telnet_gate() now enables the telnet command pass-through after negotiating
necessary options. This doesn't quite fix the problem reported with telgating
to nethack.alt.org, but it's a step in the right direction.

Create and include in packet TOREADER.EXT if QWKEsupport is enabled
(MultiMail keys of this file for QWKE support, so without, no QWKE features are
enabled in MultiMail).
Parse TODOOR.EXT if included in REP packets (for adding/dropping subs or
setting/resetting pointers).
Parse To:, From:, and Subject: QWKE kludge lines and use if/when appropriate
(e.g. to defeat QWK 25-char header field limits).
Create To:, From:, and Subject: QWKE kludge lines in QWK/REP packets when
QWKE support is enabled and those fields exceed QWK limits (25 chars).
Also, legacy SyncQNET kludge lines (@VIA, @TZ, etc.) may now exist in the
top of the message body in any order.
Note: current versions of MultiMail do not support "To" fields > 25 chars, even
in QWKE mode (though I have a patch pending) and do not (yet) support
Synchronet HEADERS.DAT file (rendering QWKE kludges unnecessary).
These are major changes in the QWK/REP creation/parsing code, so testing
(especially with QWKE-compliant offline readers) and bug reports are welcome!

commonly modified by multiple threads, potentially simultaneously, resulting in
bad values (e.g. active_clients = -1).
Updated copyright date to 2011.
Fixed Win32 compile bug in main.cpp introduced in last commit.
Eliminated some unused variables and updated some times (e.g. ulong in place
fo uint32_t).

Updated copyright dates.

make system.matchuserdata return an array of all matches if match_all==true, starting at usernumber if specified. (userdatdupe now takes an additional paremeter)

(potentially) non-blank for external JS modules configured in SCFG->External
Programs, as well as those executed from jsexec. If the startup_dir is
specified in SCFG and no path is given for the .js file, it will look in the
startup_dir first. js.startup_dir will be an empty string for all other cases
(currently).

Reading Mail prompt with multiple undeleted mails remaining, instead make
(K)ill All Mail command available, which can be still be reversed before
quitting the reading mail prompt.

Added FTN PID header field to all messages created by sbbs.

system (in a multi-computer BBS) the error occurred.
For the mail server:
- keep track of the number of times each mailproc handled a message
- display counters when blocking/filtering messages in log output
- advertise 8BITMIME in the ESMTP EHLO response
- do not route mail to mailproc when the execution requirements are not met
  (e.g. ARS, DNSBL, SPAM, etc.)

to a host name before comparing against host.can.

created:
js.exec_path - full path and filename of file executed
js.exec_dir - directory of executed file
js.exec_file - filename executed (with no path)

Configurable via ctrl/sbbs.ini file JavaScriptLoadPath key value
(comma-separate listed of search directories), default value is "load".
This list is exposed in the JS object model via the js.load_path_list array
(may be modified by scripts).
For JSexec, the default load path list may be over-ridden with the '-i' option.
For relative load paths (e.g. not beginning with '/' or '\'), the path is
assumed to be a sub-directory of the (configurable) mods or exec directories
and is searched accordingly.
So, by default, load("somefile.js") will search in this order:
mods/load/somefile.js
exec/load/somefile.js
mods/somefile.js
exec/somefile.js

This resolves JasHud's issue with question text.dat items over > 128 chars
being truncated.

instance (e.g. sbbs->useron) - this should solve the problem with an online
Guest user not getting account changes made via JS scripts (e.g. logon.js).
This may create other unforeseen problems. :-)

Added HOST and IP keywords to allow restricted access/privileges to/for
specific remote hostnames or IP addresses (wildcards allowed).
All string-argument type ARS keywords (e.g. SHELL, PROT, etc.) now support .can
style wildcards.
The current remote client is now used for protocol, host, and IP ARS checking,
when available, so this requires passing the client pointer around (which
explains why so many files are touched by this change) and takes care of a
long standing to-do item (the user's 'modem' value was used for the PROT
value checking, which was not always correct).

(non-deleted) mail in the mailbox, prompt the user to delete all the mail
waiting.

sbbs_t::rputs() now:
* Can accept an optional length argument which allows it to output strings
with embedded NULs if desired.
* If a call to outcom() fails, terminates and returns the sent number of chars.
* Performs Telnet IAC escaping when appropriate.
* Checks online value in while() loop.
* No longer output to log if "on local" - event scripts aren't expected to use
this function.
JS console.write() and writeln() now use rputs() instead of putcom() so that
Telnet IAC chars are escaped when necessary and saveline/restoreline will work.
sbbs_t::putcom() now returns the number of chars sent successfully.

argument and buffer and retruns an optimized ANSI sequence to generate
the desired attribute.
console.ansi() method now support this usage as well, so msgutils.js's
attr_to_ansi() should no longer be needed.

place to do it, but may need some special handling for HTMLterm mode.
Parse columns from ANSI cursor position reports.
Request Telnet "Negotiate Window Size" option during answer since this is more
reliable than ANSI cursor position reporting (especially with terminals that
wrap cursor positioning requests).
Improve validation of ANSI cursor position reports.

Created valid_ctrl_a_code() which returns TRUE if the passed ctrl-a code is
valid (safe) for message networking, user input.
Only QWK-export valid ctrl-a codes.

Constified the arguments.
Created exascii_to_ascii_char().
Created ctrl_a_to_ascii_char().
strip_ctrl() and prep_file_desc() now takes 2 args.
strip_ctrl(), strip_exascii() and prep_file_desc() now strdup() the input
string and if the dest arg is NULL, returns the allocated and modified copy.
sbbs_t::bputs() now handles ^AZ.
Change Ctrl-A code message network exporting:
1. ^AL, ^A<, ^A[, and ^A] are always translated to ASCII equivalents.
2. ^AZ will terminate the message body prematurely if found.
3. Ex-ASCII (CP437) chars will be translated to ASCII equivalents (rather than
'*') for "ASCII Only" subs.
For import, don't allow Ctrl-AP or Ctrl-AL.
For SBBSecho:
Added support for specifying a link's address on the command-line (e.g. for
use with -h).
Mail server now strips Ctrl-A codes from message bodies for sending via POP3
or SMTP.

kludge lines).
Created common sbbs_t member function for generating QUOTES.TXT file/path.

data/error.log file.
Also, fixed the possible infinite loop in main.cpp output_thread() when send()
fails for the system output socket.

While you can't run with a missing text.dat yet, you should be able
to run with a zero-byte one.

ansi_restore() in place of dumb macros that don't check the current terminal
support.

and context. This allows the event thread to perform its own JS cleanup without
deleting its sbbs_t instance (causing the exception on sbbs/sbbsctrl shutdown).
This replaces the change committed in rev 501.

by properly const-ifying the appropriate functions and variables.

Not yet tested on Win32

regular expression suppor).
Reverted findstr() and trashcan() to read the file a line at a time rather than
load the entire file (potentially very large) into a string list.

* Parses/consumes QWK headers.dat files
  - No more to/from/subj length limits
  - Extensive header details transferred for each message
  - IP/hostname filters (.can files) are applied to appropriate header fields
  - Code cleanup in the QWK functions
* New functions to read/parse/search filter (.can) files as string lists
  - Performance boost - no need to open/read/close .can file for each message
* More thread-safe Message-ID retrieval/generation (ftn_msgid and get_msgid)
* Better Message-ID generation for misconfigured systems (e.g. no hostname)
! These changes require the latest smblib and xpdev libraries !

A lot of strcpy->SAFECOPY and sprintf->safe_snprintf/SAFEPRINTF conversions
and other general code cleanup too.

If default value is not zero, it is the value returned if enter is
pressed with no input.  In this case, zero is allowed to be entered.

Fix DLLCALL usage.