Hyphenation routine to Spanish

[typo]

Does anyone know of any hyphenation routine to Spanish, Portuguese or even to English?

[typo]

Please answer.

[jesusr]

if you find some implemented in pascal let me know.

if nobody answer is probably because nobody knows.

... the problem is who is nobody 

[typo]

I would be satisfied with a routine that recognizes 90% of syllables without using dictionaries.

[jesusr]

There is a rudimentary hyphenator routine who use undeclared rules in LazReport. Check BreakWord routine in Lazarus/components/lazreport/source/lr_class.pas:2458

[José Mejuto]

I would be satisfied with a routine that recognizes 90% of syllables without using dictionaries.

Hello,

here you have one spanish in C http://tip.dis.ulpgc.es/es/silabas/descargar which should be quite easy to port to pascal, take care that it uses ANSI characters and maybe you need UTF8, in that case the UTF8Tools package will help you.

And here there is a description of the basic rules to decompose words in spanish http://f01.middlebury.edu/SP210A/gramatica/acentu-index.html

[theo]

I've once written a wrapper class for the hyphen lib.
http://sourceforge.net/projects/hunspell/files/ see hyphen 2.5
There are many dictionaries for it:
http://wiki.services.openoffice.org/wiki/Dictionaries#Spanish_.28Spain.2C_....29
See "Hyphenation" under "Spanish"

This is probably the most professional solution you can get.

There are different ways to use the wrapper. For ex:

Code: [Select]

procedure TForm1.Button1Click(Sender: TObject);
var
  hList: TList;
  hyp: THyphen;
  i, Start: integer;
  AWord: UTF8String;
begin
  hList := TList.Create;
  hyp := THyphen.Create('/home/theo/install/hyphen-2.5/hyph_es_ES.dic',
    '/home/theo/install/hyphen-2.5/.libs/libhyphen.so');
  AWord := 'electroencefalografistas';
  Start := 1;
  hyp.Hyphenate(UTF8LowerCase(AWord), hList);
  for i := 0 to hList.Count - 1 do
  begin
    Memo1.Lines.add(Copy(AWord, Start, PtrUInt(hList[i]) - Start + 1));
    Start := PtrUInt(hList[i]) + 1;
  end;
  Memo1.Lines.add(Copy(AWord, Start, Length(AWord) - Start + 1));
  hyp.Free;
  hList.Free;
end;

Result:

elec-
tro-
en-
ce-
fa-
lo-
gra-
fis-
tas

If you need the wrapper, tell me.

[typo]

Yes, I want it.

BTW, trying to compile uHunSpellLib.pas I receive this error message:

uHunSpellLib.pas(64,28) Error: Illegal type conversion: "ShortString" to "^Char"

in this line:

DLLHandle := LoadLibrary(PAnsiChar(libraryName));

Any suggestions?

[theo]

Yes, I want it.

OK, http://www.theo.ch/lazarus/hyphen.zip

You do not need Hunspell for Hyphenation.

[typo]

Thanks.

[typo]

Is it necessary the Hyphen.Dll? How can I get it? Is it possible to get the source code of it?

[theo]

How can I get it? Is it possible to get the source code of it?

All written above. Direct donwload: http://sourceforge.net/projects/hunspell/files/Hyphen/2.5/hyphen-2.5.tar.gz/download

Easy to build on Linux. I don't know about Windows. Probably needs mingw to build.
http://www.mingw.org/
or
http://www.bloodshed.net/devcpp.html

[theo]

Here we go: http://www.theo.ch/lazarus/hyphen.zip
The dll is now included. I don't know exactly what I did, but I built it somehow with Dev C++ and it seems to work :-)

[typo]

OK, the project is now compiling and running. Thank you very much.

[theo]

OK, the project is now compiling and running. Thank you very much.

Good. You should probably convert the *.dic files to UTF-8 so that they work properly.
You can see in the first line of the file hyph_pt_PT.dic that it is in ISO8859-1.
Convert the file and change the header to UTF-8.
This could also be solved in code, but would be more complicated.