view release on metacpan or  search on metacpan

t/features/metrics_history.feature  view on Meta::CPAN

Feature: analizo metrics-history
  As a software engineering researcher
  I want to analyse the entire history of a project
  To understand its development process

  Scenario: listing merge commits with code changes that should be analyzed
    When I explode t/samples/evolution.tar.gz
    And I run "analizo metrics-history --list ."
    Then the output must match "0a06a6fcc2e7b4fe56d134e89d74ad028bb122ed"
    And the output must match "eb67c27055293e835049b58d7d73ce3664d3f90e"
    And the output must match "aa2d0fcb7879485d5ff1cd189743f91f04bea8ce"
    And the output must match "e8faf88f0e20a193d700b6c68eeb31897dd85e53"
    And the output must match "d7f52e74dc3d8f57640e83d41c5e9f8fcf621c00"
    And the output must match "0d3c023120ad4e9f519a03fff275d048c52671ad"

  Scenario: non-code commits should not be analyzed
    When I explode t/samples/evolution.tar.gz
    And I run "analizo metrics-history --list ."
    Then the output must not match "ba62278e976944c0334103aa0044535169e1a51e"

  Scenario: merge commits without code change should not be analyzed
    When I explode t/samples/evolution.tar.gz
    And I run "analizo metrics-history --list ."
    Then the output must not match "0fdaaa7dcc8073332a957024fafc8c98f165e725"

  Scenario: actually processing merge commits
    When I explode t/samples/evolution.tar.gz
    And I run "analizo metrics-history -o metrics.csv . && cat metrics.csv"
    Then the output must match "^id,previous_commit_id,author_date,author_name,author_email,.*,sc_mean"
    And the output must match "0a06a6fcc2e7b4fe56d134e89d74ad028bb122ed,eb67c27055293e835049b58d7d73ce3664d3f90e"
    And the output must match "eb67c27055293e835049b58d7d73ce3664d3f90e,,"
    And the output must match "aa2d0fcb7879485d5ff1cd189743f91f04bea8ce,d7f52e74dc3d8f57640e83d41c5e9f8fcf621c00"
    And the output must match "e8faf88f0e20a193d700b6c68eeb31897dd85e53,d7f52e74dc3d8f57640e83d41c5e9f8fcf621c00"
    And the output must match "d7f52e74dc3d8f57640e83d41c5e9f8fcf621c00,0d3c023120ad4e9f519a03fff275d048c52671ad"

  Scenario: actually processing initial commit and first commit after a non-relevant merge
    When I explode t/samples/evolution.tar.gz
    And I run "analizo metrics-history -o metrics.csv . && cat metrics.csv"
    Then the output must match "0d3c023120ad4e9f519a03fff275d048c52671ad,,"
    And the output must match "8183eafad3a0f3eff6e8869f1bdbfd255e86825a,0a06a6fcc2e7b4fe56d134e89d74ad028bb122ed"

  Scenario: support for parallel processing
    Given I copy t/samples/evolution.tar.gz into a temporary directory
    When I run "tar xzf evolution.tar.gz"
    And I run "cd evolution && analizo metrics-history -o ../sequential.csv"
    And I run "cd evolution && analizo metrics-history -p 2 -o ../parallel.csv"
    Then the exit status must be 0
    When I run "sort sequential.csv > sequential-sorted.csv"
    And I run "sort parallel.csv > parallel-sorted.csv"
    And I run "diff -u sequential-sorted.csv parallel-sorted.csv"
    Then the output must not match "---"
    And the exit status must be 0

  Scenario: parsing git log format containing renamed files among status of changed files
    Given I copy t/samples/evolution.tar.gz into a temporary directory
    When I run "tar xzf evolution.tar.gz"
    And I run "cd evolution && git checkout doc && analizo metrics-history ."
    Then the exit status must be 0

  Scenario: language filters
    Given I copy t/samples/mixed into a temporary directory
    When I run "(cd mixed && git init && git add * && git commit -m 'initial commit')"
    And I run "analizo metrics-history --language java mixed"
    Then the output must not match "native_backend.c"