Class: AhoCorasick::KeywordTree
- Inherits:
-
Object
- Object
- AhoCorasick::KeywordTree
- Defined in:
- ext/ruby-ahocorasick.c
Class Method Summary collapse
-
.from_file ⇒ Object
Creates a new KeywordTree and loads the dictionary from a file.
Instance Method Summary collapse
-
#add_string ⇒ Object
(also: #<<)
Adds a sequence to this KeywordTree.
-
#find_all ⇒ Object
(also: #search)
Search the current tree.
-
#initialize ⇒ Object
constructor
Creates a new KeywordTree.
-
#make ⇒ Object
It freezes the current KeywordTree.
-
#size ⇒ Object
Returns the size of this KeywordTree.
Constructor Details
#initialize ⇒ Object
Creates a new KeywordTree
require 'ahocorasick'
kwt = Ahocorasick::KeywordTree.new
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 |
# File 'ext/ruby-ahocorasick.c', line 58
static VALUE
rb_kwt_init(VALUE self)
{
AC_STRUCT * tree;
struct kwt_struct_data *kwt_data;
kwt_data = ALLOC(struct kwt_struct_data);
tree = ac_alloc();
DATA_PTR(self) = kwt_data;
kwt_data->tree = tree;
kwt_data->last_id = 1;
kwt_data->dictionary_size = 0;
kwt_data->is_frozen = 0;
return self;
}
|
Class Method Details
.from_file ⇒ Object
Creates a new KeywordTree and loads the dictionary from a file
% cat dict0.txt
foo
bar
base
k= AhoCorasick::KeywordTree.from_file "dict0.txt"
k.search("basement").size # => 1
266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 |
# File 'ext/ruby-ahocorasick.c', line 266
static VALUE
rb_kwt_new_from_file(int argc, VALUE *argv, VALUE klass)
{
// TODO:
// * use rb_kwt_add_string
// * use rb_io* to handle the file
struct kwt_struct_data *kwt_data;
char word[1024];
int id;
VALUE self;
VALUE f_string;
FILE *dictionary;
rb_scan_args(argc, argv, "10", &f_string);
id = 0;
SafeStringValue( f_string );
self= rb_class_new_instance( 0, NULL, klass );
KeywordTree( self, kwt_data );
dictionary = fopen( RSTRING( f_string )->ptr, "r" );
if(dictionary == NULL)
rb_raise(rb_eRuntimeError, "Cannot open `%s\". No such file?", RSTRING(f_string)->ptr);
while(fgets(word, 1024, dictionary) != NULL) {
ac_add_string(kwt_data->tree, word, strlen(word)-1, id++);
kwt_data->dictionary_size++;
}
kwt_data->last_id= id+1;
fclose(dictionary);
return self;
}
|
Instance Method Details
#add_string ⇒ Object Also known as: <<
Adds a sequence to this KeywordTree.
kwt.add_string("foo1$21^ 98N3 ba>Z")
kwt << "bar" # using the alias
Note: you can also specify the id, a number between 1 and k
kwt.add_string "bar", 123 # => 123
This id should be unique in the context of the current tree.
Returns the id of the inserted object.
kwt.add_string("test", 18) # => 18
kwt.add_string("baz") # => 19
217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 |
# File 'ext/ruby-ahocorasick.c', line 217
static VALUE
rb_kwt_add_string(int argc, VALUE *argv, VALUE self)
{
VALUE v_string, v_id;
struct kwt_struct_data *kwt_data;
char * string;
int id;
rb_scan_args(argc, argv, "11", &v_string, &v_id);
Check_Type(v_string, T_STRING);
string= RSTRING(v_string)->ptr;
KeywordTree(self, kwt_data);
if(kwt_data->is_frozen == 1)
rb_raise(rb_eRuntimeError, "Cannot add `%s\" into a frozen tree.", string);
if(v_id == Qnil) {
id = kwt_data->last_id;
} else if(TYPE(v_id) != T_FIXNUM) {
rb_raise(rb_eRuntimeError, "Please use a number from 1 to K as id, or leave nil to auto-generate one. `%s\" given.", RSTRING(v_id)->ptr);
} else if(NUM2INT(v_id) <= 0) {
rb_raise(rb_eRuntimeError, "Please use a number from 1 to K as id, or leave nil to auto-generate one. `%d\" given.", NUM2INT(v_id));
} else {
id= NUM2INT(v_id);
}
if(ac_add_string(kwt_data->tree, string, strlen(string), id) == 0)
rb_raise(rb_eRuntimeError, "Failed to add `%s\", duplicate id `%d\"?", string, id);
kwt_data->last_id= id + 1;
kwt_data->dictionary_size++;
return INT2FIX(id);
}
|
#find_all ⇒ Object Also known as: search
Search the current tree.
It returns an array on hashes, e.g.
[ { :id => int, :value => int, :starts_at => int, :ends_at => int}, { ... } ]
Returns an empty array when the search didn’t return any result.
# assuming a valid KeywordTree kwt object:
kwt.add_string("one")
kwt.add_string("two")
kwt.search( "moved two times already" ).each do | result |
result[:id] # => 2
result[:ends_at] # => 9
result[:starts_at] # => 6
result[:value] # => two
end # => 1
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 |
# File 'ext/ruby-ahocorasick.c', line 131
static VALUE
rb_kwt_find_all(int argc, VALUE *argv, VALUE self)
{
char * remain; // returned by ac_search, the remaing text to search
int lgt, id, ends_at, starts_at; // filled in by ac_search: the length of the result, the id, and starts_at/ends_at position
VALUE v_result; // one result, as hash
VALUE v_results; // all the results, an array
VALUE v_search; // search string, function argument
struct kwt_struct_data *kwt_data;
// one mandatory argument.
rb_scan_args(argc, argv, "1", &v_search);
// it should be string.
Check_Type(v_search, T_STRING);
// get the structure
KeywordTree(self, kwt_data);
// freeze the tree, if not already
if(kwt_data->is_frozen == 0) {
if(ac_prep( kwt_data->tree ) == 0)
rb_raise(rb_eRuntimeError, "Cannot freeze the tree");
kwt_data->is_frozen = 1;
}
// prepare the return value
v_results= rb_ary_new();
// fail quickly and return the empty array
if(kwt_data->dictionary_size == 0)
return v_results;
// prepare the search
ac_search_init(kwt_data->tree, RSTRING( v_search )->ptr, RSTRING( v_search )->len);
// loop trought the results
while((remain= ac_search(kwt_data->tree, &lgt, &id, &ends_at)) != NULL) {
// this is an individual result as a hash
v_result= rb_hash_new();
rb_hash_aset( v_result, sym_id, INT2FIX(id) );
rb_hash_aset( v_result, sym_starts_at, INT2FIX( ends_at - lgt - 1 ) );
rb_hash_aset( v_result, sym_ends_at, INT2FIX( ends_at - 1 ) );
rb_hash_aset( v_result, sym_value, rb_str_new(remain, lgt) );
rb_ary_push( v_results, v_result );
}
// reopen the tree
kwt_data->is_frozen= 0;
return v_results;
}
|
#make ⇒ Object
It freezes the current KeywordTree.
Note: This method is called internally by search
require 'ahocorasick'
kwt = Ahocorasick::KeywordTree.new
kwt.add_string("one")
kwt.add_string("two")
kwt.make()
90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 |
# File 'ext/ruby-ahocorasick.c', line 90
static VALUE
rb_kwt_make(VALUE self)
{
struct kwt_struct_data *kwt_data;
KeywordTree(self, kwt_data);
if(kwt_data->is_frozen == 1)
return Qtrue;
if(ac_prep( kwt_data->tree ) == 1) {
kwt_data->is_frozen = 1;
return Qtrue;
}
rb_raise(rb_eRuntimeError, "Cannot freeze the tree");
}
|
#size ⇒ Object
Returns the size of this KeywordTree
kwt.add_string("foo")
kwt.add_string("bar")
kwt.size #=> 2
187 188 189 190 191 192 193 194 |
# File 'ext/ruby-ahocorasick.c', line 187
static VALUE
rb_kwt_size(VALUE self)
{
struct kwt_struct_data *kwt_data;
KeywordTree(self, kwt_data);
return INT2FIX(kwt_data->dictionary_size);
}
|