Class: StringEater::CTokenizer

Inherits:
Object
  • Object
show all
Defined in:
lib/c-tokenizer.rb,
ext/string-eater/c-tokenizer.c

Direct Known Subclasses

Tokenizer

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ CTokenizer

Returns a new instance of CTokenizer.



22
23
24
 
# File 'lib/c-tokenizer.rb', line 22

def initialize
  refresh_tokens
end

Class Method Details

.add_field(name, opts = {}) ⇒ Object 



8
9
10
11
 
# File 'lib/c-tokenizer.rb', line 8

def self.add_field name, opts={}
  self.tokens << StringEater::Token::new_field(name, opts)
  define_method(name) {@extracted_tokens[name]}
end

.dup_tokens ⇒ Object

This is very slow, only do it when necessary



18
19
20
 
# File 'lib/c-tokenizer.rb', line 18

def self.dup_tokens
  Marshal.load(Marshal.dump(tokens))
end

.look_for(tokens) ⇒ Object 



13
14
15
 
# File 'lib/c-tokenizer.rb', line 13

def self.look_for tokens
  self.tokens << StringEater::Token::new_separator(tokens)
end

.tokens ⇒ Object 



4
5
6
 
# File 'lib/c-tokenizer.rb', line 4

def self.tokens
  @tokens ||= []
end

Instance Method Details

#ctokenize!(string, tokens_to_find_indexes, tokens_to_find_strings, tokens_to_extract_indexes, tokens_to_extract_names) ⇒ Object 



12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
 
# File 'ext/string-eater/c-tokenizer.c', line 12

static VALUE tokenize_string(VALUE self, 
    VALUE string,
    VALUE tokens_to_find_indexes,
    VALUE tokens_to_find_strings,
    VALUE tokens_to_extract_indexes,
    VALUE tokens_to_extract_names)
{
  const char* input_string = StringValueCStr(string);
  VALUE extracted_tokens = rb_hash_new();
  VALUE curr_token;
  unsigned int curr_token_ix;
  long n_tokens_to_find = RARRAY_LEN(tokens_to_find_indexes);
  size_t str_len = strlen(input_string);
  size_t ix;
  char c;
  char looking_for;
  size_t looking_for_len;
  size_t looking_for_ix = 0;
  long find_ix = 0;
  const char*  looking_for_token;
  unsigned int n_tokens = (unsigned int)RARRAY_LEN(rb_iv_get(self, "@tokens"));

  size_t startpoint = 0;

  long n_tokens_to_extract = RARRAY_LEN(tokens_to_extract_indexes);
  long last_token_extracted_ix = 0;

  long next_token_to_extract_ix = NUM2UINT(rb_ary_entry(tokens_to_extract_indexes, last_token_extracted_ix));

  curr_token = rb_ary_entry(tokens_to_find_strings, find_ix);
  curr_token_ix = NUM2UINT(rb_ary_entry(tokens_to_find_indexes, find_ix));
  looking_for_token = StringValueCStr(curr_token);
  looking_for_len = strlen(looking_for_token);
  looking_for = looking_for_token[looking_for_ix];

  for(ix = 0; ix < str_len; ix++)
  {
    c = input_string[ix];
    if(c == looking_for)
    {
      if(looking_for_ix == 0)
      {
        /* entering new token */
        if(curr_token_ix > 0)
        {
          /* extract, if necessary */
          if((curr_token_ix - 1) == next_token_to_extract_ix)
          {
            last_token_extracted_ix++;
            if(last_token_extracted_ix < n_tokens_to_extract)
            {
              next_token_to_extract_ix = NUM2UINT(rb_ary_entry(tokens_to_extract_indexes, last_token_extracted_ix));
            }
            else
            {
              next_token_to_extract_ix = -1;
            }
            rb_hash_aset(extracted_tokens,
                rb_ary_entry(tokens_to_extract_names, curr_token_ix - 1),
                rb_usascii_str_new(input_string + startpoint,
                  ix - startpoint));
          }
        }
        startpoint = ix;
      }
      if(looking_for_ix >= looking_for_len - 1)
      {
        /* leaving token */
        if(curr_token_ix >= n_tokens-1)
        {
          break;
        }
        else
        {
          startpoint = ix + 1;
        }


        /* next token */
        find_ix++;
        if(find_ix >= n_tokens_to_find)
        {
          /* done! */
          break;
        }

        curr_token = rb_ary_entry(tokens_to_find_strings, find_ix);
        curr_token_ix = NUM2UINT(rb_ary_entry(tokens_to_find_indexes, find_ix));
        looking_for_token = StringValueCStr(curr_token);
        looking_for_len = strlen(looking_for_token);
        looking_for_ix = 0;
      }
      else
      {
        looking_for_ix++;
      }
      looking_for = looking_for_token[looking_for_ix];
    }
    else
    {
      /* reset token (or just keep looking, which is also fine) */
      looking_for_ix = 0;
      looking_for = looking_for_token[looking_for_ix];
    }
  }

  curr_token_ix = n_tokens - 1;

  if(ix < str_len && curr_token_ix == next_token_to_extract_ix)
  {
    rb_hash_aset(extracted_tokens,
        rb_ary_entry(tokens_to_extract_names, curr_token_ix),
        rb_usascii_str_new(input_string + startpoint,
          str_len - startpoint));
  }

  return extracted_tokens;
}

#describe_line ⇒ Object 



76
77
78
79
80
 
# File 'lib/c-tokenizer.rb', line 76

def describe_line
  tokens.inject("") do |desc, t|
    desc << (t.string || t.name.to_s || "xxxxxx")
  end
end

#do_extra_parsing ⇒ Object 



82
83
 
# File 'lib/c-tokenizer.rb', line 82

def do_extra_parsing
end

#extract_all_fields ⇒ Object 



30
31
32
33
34
35
 
# File 'lib/c-tokenizer.rb', line 30

def extract_all_fields
  @token_filter = lambda do |t|
    t.opts[:extract] = true if t.name
  end
  refresh_tokens
end

#extract_fields(*fields) ⇒ Object 



44
45
46
47
48
49
 
# File 'lib/c-tokenizer.rb', line 44

def extract_fields *fields
  @token_filter = lambda do |t|
    t.opts[:extract] = fields.include?(t.name)
  end
  refresh_tokens
end

#extract_no_fields ⇒ Object 



37
38
39
40
41
42
 
# File 'lib/c-tokenizer.rb', line 37

def extract_no_fields
  @token_filter = lambda do |t|
    t.opts[:extract] = false if t.name
  end
  refresh_tokens
end

#refresh_tokens ⇒ Object

This is very slow, only do it once before processing



52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
 
# File 'lib/c-tokenizer.rb', line 52

def refresh_tokens
  @tokens = self.class.dup_tokens

  if @token_filter
    @tokens.each{|t| @token_filter.call(t)}
  end

  tokens_to_find = tokens.each_with_index.map do |t, i|
    [i, t.string] if t.string
  end.compact

  @tokens_to_find_indexes = tokens_to_find.map{|t| t[0]}
  @tokens_to_find_strings = tokens_to_find.map{|t| t[1]}

  tokens_to_extract = tokens.each_with_index.map do |t, i|
    [i, t.name] if t.extract?
  end.compact

  @tokens_to_extract_indexes = tokens_to_extract.map{|t| t[0]}
  @tokens_to_extract_names = tokens.map{|t| t.name}

  @have_tokens_to_extract = (@tokens_to_extract_indexes.size > 0)
end

#tokenize!(string, &block) ⇒ Object 



85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
 
# File 'lib/c-tokenizer.rb', line 85

def tokenize! string, &block
  @string = string
  @extracted_tokens ||= {}
  @extracted_tokens.clear

  return unless @have_tokens_to_extract

  @extracted_tokens = ctokenize!(@string, 
                                 @tokens_to_find_indexes,
                                 @tokens_to_find_strings,
                                 @tokens_to_extract_indexes,
                                 @tokens_to_extract_names)

  # extra parsing hook
  do_extra_parsing

  if block_given?
    yield @extracted_tokens
  end

  # return self for chaining
  self
end

#tokens ⇒ Object 



26
27
28
 
# File 'lib/c-tokenizer.rb', line 26

def tokens
  @tokens
end