Summary: 74 instances, 70 unique

Text	Count
// TODO: Validate field types for histograms	1
// TODO: revisit - not entirely sure if this is the best thing to do, especially when there is a range query	1
// TODO: Let User specify their own filter query that is applied to the composite agg search	1
// TODO: Refactor so we can get isLastStep from somewhere besides an instantiated Action class so we can simplify this to a when block	1
// TODO: Validate field type for metrics,	1
// TODO: error handling	1
* TODO: When/if FieldCapabilitiesResponse and other subclasses package private constructors are elevated to public we can remove this logic.	1
// TODO: Is seqno/prim and id returned for all?	1
// TODO: Clean up runner	1
// TODO: This is a hacky solution to get the current start time off the job interval as job-scheduler currently does not	1
// TODO: Move this enum to Job Scheduler plugin	1
// TODO refactor these lists usage to map	1
* TODO situations:	1
// TODO: Should create a transport action to update metadata	1
// TODO: Read in the actual mappings from the source index and use that	1
// TODO: This is incomplete as more than keywords can be grouped as terms, need to figure out the correct way to do this check for now just	1
// TODO: Add indicesOptions?	1
return Instant.now().isAfter(metadata.continuous!!.nextWindowEndTime.plusMillis(rollup.delay ?: 0)) // TODO: !!	1
// TODO: How does this job matching work with roles/security?	1
// TODO: Verify if offset, missing value, min_doc_count, extended_bounds are usable in Composite histogram source	1
// TODO: Add some entry in metadata that will store index -> indexUUID for validated indices	1
// TODO what if update setting failed, cannot reset to -1/-2	1
// TODO: Remove this once the step interface is updated to pass in thread context information.	1
// TODO: Remove this once the step interface is updated to pass in user information.	1
// TODO: error handling - can RemoteTransportException happen here?	1
// TODO: Validate if field is date type: date, date_nanos?	1
// TODO: This can be moved to a common place, since this is shared between Rollup and ISMRollup	1
// TODO: Validation of fields across source and target indices overwriting existing rollup data	1
// TODO: If we have to set this manually for each aggregation builder then it means we could miss new ones settings in the future	1
return hasNextFullWindow(rollup, metadata) // TODO: Behavior when next full window but 0 docs/afterkey is null	1
// TODO clean up for actionIndex	1
// TODO: Scenario: The rollup job is finished, but I (the user) want to redo it all again	1
// TODO: when this happens is it failure or invalid?	1
// TODO: Should we reject all other timezones if they are specified in the query?: aggregationBuilder.timeZone()	1
// TODO: This is mostly used for the check in runJob(), maybe move this out and make that call "metadata == null || shouldProcessRollup"	1
// TODO: Allow filtering for [continuous, job state, metadata status, targetindex, sourceindex]	1
// TODO: Add circuit breaker checks - [cluster healthy, utilization within limit]	1
// TODO: Field and mappings validations of source and target index, i.e. reject a histogram agg on example_field if its not possible	1
// TODO: Make startTime public in Job Scheduler so we can just directly check the value	1
// TODO: Does schema_version get overwritten?	1
// TODO: Failed shouldn't process? How to recover from failed -> how does a user retry a failed rollup	1
// TODO: Source index could be a pattern but it's used at runtime so it could match new indices which weren't matched before	1
// TODO: restrict it for testing	1
// TODO: Perhaps we can do better than this for mappings... as it'll be dynamic for rest	1
// TODO: Verify the seqNo/primterm got updated	1
* // TODO: When FGAC is supported in transform should check the user has the correct permissions	1
// TODO: if metadata id exists delete the metadata doc else just delete transform	1
# TODO: Remove this before initial release, only for developmental purposes	1
// TODO: get rid of !!	1
// TODO: A cluster level setting to control how many rollups can run at once? Or should we be skipping when cpu/memory/jvm is high?	1
// TODO: Clean this up	1
// TODO: Doc counts for aggregations are showing the doc counts of the rollup docs and not the raw data which is expected...	1
// TODO: Is schema_version always present?	1
// TODO missing terms field	1
// TODO: Should update metadata and disable job here instead of allowing the rollup to keep going	1
// TODO: Should we store the value of the past successful page size (?)	1
// TODO this can be moved to job scheduler, so that all extended plugin	1
// TODO: Should we throw error instead?	1
// TODO: Could make this an extension function of DateHistogram and add to some utility file	1
// TODO: Not a fan of this.. but I can't find a way to overwrite the aggregations on the shallow copy or original	1
// TODO: Should we attempt to create a new document instead if failed to parse, the only reason this can happen is if someone deleted	1
// TODO: Catching general exceptions for now, can make more granular	4
// TODO should wrap document already exists exception	1
// TODO: Make sure the interval string is validated before getting here so we don't get errors	1
// TODO: The use of the master transport action UpdateRollupMappingAction will prevent	1
// TODO: Validate field types for terms	1
// TODO: Create namespaces to group properties together	1
// TODO: Wrap client calls in retry for transient failures	2
// TODO: Get all ManagedIndices at once or split into searchAfter queries?	1
.sort(dateHistogram.sourceField, SortOrder.ASC) // TODO: figure out where nulls are sorted	1