Parser chokes on `sobelow_skip` comment with trailing content

[mikebveil]

Consider a code snippet like this:

# sobelow_skip ["DOS.StringToAtom"] (atoms are a fixed set of table names in the DB schema)
defp get_table_sizes() do
    …
end

The explanation given at the end of the sobelow_skip line will cause parsing to silently fail for the whole file when --skip is used. This is a insidious failure, since it will seem like the comment is working but we're also now ignoring any other findings in the file.

This is because of a regex that runs during parsing to convert these comment lines to @sobelow module attribute statements. The regex replaces the first part of the line but not any trailing content, which becomes garbage at the end of the attribute statement.

I see two ways to fix the regex:

1. Don't match if sobelow_skip is followed by other content. So we fail to skip the finding but the failure is visible to the user.
2. Match the whole line and delete any trailing content. This seems the more reasonable way to go, and I've opened a PR (Issue #168: trailing content on an sobelow_ignore line silently breaks parsing #169).