Problem displaying Unicode characters in Pango/Cairo/DirectFB

I have an issue while displaying some Japanese characters using Pango. First few times the text comes properly, and after that I see the junk characters. It looks as if the correct characters are displayed, and are overwritten by something else. I have ensured that the string that we are passing to pango_layout_set_markup() is correct. So I am suspecting some internal setting in Pango/Cairo/DirectFB is causing the problem. I may be missing setting something to NULL, freeing something or alike. The sequence of calls I am having is listed below. Can anyone please look at this and see what is wrong with this?

Thanks in advance

/*********** During initialization ***********/
// Initialize DirectFB
DirectFBInit(&argc, &argv);

// Create DirectFB instance

/*********** For each set of texts ***********/
// Get the DirectFB layer
dfb->GetDisplayLayer(dfb, DLID_PRIMARY, &layer);

// Create DirectFB window
layer->CreateWindow(layer, &wDesc, &window);

// Get DirectFB surface
window->GetSurface(window, &surface);

// Create Cairo surface on DirectFB surface.
cairo_surface = cairo_directfb_surface_create(dfb, surface);

// Get Cairo handle
pCrHandle = cairo_create(cairo_surface);

/*********** For each Text information ***********/
// Get Pango Layout using Cairo handle
layout = pango_cairo_create_layout(pCrHandle);

// Set Text
pango_layout_set_markup(layout, string, -1);

// Show the layout
pango_cairo_show_layout(pCrHandle, layout);

// Free the layout

/*********** End of Text information ***********/

// Destroy Cairo handle

// Destroy Cairo surface

// Release DirectFB surface

// Destroy DirectFB window

// Reset DirectFB layer
layer = NULL;

/*********** End of set ***********/

Open in new window

LVL 37
Harisha M GAsked:
Who is Participating?
Harisha M GConnect With a Mentor Author Commented:
Finally I found that it is a bug in the Pango library itself, and have raised it in the pango bug tracker.
I'm not sure whats going on with your code specifically, however, if the layout engine you are using is written, or derived from, C and it compiled multi-byte everytime a unicode character with a leading byte of zero (0) is encountered it'll ignore that byte and junk the other.

Make sure you are sending two byte cuplets as characters, and further make sure the layout engine supports the unicode correctly.

If in doubt you may need to fudge things over by sending the two bytes of the unicode character (short) together and processing the ones over and above the first page (255) to offset to the correct unicode character yourself...

Failing all that, if you are building [as in compiling from code] your layout engine yourself try to switch the project to "unicode" in the options.  If you're just using an off the shelf library however it can be difficult to work out what byte encoding they're using.

Hope you find the answer out there somewhere, I'm just pointing out the obvious...
Harisha M GAuthor Commented:
The problem wasn't solved, and I found that it's a bug with Pango library itself.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.