Behdad Esfahbod
Behdad Esfahbod
Duplicate of https://github.com/fonttools/fonttools/issues/2937
While having race conditions is bad, I don't see this change happening any time soon.
> Does the fuzzer literally generates a random string? That probably won't even ever pass the very early loading of the sfnt font header.. The fuzzer is smart... You'd be...
> @behdad would you like me to add yours too? e.g. https://github.com/google/oss-fuzz/blob/39eebf89b2bd2522935724d43e4e245a8b59aae9/projects/harfbuzz/project.yaml#L5 ? I just submitted https://github.com/google/oss-fuzz/pull/9926 But yeah now I remembered why I used my gmail address before because...
TTFont in general does not guarantee loading broken fonts. One way I can think of as useful fuzzing it is cross-fuzzing with ots-sanitize. That is, first sanitize font with ots-sanitize...
> Am not sure how much work this would require though. It would require building OTS with fuzzing instrumentation and driving the fuzzing with the C fuzzer infrastructure though, so...
Not currently. But I agree we should make that a module. Maybe try to unify with the CFF-to-CFF2 as well.
> > But I cannot find any lightweight CLI way to losslessly convert TTF TrueType to OTF PostScript. > > A quick update with the solution to my question since...
@khaledhosny What's the consensus here?