====== JavaScript ======

Synchronet uses Mozilla's [[https://developer.mozilla.org/en-US/docs/Mozilla/Projects/SpiderMonkey|JavaScript]] engine (a.k.a. JavaScript-C or "SpiderMonkey") for its preferred local scripting environment.

===== Versions =====

  * Synchronet v3.14 used JavaScript v1.5.0
  * Synchronet v3.15 used JavaScript v1.7.0
  * Synchronet v3.16 uses JavaScript v1.8.5

===== About =====

You can learn about the core JavaScript language and object model from the following documents: 
  * [[https://developer.mozilla.org/en-US/docs/Web/JavaScript|Core JavaScript Guide]] ([[http://developer.mozilla.org/docs/Core_JavaScript_1.5_Guide|Historic: 1.5]])
  * [[https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference|Core JavaScript Reference]] ([[http://developer.mozilla.org/docs/Core_JavaScript_1.5_Reference|Historic: 1.5]])

JavaScript is an established, mature scripting language syntactically similar to C++ and Java. 

[[https://developer.mozilla.org/en/Core_JavaScript_1.5_Guide/JavaScript_Overview#JavaScript_and_Java|JavaScript is not Java]]. 

The ECMA and ISO standards organizations have standardized the core JavaScript language in [[https://262.ecma-international.org/5.1/ECMA-262.pdf|ECMA-262]] (ECMAScript) and ISO-16262. 

===== Baja and JavaScript =====

[[http://synchro.net/docs/baja.html|Baja]] is the original scripting language of Synchronet (introduced in v2.0, 1994), used to create Synchronet-specific modules and command shells. Baja was originally designed as a simple BASIC-like language for controlling the display of menus and command prompts, accepting commands from the user and passing on control to high-level BBS functions. Over the years, Baja has been extended and enhanced to allow a high-level of functionality, but it was never going to reach the power and flexibility of JavaScript.

For example Baja modules and command shells, see the ''.src'' files in your Synchronet ''[[dir:exec]]'' directory.

For the foreseeable future Baja modules and command-shells will continue to be supported in Synchronet, but sysops and developers are encouraged to use JavaScript instead of Baja moving forward. With very few exceptions, everything that can be done in a Baja module can be done in a JavaScript module, and much much more.

Eventually, all the stock command-shells and external modules will be converted to JavaScript.

===== JavaScript Files =====

JavaScript files are just ASCII text files. They are normally named with a ''.js'' file extension and located in your Synchronet ''[[dir:exec]]'' directory or with a ''.ssjs'' file extension and located in your Synchronet web hierarchy. JavaScript files do not need to be compiled. JavaScript files are loaded into memory at the time of execution, so a change to a JavaScript file will take effect the next time that file is executed (no recycling of servers is normally required).

For example JavaScript modules and services, see the ''.js'' files in your Synchronet ''[[dir:exec]]'' directory. 

Modified stock ''.js'' files should be placed in your Synchronet ''[[dir:mods]]'' directory to prevent over-writing by future upgrades. 

===== Invocation =====

JavaScript files can be executed / invoked from:
  * [[:server:terminal|Terminal Server]]
as a timed event, external program (door), login/logon/newuser module, via the ''EXEC'' sysop command
((basically anywhere a Baja module or native or MS-DOS executable can be launched))
  * [[:server:web|Web Server]]
dynamically generates HTTP responses (e.g. HTML/CSS content),
see ''[[dir:web]]/root/*.ssjs''
  * [[:server:ftp|FTP Server]]
dynamically generates HTML index files,
see ''[[dir:exec]]/ftp-html.js'' and ''[[dir:exec]]/ftp-web-html.js''
  * [[:service:|Services]]
all services at this time (both static and dynamic) are written in JavaScript,
see ''[[dir:exec]]/*service.js'' and ''[[dir:ctrl]]/[[config:services.ini]]''
  * [[:server:mail|Mail Server]]
inbound mail processors may be written in JavaScript, 
see ''[[dir:exec]]/[[module:mailproc_example]].js'' and ''[[dir:ctrl]]/[[config:mailproc.ini]]''
  * [[:util:JSexec]]
some script files may be executed outside of Synchronet (e.g. as a CGI script or daemon) using [[util:JSexec]],
examples: ''ircd.js'', ''newslink.js''

From the Terminal Server, a JavaScript file is executed on a native command-line by placing a question mark (''?'') or asterisk (''*'') at the beginning of the command-line before the JavaScript file name (in [[util:SCFG]]). It is not necessary to specify the ''.js'' portion of the file name on the command-line. For example, the command-line to execute the file ''[[dir:exec]]/newslink.js'' would be "''?newslink''" or "''*newslink''".

From within a Baja module, a JavaScript file may be executed using the following Baja code:
<code baja>
exec "?modname" # where modname is the JavaScript filename or base filename.
</code>
or:
<code baja>
exec "*modname" # where modname is the JavaScript filename or base filename.
</code>
or:
<code baja>
exec_bin "modname" # where modname is the JavaScript filename or base filename.
</code>
===== Object Model =====

Synchronet has its own constantly evolving JavaScript object model (containing classes, objects, methods, and properties), not to be confused with the Document Object Model (DOM) used in web browsers. In order to fully understand the capabilities of JavaScript modules in Synchronet, you must familiarize yourself with [[http://developer.mozilla.org/docs/Core_JavaScript_1.5_Reference|Core JavaScript]] as well as the [[http://synchro.net/docs/jsobjs.html|Synchronet JavaScript Object Model]].

==== load ==== 

The Synchronet JavaScript object model includes a global method: ''load()'' (a closely-related ''require()'' method was added in v3.17).

The ''load'' method is used to compile and execute an external script from within a //parent// script. It is most often used to load numeric constants and object definitions from files in the ''[[dir:exec]]/[[dir:load]]'' directory as these definitions (e.g. ''sbbsdefs.js'') and object libraries (e.g. ''graphic.js'') enable code re-use and sharing among modules, saving us JS programmers and lot of typing and redundancy.

There are 3 primary ways to use the ''load'' method:

1. This example just runs the referenced script name (''myscript.js'') and passes the argument values (''1, 2, 3'') to the script, saving the result (the last expression evaluated in ''myscript.js'') in the ''result'' variable:
<code>
var result = load('myscript.js', 1, 2, 3);
</code>

2. This example runs the referenced script in a background/child thread, passing the argument values (''1, 2, 3'') and saving the bi-directional [[http://synchro.net/docs/jsobjs.html#Queue|queue]] that may be used to communicate with the child thread in the ''queue'' variable:
<code>
var queue = load(true, 'myscript.js', 1, 2, 3);
var value = queue.read(1000);
writeln(value);
</code>

3. This example runs the referenced script in the scope of the specified object (passing no arguments, though that is supported):
<code>
var obj = new Object;
load(obj, 'myscript.js');
</code>

And a short-hand, preferred, syntax (same result):
<code>
var lib = load({}, 'myscript.js');
</code>

JavaScript files in the ''load'' directory that end in (last line is) a ''this;'' statement are intended to be ''load''ed / used in this fashion.

Once a file is loaded in this fashion, any methods, variables, constants or objects defined within may be referenced from the parent script like so:
<code>
var lib = load({}, 'mylib.js');
lib.do_thing();
if(lib.status_good == true)
    writeln('good');
else
    alert('bad');
</code>

This usage expressly places all items created by the loaded script into the scope of the passed variable (in this case, a newly created empty //Object//, just for this purpose). This means there should be no naming conflicts between the parent script and the loaded script (e.g. they may have methods or variables defined with the **same name** and there will be [[https://www.youtube.com/watch?v=2CifYWxJXaI|no confrict]]).

==== Output ==== 
There are many different text output methods supported by the Synchronet JavaScript Object Model and knowing which method (function) to use in what situation can be confusing. The following table should help:

^ Method          ^ BBS((Terminal Server)) ^ JSexec ^ Newline ^ Xlat ((Character set translation)) ^ Expands/Decodes    ^ Multiple Values((as function arguments)) ^ Notes ^
| ''write()''     | Yes                    | Yes    | No           | Yes | Telnet-IAC, [[custom:ctrl-a_codes|Ctrl-A]] | Yes                | If user not online, same as ''log(LOG_INFO, ...)'' |
| ''write_raw()'' | Yes                    | No     | No           | No  | Telnet-IAC         | Yes                | Value may contain NULs, no charset translation |
| ''writeln()''   | Yes                    | Yes    | Yes          | Yes | Telnet-IAC, Ctrl-A | Yes                | aka ''print()'' |
| ''printf()''    | Yes                    | Yes    | No           | Yes | Telnet-IAC, Ctrl-A | Yes((using C printf format syntax))  | ala C ''printf()''  |
| ''alert()''     | Yes                    | Yes    | Yes          | Yes | Telnet-IAC, Ctrl-A | No                 | If user not online, same as ''log(LOG_WARNING, ...)'' |
| ''log()''             | Yes              | Yes    | N/A          | N/A |                    | Yes((following the log-level argument)) | Displayed/logged on the server only |
| ''console.print()''   | Yes              | No     | No           | Yes | Telnet-IAC, Ctrl-A | Yes | |
| ''console.write()''   | Yes              | No     | No           | Yes | Telnet-IAC         | Yes | No line-counting/auto-pause |
| ''console.writeln()'' | Yes              | No     | Yes          | Yes | Telnet-IAC         | Yes | No line-counting/auto-pause |
| ''console.putmsg()''  | Yes              | No     | No           | Yes | Telnet-IAC, Ctrl-A, [[custom:atcodes|@-Codes]], [[custom:colors|Extra Attributes]] | No        | Typically used for displaying text/menu files |
| ''console.center()''  | Yes              | No     | Yes          | Yes | Telnet-IAC, Ctrl-A | No | Text centered on screen |
| ''console.mnemonics()'' | Yes            | No     | No           | Yes | Telnet-IAC, Ctrl-A, @-Codes, ~ | No | Hotkey highlights |
| ''console.putbyte()'' | Yes              | No     | No           | No  | Telnet-IAC         | No | No charset translation |
| ''client.socket.send()'' | Yes           | No     | No           | No  |                    | No | Avoid using |


=== Carriage Returns ===
If you wish to send a traditional "carriage return" character to the client terminal (i.e. move the cursor to the far left column without advancing to the next line) it is recommended that you use an output method/function that supports [[custom:Ctrl-A codes]] and you send the Ctrl-A ''[''  sequences //instead of// an ASCII 13 (''\r'') character. This is required for PETSCII terminal compatibility, where an ASCII 13 character performs a complete "newline" sequence (the equivalent of ''\r\n'') on the client terminal.

==== Input ==== 

^ Method                    ^ BBS((Terminal Server)) ^ JSexec ^ Notes ^
| ''read()''                | Yes                    | Yes    | Read up to maxlen (default 128) characters of input  |
| ''readln()''              | Yes                    | Yes    | Read a string (line of input) from user, up to maxlen chars (default: 128) |
| ''console.inkey()''       | Yes                    | No     | Wait for a key-press or an elapsed timeout duration |
| ''console.getstr()''      | Yes                    | No     | Get a string of characters from user  |
| ''console.getkey()''      | Yes                    | No     | Get a single key-press       |
| ''console.getbyte()''     | Yes                    | No     | Get an unprocessed input byte from remote terminal as a number |
| ''console.getnum()''      | Yes                    | No     | Get a number (between 1 and maxnum) from user     |
| ''console.getkeys()''     | Yes                    | No     | Get a key from a list of valid keys, or a number  |
| ''console.gettemplate()'' | Yes                    | No     | Get a string conforming to a provided input template  |

==== Prompt ==== 

Often you'll want your script to prompt the user for some input (e.g. a string of text or a single key). A prompt involves both //output//  (first, from the BBS) and //input// (from the user):

^ Method              ^ BBS((Terminal Server)) ^ JSexec ^ Notes ^
| ''prompt()''        | Yes                    | Yes    |       |
| ''confirm()''       | Yes                    | Yes    |       |
| ''deny()''          | Yes                    | Yes    |       |
| ''console.yesno()'' | Yes                    | No     |       |
| ''console.noyes()'' | Yes                    | No     |       |
| ''console.mnemonics()'' | Yes                    | No     |       |

===== See Also =====
  * [[http://synchro.net/docs/jsobjs.html|Synchronet JavaScript Object Model]]
  * [[http://nix.synchro.net/jsobjs.html|Synchronet JavaScript Object Model (preliminary for v3.17a)]]
  * [[:util:JSexec]]
  * [[:util:Baja]] 
  * [[:custom:javascript:lib:|JavaScript Libraries]]
  * [[:custom:index|Customization]]

{{tag>javascript petscii}}