Name List Manager
This program automatically generates code and documentation for Fortran programs that manage their input using namelists. To use it, you first write an xml file which describes all the input parameters. An example input file is available in example/input.xml . This xml file describes all the parameters, as well as their data types and purposes. Then, you run this program on the xml file, and it automatically generates:
- Restructered text documenting the parameters, including default values.
- Module files to store all the input variables.
- Reader subroutines that read in data from a namelist specifying input values.
You install the program as:
pip install .
And then run it like:
namelistmanager input.xml doc_path mod_path read_path
You can run the examples using the following command:
namelistmanager example/input.xml example/doc example/module example/readers
The documentation, module files, and readers will then be available in the example directory. The documentation files are restructed text, which means you could easily add them to a Sphinx project. There is also a sample driver program and makefile available in the examples directory, so you can test out the code.
Requirements
This package requires Python 3.0 and up. It also depends on the lxml package, which pip will automatically install.
XML File Details
The XML is used to describe the different input variables available in the namelist. To begin your xml file, wrap everything in the input tag:
<input name="project_name" lang="en" mpi="false"> ... </input>
Inside <input> you can place different groups. For each group, a module will be created to store the data of that group, and a reader for that group as well. Each group is identified by the name attribute:
<group name="matrixmultiplication"> </group>
Inside each group, you need to provide a list of descriptions of that group:
<description_list> <description xml:lang="en"> The parameters for performing matrix multiplication. </description> <description xml:lang="ja"> 説明 </description> </description_list>
Next you specify the elements of that group. The variable name associated with that element is based on the name attribute. Each element also requires a description of that variable. You also need to specify the data type and the default value:
<element_list> <element name="block_size"> <description_list> <description xml:lang="en">Size of blocks.</description> </description_list> <datatype>integer</datatype> <default>64</default> </element> </element_list>
You can optionally add a description of what type of values are valid:
<valid_list> <valid name="1">Block Size 1</valid> <valid name="2">Block Size 2</valid> <valid name="4">Block Size 4</valid> </valid_list>
Any XML file you create can be validated against the XML schema stored in namelistmanager/namelist.xsd .
Additional Options
The default language can be specified using the lang attribute in the input element. You can also specify whether to generate MPI safe input handling. This mainly affects how error handling is performed.