update the path info data and re-extract the filename so that widcards in
section names can match path info index files.

longer than 127 bytes.

the script until the entire POST data is read, so is sub-optimal, but it
works.

Now tested on FreeBSD (should work for any *nix).
CGI and FastCGI still needs testing on Win32.

Add support for FastCGI with an eye to making PHP significantly faster.  In
the webctl.ini file, add something like the following:

[*.php]
FastCGI=localhost:9000

To pass all PHP requests under that directory through to a PHP fastCGI server.
You can start the PHP FastCGI server with a command something like this:
php-cgi -q -b localhost:9000

This change has also refactored normal CGI support, especially the Win32
support.  Since this patch is currently untested, it's best to assume that
it breaks both *nix and Win32 CGI support and doesn't work for FastCGI either
at this point.

Also set the web server status to "Listening" once initializating is complete.
I'm not sure what the newly-introduced "terminated" flag is supposed to
indicate, but it's beign set to TRUE during recycle, so it's probably not
working as intended. Something cleaner probably should be done, but I don't
know exactly what Deuce was trying to accomlish with this variable.

cause of Robb Starr's reported "ERROR: 400 Bad Request" problem.

testing the Host value (vhost value already has port stripped).

Zero-length host names are not legal either.

If the TLS context couldn't be created, don't listen on the TLS interfaces.

TLS (HTTPS) will not be enabled/supported in the web server.
Lowered the log level of the cryptlib failures in handle_crypt_call() from
LOG_ERR to LOG_WARNING since these apparently are expected failures when
interacting with misbehaving HTTPS/TLS clients.

seperators) and return a 400 error otherwise.  With this check before the
vhost is set, eliminates the need to check when setting the log file name.

Also, lower-case hostnames for consistency and simpler validity checking.

filename if it contains any path delimiters.

Create a new function to close without one and use it instead until the
ring buffer is created.

This means moving ssl.o into libsbbs.so and out of libwebsrvr.so.
I'll fix Windows in a minute (?).

allocated attribute string values or NULL for an error.

Also, add free_crypt_attrstr() because Windows sucks.

This needs to be moved elsewhere and used everywhere an attribute is read.

addresses specified in the Interface line of sbbs.ini for the server.

before re-reading them.

xpms_add_chararray_list instead.  Prevents having to free string lists
in the sbbs.ini parsing code.

the same as startup->port anymore.

Was broken on IPv6 literals such as http://[dead::beef]/ as well as any other
URI containing a colon.

New Features:
- Multiple bindings for each service
  Use comma-separated interfaces on Interface= lines in the ini file.
  Default is now "0.0.0.0,::"
- IPv6 support
- TLS support for the webserver and (non-static) services
  New TLS option in services.ini (ie: Options=TLS)
- Decrease LEN_SCAN_CMD to 35 chars, increase the CID field to 45 chars,
  and rename the MAIL_CMD string to IPADDR.  I think this frees up the
  note field for SysOp use.

for HEAD requests.

Content-Length header for static files.

scripts global scope (as opposed to js.global which is the instance global
scope), and having exit() define exit_code in js.scope instead of js.global.

This also sets exit_code in js.scope to null when preparing to execute a new
script.  If a new script starts in the same scope as an old one, the old
exit_code value will be destroyed.

This should only impact scripts where js.global != js.scope (bbs.exec()ed,
and mailsrvr)