bool SpellChecker::checkWordSpelling(const QString &word)
{
QString tmp = word;
tmp = tmp.remove(QString::fromUtf8("»"));
tmp = tmp.remove(QString::fromUtf8("«"));
QByteArray ba = tmp.toUtf8();
return (aspell_speller_check(spell_checker1, ba.data(), ba.size()) != 0) ||
(aspell_speller_check((spell_checker2 != NULL ? spell_checker2 : spell_checker1), ba.data(), ba.size()) != 0);
}
void permute(char* a, int l, int r)
{
int i;
if (l == r) {
int correct = aspell_speller_check(spell_checker, a, -1);
if (correct) {
printf("%s\n", a);
}
else if (!correct) {
}
else {
fprintf(stderr, "Error: %s\n",
aspell_speller_error_message(spell_checker));
}
}
else {
for (i = l; i <= r; i++) {
swap((a + l), (a + i));
permute(a, l + 1, r);
swap((a + l), (a + i)); //backtrack
}
}
}
int
weechat_aspell_check_word (struct t_aspell_speller_buffer *speller_buffer,
const char *word)
{
int i;
/* word too small? then do not check word */
if ((weechat_config_integer (weechat_aspell_config_check_word_min_length) > 0)
&& ((int)strlen (word) < weechat_config_integer (weechat_aspell_config_check_word_min_length)))
return 1;
/* word is a number? then do not check word */
if (weechat_aspell_string_is_simili_number (word))
return 1;
/* check word with all spellers (order is important) */
if (speller_buffer->spellers)
{
for (i = 0; speller_buffer->spellers[i]; i++)
{
#ifdef USE_ENCHANT
if (enchant_dict_check (speller_buffer->spellers[i], word, strlen (word)) == 0)
#else
if (aspell_speller_check (speller_buffer->spellers[i], word, -1) == 1)
#endif /* USE_ENCHANT */
return 1;
}
}
/* misspelled word! */
return 0;
}
int
spelling_correct(char *word, struct spelling *s)
{
int ret;
char *conv_word;
char *p;
size_t conv_in, conv_out;
conv_word = spelling_conv(s->conv, word);
if (isdigit(conv_word[0])) {
int i, digit;
digit = 1;
for (i = 1; conv_word[i] != '\0'; i++) {
if (!isdigit(conv_word[i])) {
digit = 0;
break;
}
}
if (digit)
return 1;
}
ret = aspell_speller_check(s->speller, conv_word, -1);
free(conv_word);
return ret;
}
bool KAspellChecker::checkWord(const QString &word)
{
if (!m_speller)
return true;
return aspell_speller_check(m_speller, word.toLocal8Bit().data(), -1);
}
bool CheckerString::next_misspelling()
{
if (off_end(cur_line_)) return false;
if (has_repl_) {
has_repl_ = false;
CharVector word;
bool correct = false;
// FIXME: This is a hack to avoid trying to check a word with a space
// in it. The correct action is to reparse to string and
// check each word individually. However doing so involves
// an API enhancement in Checker.
for (int i = 0; i != real_word_size_; ++i) {
if (asc_isspace(*(real_word_begin_ + i)))
correct = true;
}
if (!correct)
correct = aspell_speller_check(speller_, &*real_word_begin_, real_word_size_);
diff_ += real_word_size_ - tok_.len;
tok_.len = real_word_size_;
if (!correct)
return true;
}
while ((tok_ = checker_->next_misspelling()).len == 0) {
next_line(cur_line_);
diff_ = 0;
if (off_end(cur_line_)) return false;
checker_->process(cur_line_->real.data(), cur_line_->real.size());
}
real_word_begin_ = cur_line_->real.begin() + tok_.offset + diff_;
real_word_size_ = tok_.len;
fix_display_str();
return true;
}
bool ASpellChecker::isCorrect(const QString& word)
{
if(speller_) {
int correct = aspell_speller_check(speller_, word.toUtf8().constData(), -1);
return (correct != 0);
}
return true;
}
bool SpellCheck::ok(const QString &word){
if (!spell_checker || word.isEmpty())
return true;
int correct = aspell_speller_check(spell_checker, word.toAscii().constData(), -1);
return (correct != 0);
}
//__________________________________________________________________________
bool Speller::Aspell::Suggest::checkWord(const std::string& word)
{
bool status = true;
if( aspell_speller_check( fspeller, word.c_str(), -1 ) == 0 )
{
status = false;
}
return status;
}
static void
check_word(GtkSpell *spell, GtkTextBuffer *buffer,
GtkTextIter *start, GtkTextIter *end) {
char *text;
text = gtk_text_buffer_get_text(buffer, start, end, FALSE);
if (debug) g_print("checking: %s\n", text);
if (aspell_speller_check(spell->speller, text, -1) == FALSE)
gtk_text_buffer_apply_tag(buffer, spell->tag_highlight, start, end);
g_free(text);
}
/**
* Ask the spell-checker if the spelling looks good.
* Return true if the spelling is good, else false.
*/
Boolean spellcheck_test(void * chk, const char * word)
{
int val = 0;
struct linkgrammar_aspell *aspell = (struct linkgrammar_aspell *)chk;
if (aspell && aspell->speller) {
/* this can return -1 on failure */
val = aspell_speller_check(aspell->speller, word, -1);
}
return (val == 1) ? TRUE : FALSE;
}
bool ASpellChecker::isCorrect(const QString& word)
{
if(spellers_.isEmpty())
return true;
foreach(AspellSpeller* speller, spellers_) {
if (aspell_speller_check(speller, word.toUtf8().constData(), -1) != 0)
return true;
}
return false;
}
/**
* Check a given word for correctness.
* @param word the word to check
* @return true if the word is correct, otherwise false.
*/
static VALUE aspell_check(VALUE self, VALUE word) {
AspellSpeller *speller = get_speller(self);
VALUE result = Qfalse;
int code = aspell_speller_check(speller, STR2CSTR(word), -1);
if (code == 1)
result = Qtrue;
else if (code == 0)
result = Qfalse;
else
rb_raise( cAspellError, "%s", aspell_speller_error_message(speller));
return result;
}
int
weechat_aspell_check_word (struct t_gui_buffer *buffer,
struct t_aspell_speller_buffer *speller_buffer,
const char *word)
{
const char *buffer_type, *buffer_nick, *buffer_channel;
int i;
/* word too small? then do not check word */
if ((weechat_config_integer (weechat_aspell_config_check_word_min_length) > 0)
&& ((int)strlen (word) < weechat_config_integer (weechat_aspell_config_check_word_min_length)))
return 1;
/* word is a number? then do not check word */
if (weechat_aspell_string_is_simili_number (word))
return 1;
/* word is a nick of nicklist on this buffer? then do not check word */
if (weechat_nicklist_search_nick (buffer, NULL, word))
return 1;
/* for "private" buffers, ignore self and remote nicks */
buffer_type = weechat_buffer_get_string (buffer, "localvar_type");
if (buffer_type && (strcmp (buffer_type, "private") == 0))
{
/* check self nick */
buffer_nick = weechat_buffer_get_string (buffer, "localvar_nick");
if (buffer_nick && (weechat_strcasecmp (buffer_nick, word) == 0))
return 1;
/* check remote nick */
buffer_channel = weechat_buffer_get_string (buffer, "localvar_channel");
if (buffer_channel && (weechat_strcasecmp (buffer_channel, word) == 0))
return 1;
}
/* check word with all spellers for this buffer (order is important) */
if (speller_buffer->spellers)
{
for (i = 0; speller_buffer->spellers[i]; i++)
{
#ifdef USE_ENCHANT
if (enchant_dict_check (speller_buffer->spellers[i], word, strlen (word)) == 0)
#else
if (aspell_speller_check (speller_buffer->spellers[i], word, -1) == 1)
#endif
return 1;
}
}
/* misspelled word! */
return 0;
}
QStringList SpellChecker::suggestions()
{
QStringList sl;
if ((spell_checker1 == 0) || (spell_checker2 == 0))
return sl;
QTextCursor cursor = m_textEdit->textCursor();
cursor.select(QTextCursor::WordUnderCursor);
QString word = cursor.selectedText();
QByteArray ba = word.toUtf8();
if ((aspell_speller_check(spell_checker2, ba.data(), ba.size()) != 0)||(aspell_speller_check(spell_checker1, ba.data(), ba.size()) != 0))
return sl;
const struct AspellWordList *awl = aspell_speller_suggest(spell_checker1, ba.data(), ba.size());
if (aspell_word_list_size(awl) > 0) {
struct AspellStringEnumeration *ase = aspell_word_list_elements(awl);
int i = 0;
while ((!aspell_string_enumeration_at_end(ase))&&(i < 10)) {
const char *text = aspell_string_enumeration_next(ase);
sl << QString::fromUtf8(text);
i++;
}
delete_aspell_string_enumeration(ase);
}
return sl;
}
bool SpellChecker::check(QString word)
{
if(word.isEmpty()) return true;
AspellSpeller * checker;
if(word[0].toLower()>=QChar('a') && word[0].toLower()<=QChar('z')){
checker = m_spell_checker_en;
}
else{
checker = m_spell_checker_ru;
}
if(!checker) return true;
return (bool)aspell_speller_check(checker, m_codec->fromUnicode(word).data(), -1);
}
int
weechat_aspell_check_word (struct t_gui_buffer *buffer, const char *word)
{
struct t_aspell_speller *ptr_speller;
int rc;
rc = 0;
/* word too small? then do not check word */
if ((weechat_config_integer (weechat_aspell_config_check_word_min_length) > 0)
&& ((int)strlen (word) < weechat_config_integer (weechat_aspell_config_check_word_min_length)))
rc = 1;
else
{
/* word is URL? then do not check word */
if (weechat_aspell_string_is_url (word))
rc = 1;
else
{
/* word is a number? then do not check word */
if (weechat_aspell_string_is_simili_number (word))
rc = 1;
else
{
/* word is a nick of nicklist on this buffer? then do not check word */
if (weechat_nicklist_search_nick (buffer, NULL, word))
rc = 1;
else
{
/* check word with all spellers for this buffer (order is important) */
for (ptr_speller = weechat_aspell_spellers; ptr_speller;
ptr_speller = ptr_speller->next_speller)
{
if (aspell_speller_check (ptr_speller->speller, word, -1) == 1)
{
rc = 1;
break;
}
}
}
}
}
}
return rc;
}
void checkTheWord(char* word,int checkDoc)
{
#if 1
int correct;
AspellWordList* suggestions;
AspellStringEnumeration* elements;
const char* suggestedword;
int wordcnt=0;
char* wordlist[100];
char* labeltext[512];
correct=aspell_speller_check(spellChecker,word,-1);
if(!correct)
{
badWord=word;
cancelCheck=false;
if(spellCheckWord==NULL)
buildWordCheck(checkDoc);
else
{
for(int j=0; j<numWords; j++)
gtk_combo_box_text_remove((GtkComboBoxText*)wordListDropbox,0);
sprintf((char*)&labeltext,"Change <i><b>%s</b></i> to: ",badWord);
gtk_label_set_text((GtkLabel*)badWordLabel,(char*)&labeltext);
gtk_label_set_use_markup((GtkLabel*)badWordLabel,true);
}
suggestions=(AspellWordList*)aspell_speller_suggest(spellChecker,word,-1);
elements=aspell_word_list_elements(suggestions);
while((suggestedword=aspell_string_enumeration_next(elements))!=NULL)
{
wordlist[wordcnt]=strdup(suggestedword);
gtk_combo_box_text_append_text((GtkComboBoxText*)wordListDropbox,wordlist[wordcnt]);
wordcnt++;
}
numWords=wordcnt;
delete_aspell_string_enumeration(elements);
gtk_combo_box_set_active((GtkComboBox*)wordListDropbox,0);
gtk_widget_show_all(spellCheckWord);
gtk_dialog_run((GtkDialog *)spellCheckWord);
}
#endif
}
bool SpellChecker::checkWord(QString word)
{
bool isWordValid = checkers.size() == 0;
if (word.indexOf(QRegExp("\\D")) == -1)
isWordValid = true;
else
{
for (Checkers::Iterator it = checkers.begin(); it != checkers.end(); it++)
{
if (aspell_speller_check(it.value(), word.toUtf8(), -1))
{
isWordValid = true;
break;
}
}
}
return isWordValid;
}
/*
* Class: calliope_AeseSpeller
* Method: hasWord
* Signature: (Ljava/lang/String;Ljava/lang/String;)Z
*/
JNIEXPORT jboolean JNICALL Java_calliope_AeseSpeller_hasWord
(JNIEnv *env, jobject obj, jstring jword, jstring lang)
{
int correct = 0;
jboolean copied1, copied2;
checker *c = checkers;
const char *word = load_string( env, jword, &copied1 );
const char *language = load_string( env, lang, &copied2 );
while ( c != NULL )
if ( strcmp(language,c->lang)!=0 )
c = c->next;
else
break;
if ( c == NULL )
{
c = checker_create( language );
if ( c != NULL )
{
if ( checkers == NULL )
checkers = c;
else
{
checker *temp = checkers;
while ( temp->next != NULL )
temp = temp->next;
temp->next = c;
}
}
}
if ( c != NULL )
{
correct = aspell_speller_check(c->spell_checker, word,
strlen(word));
}
else
fprintf(stderr,"checker: no dict for language %s\n",language);
if ( copied1 )
unload_string( env, jword, word, copied1 );
if ( copied2 )
unload_string( env, lang, language, copied2 );
return correct;
}
/* method:check ***************************************************************/
static PyObject* m_check(PyObject* self, PyObject* args) {
char* word;
int length;
if (!PyArg_ParseTuple(args, "s#", &word, &length)) {
PyErr_SetString(PyExc_TypeError, "a string is required");
return NULL;
}
if (!length)
return Py_BuildValue("i", 1);
switch (aspell_speller_check(Speller(self), word, length)) {
case 0:
return Py_BuildValue("i", 0);
case 1:
return Py_BuildValue("i", 1);
default:
PyErr_SetString(_AspellSpellerException, aspell_speller_error_message(Speller(self)));
return NULL;
}
}
int main(int argc, const char *argv[])
{
AspellCanHaveError * ret;
AspellSpeller * speller;
int have;
char word[81];
char * p;
char * word_end;
AspellConfig * config;
if (argc < 2) {
printf("Usage: %s <language> [<size>|- [[<jargon>|- [<encoding>]]]\n", argv[0]);
return 1;
}
config = new_aspell_config();
aspell_config_replace(config, "lang", argv[1]);
if (argc >= 3 && argv[2][0] != '-' && argv[2][1] != '\0')
aspell_config_replace(config, "size", argv[2]);
if (argc >= 4 && argv[3][0] != '-')
aspell_config_replace(config, "jargon", argv[3]);
if (argc >= 5 && argv[4][0] != '-')
aspell_config_replace(config, "encoding", argv[4]);
ret = new_aspell_speller(config);
delete_aspell_config(config);
if (aspell_error(ret) != 0) {
printf("Error: %s\n",aspell_error_message(ret));
delete_aspell_can_have_error(ret);
return 2;
}
speller = to_aspell_speller(ret);
config = aspell_speller_config(speller);
fputs("Using: ", stdout);
fputs(aspell_config_retrieve(config, "lang"), stdout);
fputs("-", stdout);
fputs(aspell_config_retrieve(config, "jargon"), stdout);
fputs("-", stdout);
fputs(aspell_config_retrieve(config, "size"), stdout);
fputs("-", stdout);
fputs(aspell_config_retrieve(config, "module"), stdout);
fputs("\n\n", stdout);
puts("Type \"h\" for help.\n");
while (fgets(word, 80, stdin) != 0) {
/* remove trailing spaces */
word_end = strchr(word, '\0') - 1;
while (word_end != word && (*word_end == '\n' || *word_end == ' '))
--word_end;
++word_end;
*word_end = '\0';
putchar('\n');
switch (word[0]) {
case '\0':
break;
case 'h':
puts(
"Usage: \n"
" h(elp) help\n"
" c <word> check if a word is the correct spelling\n"
" s <word> print out a list of suggestions for a word\n"
" a <word> add a word to the personal word list\n"
" i <word> ignore a word for the rest of the session\n"
" d <file> spell checks a document\n"
" p dumps the personal word list\n"
" P dumps the session word list\n"
" m dumps the main word list\n"
" o <option> <value> sets a config option\n"
" r <option> retrieves a config option\n"
" l <option> retrieves a config option as a list\n"
" S saves all word lists\n"
" C clear the curent sesstion word list\n"
" x quite\n" );
break;
case 'p':
print_word_list(speller,
aspell_speller_personal_word_list(speller), '\n');
break;
case 'P':
print_word_list(speller,
aspell_speller_session_word_list(speller), '\n');
break;
case 'm':
print_word_list(speller,
aspell_speller_main_word_list(speller), '\n');
break;
case 'S':
aspell_speller_save_all_word_lists(speller);
check_for_error(speller);
break;
case 'C':
aspell_speller_clear_session(speller);
check_for_error(speller);
break;
case 'x':
goto END;
case 'c':
if (strlen(word) < 3) {
printf("Usage: %c <word>\n", word[0]);
} else {
have = aspell_speller_check(speller, word + 2, -1);
if (have == 1)
puts("correct");
else if (have == 0)
puts("incorrect");
else
printf("Error: %s\n", aspell_speller_error_message(speller));
}
break;
case 's':
if (strlen(word) < 3) {
printf("Usage: %c <word>\n", word[0]);
} else {
print_word_list(speller,
aspell_speller_suggest(speller, word + 2, -1), '\n');
}
break;
case 'a':
if (strlen(word) < 3) {
printf("Usage: %c <word>\n", word[0]);
} else {
aspell_speller_add_to_personal(speller, word + 2, -1);
check_for_error(speller);
}
break;
case 'i':
if (strlen(word) < 3) {
printf("Usage: %c <word>\n", word[0]);
} else {
aspell_speller_add_to_session(speller, word + 2, -1);
check_for_error(speller);
}
break;
case 'o':
word[80] = '\0'; /* to make sure strchr doesn't run off end of string */
p = strchr(word + 3, ' ');
if (strlen(word) < 3 || p == 0) {
printf("Usage: %c <option> <value>\n", word[0]);
} else {
*p = '\0';
++p;
aspell_config_replace(config, word + 2, p);
check_for_config_error(config);
}
break;
case 'r':
if (strlen(word) < 3) {
printf("Usage: %c <option>\n", word[0]);
} else {
const char * val = aspell_config_retrieve(config, word + 2);
check_for_config_error(config);
if (val)
printf("%s = \"%s\"\n", word + 2, val);
}
break;
case 'l':
if (strlen(word) < 3) {
printf("Usage: %c <option>\n", word[0]);
} else {
AspellStringList * lst = new_aspell_string_list();
AspellMutableContainer * lst0
= aspell_string_list_to_mutable_container(lst);
AspellStringEnumeration * els;
const char * val;
aspell_config_retrieve_list(config, word + 2, lst0);
check_for_config_error(config);
els = aspell_string_list_elements(lst);
printf("%s:\n", word + 2);
while ( (val = aspell_string_enumeration_next(els)) != 0)
printf(" %s\n", val);
delete_aspell_string_enumeration(els);
delete_aspell_string_list(lst);
}
break;
case 'd':
if (strlen(word) < 3) {
printf("Usage: %c <file>\n", word[0]);
} else {
check_document(speller, word + 2);
printf("\n");
}
break;
default:
printf("Unknown Command: %s\n", word);
}
putchar('\n');
}
END:
delete_aspell_speller(speller);
return 0;
}
/**
* Does the current spell-checker have this word?
* @param u the userdata object
* @param word the word to lookup
* @return 1 if it was there else 0
*/
int userdata_has_word( userdata *u, XML_Char *word )
{
int correct = aspell_speller_check(u->spell_checker, word,
strlen((char*)word));
return correct;
}
bool AspellAdapterImpl::isWordCorrect(const std::string & word) {
int correct = aspell_speller_check(aspellSpeller_, word.c_str(), -1);
return (bool) correct;
}
bool Speller::add(const char *word)
{
if (speller == NULL)
return false;
return aspell_speller_check(speller, word, strlen(word)) != 0;
}
int Speller::check(const char *word)
{
if (speller == NULL)
return -1;
return aspell_speller_check(speller, word, strlen(word));
}
// runs in O(t^2) time where t is the number of tokens in the input corpus
// We consider maxK to be fairly constant
void rawr::compile(int maxK)
{
_maxK = maxK;
std::vector<std::vector<token_id>> tokens;
std::set<std::string> thashtags;
std::set<std::string> fv_emoticons;
std::ifstream fvefile("emoticons.txt");
if (fvefile)
{
std::string line;
while (getline(fvefile, line))
{
fv_emoticons.insert(line);
emoticons.forms.add(line);
}
}
fvefile.close();
std::map<std::string, std::string> canonical_form;
AspellConfig* spell_config = new_aspell_config();
AspellCanHaveError* possible_err = new_aspell_speller(spell_config);
if (aspell_error_number(possible_err) != 0)
{
std::cout << "aspell error: " << aspell_error_message(possible_err) << std::endl;
exit(1);
}
AspellSpeller* spell_checker = to_aspell_speller(possible_err);
std::cout << "Reading emojis..." << std::endl;
prefix_search emojis;
std::ifstream emoji_file("emojis.txt");
if (emoji_file)
{
while (!emoji_file.eof())
{
std::string rawmojis;
getline(emoji_file, rawmojis);
if (rawmojis.back() == '\r')
{
rawmojis.pop_back();
}
emojis.add(rawmojis);
}
emoji_file.close();
}
std::cout << "Tokenizing corpus... 0%" << std::flush;
int len = 0;
for (auto c : _corpora)
{
len += c.length();
}
int startper = 0;
int per = 0;
int perprime = 0;
std::cout.fill(' ');
for (int i = 0; i < _corpora.size(); i++)
{
size_t start = 0;
int end = 0;
std::vector<token_id> tkcor;
while (end != std::string::npos)
{
perprime = (startper + end) * 100 / len;
if (perprime != per)
{
per = perprime;
std::cout << "\b\b\b\b" << std::right;
std::cout.width(3);
std::cout << per << "%" << std::flush;
}
end = _corpora[i].find_first_of(" \n", start);
bool emoji = false;
std::string te = _corpora[i].substr(start, (end == std::string::npos) ? std::string::npos : end - start + 1);
std::string t = "";
if (te.compare("") && te.compare(".") && te.compare(" "))
{
if (te.back() == ' ')
{
te.pop_back();
}
// Extract strings of emojis into their own tokens even if they're not space delimited
int m = emojis.match(te);
emoji = m > 0;
if (m == 0) m = 1;
t = te.substr(0,m);
te = te.substr(m);
while (!te.empty())
{
m = emojis.match(te);
if (emoji == (m > 0))
{
if (m == 0) m = 1;
t += te.substr(0,m);
te = te.substr(m);
} else {
end = start + t.length() - 1;
break;
}
}
std::string tc(t);
std::transform(tc.begin(), tc.end(), tc.begin(), ::tolower);
int pst = tc.find_first_not_of("\"([*");
int dst = tc.find_last_not_of("\")]*.,?!\n;:");
std::string canonical("");
if ((pst != std::string::npos) && (dst != std::string::npos))
{
canonical = std::string(tc, pst, dst - pst + 1);
}
word& w = ([&] () -> word& {
// Hashtag freevar
if (canonical[0] == '#')
{
thashtags.insert(canonical);
return hashtags;
}
// Emoticon freevar
if (emoji)
{
emoticons.forms.add(canonical);
return emoticons;
}
if ((pst != std::string::npos) && (dst != std::string::npos))
{
std::string emoticon_canon(t, pst, t.find_last_not_of("\"]*\n.,?!") - pst + 1);
if (fv_emoticons.count(emoticon_canon) == 1)
{
emoticons.forms.add(emoticon_canon);
return emoticons;
}
}
// Basically any other word
if (canonical_form.count(canonical) == 0)
{
if (
// Legacy freevars should be distinct from tokens containing similar words
(canonical.find("$name$") != std::string::npos)
// Words with no letters will be mangled by the spell checker
|| (canonical.find_first_of("ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz") == std::string::npos)
)
{
canonical_form[canonical] = canonical;
words.emplace(canonical, canonical);
} else {
int correct = aspell_speller_check(spell_checker, canonical.c_str(), canonical.size());
if (correct)
{
words.emplace(canonical, canonical);
canonical_form[canonical] = canonical;
} else {
const AspellWordList* suggestions = aspell_speller_suggest(spell_checker, canonical.c_str(), canonical.size());
AspellStringEnumeration* elements = aspell_word_list_elements(suggestions);
const char* replacement = aspell_string_enumeration_next(elements);
if (replacement != NULL)
{
std::string sugrep(replacement);
canonical_form[canonical] = sugrep;
if (words.count(sugrep) == 0)
{
words.emplace(sugrep, sugrep);
}
} else {
words.emplace(canonical, canonical);
canonical_form[canonical] = canonical;
}
delete_aspell_string_enumeration(elements);
}
}
}
word& tw = words.at(canonical_form.at(canonical));
tw.forms.add(canonical);
return tw;
})();
token tk(w);
tk.raw = t;
for (char c : t)
{
if (c == '*')
{
tk.delimiters[{parentype::asterisk, doublestatus::opening}]++;
} else if (c == '[')
{
tk.delimiters[{parentype::square_bracket, doublestatus::opening}]++;
} else if (c == '(')
{
tk.delimiters[{parentype::paren, doublestatus::opening}]++;
} else if (c == '"')
{
tk.delimiters[{parentype::quote, doublestatus::opening}]++;
} else {
break;
}
}
int backtrack = t.find_last_not_of(".,?!])\"*\n;:") + 1;
if (backtrack != t.length())
{
std::string ending = t.substr(backtrack);
std::string suffix;
bool newline = false;
bool terminating = false;
for (char c : ending)
{
if ((c == '.') || (c == ',') || (c == '?') || (c == '!') || (c == ';') || (c == ':'))
{
suffix += c;
terminating = true;
continue;
} else if (c == '\n')
{
newline = true;
terminating = true;
continue;
}
parentype pt = ([&] {
switch (c)
{
case ']': return parentype::square_bracket;
case ')': return parentype::paren;
case '*': return parentype::asterisk;
case '"': return parentype::quote;
}
})();
if (tk.delimiters[{pt, doublestatus::opening}] > 0)
{
tk.delimiters[{pt, doublestatus::opening}]--;
tk.delimiters[{pt, doublestatus::both}]++;
} else {
tk.delimiters[{pt, doublestatus::closing}]++;
}
}
if (terminating)
{
if ((suffix == ",") && (!newline))
{
tk.suffix = suffixtype::comma;
} else {
tk.suffix = suffixtype::terminating;
if (!newline)
{
w.terms.add({suffix, false});
} else {
w.terms.add({".", false});
}
}
}
}
tkcor.push_back(_tokenstore.add(tk));
}
start = ((end > (std::string::npos - 1) ) ? std::string::npos : end + 1);
}
tokens.push_back(tkcor);
startper += _corpora[i].length();
}
std::cout << "\b\b\b\b100%" << std::endl;
delete_aspell_speller(spell_checker);
delete_aspell_config(spell_config);
std::cout << canonical_form.size() << " distinct forms" << std::endl;
std::cout << words.size() << " distinct words" << std::endl;
// Time to condense the distribution stuff for the words
std::cout << "Compiling token histograms..." << std::endl;
for (auto& it : words)
{
it.second.forms.compile();
it.second.terms.compile();
}
// Hashtag freevar is not frequency distributed
for (auto& it : thashtags)
{
hashtags.forms.add(it);
}
hashtags.forms.compile();
hashtags.terms.compile();
// Compile other freevars
emoticons.forms.compile();
emoticons.terms.compile();
// Compile the interned tokens.
_tokenstore.compile();
// kgram distribution
std::cout << "Creating markov chain... 0%" << std::flush;
std::map<kgram, std::map<token_id, token_data> > tstats;
len = 0;
for (auto c : tokens)
{
len += (maxK-1) * c.size();
}
startper = 0;
per = 0;
perprime = 0;
int corpid = 0;
for (auto corpus : tokens)
{
for (int k=0; k<maxK && k<corpus.size(); k++)
{
// The zero'th token should be a terminator.
token_id fid = corpus[k];
const token& f = _tokenstore.get(fid);
kgram term_prefix(corpus.begin(), corpus.begin()+k);
term_prefix.push_front(wildcardQuery);
if (tstats[term_prefix].count(fid) == 0)
{
tstats[term_prefix].emplace(fid, fid);
}
token_data& td2 = tstats[term_prefix].at(fid);
td2.all++;
td2.corpora.insert(corpid);
if (std::find_if(f.raw.begin(), f.raw.end(), ::islower) == f.raw.end())
{
td2.uppercase++;
} else if (isupper(f.raw[0]))
{
td2.titlecase++;
}
}
for (int k=1; k<maxK && k<corpus.size(); k++)
{
for (int i=0; i<(corpus.size() - k); i++)
{
perprime = (startper+i) * 100 / len;
if (perprime != per)
{
per = perprime;
std::cout << "\b\b\b\b" << std::right;
std::cout.width(3);
std::cout << per << "%" << std::flush;
}
kgram prefix(corpus.begin()+i, corpus.begin()+i+k);
token_id fid = corpus[i+k];
const token& f = _tokenstore.get(fid);
if (tstats[prefix].count(fid) == 0)
{
tstats[prefix].emplace(fid, fid);
}
token_data& td = tstats[prefix].at(fid);
td.all++;
td.corpora.insert(corpid);
if (std::find_if(f.raw.begin(), f.raw.end(), ::islower) == f.raw.end())
{
td.uppercase++;
} else if (isupper(f.raw[0]))
{
td.titlecase++;
}
const token& startTok = _tokenstore.get(std::begin(prefix)->tok);
if (startTok.suffix == suffixtype::terminating)
{
kgram term_prefix(prefix);
term_prefix.pop_front();
term_prefix.push_front(wildcardQuery);
if (tstats[term_prefix].count(fid) == 0)
{
tstats[term_prefix].emplace(fid, fid);
}
token_data& td2 = tstats[term_prefix].at(fid);
td2.all++;
td2.corpora.insert(corpid);
if (std::find_if(f.raw.begin(), f.raw.end(), ::islower) == f.raw.end())
{
td2.uppercase++;
} else if (isupper(f.raw[0]))
{
td2.titlecase++;
}
}
}
startper += corpus.size();
}
corpid++;
}
std::cout << "\b\b\b\b100%" << std::endl;
// Condense the kgram distribution
std::cout << "Compiling kgram distributions... 0%";
len = tstats.size();
per = 0;
perprime = 0;
int indicator = 0;
for (auto& it : tstats)
{
indicator++;
perprime = indicator * 100 / len;
if (per != perprime)
{
per = perprime;
std::cout << "\b\b\b\b" << std::right;
std::cout.width(3);
std::cout << per << "%" << std::flush;
}
kgram klist = it.first;
auto& probtable = it.second;
auto& distribution = _stats[klist];
int max = 0;
for (auto& kt : probtable)
{
max += kt.second.all;
distribution.emplace(max, kt.second);
}
}
std::cout << "\b\b\b\b100%" << std::endl;
_compiled = true;
}
static int get_aspell_hits(const char *word, int len)
{
if (len < 2)
{
log_message(DEBUG, " [-]Skip aspell checking (word is very short)");
return NO_LANGUAGE;
}
AspellConfig *spell_config = new_aspell_config();
for (int lang = 0; lang < xconfig->total_languages; lang++)
{
char *lang_word = (char *) malloc(1 * sizeof(char));
if (lang_word == NULL)
continue;
lang_word[0] = NULLSYM;
for (int i = 0; i < len; i++)
{
KeyCode kc;
int modifier;
main_window->xkeymap->char_to_keycode(main_window->xkeymap, word[i], &kc, &modifier);
char *symbol = keycode_to_symbol(kc, lang, modifier);
if ((symbol == NULL) || (lang_word == NULL))
continue;
lang_word = (char *) realloc(lang_word, (strlen(lang_word) + strlen(symbol) + 1) * sizeof(char));
if (lang_word != NULL)
strcat(lang_word, symbol);
free(symbol);
}
if (lang_word == NULL)
continue;
aspell_config_replace(spell_config, "lang", xconfig->languages[lang].dir);
AspellCanHaveError *possible_err = new_aspell_speller(spell_config);
AspellSpeller *spell_checker = 0;
if (aspell_error_number(possible_err) == 0)
{
spell_checker = to_aspell_speller(possible_err);
int correct = aspell_speller_check(spell_checker, lang_word, strlen(lang_word));
if (correct)
{
log_message(DEBUG, " [+]Found this word in %s aspell dictionary", xconfig->get_lang_name(xconfig, lang));
delete_aspell_speller(spell_checker);
free(lang_word);
return lang;
}
}
else
{
log_message(DEBUG, " [!]Error aspell checking for %s aspell dictionary", xconfig->get_lang_name(xconfig, lang));
}
delete_aspell_speller(spell_checker);
free(lang_word);
}
log_message(DEBUG, " [-]This word has no hits for all aspell dictionaries");
return NO_LANGUAGE;
}
