Unix pipelines

 (Ritchie & Thompson, 1974) provide a global reuse mechanism: the input and output of individual Unix programs can be connected together to form a pipeline of programs that carry out a certain task. Typical pipelines perform sequences of operations for searching, replacing, transforming, and formatting text streams. Each program in the pipeline may be written in a different language, the data exchanged between programs has the form of a list of strings, and programs can only be combined in a strictly sequential fashion. As a prelude to discussions later (in Section 2.3), we observe that the pipe mechanism has two fundamental properties: (i) a standardized format to exchange data between programs; and (ii) a sequential composition operator to connect programs.