René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Redesign contest estimator entry points","articleBody":"Currently there 3 differents ways to start `Contest`:\r\n1. `UtBotTool2` - as separate process for SBFT, calls `ContestKt.main`\r\n2. `ContestEstimatorKt.main` - calls `runEstimator`\r\n3. `StatisticsMonitoringKt.main` - calls `runEstimator`\r\n\r\nAll of these tools set options for Contest Estimator differently, which causes divergence in Grafana monitoring, local testing and result in SBFT. it is hard to repro bug occured in nightly statistics on local machine as settings differ and Grafana shows incorrect results.\r\n\r\nEntry points should be reduced to single one, which will be used for all scenarios with same arguments. \r\n\r\nProposed solution: \r\n1. Merge `ContestEstimator` and `StatisticsMonitoring` in a single class, which will be used as an entry point both at CI and CLI.\r\n2. New entry point should have clean and simple way to configurate execution. Default arguments should correspond to statistics monitoring scenario and should work both from CI and CLI(for local repro).\r\n-  Nice to have, but not mandatory: New entry point should use `UtBotTool2` for generation. Monitoring will work as it is used in SBFT.\r\n\r\nThis would allow run this `utbot-junit-contest` in a single way as from CLI and Gradle. \r\n\r\nAdditional:\r\nIt is important either not to break current monitoring scripts or completely rewrite them.","author":{"url":"https://github.com/Domonion","@type":"Person","name":"Domonion"},"datePublished":"2023-01-24T10:38:16.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/1741/UTBotJava/issues/1741"}