[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GSOC Project
From: |
Werner LEMBERG |
Subject: |
Re: GSOC Project |
Date: |
Thu, 18 Mar 2021 07:32:48 +0100 (CET) |
Hello Aman,
> I am interested in the project -_"Develop a test framework for
> checking Freetype's rendering output"
Great!
> 1. What signifies the meaning of word "font corpora" as i haven't
> found anything about by googling it.
A good and valid question.
'Corpora' is the plural form of the Latin word 'corpus'. A 'font
corpus' is a set of fonts that kind-of belong together, or have been
collected for a common purpose. For example, each time the FreeType
fuzzer finds a problem, the offending fuzzed font gets added to such a
corpus.
For the problem of checking the rendering output, collecting fonts for
a similar corpus is necessary. Ideally, the elements of this corpus
should be tiny fonts of various font formats that either
(1) increase the code coverage of the FreeType (mainly to check error
handling)
or
(2) test all aspects of the rendering pipeline, covering as much
features as possible.
For the GSoC task at hand, (1) can be ignored more or less, since it
doesn't contribute to visual differences.
A serious problem is that many important fonts are commercial fonts,
which can't be directly added to a freely available corpus. Most
fonts, however, allow subsetting, and collecting subsetted fonts is
similar to collecting PDF files that contain them and thus legally
allowed.[*]
Let's assume that glyph 'foo' of font 'bar' causes a rendering
problem. A logical step would be to create a subsetted font of 'bar'
that contains just this single 'foo' glyph (more or less).
> 2. Where can i find instructions to build free type
Well, as with almost all source code package you should start with the
`README` file :-)
https://gitlab.freedesktop.org/freetype/freetype
Werner
[*] At least this is what I think. It should be otherwise possible to
construct a minimum PDF file around a subsetted font.
- GSOC Project, Aman Kapoor, 2021/03/17
- Re: GSOC Project,
Werner LEMBERG <=