How to plan additional tests performed by an external sub?

Question

We wrote parsers for different formats of scientific data in Perl. I recently added a test suite with the parser_*.t file for each format and subformat.

Of course, the parser APIs are exactly the same, as soon as the data read from the sample files that are used to test the parsing is different. To simplify the test files, I wrote a sub that receives the passed parser object and a hash structure representing the expected data. He looks like

 my $parser = new MyApp::Parser($file); test_nested_objects = ($parser, { property1 => "value", property2 => 123, subobject_accessor => { property3 => "foobar", } } 

Sub test_nested_objects goes through the hash and runs the tests for all properties defined in the hash, for example. if subobject_accessor can be called, returns an object and this object can be called property3 .

I checked how many tests are performed by the whole *.t file and tests => 123 added to all *.t files. Now I have added some checks to the general function, and all plans are wrong.

How to make my plan aware of subtests? I would like to do the following:

• number of tests given before running them to view progess
• the total number increased automatically → without manually changing numbers when editing sub
• individual tests for sub visible at startup prove (hiding tests in sub and returning only 0 or 1 is not acceptable, because I really need to know what is wrong with the data being analyzed)

I hope you understand. Sorry for the long story, but I thought that people probably would not understand without any basic knowledge.

Answer 1

The latest trend in test planning is simply not to. At the end of the test file, you declare that you have reached the end:

  done_testing(); 

Many applications of the test plan exist only to complete the testing of the script. If he saw the correct number of tests, then the script should have finished everything. This is not entirely true, but it is a convenient and workable fantasy.

If you need to do some calculations to determine the number of tests, for example, measuring the size of a data structure, do this and then call plan later:

  use Test::More; # don't declare the plan yet my $test_count = ...; plan tests => $test_count; 

Instead of making big plans, you can split your tests into subtests. Each subtest should only know how many tests it runs, and the script should know only the number of subtests:

  subtest some_label => sub { plan tests => $n; ... }; 

Each of your calls to the test function will be a single test, and the function will use a test to do whatever it wants, without undue impact on the caller.

Answer 2

If the number of auxiliary tests for each format is known, you can perform test calculations in the BEGIN block before using Test :: More.