Move the cluster test inside the space-to-read check with the other tests.

We'll sync keyframes on that track next commit.

* Loop over elements in input buffer.
* Most are literally copied with their contents without
  inspection into the header or data buffers as appropriate.
* Some only copy the element header, to allow inspecting
  children elements.
* Cluster elements are identified and used as sync points.

No probing is done for keyframes *yet*

This will allow leaving unparsed data in the input buffer to wait for completion.

This simplifies some fragile "was the last chunk a sync point?" logic.

This should help make future changes to the code clearer.

This adds support for allow-deny kind of rules with the matchfile API.

See: #2119

ban and allow file support has been moved into a generic implementation
that can be re-used by later code such as proxy matching or other
blacklisting.
See: #1959

When unset no default is used but the feature is disabled.
(The Shoutcast listen socket is not opened at all.)

Thanks to Jeremiah Rogers for pointing out.

This fixes a bug that could cause the global listener count of Icecast to
be negative. This was caused due to wrong handling of errors in the
format_prepare_headers function.

Fixes: #2199

Thanks a lot to Simeon Völkel (0xBD4E031CDB4043C9) for reporting
and investigating the bug.

This fixes some possible segfaults that could happen if the header
(re)allocation fails.

Also adjust formatting a bit, kill some trailing spaces.

Fixes: #2198

If there is a single "-" or it is only accompanied by whitespace, then
it wouldn't get escaped as it was detected as as a number. This would
produce invalid JSON.
This makes the numbers part ignore such and thus ensures it gets escaped
as text.

There might be other corner cases related to number parsing, this should be investigated.

https://github.com/doekman/xml2json-xslt/pull/20

 Some numeric values would not get formatted to valid JSON. This
 incorporates the suggestions from issue #15.

Also fixed whitespace of original patch.
Additional fix for related issue will follow.

This (hopefully) corrects the usage of realloc() as free().
This patch is part of the series based on Denis Denisov's input.

This let util_url_escape() handle NULL passed as parameter.
In case the parameter is NULL it will also return NULL.

This patch also does some cleanup of the code such as migration
away from int and thus avoiding future failtures.

This is a re-apply of 3baa4e46
as git refuses to cherry-pick.

This is a first commit to correct minor problmes reported by
Denis Denisov. There is likely another commit to correct some more
such problems.

Thank you Denis Denisov.