Skip to content

Latest commit

 

History

History
142 lines (121 loc) · 3.97 KB

example5.md

File metadata and controls

142 lines (121 loc) · 3.97 KB
Raw

Example 5: Store Results in Database using DbBot

This example is based on the documentation for DbBot on Github.

To run this example cd to the example5 folder

cd /vagrantsync/examples/example5

And run the provisioning script

scripts/provision_example5.sh

Then run the example testsuites

robot --outputdir logs/5 example5.robot

The default database is a file named robot_results.db. But here we store the db file in /tmp/ since during the creation of this example I always had a sqlite3.OperationalError: disk I/O error when I tried to store the db in the /vagrantsync/ folder. So to store the db just run:

python -m dbbot.run  -b /tmp/my_own_database.db /vagrantsync/examples/example5/logs/5/output.xml

In order to show the current content of the db run the following commands:

$ sqlite3 /tmp/my_own_database.db

sqlite> .tables
arguments        suite_status     test_run_errors  tests
keyword_status   suites           test_run_status
keywords         tag_status       test_runs
messages         tags             test_status

sqlite> SELECT count(), tests.id, tests.name
        FROM tests, test_status
        WHERE tests.id == test_status.test_id AND
        test_status.status == "FAIL"
        GROUP BY tests.name;

The output of the query will lokk like this:

1|1|Failing Example
1|2|Good Example

Type .quit or use Ctrl+D to leave sqlite3 console

robot --outputdir logs/5.1 example5.robot

python -m dbbot.run -b /tmp/my_own_database.db /vagrantsync/examples/example5/logs/5.1/output.xml

$ sqlite3 /tmp/my_own_database.db

sqlite> .tables
arguments        suite_status     test_run_errors  tests
keyword_status   suites           test_run_status
keywords         tag_status       test_runs
messages         tags             test_status

sqlite> SELECT count(), tests.id, tests.name
        FROM tests, test_status
        WHERE tests.id == test_status.test_id AND
        test_status.status == "FAIL"
        GROUP BY tests.name;

The output of the query will now look like this:

2|1|Failing Example
2|2|Good Example

This is because now we have two results for each test case.

You can also show the suite name to the test case name:

sqlite> SELECT count(), tests.id, tests.name, suites.name
        FROM tests, test_status, suites
        WHERE tests.id == test_status.test_id AND
        test_status.status == "FAIL"
        GROUP BY tests.name;

With the output:

2|1|Failing Example|Example5
2|2|Good Example|Example

robot --outputdir logs/5.2 example5.2.robot

python -m dbbot.run -b /tmp/my_own_database.db /vagrantsync/examples/example5/logs/5.2/output.xml

Now when we have a second suite, run the command again:

sqlite> SELECT count(), tests.id, tests.name, suites.name
        FROM tests, test_status, suites
        WHERE tests.id == test_status.test_id AND
        test_status.status == "FAIL"
        GROUP BY tests.name;

But the output looks odd:

6|3|Failing Example|Example5.2
6|4|Good Example|Example5.2

Adapt the query to

sqlite> SELECT count(), suites.name
        FROM tests, test_status, suites
        WHERE tests.id == test_status.test_id AND test_status.status == "FAIL" AND tests.suite_id == suites.id
        GROUP BY suites.name;

This for example will show us how many results we have for each suite:

4|Example5
2|Example5.2

Now as a last example, show the number of test runs:

sqlite> SELECT count(), test_runs.source_file, test_runs.started_at, test_runs.finished_at
        FROM test_runs
        GROUP BY test_runs.id;

The number of test runs per imported file with start and end time:

1|/vagrantsync/examples/example5/logs/5/output.xml|2018-07-30 19:10:05.896000|2018-07-30 19:10:05.913000
1|/vagrantsync/examples/example5/logs/5.1/output.xml|2018-07-30 19:53:36.684000|2018-07-30 19:53:36.702000
1|/vagrantsync/examples/example5/logs/5.2/output.xml|2018-07-30 20:03:30.761000|2018-07-30 20:03:30.779000