && (conf.mmaped_dicfile != NULL)) {
//free mem of previously used dicfile
munmap(conf.mmaped_dicfile->mem, conf.mmaped_dicfile->size);
conf.mmaped_dicfile->mem = NULL;
conf.mmaped_dicfile = NULL;
}
if (conf.mmaped_dicfile == NULL) {
//mmap dicfile into memory
conf.mmaped_dicfile = dicfile;
dicfile->mem = (gchar *) mmap(NULL, dicfile->size, PROT_READ, MAP_SHARED, dicfile->file, 0);
if (dicfile->mem == NULL) gjiten_abort_with_msg("mmap() failed\n");
conf.mmaped_dicfile = dicfile;
}
if (srchtype == SRCH_START) {
linsrchptr = dicfile->mem;
}
bad_hit:
search_result = SRCH_FAIL; // assume search fails
linsrchptr = strstr(linsrchptr, srchstrg);
if (linsrchptr != NULL) { // if we have a match
linestart = linsrchptr;
while ((*linestart != '\n') && (linestart != dicfile->mem)) linestart--; // find beginning of line
if (linestart == dicfile->mem) {
if ((isKanjiChar(g_utf8_get_char(linestart)) == FALSE) && (isKanaChar(g_utf8_get_char(linestart)) == FALSE)) {
linsrchptr++;
goto bad_hit;
}
}
linestart++;
lineend = linestart;
*hit_pos = linsrchptr - linestart;
while (*lineend != '\n') { // find end of line
lineend++;
if (lineend >= dicfile->mem + dicfile->size) {
printf("weird.\n");
break;
}
}
linsrchptr++;
if ((lineend - linestart + 1) < 1023) copySize = lineend - linestart + 1;
else copySize = 1023;
strncpy(res_str, linestart, copySize);
res_str[copySize] = 0;
*res_index = (guint32)linestart;
search_result = SRCH_OK; // search succeeded
}
return search_result;
}
GList* load_radkfile(GHashTable **pp_rad_info_hash,
GHashTable **pp_kanji_info_hash,
GList *rad_info_list) {
int error = FALSE;
gint rad_cnt = 0;
gchar *radkfile_ptr;
gchar *radkfile_end;
RadInfo *rad_info = NULL;
KanjiInfo *kanji_info;
gchar *radkfile = NULL;
GHashTable *rad_info_hash = *pp_rad_info_hash;
GHashTable *kanji_info_hash = *pp_kanji_info_hash;
radkfile = read_file(RADKFILE_NAME);
if (radkfile == NULL) gjiten_abort_with_msg("failed to read radkfile %s\n", RADKFILE_NAME);
if (error == TRUE) {
gjiten_print_error("Error opening %s.\n " \
"Check your preferences or read the documentation!",
RADKFILE_NAME);
return;
}
radkfile_end = radkfile + strlen(radkfile); //FIXME: lseek
radkfile_ptr = radkfile;
//parse the content of the file
while((radkfile_ptr < radkfile_end) && (radkfile_ptr != NULL)) {
//if comment (first char on this line is #), skip this line
if (*radkfile_ptr == '#') {
radkfile_ptr = get_eof_line(radkfile_ptr, radkfile_end);
continue;
}
//if radical info line (first char on this line is $)
if (*radkfile_ptr == '$') {
rad_cnt++; //Increase number of radicals found
radkfile_ptr = g_utf8_next_char(radkfile_ptr);
//move the pointer forward until the character is wide (kanji)
while (g_unichar_iswide(g_utf8_get_char(radkfile_ptr)) == FALSE) {
radkfile_ptr = g_utf8_next_char(radkfile_ptr);
}
//new rad_info to be stored in the rad_info_hash and rad_info_list
rad_info = g_new0(RadInfo, 1);
rad_info->kanji_info_list = NULL;
rad_info_list = g_list_prepend(rad_info_list, rad_info);
//store radical character
//the characters in the file are in UTF8 format. We need unicode.
gunichar utf8radical = g_utf8_get_char(radkfile_ptr);
gunichar *p_str_radical = g_new0(gunichar, 1);
g_unichar_to_utf8(utf8radical, (gchar*)p_str_radical);
rad_info->radical = p_str_radical;
//Find stroke number (move until digit detected)
while (g_ascii_isdigit(*radkfile_ptr) == FALSE) {
radkfile_ptr = g_utf8_next_char(radkfile_ptr);
}
//Store the stroke number
rad_info->strokes = atoi(radkfile_ptr);
//insert this radical as key and the info as value
g_hash_table_insert(rad_info_hash, (gpointer)rad_info->radical, rad_info);
//Goto next line
radkfile_ptr = get_eof_line(radkfile_ptr, radkfile_end);
}
else {
//search the kanji to be stored in the list of the kanji key / radical info list
//the kanji are located between radical $ markers and the radical info
while ((*radkfile_ptr != '$') && (radkfile_ptr < radkfile_end)) {
if (*radkfile_ptr == '\n') {
radkfile_ptr++;
continue;
}
gunichar utf8kanji = g_utf8_get_char(radkfile_ptr);
gchar *kanji = g_new0(gchar, sizeof(gunichar));
g_unichar_to_utf8(utf8kanji, kanji);
//search in the kanji infohash if this kanji is alderly present,
//if not, create a new kanji and add it
kanji_info = g_hash_table_lookup(kanji_info_hash, kanji);
if (kanji_info == NULL) {
kanji_info = g_new0(KanjiInfo, 1);
kanji_info->rad_info_list = NULL;
kanji_info->kanji = kanji;
//insert this kanji as a key and the kanji info as value
g_hash_table_insert(kanji_info_hash, (gpointer) kanji, (gpointer) kanji_info);
}
//add the kanji and the radical info in their respective lists
kanji_info->rad_info_list = g_list_prepend(kanji_info->rad_info_list, rad_info);
rad_info->kanji_info_list = g_list_prepend(rad_info->kanji_info_list, kanji_info);
//navigate to next character
radkfile_ptr = g_utf8_next_char(radkfile_ptr);
}
}
}
return rad_info_list;
}