summaryrefslogblamecommitdiff
path: root/doc/packet_struct_notation.txt
blob: 4ba30cb00a152d4f90a157428fb4098df455da43 (plain) (tree)




















































































                                                                    
//===== Athena Doc ========================================
//= Packet Structure Notation
//===== By ================================================
//= Ai4rei
//===== Version ===========================================
//= 1.0
//=========================================================
//= 1.0 - Initial version.
//===== Description =======================================
//= Explanation how packets are and should be documented.
//=========================================================

This document specifies how packets are and should be documented, to
keep packet structure comments consistent in the entire codebase. It
also serves as a guide to those, who are unfamiliar with the general
packet layout.

All mentioned data types are assumed to be little-endian (least-
significant byte first, least significant bit last) and of same size
regardless of architecture.

= Typical description of a packet =

 /// Notifies the client about entering a chatroom (ZC_ENTER_ROOM).
 /// 00db <packet len>.W <chat id>.L { <role>.L <name>.24B }*
 /// role:
 ///     0 = owner (menu)
 ///     1 = normal

The first line contain a brief description of what the packet does,
or what it is good for, followed by it's AEGIS name in parentheses;
first two letters of the AEGIS name specify origin (first letter)
and destination (second letter) of the packet. If the packet's name
is not known or is not applicable (eAthena server-server packets),
specify at least these two letters to indicate the direction of the
packet. Do not use S(end)/R(ecv) for this, as it is inaccurate and
location dependent (if the description is copied to different server
or other RO-related projects, it might change it's meaning).

If there are multiple versions of the packet, the AEGIS name is
appended to the end of the packet's structure instead. If the name
did not change between versions, a PACKETVER expression is appended,
such as (PACKETVER >= 20111111).

Second line describes the packet's field structure, beginning with a
%04x formatted packet type, followed by the individual fields and
their types. Each field begins with it's name enclosed in angle
brackets ( <field name> ) followed by a dot and the data size type.
Field names should be lower-case and without underscores. If other
packets already have a field in common, use that name, rather than
inventing your own (ex. "packet len" and "account id"). Repeated and
optional fields are designated with curly and square brackets
respectively, padded with a single space at each side.

Further lines are optional and either include details about the
the packet's mechanics or further explanation on the packet fields'
values.

= Packet field data size type =

 B = 1 byte (byte)
 W = 2 bytes (word)
 L = 4 bytes (dword)
 Q = 8 bytes (quad)

 nB = n bytes
 ?B = variable/unknown amount of bytes
 nS = n bytes, zero-terminated
 ?S = variable/unknown amount of bytes, zero-terminated

= Repetition of packet fields =

 {} = repeated block
 {}* = variable/unknown amount of consecutive blocks
 {}*n = n times repeated block
 [] = optional fields

= Packet origin and destination letters =

 A = Account (Login)
 C = Client
 H = Character
 I = Inter
 S = Server (any type of server)
 Z = Zone (Map)