Natools

Check-in [eb0d515075]
Login
Overview
Comment:tools/smaz: add compression of an input list of strings
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: eb0d5150754288735bfd50ff4b7c28f67d34739f
User & Date: nat on 2016-09-23 20:30:38
Other Links: manifest | tags
Context
2016-09-24
21:53
tools/smaz: add the output of statistics about compression check-in: 939809f44a user: nat tags: trunk
2016-09-23
20:30
tools/smaz: add compression of an input list of strings check-in: eb0d515075 user: nat tags: trunk
2016-09-22
21:26
smaz-tools: add a function for Hash in dynamically created dictionaries check-in: 418b41afc4 user: nat tags: trunk
Changes

Modified tools/smaz.adb from [bd2e8cf2c0] to [eaae217532].

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

with Ada.Command_Line;
with Ada.Streams;
with Ada.Strings.Unbounded;
with Ada.Text_IO.Text_Streams;
with Natools.Getopt_Long;
with Natools.S_Expressions.Parsers;

with Natools.Smaz.Tools;
with Natools.Smaz.Tools.GNAT;

procedure Smaz is






   package Options is
      type Id is
        (Help,
         Output_Ada_Dictionary,

         Output_Hash);

   end Options;

   package Getopt is new Natools.Getopt_Long (Options.Id);

   type Callback is new Getopt.Handlers.Callback with record
      Display_Help : Boolean := False;
      Need_Dictionary : Boolean := False;

      Ada_Dictionary : Ada.Strings.Unbounded.Unbounded_String;
      Hash_Package : Ada.Strings.Unbounded.Unbounded_String;
   end record;

   overriding procedure Option
     (Handler  : in out Callback;
      Id       : in Options.Id;







>




>
>
>
>
>
>


<
|
>
|
>







>







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

with Ada.Command_Line;
with Ada.Streams;
with Ada.Strings.Unbounded;
with Ada.Text_IO.Text_Streams;
with Natools.Getopt_Long;
with Natools.S_Expressions.Parsers;
with Natools.S_Expressions.Printers;
with Natools.Smaz.Tools;
with Natools.Smaz.Tools.GNAT;

procedure Smaz is
   package Actions is
      type Enum is
        (Nothing,
         Encode);
   end Actions;

   package Options is
      type Id is

        (Output_Ada_Dictionary,
         Encode,
         Output_Hash,
         Help);
   end Options;

   package Getopt is new Natools.Getopt_Long (Options.Id);

   type Callback is new Getopt.Handlers.Callback with record
      Display_Help : Boolean := False;
      Need_Dictionary : Boolean := False;
      Action : Actions.Enum := Actions.Nothing;
      Ada_Dictionary : Ada.Strings.Unbounded.Unbounded_String;
      Hash_Package : Ada.Strings.Unbounded.Unbounded_String;
   end record;

   overriding procedure Option
     (Handler  : in out Callback;
      Id       : in Options.Id;
79
80
81
82
83
84
85




86
87
88
89
90
91
92
      Id       : in Options.Id;
      Argument : in String) is
   begin
      case Id is
         when Options.Help =>
            Handler.Display_Help := True;





         when Options.Output_Ada_Dictionary =>
            Handler.Need_Dictionary := True;

            if Argument'Length > 0 then
               Handler.Ada_Dictionary
                 := Ada.Strings.Unbounded.To_Unbounded_String (Argument);
            else







>
>
>
>







88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
      Id       : in Options.Id;
      Argument : in String) is
   begin
      case Id is
         when Options.Help =>
            Handler.Display_Help := True;

         when Options.Encode =>
            Handler.Need_Dictionary := True;
            Handler.Action := Actions.Encode;

         when Options.Output_Ada_Dictionary =>
            Handler.Need_Dictionary := True;

            if Argument'Length > 0 then
               Handler.Ada_Dictionary
                 := Ada.Strings.Unbounded.To_Unbounded_String (Argument);
            else
104
105
106
107
108
109
110

111
112
113
114
115
116
117

   function Getopt_Config return Getopt.Configuration is
      use Getopt;
      use Options;
      R : Getopt.Configuration;
   begin
      R.Add_Option ("ada-dict", 'A', Optional_Argument, Output_Ada_Dictionary);

      R.Add_Option ("hash-pkg", 'H', Required_Argument, Output_Hash);
      R.Add_Option ("help",     'h', No_Argument,       Help);

      return R;
   end Getopt_Config;









>







117
118
119
120
121
122
123
124
125
126
127
128
129
130
131

   function Getopt_Config return Getopt.Configuration is
      use Getopt;
      use Options;
      R : Getopt.Configuration;
   begin
      R.Add_Option ("ada-dict", 'A', Optional_Argument, Output_Ada_Dictionary);
      R.Add_Option ("encode",   'e', No_Argument,       Encode);
      R.Add_Option ("hash-pkg", 'H', Required_Argument, Output_Hash);
      R.Add_Option ("help",     'h', No_Argument,       Help);

      return R;
   end Getopt_Config;


174
175
176
177
178
179
180





181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208

         case Id is
            when Options.Help =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Display this help text");






            when Options.Output_Ada_Dictionary =>
               Put_Line (Output, "=[filename]");
               Put_Line (Output, Indent & Indent
                 & "Output the current dictionary as Ada code in the given");
               Put_Line (Output, Indent & Indent
                 & "file, or standard output if filename is ""-""");

            when Options.Output_Hash =>
               Put_Line (Output, " <Hash Package Name>");
               Put_Line (Output, Indent & Indent
                 & "Build a package with a perfect hash function for the");
               Put_Line (Output, Indent & Indent
                 & "current dictionary.");
         end case;
      end loop;
   end Print_Help;


   Opt_Config : constant Getopt.Configuration := Getopt_Config;
   Handler : Callback;
   Input_List : Natools.Smaz.Tools.String_Lists.List;
begin
   Process_Command_Line :
   begin
      Opt_Config.Process (Handler);
   exception
      when Getopt.Option_Error =>
         Print_Help (Opt_Config, Ada.Text_IO.Current_Error);







>
>
>
>
>

















<


|







188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216

217
218
219
220
221
222
223
224
225
226

         case Id is
            when Options.Help =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Display this help text");

            when Options.Encode =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Read a list of strings and encode them");

            when Options.Output_Ada_Dictionary =>
               Put_Line (Output, "=[filename]");
               Put_Line (Output, Indent & Indent
                 & "Output the current dictionary as Ada code in the given");
               Put_Line (Output, Indent & Indent
                 & "file, or standard output if filename is ""-""");

            when Options.Output_Hash =>
               Put_Line (Output, " <Hash Package Name>");
               Put_Line (Output, Indent & Indent
                 & "Build a package with a perfect hash function for the");
               Put_Line (Output, Indent & Indent
                 & "current dictionary.");
         end case;
      end loop;
   end Print_Help;


   Opt_Config : constant Getopt.Configuration := Getopt_Config;
   Handler : Callback;
   Input_List, Input_Data : Natools.Smaz.Tools.String_Lists.List;
begin
   Process_Command_Line :
   begin
      Opt_Config.Process (Handler);
   exception
      when Getopt.Option_Error =>
         Print_Help (Opt_Config, Ada.Text_IO.Current_Error);
216
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

   if not Handler.Need_Dictionary then
      return;
   end if;

   Read_Input_List :
   declare


      Input : constant access Ada.Streams.Root_Stream_Type'Class
        := Ada.Text_IO.Text_Streams.Stream (Ada.Text_IO.Current_Input);
      Parser : Natools.S_Expressions.Parsers.Stream_Parser (Input);
   begin
      Parser.Next;
      Natools.Smaz.Tools.Read_List (Input_List, Parser);





   end Read_Input_List;


   Build_Dictionary :
   declare
      Dictionary : constant Natools.Smaz.Dictionary
        := Natools.Smaz.Tools.To_Dictionary (Input_List, True);


      Ada_Dictionary : constant String
        := Ada.Strings.Unbounded.To_String (Handler.Ada_Dictionary);
      Hash_Package : constant String
        := Ada.Strings.Unbounded.To_String (Handler.Hash_Package);
   begin



      if Ada_Dictionary'Length > 0 then
         Print_Dictionary (Ada_Dictionary, Dictionary, Hash_Package);
      end if;

      if Hash_Package'Length > 0 then
         Natools.Smaz.Tools.GNAT.Build_Perfect_Hash (Input_List, Hash_Package);
      end if;











   end Build_Dictionary;
end Smaz;







>
>






>
>
>
>
>





|

>
>





>
>
>







>
>
>
>
>
>
>
>
>
>
>


234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
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

   if not Handler.Need_Dictionary then
      return;
   end if;

   Read_Input_List :
   declare
      use type Actions.Enum;

      Input : constant access Ada.Streams.Root_Stream_Type'Class
        := Ada.Text_IO.Text_Streams.Stream (Ada.Text_IO.Current_Input);
      Parser : Natools.S_Expressions.Parsers.Stream_Parser (Input);
   begin
      Parser.Next;
      Natools.Smaz.Tools.Read_List (Input_List, Parser);

      if Handler.Action /= Actions.Nothing then
         Parser.Next;
         Natools.Smaz.Tools.Read_List (Input_Data, Parser);
      end if;
   end Read_Input_List;


   Build_Dictionary :
   declare
      Dictionary : Natools.Smaz.Dictionary
        := Natools.Smaz.Tools.To_Dictionary (Input_List, True);
      Output : Natools.S_Expressions.Printers.Canonical
        (Ada.Text_IO.Text_Streams.Stream (Ada.Text_IO.Current_Output));
      Ada_Dictionary : constant String
        := Ada.Strings.Unbounded.To_String (Handler.Ada_Dictionary);
      Hash_Package : constant String
        := Ada.Strings.Unbounded.To_String (Handler.Hash_Package);
   begin
      Dictionary.Hash := Natools.Smaz.Tools.Linear_Search'Access;
      Natools.Smaz.Tools.List_For_Linear_Search := Input_List;

      if Ada_Dictionary'Length > 0 then
         Print_Dictionary (Ada_Dictionary, Dictionary, Hash_Package);
      end if;

      if Hash_Package'Length > 0 then
         Natools.Smaz.Tools.GNAT.Build_Perfect_Hash (Input_List, Hash_Package);
      end if;

      case Handler.Action is
         when Actions.Nothing => null;

         when Actions.Encode =>
            Output.Open_List;
            for S of Input_Data loop
               Output.Append_Atom (Natools.Smaz.Compress (Dictionary, S));
            end loop;
            Output.Close_List;
      end case;
   end Build_Dictionary;
end Smaz;