Natools

Check-in [db2278efbb]
Login

Check-in [db2278efbb]

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:tools/smaz: replace "word list" with clearer "[sample] text list"
Timelines: family | ancestors | descendants | both | trunk
Files: files | file ages | folders
SHA1: db2278efbbf160585552b8542635b45ffbf9e766
User & Date: nat 2016-10-22 19:21:42.017
Context
2016-10-23
21:23
smaz-tools: new primitive to remove an entry from a dictionary check-in: 134a6f8380 user: nat tags: trunk
2016-10-22
19:21
tools/smaz: replace "word list" with clearer "[sample] text list" check-in: db2278efbb user: nat tags: trunk
2016-10-21
19:32
tools/smaz: new command-line option to filter substrings by count check-in: 637ebd90fa user: nat tags: trunk
Changes
Unified Diff Ignore Whitespace Patch
Changes to tools/smaz.adb.
︙︙
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

         Encode,
         Evaluate);
   end Actions;

   package Dict_Sources is
      type Enum is
        (S_Expression,
         Word_List);
   end Dict_Sources;

   package Options is
      type Id is
        (Output_Ada_Dict,
         Dictionary_Input,
         Decode,
         Encode,
         Evaluate,
         Filter_Threshold,
         Output_Hash,
         Job_Count,
         Help,
         Sx_Dict_Output,
         Min_Sub_Size,
         Max_Sub_Size,
         Stat_Output,
         No_Stat_Output,
         Word_List_Input,
         Max_Word_Size,
         Sx_Output,
         No_Sx_Output);
   end Options;

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









|


















|








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

         Encode,
         Evaluate);
   end Actions;

   package Dict_Sources is
      type Enum is
        (S_Expression,
         Text_List);
   end Dict_Sources;

   package Options is
      type Id is
        (Output_Ada_Dict,
         Dictionary_Input,
         Decode,
         Encode,
         Evaluate,
         Filter_Threshold,
         Output_Hash,
         Job_Count,
         Help,
         Sx_Dict_Output,
         Min_Sub_Size,
         Max_Sub_Size,
         Stat_Output,
         No_Stat_Output,
         Text_List_Input,
         Max_Word_Size,
         Sx_Output,
         No_Sx_Output);
   end Options;

   package Getopt is new Natools.Getopt_Long (Options.Id);
︙︙
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199


         when Options.Sx_Output =>
            Handler.Sx_Output := True;

         when Options.Dictionary_Input =>
            Handler.Dict_Source := Dict_Sources.S_Expression;

         when Options.Word_List_Input =>
            Handler.Dict_Source := Dict_Sources.Word_List;

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

         when Options.Min_Sub_Size =>
            Handler.Min_Sub_Size := Positive'Value (Argument);








|
|








184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199


         when Options.Sx_Output =>
            Handler.Sx_Output := True;

         when Options.Dictionary_Input =>
            Handler.Dict_Source := Dict_Sources.S_Expression;

         when Options.Text_List_Input =>
            Handler.Dict_Source := Dict_Sources.Text_List;

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

         when Options.Min_Sub_Size =>
            Handler.Min_Sub_Size := Positive'Value (Argument);
︙︙
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243

      R.Add_Option ("hash-pkg",      'H', Required_Argument, Output_Hash);
      R.Add_Option ("jobs",          'j', Required_Argument, Job_Count);
      R.Add_Option ("sx-dict",       'L', No_Argument,       Sx_Dict_Output);
      R.Add_Option ("min-substring", 'm', Required_Argument, Min_Sub_Size);
      R.Add_Option ("max-substring", 'M', Required_Argument, Max_Sub_Size);
      R.Add_Option ("stats",         's', No_Argument,       Stat_Output);
      R.Add_Option ("no-stats",      'S', No_Argument,       No_Stat_Output);
      R.Add_Option ("word-list",     'w', No_Argument,       Word_List_Input);
      R.Add_Option ("max-word-len",  'W', Required_Argument, Max_Word_Size);
      R.Add_Option ("s-expr",        'x', No_Argument,       Sx_Output);
      R.Add_Option ("no-s-expr",     'X', No_Argument,       No_Sx_Output);

      return R;
   end Getopt_Config;









|








229
230
231
232
233
234
235
236
237
238
239
240
241
242
243

      R.Add_Option ("hash-pkg",      'H', Required_Argument, Output_Hash);
      R.Add_Option ("jobs",          'j', Required_Argument, Job_Count);
      R.Add_Option ("sx-dict",       'L', No_Argument,       Sx_Dict_Output);
      R.Add_Option ("min-substring", 'm', Required_Argument, Min_Sub_Size);
      R.Add_Option ("max-substring", 'M', Required_Argument, Max_Sub_Size);
      R.Add_Option ("stats",         's', No_Argument,       Stat_Output);
      R.Add_Option ("no-stats",      'S', No_Argument,       No_Stat_Output);
      R.Add_Option ("text-list",     't', No_Argument,       Text_List_Input);
      R.Add_Option ("max-word-len",  'W', Required_Argument, Max_Word_Size);
      R.Add_Option ("s-expr",        'x', No_Argument,       Sx_Output);
      R.Add_Option ("no-s-expr",     'X', No_Argument,       No_Sx_Output);

      return R;
   end Getopt_Config;
︙︙
440
441
442
443
444
445
446
447
448
449
450

451
452
453
454
455
456
457

                 & "Output filtered results in a S-expression");

            when Options.Dictionary_Input =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Read dictionary directly in input S-expression (default)");

            when Options.Word_List_Input =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Compute dictionary from word list in input S-expression");


            when Options.Sx_Dict_Output =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Output the dictionary as a S-expression");

            when Options.Min_Sub_Size =>








|


|
>








440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458

                 & "Output filtered results in a S-expression");

            when Options.Dictionary_Input =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Read dictionary directly in input S-expression (default)");

            when Options.Text_List_Input =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Compute dictionary from sample texts"
                 & " in input S-expression");

            when Options.Sx_Dict_Output =>
               New_Line (Output);
               Put_Line (Output, Indent & Indent
                 & "Output the dictionary as a S-expression");

            when Options.Min_Sub_Size =>
︙︙
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510

   is
      use type Natools.Smaz.Tools.String_Count;
   begin
      case Handler.Dict_Source is
         when Dict_Sources.S_Expression =>
            return Natools.Smaz.Tools.To_Dictionary (Input, True);

         when Dict_Sources.Word_List =>
            declare
               Counter : Natools.Smaz.Tools.Word_Counter;
            begin
               for S of Input loop
                  Natools.Smaz.Tools.Add_Substrings
                    (Counter, S, Handler.Min_Sub_Size, Handler.Max_Sub_Size);









|








497
498
499
500
501
502
503
504
505
506
507
508
509
510
511

   is
      use type Natools.Smaz.Tools.String_Count;
   begin
      case Handler.Dict_Source is
         when Dict_Sources.S_Expression =>
            return Natools.Smaz.Tools.To_Dictionary (Input, True);

         when Dict_Sources.Text_List =>
            declare
               Counter : Natools.Smaz.Tools.Word_Counter;
            begin
               for S of Input loop
                  Natools.Smaz.Tools.Add_Substrings
                    (Counter, S, Handler.Min_Sub_Size, Handler.Max_Sub_Size);
︙︙