Overview of changes:

 *  Removed source-highlight requirement

 *  Fixed sfdoc dependencies

 *  sfdoc will now fail gracefully if man version is too old

man version 1.x does not support local file mode.  sfdoc is mostly for
development purposes, so instead of jumping through hoops to make it
work; we assume user can view the pages in other format or on a less
obsolete distro.

Turns out moving sfdoc code to dedicated module *and* having the binary
in main package created cyclic dependecy.  However, there's easy way out;
let's just run with it as long as it works on most distros.

source-highlight is not necessary (-s|--src will fall back to plain text
if it's missing) and requiring it strictly breaks RHEL6 build.

deb users can follow Suggests field; rpm users will have to find out about
it and install it separately.  (Optional requires do exist in newer RPM,
but it's not worth the effort/ifdef bloat.)

Overview of changes:

 *  Many docstrings fixed, updated and added

 *  Fixed problems with POD encoding

 *  sfdoc now supports --which and -s|--src to help with debugging

 *  Improved error handling in hidden object filtering in sfdoc.sh

 *  sfembed will not force debugging mode anymore

 *  sfembed now removes its temp directory

 *  Codebase maintenance

This has been disabled during early debugging phase and apparently I
just forgot to re-enable it.

Morale: don't use `:` to disable stuff; use `#` (ideally at column 0),
which stands out much more.

In some cases the file is not slurped so declaring encoding at the tail
is useless.

Overview of changes:

 *  Revamped version directive format

    This is a backwards-incompatible change, but is an exception because
    (well, we're in 0.x.y phase and) it's only used in manpage
    generator, so at worst people will lose the version info from
    manpage footer.  Also it hasn't been around for a long.

 *  Moved most sfdoc code to separate module

    Yet another module has been born.  This enables better design, reuse
    and also easier testing.

 *  mkusage will not force empty line before indent mode (-i)

    This way it's much more useful; one can produce much more concise
    listings (*and* have them indented, *and* read them from array).

 *  Added -o|--only option to sfdoc for single-directory sets

 *  sfdoc will now correctly report module version (not Shellfu version)

 *  sfpath now supports -V, short version of --version-semver

 *  sfpath now makes sure to sort items consistently under C locale

 *  Added __usage sub-command to shellfu()

 *  Moved perl dependency correctly to shellfu-bash-inigrep

 *  Fixed bad shebang line (thx lintian)

 *  Solved or muted last ShellCheck issues

 *  Revised coding style an docstrings

 *  Codebase maintenance

Useful when working with a private set of modules; to see what's inside,
etc.

Allow for re-use and more independent testing.

This was not intended,  only indentation was intended (no pun indented).

Seriously, omitting the newline makes the -i mode more useful as we can
now use syntax like:

    mkusage "topic" \
      --
      "topics:" \
      "${topics[@]}"

That is much less code, especially if the list we wanted to include is
already in array.

Use CamelCase for inheritable variable names, avoid re-hashing each
module twice and add function and variable docstrings, remove unnecessary
whitespace

...or '(no version)', if version is not present.

This also changes format; the 'K: V' format is problematic as it makes
it hard to separate from other directives.  'K=V' is now the only
accepted format.

Note that essential functions should be as fast as possible so they
probably should not inspect their arguments.

Overview of changes:

 *  Added proper obsolete declarations

Tested on Fedora, praying for it to work on Debian.

Overview of changes:

 *  Added missing dependencies

 *  Removed unnecessary dependency

 *  Codebase maintenance

Overview of changes:

 *  Override SHELLFU_INCLUDE in test runs

    Otherwise modules are actually loaded from system, if they are
    installed.

 *  SHELLFU_PATH now has preference over SHELLFU_INCLUDE

    This should have been all along; SHELLFU_PATH would not be able to
    force a particular module version in cases like testing.

 *  Added shellfu() docstring

 *  Minor co

 *  Improved and updated test header

Should have been for a long time now.

Otherwise we're not really testing "this" shellfu but the installed one.

This should have been all along; if SHELLFU_PATH is not preferred,
one cannot override it eg. in test runs.