The title area gives the name and reference number of the algorithm. The following area is used for the declaration of global variables. Then come the statements of the processes. For two processes, which are named p and q, the statements are written in two columns. Algorithms for N processes are written in a single column. The code for each process is identical except for the ID of each process, written i in the concurrent algorithms and myID in the distributed algorithms.

In the distributed algorithms, the algorithm is given for each node. The node itself may contain several processes which are denoted by giving them titles; see Algorithm 11.3 for an example.

Monitors and other global subprograms that can be executed by any process appear between the global data declarations and the statements of the processes.

Text from the symbol // until the end of a line is a comment.

Data declarations are of the form:

type-name variable-name ← initial value

The most common type is integer, which represents any integer type; implementations can use other types such as byte or short. In some algorithms, like producer–consumer algorithms, a data type for arbitrary elements is needed, but its precise specification is not important. In these cases, a name ending in Type is used, usually dataType.

Arrays are declared by giving the element type, the index range and the initial values:

integer array [1.. n] number ← [0,. . . ,0]

Elementary abstract data types like queues and sets are used as needed.

The arrow ← is used in assignments, so that we don’t have to choose between languages that use = and those that use :=. In expressions, mathematical symbols like =, ≠ and ≤ are used.

if statements are conventional. We are quite flexible in the syntax of loop statements, from conventional while statements to abbreviations like do two times. loop forever is used in most concurrent algorithms to denote the indefinite looping of a process.

Numbered statements represent atomic operations as in Promela; if a continuation line is needed, it is not numbered.

Assignment statements and expression evaluations are atomic statements. The atomicity of evaluating the expression in a condition includes the decision to take one branch or another; for example, in the statement if a<b then s1 else s2, if the value of a is in fact less than the value of b, evaluating a<b will cause the control pointer of the process to point to s1.

Note that loop forever, else and goto are not numbered; they serve only to specify the control flow and do not denote executable statements.

The statement await boolean-valued-expression is an implementation-independent notation for a statement that waits until the expression becomes true. This can be implemented (albeit inefficiently) by a busy-wait loop that does nothing until the condition is true. Note that in Promela, an await statement can be implemented by simply writing an expression:

turn == 1

The notation ch⇐x indicates that the value of x is sent on the channel ch, and similarly, the notation ch⇒y means that the value of the message received from the channel ch is assigned to the variable y. In CSP and Promela these are denoted ch!x and ch?y, but the use of arrows clarifies the data flow.