Problems with extended Postscript table

Hello everybody,

I’m having troubles with the Postscript table. I’m trying to write to a .ps or .eps file letters like çãá, but the 2nd table presented by pstable.C in a ROOT canvas is completely different from the generated For example, I can not find the ã which should be \343 by the ROOT canvas. Even other symbols are completely strange.
But, my generated is exactly the same as the image showed in the TPostscript reference: [link][/link]

What is right the showed in a ROOT canvas or the generated postscript?
Where is the ã symbol?

Cheers, Luís

I guess you are using TText. I would recommend to use TLatex which provides the following accents:

// ** Accents
// ----------
// #hat{a} = hat
// #check = inversed hat
// #acute = acute
// #grave = agrave
// #dot = derivative
// #ddot = double derivative
// #tilde = tilde

Hello Olivier,

It would be a good workaround, but I get “Posiç ão” instead of “Posição” in my histogram title. The histogram tiltle is set by:
prim->SetTitle (“Banda 1 - Cosseno de Posi\321#tilde{a}o”);

Do you have any other way?

Cheers, Luís

Hi Olivier,

I have found out other problems with PostScript output using TLatex. See in the attached file the results for #acute{i}, which comes over i’s dot (it looks bad). Almost the same problem have been observed for #acute {o}, wich is over the letter ‘o’. Another problem happens when I write the title for Xaxis with:

prim->SetXTitle (“Energia de sa#acute{i}da (E_{sa#acute{i}da}/E_{inicial})”);

Is this a ROOT bug or am I doing something wring?

Thanks, Luís
histograms.tar.gz (11.4 KB)

I see your point. Using code with “” in TLatex doesn’t work very well because their size is consider to be the size of four characters. That’s why you get this white spaces. So, in your case, it is better to use TText. But you will miss “ã” . I can add it in the available PS characters. There is still some free spaces. I can also add í … is there some other you would like to have in PS like ì ? let me know.

Hello Olivier,

Currently I will need these letters: á, é, í, ó, ã, õ, â, ê, ô and ç (small and capital).
I need those letters because now I am writing my Ph.D. Thesis in Portuguese and I do not need the complete Latin-1 table, but would be nice to get it implemented one day.

Let me know when it is implemented and I will get the it from the CVS.

Thanks, Luís

Note that some of these characters are already there:

é = \323 É = \324
á = \366 Á = \367
ç = \321 Ç = \322
ê = \327 Ê = \330
â = \276 Â = \300
ô = \342 Ô = \344

Are missing:

í, ó, ã, õ, Í, Ó, Ã, Õ

I will add them.


I forgot to mention but I need the ú and Ú too.

Thanks, Luís

Hi Luis,

I have added these new symbols in the new TPostScript.cxx file available in the CVS head. I have also modified the macro pstable.C in order to show them.

Regards, Olivier

Hello Olivier,

Finally I could compile and install the latest ROOT version from CVS.
Some of results are good, but when I mix those characters with TLatex I got weird results. See what I mean in the attached file.
In the b1-cos.eps, the x title has a little shift in the ‘kugel’ word toward to left just above R. The same can be seen in the b1_en.eps, with Esaída in the x title. Another problem can be seen in the title of b1_x.eps, with dcp crossing the box. But those can be workarounded by the user.
The worst result can be seen in the b1_mu.eps, where I used the string “Trajet\201ria de sa\203da cos(#mu)” as graphic’s title and I got a misplaced ‘mu’ and a colapsed ‘()’.

I guess this is happening because the characters is not well displayed by ROOT, so it can not calculate correctly the needed space and position.

Cheers, Luís
results.tar.gz (6.17 KB)

I told you, you should not use these characters in TLatex. They are usable in TText only (see previous posts)