namelist

A namelist is a data input mechanism used primarily in Fortran and related scientific computing environments. It provides a way to group a set of program variables into a single named collection that can be read from a text file or standard input at runtime. Namelist-style I/O is designed to let users override configuration parameters without recompiling.

Originating in Fortran as a language extension for convenient input of configuration parameters, namelists are supported by many compilers. The exact syntax varies by implementation, but the common idea is that variables are organized into named groups. Some systems use a directive in source code to declare the group, while others use a block form in the input file that begins with a group header and ends with a terminator. Names within a group can reference scalars, arrays, and basic data types such as integer, real, logical, and character.

Structure and usage: A namelist consists of one or more variables grouped under a single group name. In a typical input file, the group is opened, assignments are provided for the desired variables, and the group is closed. For example, a block might specify values for grid dimensions and physical parameters. Variables not mentioned in the namelist usually retain their existing values in the program. The approach supports single values and arrays and relies on compiler-specific rules for formatting, comments, and delimiters.

Advantages and limitations: Namelists offer a human-friendly way to configure runs, enabling easy parameter sweeps and testing without changing source code. They are convenient in high-performance computing workflows but are less portable across languages and depend on compiler support. For broader configuration needs, formats such as JSON, YAML, or XML are often used alongside or instead of namelists.