Thanks guys..my project and my many problems

Thanks guys..my project and my many problems

[Paul Gregory]

Thank-you Larry and Preban for your kind help.

I'm literally new to ADA and I'm finding it quite difficult if anybody
can help I would be so grateful its untrue!!.

I'm getting really stressed by this - I suppose when you were novices
you were pulling your hair out too ;-)
---------------------------------------------------------------------------------

My project is this

You are invited to design and implement a translation program which will

take an English sentence and translate it into French (and optionally
translate from French
to English).

Your program will use a simple dictionary containing the following
words:

                                                       English
                                                              French
                                                       big
                                                              grand
                                                       black
                                                              noir
                                                       cat
                                                              chat
                                                       dog
                                                              chien
                                                       is
                                                              est
                                                       small
                                                              petit
                                                       the
                                                              le
                                                       white
                                                              blanc


Example 1

Do you wish to add words to the dictionary?: N

Please enter a sentence:
The cat is white.

The French translation of your sentence is:
Le chat est blanc.

Example 2

Do you wish to add words to the dictionary?: N

Please enter a sentence:
The cat is dead.

I cannot translate that sentence.

Example 3

Do you wish to add words to the dictionary?: Y
English word: horse
Equivalent French word: cheval

Do you wish to add more words?: N

Please enter a sentence:
Le cheval est grand.

The English translation of your sentence is:
The horse is big.

______________________________________________________________
My code so far is this..

--
-- Read the English and translate it into French
--
-- Constant array.
--
with Ada.Text_io;
with Nice_IO; use Nice_IO;
with String_tokenizer; use String_tokenizer;
procedure translate6 is
  type English is
    (big, black, cat, dog, es, small, the, white);

  type French is (grand, noir, chat, chien, est, petit, le, blanc);

  English_to_French: constant array(English) of French:=
    ( big                 => grand,
      black          => noir,
      cat               => chat,
      dog => chien,
      es     => est,
      small             => petit,
      the => le,
      white        => blanc);

  S:       String(1..80);   -- Input string
  Last:    Integer;         -- Last character of input string
  S2:       string (1..2);
  First:   Integer;
  Eng:     English;
  my_token : String_tokenizer.Token;
  length : integer;
begin
loop
  Put ("Do you wish to add words to the dictionary? : ");
Ada.Text_IO.Get_Line (S, last);
    Put("Please enter a sentence: ");
    Ada.Text_io.Get_Line(S, Last);
    exit when Last = 0;
    set_string(s(1..Last));  -- this sets up the tokenizer
    while has_more_tokens loop
      my_token := next_token;  -- my_token has the next token in the
string

      length := my_token.token_string_length;  -- length has the length
of this token string
      Eng := English'Value(my_token.token_string(1..length));

    Put_Line ("The French translation of your sentence is: " &
French'Image(English_to_French(Eng)));

    end loop;
end loop;
exception
  when Constraint_Error =>
    Put_Line(S(1..Last) & " cannot be translated");
end translate6;
________________________________________________________________

My problems are the following (deadline Friday - I'm getting really
worried and stressed as I'm reading but finding it hard to understand /
get the information I want for these paticular circumstance...

Problem 1 - the UPPER CASE problem as stated before

(where about in the program would I insert the character.handling bit?)

Problem 2 - When the program translates the tokens it does it as
follows..

When the program translates the tokens it does it like this...

Please enter a sentence: the cat dog

The French translation is : LE
The French translation is: CHAT
The French translation is: CHIEN

----> instead of

The French translation is:  Le chat chien.

Any ideas how I can get round this ?

Problem 3

is that one of the words I have to translate is 'Is' but as this is a
reserved word it causes an error message everytime I run it.

Problem 4

Is that as you can see from the question I have to "add words to the
dictionary" - I'm not overly sure how to do this and presume I'd have to
write some sort of IF statement. Everytime I try and write one it raises
and untold amount of errors.

Basically if they say 'N' I want it to go straight to the translator bit
but if they say 'Y' then they must add a English word, then add the
French version of the word then if I type the added word in a sentennce
it must be able to recognise it....

Nurse...pass me the pills !!

Cheers,

PG

Re: Thanks guys..my project and my many problems

[Preben Randhol]

Paul Gregory wrote:
> Thank-you Larry and Preban for your kind help.
                          e ;-)
>   type English is
>     (big, black, cat, dog, es, small, the, white);
> 
>   type French is (grand, noir, chat, chien, est, petit, le, blanc);
> 
>   English_to_French: constant array(English) of French:=
>     ( big                 => grand,
>       black          => noir,
>       cat               => chat,
>       dog => chien,
>       es     => est,
>       small             => petit,
>       the => le,
>       white        => blanc);
> 

Why do you use enumerates? Is it required? If you are supposed to be
able to add words and possibly store them too, why don't you rather make
something like:

   type Max_Word_Length : constant Natural := 80;

   type Word_Record is
      record
         english : String (1 .. Max_Word_Length);
         french  : String (1 .. Max_Word_Length);
      end record;

of course you have to build a linked list or an array and a way to
search for the correct word. 

Another way could be to use arrays (you don't have to deal with
the linked list problems) but have the design a bit more flexible so
you can add more languages if you want:

   type Dictionary is
      record
         word_id : Natural := 0;
         word    : String (1 .. Max_Word_Length);
      end record;

   type English_List is array (1 .. 1000) of Dictionary;
   type French_List is array (1 .. 1000) of Dictionary;

then when you add you generate a unique word_id /= 0 for each word and
you put this in the English list along with the English word and in the
French list with the french word. So that when you look up a word you
get the word_id and you use this to find the same word in the other
language. Of course you don't have to do this if you know that word
number 3 in the English list will correspond with word number 3 in the
French.

Please note I haven't checked my code so there might be errors in it.

Hope you can get further on your project with this.

-- 
Preben Randhol ---------------- http://www.pvv.org/~randhol/ --
"Violence is the last refuge of the incompetent", Isaac Asimov

Re: Thanks guys..my project and my many problems

[Hyman Rosen]

Paul Gregory wrote:
>   type English is (big, black, cat, dog, es, small, the, white);
>   type French is (grand, noir, chat, chien, est, petit, le, blanc);

This is the core of your difficulties. Since the dictionary is
meant to be extensible (as you yourself have noticed), using
enumerations to encode the words is inappropriate - these types
are fixed at compile-time, and you cannot add to them at runtime.

You need to represent your words as strings, and maintain a mapping
table between English and French that way. It would be easy in C++
or Perl, which come with built in map containers. In Ada, you'll
have to roll your own.

Re: Thanks guys..my project and my many problems

[Preben Randhol]

Hyman Rosen wrote:
> table between English and French that way. It would be easy in C++
> or Perl, which come with built in map containers. In Ada, you'll

Oh, is Safe STL standard now?

   http://www.horstmann.com/safestl.html

-- 
Preben Randhol ---------------- http://www.pvv.org/~randhol/ --
"Violence is the last refuge of the incompetent", Isaac Asimov

Re: Thanks guys..my project and my many problems

[Hyman Rosen]

Preben Randhol wrote:
> Hyman Rosen wrote:
>>table between English and French that way. It would be easy in C++
>>or Perl, which come with built in map containers. In Ada, you'll
> Oh, is Safe STL standard now?
>    http://www.horstmann.com/safestl.html

What does easy have to do with safe? In any case, the site that you
point to states that Safe STL has the same interface as normal STL,
so any implementation may choose to supply it as a conforming option.

Re: Thanks guys..my project and my many problems

[Preben Randhol]

Hyman Rosen wrote:
> What does easy have to do with safe? 

This is the exact problem with the software industry. Doing it the easy
way and ignoring the safty.

-- 
Preben Randhol ---------------- http://www.pvv.org/~randhol/ --
"Violence is the last refuge of the incompetent", Isaac Asimov

Re: Thanks guys..my project and my many problems

[Hyman Rosen]

Preben Randhol wrote:
> This is the exact problem with the software industry.
 > Doing it the easy way and ignoring the safty.

Huh? In C++ I can declare a map<string,string> and store
my dictionary in it as simply as 'EtoF["cat"] = "chat";'.
In Perl, I can do '$EtoF{cat} = "chat";'. Ada doesn't
come with a containers library, so you have to build,
buy, or find one before you can use it. What does any
of this have to do with safety?

Re: Thanks guys..my project and my many problems

[Simon Wright]

Hyman Rosen <hyrosen@mail.com> writes:

> You need to represent your words as strings, and maintain a mapping
> table between English and French that way. It would be easy in C++
> or Perl, which come with built in map containers. In Ada, you'll
> have to roll your own.

There are several Ada container libraries, but none of them are
standardised.

Paul, would you be marked down for using a container library? if not,
you could check out the Booch Components at
http://www.pushface.org/components/bc/ or the other container
libraries at http://www.pushface.org/components/ .

Re: Thanks guys..my project and my many problems

[Jano]

En el mensaje <x7vk7fpr0ch.fsf@smaug.pushface.org>, simon@pushface.org 
dice...

> Paul, would you be marked down for using a container library? if not,
> you could check out the Booch Components at
> http://www.pushface.org/components/bc/ or the other container

Uhm, being a novice I think he will find the Booch Components very 
difficult.

I would go for two static arrays of (un)bounded strings, and a counter 
for words in use. I know, the maximum capacity would be bounded, but...

I guess is a more novice approach.

-- 
-------------------------
Jano
402450[at]cepsz.unizar.es
-------------------------

Re: Thanks guys..my project and my many problems

[Matthew Heaney]

Jano <402450@cepsz.unizar.es> wrote in message news:<MPG.18c4c14ba0317dc99896a9@News.CIS.DFN.DE>...
> 
> I would go for two static arrays of (un)bounded strings, and a counter 
> for words in use. I know, the maximum capacity would be bounded, but...

The Charles library has dedicated map containers that have type String as the key.

http://home.earthlink.net/~matthewjheaney/charles/

Re: Thanks guys..my project and my many problems

[Simon Wright]

Simon Wright <simon@pushface.org> writes:

> Paul, would you be marked down for using a container library? if not,
> you could check out the Booch Components at
> http://www.pushface.org/components/bc/ or the other container
> libraries at http://www.pushface.org/components/ .

Sorry, what I meant by the last line was "the other container
libraries that I've listed at http://www.pushface.org/components/"

Re: Thanks guys..my project and my many problems

[Rodrigo García]

Hyman Rosen wrote:
> Paul Gregory wrote:
> 
>>   type English is (big, black, cat, dog, es, small, the, white);
>>   type French is (grand, noir, chat, chien, est, petit, le, blanc);
> 
> 
> This is the core of your difficulties. Since the dictionary is
> meant to be extensible (as you yourself have noticed), using
> enumerations to encode the words is inappropriate - these types
> are fixed at compile-time, and you cannot add to them at runtime.
> 
> You need to represent your words as strings, and maintain a mapping
> table between English and French that way. It would be easy in C++
> or Perl, which come with built in map containers. In Ada, you'll
> have to roll your own.
> 

   I prefer Python for this. Not surprisingly, in Python they call these 
kind of containers "Dictionaries".

   If you are using GNAT, maybe you can try its hash table...

Rodrigo

Re: Thanks guys..my project and my many problems

[Matthew Heaney]

Hyman Rosen <hyrosen@mail.com> wrote in message news:<1046101109.45189@master.nyc.kbcfp.com>...
> 
> You need to represent your words as strings, and maintain a mapping
> table between English and French that way. It would be easy in C++
> or Perl, which come with built in map containers. In Ada, you'll
> have to roll your own.

No, you won't, because the STL has already been ported to Ada.  

The library is called "Charles", in honor of Charles Babbage, with
whom Lady Ada was affiliated.

http://home.earthlink.net/~matthewjheaney/charles/

Re: Thanks guys..my project and my many problems

[Hyman Rosen]

Matthew Heaney wrote:
> No, you won't, because the STL has already been ported to Ada.

I know, actually, and good work it is. But it won't be
readily at hand for the struggling student doing his
homework.

Re: Thanks guys..my project and my many problems

[Preben Randhol]

Hyman Rosen wrote:
> Matthew Heaney wrote:
>> No, you won't, because the STL has already been ported to Ada.
> 
> I know, actually, and good work it is. But it won't be
> readily at hand for the struggling student doing his
> homework.

Well, depends on the assignment. However if one were allowed to use
charles or STL I think the assignment is a bit too trivial and would
probably be bigger.

-- 
Preben Randhol ---------------- http://www.pvv.org/~randhol/ --
"Violence is the last refuge of the incompetent", Isaac Asimov

Re: Thanks guys..my project and my many problems

[Matthew Heaney]

Paul Gregory <pg16@bton.ac.uk> wrote in message news:<3E5A2381.21FD3F14@bton.ac.uk>...
> 
> My project is this
> 
> You are invited to design and implement a translation program which will
> take an English sentence and translate it into French (and optionally
> translate from French to English).

You could use Charles, which contains data structures that are exactly
suited to your problem.

package Translation_Tables is
   new Charles.Maps.Sorted.Strings.Unbounded 
     (Element_Type => Charles.Strings.Unbounded.Container_Type);

E2F : Translation_Tables.Container_Type;

Insert (E2F, "big", Element => To_Container ("grand"));
Insert (E2F, ...);

procedure Lookup (E : String) is
   I : constant Iterator_Type := Find (E2F, Key => E);
begin
   if I = Back (E2F) then
      --not found
   else
      declare
        F : Element_Subtype renames To_Access (I);  --no copy
        --or F : constant Element_Subtype := Element (I); --copy
      begin
        --do something with F
      end;
   end if;
end Lookup;

The string map above accepts a key comparison operator as a generic
formal subprogram, so you could plug one in that provides a
case-insensitive compare.

There's also a map version implemented as a hash table, and there's a
hash function provided for type String.  (The sorted version is
implemented as a red-black tree.)

I used the new Charles string container as the map element, but you
could just as easily use any other non-limited type, such as
Ada.Strings.Unbounded.  A string pointer would also suffice, too.

http://home.earthlink.net/~matthewjheaney/charles/

The most recent release occurred on 18 Feb 2003.

Drop me a line if you have any questions.

Matt

Re: Thanks guys..my project and my many problems

[chris.danx]

Matthew Heaney wrote:
> Paul Gregory <pg16@bton.ac.uk> wrote in message news:<3E5A2381.21FD3F14@bton.ac.uk>...
> 
>>My project is this
>>
>>You are invited to design and implement a translation program which will
>>take an English sentence and translate it into French (and optionally
>>translate from French to English).
> 
> 
> You could use Charles, which contains data structures that are exactly
> suited to your problem.

If it's a uni assignment, chances are that Paul isn't allowed to use 
other peoples code in their exercise (especially at or near the 
beginning of learning to program, and in data structures courses)!  At 
our uni (glasgow), we where unable to use others code until 3rd year and 
then only under authorisation by the lecturer who set the exercise!

I always found it strange that they rammed Code Reuse down our throats 
the whole of second year, yet we weren't allowed to practise it!


Danx

Re: Thanks guys..my project and my many problems

[Chad R. Meiners]

"chris.danx" <spamoff.danx@ntlworld.com> wrote in message
news:uUP6a.4405$EN3.34753@newsfep4-glfd.server.ntli.net...
> I always found it strange that they rammed Code Reuse down our throats
> the whole of second year, yet we weren't allowed to practise it!

Perhaps they wanted you to write your code so you could reuse it in other
projects.   Data structure packages I wrote in other classes came in handy
when doing other assignments when I was an undergraduate.  Of course your
milage could vary depending on your university.

-CRM

Re: Thanks guys..my project and my many problems

[chris.danx]

Chad R. Meiners wrote:
> "chris.danx" <spamoff.danx@ntlworld.com> wrote in message
> news:uUP6a.4405$EN3.34753@newsfep4-glfd.server.ntli.net...
> 
>>I always found it strange that they rammed Code Reuse down our throats
>>the whole of second year, yet we weren't allowed to practise it!
> 
> 
> Perhaps they wanted you to write your code so you could reuse it in other
> projects.   Data structure packages I wrote in other classes came in handy
> when doing other assignments when I was an undergraduate.  Of course your
> milage could vary depending on your university.

Yeah, I started doing that at the beginning of my first second year (I 
repeated second year).  The problem was that by the time I had amassed a 
number of data structures, I'd learned new ways to make them better and 
improve their interfaces.  Now they just all look ugly to me.

One of the other problems I had was that the university has an 
assignment management system which is truely crap.  Unless they'd 
specifically set it up to grab everything in a directory, it would only 
grab the files it knew about so we couldn't submit our data structures 
with our assignments.  In first and second year, I usually had to moan 
at my supervisor to get packages sumbitted as I often liked to use the 
main program simply as a driver rather than plonk all the code in it 
once it got to 400+ lines of code.  Really frustrating :(


Also we can't really use them now, Ada was only used in term 1 and now 
it's all Java and C.  Java has loads of data structures, so writing them 
for that is not really worth it and C, well, I don't want to go there... :)


Chris

Re: Thanks guys..my project and my many problems

[Paul Gregory]

Thanks a million Matthew and everybody else...however I went to see my tutor today and he
refused me permission to use your package. Apparantly I MUST use a package that the
University supplies called "string_tokenizer" to read the tokens in my array when
translation...so It's back for the drawing board for me unfortunately :-(

---->
----------------------------------------------------------------------------------------
package body String_tokenizer is

-- Uses recursive descent.
-- Each syntactic token has a recognizer, which returns its token iff
--   it is recognized at the current position in the input string.
-- If a recognizer recognizes its token, it consumes the corresponding
--   portion of the input string; if not, it leaves the string for others
--   to attempt to recognize.
-- Higher level syntactic recognizers call lower level recognizers,
--   possibly recursively.

-- Token string syntax
----------------------

--  NO_MORE_TOKENS ::= end of input string
--  WORD_CHAR      ::= A..Z | a..z
--  WORD           ::= WORD_CHAR | WORD_CHAR WORD
--  PUNCTUATION    ::= . | , | : | ; | ' | " | `
--  NUMBER_CHAR    ::= 0..9
--  NUMBER         ::= NUMBER_CHAR | NUMBER_CHAR NUMBER
--  OPERATOR       ::= + | - | * | / | **
--  BRACKET        ::= ( | ) | { | } | [ | }
--  WHITE_SPACE    ::= space_character | space_character WHITE_SPACE
--  OTHER          ::= characters other than those mentioned above
--  TOKEN          ::= NO_MORE_TOKENS | WORD | PUNCTUATION | NUMBER |
--                       OPERATOR | BRACKET | OTHER
--  STRING_TOKEN   ::= WORD | PUNCTUATION | NUMBER | OPERATOR | BRACKET |
--                       OTHER
--  NON_WHITE_SPACE_TOKEN_STRING ::= NO_MORE_TOKENS | STRING_TOKEN TOKEN_STRING
--  TOKEN_STRING   ::= NON_WHITE_SPACE_TOKEN_STRING |
--                     WHITE_SPACE NON_WHITE_SPACE_TOKEN_STRING


  WHITE_SPACE : constant integer := 7;  -- needed internally, not for client
  FAIL        : constant integer := 8;  -- needed internally, not for client

  function char_type(c : Character) return Integer is

  -- post: returns type code for character c

  -- test:
  --   space
  --   all punctuation characters
  --   brackets
  --   a,z,A,Z
  --   0,9
  --   all operators
  --   other character

    char_type_table : array (Character range' '..'~') of Integer :=
    (WHITE_SPACE,PUNCTUATION,PUNCTUATION,OTHER,OTHER,OTHER,OTHER,PUNCTUATION,      -- '
'..'''
     BRACKET,BRACKET,OPERATOR,OPERATOR,PUNCTUATION,OPERATOR,PUNCTUATION,OPERATOR,  --
'('..'/'
     NUMBER,NUMBER,NUMBER,NUMBER,NUMBER,NUMBER,NUMBER,NUMBER,                      --
'0'..'7'
     NUMBER,NUMBER,PUNCTUATION,PUNCTUATION,OPERATOR,OPERATOR,OPERATOR,PUNCTUATION, --
'8'..'?'
     OTHER,WORD,WORD,WORD,WORD,WORD,WORD,WORD,                                     --
'@'..'G'
     WORD,WORD,WORD,WORD,WORD,WORD,WORD,WORD,                                      --
'H'..'O'
     WORD,WORD,WORD,WORD,WORD,WORD,WORD,WORD,                                      --
'P'..'W'
     WORD,WORD,WORD,BRACKET,OTHER,BRACKET,OTHER,OTHER,                             --
'X'..'_'
     PUNCTUATION,WORD,WORD,WORD,WORD,WORD,WORD,WORD,                               --
'`'..'g'
     WORD,WORD,WORD,WORD,WORD,WORD,WORD,WORD,                                      --
'h'..'o'
     WORD,WORD,WORD,WORD,WORD,WORD,WORD,WORD,                                      --
'p'..'w'
     WORD,WORD,WORD,BRACKET,OTHER,BRACKET,OTHER);                                  --
'x'..'~'

  begin
    if c not in ' '..'~' then return OTHER; end if;
    return char_type_table(c);
  end;


  IN_STRING_SIZE : constant integer := 1000;
  in_string : String(1..IN_STRING_SIZE);
  in_string_pos : integer := IN_STRING_SIZE;

  blank_token : constant Token := ("                    ",0,0,NO_MORE_TOKENS);
  fail_token : constant Token  := ("                    ",0,0,FAIL);

  current_token : Token := blank_token;


  -- methods to access current character and move to next character
  --   in input string

  function has_current_char return Boolean is

  --test:
  --  just before end of string
  --  end of string

  begin
    return in_string_pos <= IN_STRING_SIZE;
  end;


  function current_char return Character is

  -- pre:  has_current_char

  -- post: returns current char

  -- test: once thru

  begin
    return in_string(in_string_pos);
  end;


  procedure inc is

  -- test : once thru

  begin
    in_string_pos := in_string_pos + 1;
  end;


  -- recognizer methods
  ---------------------

  function recognize_char(i : Integer) return Token is

  -- pre:  i = character type to be recognized

  -- post: returns character token if recognized, fail token if not

  -- test:
  --   end of string
  --   current char is required char
  --   current char is not required char

    my_token : Token;
  begin
    if not has_current_char then return fail_token; end if;
    if char_type (current_char) = i then
      my_token := (current_char &  "                   ",1,0,i);
      inc;
      return my_token;
    else
      return fail_token;
    end if;
  end;


  function recognize_sequence(i : Integer) return Token is

  -- pre:  i = char type of sequence to be recognized

  -- post: returns token for recognized sequence,
  --         or fail token if not found

  -- test:
  --   not a recognized sequence
  --   sequence of 1 char
  --   sequence of 3 chars

  my_token : Token;
  next_Token : Token;
  word_pos :Integer;

  begin
    my_token := recognize_char(i);
    if my_token.token_type = FAIL then
      return my_token;
    end if;
    word_pos := 2; -- start position for rest of word
    loop
      next_Token := recognize_char(i);
      if next_token.token_type = FAIL then
        return my_token;
      end if;
      if word_pos <= TOKEN_STRING_SIZE then
        my_token.token_string(word_pos) := next_Token.token_string(1);
        word_pos := word_pos + 1;
        my_token.token_string_length := my_token.token_string_length + 1;
      end if;
    end loop;
  end;


  function recognize_no_more_tokens return Token is

  -- test:
  --   not end of string
  --   end of string

  begin
    if not has_current_char then return blank_token;
    else
      return fail_token;
    end if;
  end;


  function recognize_word_char return Token is
  begin
    return recognize_char(WORD);
  end;


  function recognize_word return Token is
  begin
    return recognize_sequence(WORD);
  end;


  function recognize_punctuation return Token is
  begin
    return recognize_char(PUNCTUATION);
  end;


  function recognize_number_char return Token is
  begin
    return recognize_char(NUMBER);
  end;


  function recognize_number return Token is

  -- test:
  --   not a number
  --   2 char number

  my_token : Token;

  begin
    my_token := recognize_sequence(NUMBER);
    if my_token.token_type = FAIL then
      return my_token;
    end if;
    for i in 1..my_token.token_string_length loop
    my_token.token_value := my_token.token_value * 10 +
      character'pos(my_token.token_string(i)) - character'pos('0');
    end loop;
    return my_token;
  end;


  function recognize_operator return Token is

  -- test:
  --   /
  --   */
  --   **/
  --   * at end of string

  my_token : Token;
  begin
    my_token := recognize_char(OPERATOR);
    if my_token.token_string(1) = '*' then
      if has_current_char and then current_char = '*' then
        my_token.token_string(2) := '*';
        my_token.token_string_length := 2;
        inc;
      end if;
    end if;
    return my_token;
  end;


  function recognize_bracket return Token is
  begin
    return recognize_char(BRACKET);
  end;


  function recognize_white_space return Token is
  begin
    return recognize_sequence(WHITE_SPACE);
  end;


  function recognize_other return Token is
  begin
    return recognize_char(OTHER);
  end;


  function recognize_string_token return Token is

  -- test:
  --   word
  --   punctuation
  --   number
  --   operator
  --   bracket
  --   other
  --   none of these

    my_token : Token;
  begin
    my_token := recognize_word;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;

    my_token := recognize_punctuation;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;

    my_token := recognize_number;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;

    my_token := recognize_operator;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;

    my_token := recognize_bracket;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;

    my_token := recognize_other;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;
    return fail_token;
  end;


  function recognize_token_string_token return Token is

  -- test:
  --   NO_MORE_TOKENS preceded by whitespace
  --   string token preceded by whitespace

    my_token : Token;
    RECOGNIZE_FAILURE : exception;

  begin
    my_token := recognize_white_space;
    -- ignore whether there or not

    my_token := recognize_no_more_tokens;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;

    my_token := recognize_string_token;
    if my_token.token_type /= FAIL then
      return my_token;
    end if;
    raise RECOGNIZE_FAILURE;
  end;


  procedure set_string(s : String) is

  -- pre:  s = string to be tokenized
  -- post: in_string has been set up (space-padded)
  --       current_token is set to first token of string

  -- test:
  --   once thru with non-null s

  begin
    in_string_pos := 1;
    for i in s'range loop
      in_string(in_string_pos) := s(i);
      in_string_pos := in_string_pos + 1;
    end loop;
    while in_string_pos <= IN_STRING_SIZE loop
      in_string(in_string_pos) := ' ';
      in_string_pos := in_string_pos + 1;
    end loop;
    in_string_pos := in_string'first;
    current_token := recognize_token_string_token;
  end;


  function has_more_tokens return boolean is

  -- post: returns true iff there are more tokens in the string

  -- test: once thru

  begin
    return current_token.token_type /= NO_MORE_TOKENS;
  end;


  function next_token return Token is

  -- post: returns next token of string, or NO_MORE_TOKENS token if none

  -- test: twice thru

  my_token : Token;

  begin
    my_token := current_token;
    current_token := recognize_token_string_token;
    return my_token;
  end;

end;

Re: Thanks guys..my project and my many problems

[tmoran]

>package body String_tokenizer is
  Why in heavens name did you post this package *body*, which is of zero
interest, just obfuscates things, and is 400 lines long?  In Ada, all the
information that should be needed by the user of a package lies in the
package *specification*.  Neither you nor any other user of
String_tokenizer should have any need to see its body.

Update - PLEASE SOMEBODY HELP!!!!

[Paul Gregory]

Well...the deadline is now 24 hours away....

I have the excellent program which Dennis Lee Bieber helped me with (thanks a lot Dennis you
are a great guy!) detailed below which works excellent on the GNAT compiler.

The problem is that the project is submitted online and run through a computer system and it
has to translate EXACTLY to the letter to gain a pass (and it is pass or fail...you could
have a program 100x better that does exactly the same but if it missed a full stop then it
won't accept it etc)

Now I need the program to compile like this for each of the four tests...

eg).

----------Sorry expected answer was-------------------------------
Do you wish to add words to the dictionary?: N

Please enter a sentence:
big.

The French translation of your sentence is:
grand.

----------Your answer however was---------------------------------
.
eg 2

----------Sorry expected answer was-------------------------------
Do you wish to add words to the dictionary?: N

Please enter a sentence:
Big cat is black.

The French translation of your sentence is:
Grand chat est noir.

eg3
----------Sorry expected answer was-------------------------------
Do you wish to add words to the dictionary?: Y
English word: red
Equivalent French word: rouge

Do you wish to add more words?: N

Please enter a sentence:
The dog is red.

The French translation of your sentence is:
Le chien est rouge.

eg4
----------Sorry expected answer was-------------------------------
Do you wish to add words to the dictionary?: Y
English word: blue
Equivalent French word: bleu

Do you wish to add more words?: Y

English word: sky
Equivalent French word: ciel

Do you wish to add more words?: N

Please enter a sentence:
Le ciel est bleu.

The English translation of your sentence is:
The sky is blue.
--------------------

However due to the way my program is structured it asks for a "source language" and a "target
language", yet I can't have this for the system.

Eg it compiles on GNAT like so...
-------------------------------------------------------------------------
Do you wish to add words to the dictionary? : n

Enter the Source Language:  English
Enter the Target Language: French
Please enter a sentence:
The big cat

The FRENCH translation of your sentence is:
Le grand chat


Enter the Source Language:
--------------------------------------------------------------------------

So the problems are - French in capitals - uses 'image is there a way round this ?
Don't want it to ask me for a source and target language.
Need a full stop after every sentence.

I'd be the happiest person in the world if somebody could help me resolve these problems and
I'd repay you in whatever favours, cash whatever you require as I've hit the wall and the
deadline is approaching and I'm going to almost certainly fail. :-(

Thanks for your time,

I know I'm very cheeky and as thick as a plank of wood...

but I can't stress how stressed out, physically ill and depressed this damn project is making
me feel.

Even if you can only work out how to do the first test Ie... just a straight one word array
comparison - your help would be greatfully appreciated.

Thanks a million,

Paul (code below, for anyone who is bored/kind enough to help)....

with Ada.Text_IO; use Ada.Text_IO;
with Ada.Strings; use Ada.Strings;
with Ada.Strings.Fixed; use Ada.Strings.Fixed;
with Ada.Characters.Handling; use Ada.Characters.Handling;
procedure main is

        MaxWord : constant := 40;
        MaxLine : constant := 120;
        MaxTranslations : constant := 100;

        type Languages is (English, French);    -- define known languages
        --      For each language named here, the Translations array will
        --      need another Head("word_in_new", MaxWord) added to each
        --      predefined set.

        subtype aWord is String(1..MaxWord);    -- assume words are smaller
        subtype aLine is String(1..MaxLine);

        LineIn : aLine;                 -- lots of poorly named scratch variables
        WordIn : aWord;
        WordOut : aWord;

        Source : Languages;
        Target : Languages;

        SOL : boolean;  -- Start of Line flag

        LineLen : integer range 0..MaxLine;
        Blnk : integer range 0..MaxLine;

        type TranslationSet is array(Languages) of aWord;

        --      use Head(string, size) with default pad to fill short
        --      strings to word length fixed strings
        Translations : array (1..MaxTranslations) of TranslationSet :=
                (       (Head("big", MaxWord), Head("grand", MaxWord)),
                        (Head("black", MaxWord), Head("noir", MaxWord)),
                        (Head("cat", MaxWord), Head("chat", MaxWord)),
                        (Head("dog", MaxWord), Head("chien", MaxWord)),
                        (Head("is", MaxWord), Head("est", MaxWord)),
                        (Head("small", MaxWord), Head("petit", MaxWord)),
                        (Head("the", MaxWord), Head("le", MaxWord)),
                        (Head("white", MaxWord), Head("blanc", MaxWord)),
                        (Head("Brighton", MaxWord),
                                Head("Brighton", MaxWord)),
                        others => (Head(" ", MaxWord), Head(" ", MaxWord))      );

        LastTranslation : integer range 1..MaxTranslations := 9;


begin

        -- allow user to add words to translation dictionary as long as space
        -- is available (note: dictionary is not saved between program runs!!)
        while LastTranslation < MaxTranslations loop
                Put("Do you wish to add words to the dictionary? : ");
                Get_Line(Item => LineIn, Last => LineLen);
                exit When To_Upper(LineIn(1)) /= 'Y';

                LastTranslation := LastTranslation + 1;
                for l in Languages loop
                        Put("Enter the new word in ");
                        Put(Languages'Image(l));
                        Put(": ");
                        Get_Line(Item => WordIn, Last => LineLen);
                        Translations(LastTranslation)(l) :=
                                Head(To_Lower(WordIn(1..LineLen)), MaxWord);
                end loop;
                New_Line;
        end loop;

        -- loop until user enters a blank line for Source language

        loop
                New_Line;

                Put("Enter the Source Language:  ");
                Get_Line(Item => WordIn, Last => LineLen);

                exit when LineLen = 0;

                Source := Languages'Value(WordIn(1..LineLen));

                Put("Enter the Target Language: ");
                Get_Line(Item => WordIn, Last => LineLen);
                Target := Languages'Value(WordIn(1..LineLen));

                Put_Line("Please enter a sentence: ");
                Get_Line(Item => LineIn, Last => LineLen);

                LineIn := Head(To_Lower(LineIn(1..LineLen)), MaxLine);

                New_Line;
                Put("The ");
                Put(Languages'Image(Target));
                Put_Line(" translation is: ");

                --      loop until the input line has been consumed.

                SOL := True;
                while LineIn(1) /= ' ' loop
                        Blnk := Index(LineIn, " ");
                        if Blnk > 1 and Blnk <= MaxWord then
                                WordIn := Head(LineIn(1..Blnk-1), MaxWord);
                                LineIn := Head(LineIn(Blnk+1..MaxLine), MaxLine);
                        else
                                WordIn := Head(LineIn(1..MaxWord), MaxWord);
                                LineIn := Head(" ", MaxLine);
                        end if;

                        WordOut := Head("<" & Trim(WordIn, BOTH) & ">", MaxWord);
                                                -- if no translation, output input value
                        for i in 1..LastTranslation loop
                                if WordIn = Translations(i)(Source) then
                                        WordOut := Translations(i)(Target);
                                        exit;
                                end if;
                        end loop;

                        if SOL then
                                WordOut(1) := To_Upper(WordOut(1));
                                SOL := False;
                        end if;

                        Put(Item => Trim(WordOut, BOTH));
                        Put(" ");

                end loop;
                New_Line;
                New_Line;
        end loop;
end main;

Re: Update - PLEASE SOMEBODY HELP!!!!

[Georg Bauhaus]

Dennis Lee Bieber <wlfraed@ix.netcom.com> wrote:
: Paul Gregory fed this fish to the penguins on Thursday 27 February 2003 
: 09:12 am:
: 
:        Addendum to my previous post...
: 
:        I /would/ be interested in seeing the spec file for the tokenizer...

Maybe you could ask John English, as Gregory appears to be in Brighton.
Gregory, have you seen him and asked him about the whys and hows of the
assignment?

Re: Update - PLEASE SOMEBODY HELP!!!!

[Paul Gregory]

It's getting hold of him !

A very busy man.



Georg Bauhaus wrote:

> Dennis Lee Bieber <wlfraed@ix.netcom.com> wrote:
> : Paul Gregory fed this fish to the penguins on Thursday 27 February 2003
> : 09:12 am:
> :
> :        Addendum to my previous post...
> :
> :        I /would/ be interested in seeing the spec file for the tokenizer...
>
> Maybe you could ask John English, as Gregory appears to be in Brighton.
> Gregory, have you seen him and asked him about the whys and hows of the
> assignment?

Re: Update - PLEASE SOMEBODY HELP!!!!

[Simon Wright]

Paul Gregory <pg16@bton.ac.uk> writes:

> It's getting hold of him !
> 
> A very busy man.

But nevertheless known to read c.l.a!