Bounding box for text

[de-code]

Is it possible to get the bounding box for text / characters (for the original / embedded font).
The SVG includes the position of the characters but not the bounding box.

There is an svgx:width which appears to be using a different unit (and no height, although there is the font-size).

[petermr]

Greetings

On Fri, Apr 21, 2017 at 3:04 PM, Daniel Ecer ***@***.***> wrote: Is it possible to get the bounding box for text / characters (for the original / embedded font). The SVG includes the position of the characters but not the bounding box. There is an *svgx:width* which appears to be using a different unit (and no height, although there is the *font-size*).

The actual BoundingBox will depend on the precise characters and glyphs you are using - e.g. bold might push the height up a few fractions of a pixel. At present the height of the bounding box is heuristically calculated from the font-size by a simple multiplier. The width is a parameter reported by PDFBox and is given by the document/font. It's the guideline for whether to include spaces between characters - the average x-distance to the origin of the next character as determined by the font setter. In principle it could vary with kerning but in practice is a single integer for each character. To get the intercharacter spacing = svgx:width / 1000. * font-size (the svgx is because SVG has no attribute for character width). It seems to work most of the time. If svgx:width is not reported (e.g. in diagrams we have to guess in some way).

…

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#2>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAsxSzvLtqXTkCOZ-5SEvt9ZKnl2QCMUks5ryLdggaJpZM4NEWHY> .

-- Peter Murray-Rust Reader Emeritus University of Cambridge +44-1223-763069 and ContentMine Ltd

[de-code]

Thank you for the response Peter.

Okay, I thought the exact bounding box of the glyph might be available when it is 'rendered'. I will for now estimate it as you suggested.

[petermr]

On Fri, Apr 21, 2017 at 5:06 PM, Daniel Ecer ***@***.***> wrote: Thank you for the response Peter. Okay, I thought the exact bounding box of the glyph might be available when it is 'rendered'.

All PDF documents must contain the glyph information in some form. Sometimes it is one of the 14 Allowed fonts which every PDF must honour. Sometimes it gets Fonts from the host, but it also has a font library and one can retrieve the font. This can be either a bitmap (non-scalable) , or a path (scalable) in SVG.

I will for now estimate it as you suggested.

If you are doing very precise kerning you will need the glyphs. I spent some years working with M$ on chemistry and maths and they have very intricate systems. But if you just want to see whether one word overlaps another then the font-height will probably be fine.

…

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAsxS49DeR_Cwx6442Leik-_W7hHcHM7ks5ryNPpgaJpZM4NEWHY> .

-- Peter Murray-Rust Reader Emeritus University of Cambridge +44-1223-763069 and ContentMine Ltd