Nick Wellnhofer

AFAICS, reproduction now uses `TEST_TIMEOUT-5` which results in 60 seconds on OSS-Fuzz. If I specify a custom timeout in the `.options` file, for example 20 seconds, then a test case...

Most of these warnings seem to come from `-Wconversion` which I wouldn't recommend personally.

> Maybe you want to explain why you do not recommend that specific flag? `-Wconversion` typically generates hundreds of warnings which are subsequently "fixed" by adding explicit casts. But adding...

This is a feature I'd use for the Apache Clownfish documentation generator as well. As @foonathan mentioned, if a single output format doesn't support HTML, all HTML blocks and inlines...

On 04/01/2017 17:29, John MacFarlane wrote: > The security issue is orthogonal. We have the --safe option > (and corresponding option for the C API). That will ensure > that...

> If you use a custom renderer you can always do the sme thing and ignore raw HTML, right? Yes, of course. I'd still find an option to ignore all...

@foonathan The spec doesn't care whether an output format supports HTML or not. It's only about the text -> AST transformation and ignoring HTML results in a different AST.

There's also issue #121 but this is caused by a GCC bug. The biggest problem I see is that re2c generates needlessly repetitive code for quantifiers like `{1,31}` which is...

That's bascially a bug in cmark's XML renderer. Markdown allows C0 control characters, but XML 1.0 doesn't (except some whitespace). The only option I see is to encode control characters...

You'd have to strip all control codes between 0x00 and 0x1F, except 0x09, 0x0A and 0x0D. For completeness, you should also strip Unicode code points 0xFFFE and 0xFFFF: https://www.w3.org/TR/REC-xml/#charsets