Refactor reconciliation step for conciseness

[tilacog]

There are seemingly redundant conditionals in the reconciliation loop code that could be improved for conciseness.

indexer/packages/indexer-agent/src/agent.ts

Lines 1006 to 1042 in d9f2d7c

	// Log details if active deployments are different from target deployments
	const isReconciliationNeeded = !isEqual(
	deploymentIDSet(activeDeployments),
	deploymentIDSet(targetDeployments),
	)
	if (isReconciliationNeeded) {
	// TODO: Return early in here case reconciliation is not needed
	this.logger.debug('Reconcile deployments', {
	syncing: activeDeployments.map(id => id.display),
	target: targetDeployments.map(id => id.display),
	withActiveOrRecentlyClosedAllocation: eligibleAllocationDeployments.map(
	id => id.display,
	),
	})
	}
	// Identify which subgraphs to deploy and which to remove
	const deploy = targetDeployments.filter(
	deployment => !deploymentInList(activeDeployments, deployment),
	)
	const remove = activeDeployments.filter(
	deployment =>
	!deploymentInList(targetDeployments, deployment) &&
	!deploymentInList(eligibleAllocationDeployments, deployment),
	)
	// TODO: Same as before: Should we return early in here case reconciliation is not needed? Also,
	// this conditional seems the be doing the same work as the previous one, we should consolidate
	// them.
	if (deploy.length + remove.length !== 0) {
	this.logger.info('Deployment changes', {
	deploy: deploy.map(id => id.display),
	remove: remove.map(id => id.display),
	})
	} else {
	this.logger.debug('No deployment changes are necessary')
	}