Using constant value of "expected" metric count instead of an expression

[MoromugHSE]

The test "test-measure-file.sh" has an integer literal as a total number of metrics (line 48). Would be probably better to change it to be evaluated on test run.

[timur-harin]

@yegor256
I will be glad to help

Can you please, as the author of the repository, tell me where the original source is, how many metrics are in the repository?
Because in README.md I see only 48 metrics, but in script test-measure-file.sh number is 52

[timur-harin]

@yegor256
Could you please help me?

[timur-harin]

@yegor256
Please answer the question above

[yegor256]

@timur-harin this is not an easy ticket. We don't know how to do this right. Somehow, the script should know what is the "right" number of metrics to be calculated by all scripts in the metrics/ directory. Now, we simply hardcoded this number, which is not convenient. But how to calculate it in runtime - I'm not sure yet. Maybe you can suggest something?

[timur-harin]

@yegor256
How about this:

• In ast.py, we can measure metric.write or in others py files m.write
• In bash scripts, find the number of unique names of metrics with capital letter after cat or echo or puts?

[yegor256]

@timur-harin let's do the following instead. We add a new test into the test-measure-file.sh. It will first run measure-file.sh with a very simple java file class Foo {} and see how many metrics were created. Then, it will run the same script, but with a very complicated file, and also check how many metrics were collected. If the numbers match, we are good.

[timur-harin]

@yegor256
Please check PR #341

[timur-harin]

@yegor256
I think we need to close this issue, because PR merged success

[yegor256]

@MoromugHSE the issue was resolved, I'm closing this ticket