Letter GlyphRectangle with Height == 0

[simonedd]

Hi,

In this document, all the letters have a GlyphRectangle with a Height equals to zero.
A1-RU-101-S.pdf

[BobLd]

Hi @simonedd, have you tried compiling and testing this document with the current master? Some fixes have been push to fix bbox height recently - but your document might be another case of this issue

[simonedd]

Yes, I tested with the latest version.

[BobLd]

The issue comes from the Type0CidFont class (in UglyToad.PdfPig.PdfFonts.CidFonts).

I've added the following property private readonly double scale; it the top, that is set in the constructor:
scale = 1 / (double)(fontProgram?.GetFontMatrixMultiplier() ?? 1000);

Then in the GetBoundingBox(int characterIdentifier) function:

PdfPig/src/UglyToad.PdfPig/PdfFonts/CidFonts/Type0CidFont.cs

Lines 89 to 118 in 68dd3f5

	public PdfRectangle GetBoundingBox(int characterIdentifier)
	{
	// TODO: correct values
	if (characterIdentifier < 0)
	{
	throw new ArgumentException($"The provided character identifier was negative: {characterIdentifier}.");
	}
	if (fontProgram == null)
	{
	return Descriptor?.BoundingBox ?? new PdfRectangle(0, 0, 1000, 0);
	}
	if (fontProgram.TryGetBoundingBox(characterIdentifier, out var boundingBox))
	{
	return boundingBox;
	}
	if (Widths.TryGetValue(characterIdentifier, out var width))
	{
	return new PdfRectangle(0, 0, width, 0);
	}
	if (defaultWidth.HasValue)
	{
	return new PdfRectangle(0, 0, defaultWidth.Value, 0);
	}
	return new PdfRectangle(0, 0, 1000, 0);
	}

I've amended the code as follow:

        public PdfRectangle GetBoundingBox(int characterIdentifier)
        {
            // TODO: correct values
            if (characterIdentifier < 0)
            {
                throw new ArgumentException($"The provided character identifier was negative: {characterIdentifier}.");
            }

            if (fontProgram == null)
            {
                return Descriptor?.BoundingBox ?? new PdfRectangle(0, 0, 1000, 1.0 / scale);
            }

            if (fontProgram.TryGetBoundingBox(characterIdentifier, out var boundingBox))
            {
                return boundingBox;
            }

            if (Widths.TryGetValue(characterIdentifier, out var width))
            {
                return new PdfRectangle(0, 0, width, 1.0 / scale);
            }

            if (defaultWidth.HasValue)
            {
                return new PdfRectangle(0, 0, defaultWidth.Value, 1.0 / scale);
            }

            return new PdfRectangle(0, 0, 1000, 1.0 / scale);
        }

This is the result, which seems to be in line with what Acrobat reader displays (Letters in blue, words in red):

I am really not sure it would work for every document. I'll try to push a PR if I manage to have a better understanding.

[BobLd]

closed as fixed in 0.1.5

[mayurjansari]

unicode character has height 0