Usage

Singleton

For use as a singleton (the same object for all modules):

Place these lines in your project's .config:

<singleton-services>
  <service name="SimpleCsvOutput" need="required" />
</singleton-services>

And in your module(s):

#include <arcane/core/ISimpleTableOutput.h>
 
using namespace Arcane;
 
ISimpleTableOutput* table = ServiceBuilder<ISimpleTableOutput>(subDomain()).getSingleton();
table->init("Example_Name", "example"); // Must only be done by a single module.
// Using the service...
table->writeFile(); // Must only be done by a single module (unless you know what you are doing).

Service

For use as a service (a different object for each module):

Place these lines in your module's .axl:

<!-- <options> -->
  <service-instance name="simple-table-output" type="Arcane::ISimpleTableOutput">
    <description>Service implementing ISimpleTableOutput</description>
  </service-instance>
<!-- </options> -->

In the .arc, you can configure the service options. For example:

<!-- <mon-module> -->
  <simple-table-output name="SimpleCsvOutput">
    <!-- The name of the directory to create/use. -->
    <tableDir>example_dir</tableDir>
    <!-- The name of the file to create/overwrite. -->
    <tableName>Results_Example</tableName>
 
    <!-- Finally, we will have a file with the path: 
    ./output/csv/example_dir/Results_Example.csv -->
  </simple-table-output>
<!-- </mon-module> -->

And in your module:

#include <arcane/ISimpleTableOutput.h>
 
using namespace Arcane;
 
options()->simpleCsvOutput()->init();
//...
options()->simpleCsvOutput()->writeFile();

You can also use the service in both ways at the same time, depending on your needs.

(For a more concrete example, see the following pages)

Naming Symbols for Parallel Execution (CSV Implementation)

In the directory name or the table name, whether in singleton mode or service mode, it is possible to add symbols that will be replaced during execution.

The available symbols are:

• @proc_id@: Will be replaced by the process rank.
• @num_procs@: Will be replaced by the total number of processes.

For example, if we have:

<tableDir>N_@num_procs@</tableDir>
<tableName>Results_P@proc_id@</tableName>

or when initializing the service:

...
table->init("Results_P@proc_id@", "N_@num_procs@");
...

And if we run the program with 2 processes (ID = 0 and 1), we will get two CSV files with the path:

• ./output/csv/N_2/Results_P0.csv
• ./output/csv/N_2/Results_P1.csv

(sequentially, we will have ./output/csv/N_1/Results_P0.csv)

This allows, among other things, to:

• create a table per process and name them easily,
• create "generic" .arc files where the number of processes does not matter,
• have a different name for each table, in the case where a cat is performed (reminder: tableName gives the name of the csv file but is also placed in the first cell of the table).

---

Service SimpleCsvOutput Examples: generalities