publication_service Class Reference

Service for publishing datasets to production tables. More...

#include <publication_service.hpp>

Collaboration diagram for publication_service:

Public Types
using	context = ores::database::context

Public Member Functions
	publication_service (context ctx)
	Constructs a publication_service.
std::vector< domain::publication_result >	publish (const std::vector< boost::uuids::uuid > &dataset_ids, domain::publication_mode mode, const std::string &published_by, bool resolve_dependencies=true)
	Publishes one or more datasets to production tables.
messaging::publish_bundle_response	publish_bundle (const std::string &bundle_code, domain::publication_mode mode, const std::string &published_by, bool atomic=true)
	Publishes all datasets in a bundle.
std::vector< domain::dataset >	resolve_publication_order (const std::vector< boost::uuids::uuid > &dataset_ids)
	Resolves the publication order for datasets.
std::vector< domain::publication >	get_publication_history (const boost::uuids::uuid &dataset_id)
	Gets the publication history for a dataset.
std::vector< domain::publication >	get_recent_publications (std::uint32_t limit=100)
	Gets recent publication history across all datasets.

Detailed Description

Service for publishing datasets to production tables.

This service handles the publication workflow:

1. Resolves dataset dependencies using a directed graph
2. Determines publication order (dependencies first)
3. Calls appropriate population functions for each dataset
4. Records publication history for auditing

The service can be used from both the binary protocol handler and HTTP endpoints.

Constructor & Destructor Documentation

publication_service()

publication_service ( context  ctx )

explicit

Constructs a publication_service.

Parameters

ctx	The database context for executing queries.

Member Function Documentation

publish()

std::vector< domain::publication_result > publish	(	const std::vector< boost::uuids::uuid > &	dataset_ids,
		domain::publication_mode	mode,
		const std::string &	published_by,
		bool	resolve_dependencies = true
	)

Publishes one or more datasets to production tables.

This is the main entry point for dataset publication. It:

1. Validates that all dataset IDs exist
2. Resolves dependencies and determines publication order
3. Publishes each dataset in dependency order
4. Records publication history for auditing

Parameters

dataset_ids	The IDs of the datasets to publish.
mode	How to handle conflicts with existing data.
published_by	Username of the person initiating publication.
resolve_dependencies	If true, automatically include dependencies.

Returns

Results for each dataset published.

publish_bundle()

messaging::publish_bundle_response publish_bundle	(	const std::string &	bundle_code,
		domain::publication_mode	mode,
		const std::string &	published_by,
		bool	atomic = true
	)

Publishes all datasets in a bundle.

Calls the SQL dq_bundles_publish_fn() which:

1. Processes datasets in display_order to respect dependencies
2. Publishes each dataset individually
3. Records publication history for auditing

Parameters

bundle_code	The bundle to publish (e.g., 'base', 'solvaris')
mode	How to handle conflicts with existing data.
published_by	Username of the person initiating publication.
atomic	If true, first failure causes entire bundle to rollback.

Returns

Publication result with per-dataset details.

resolve_publication_order()

std::vector< domain::dataset > resolve_publication_order

(

const std::vector< boost::uuids::uuid > &

dataset_ids

)

Resolves the publication order for datasets.

Uses boost.graph to build a dependency graph and performs topological sort to determine the correct order. Dependencies are placed before the datasets that depend on them.

Parameters

dataset_ids

The IDs of the datasets to order.

Returns

Ordered list of datasets (dependencies first).

Here is the caller graph for this function:

get_publication_history()

std::vector< domain::publication > get_publication_history

(

const boost::uuids::uuid &

dataset_id

)

Gets the publication history for a dataset.

Parameters

dataset_id

The ID of the dataset.

Returns

All publication records for this dataset, newest first.

get_recent_publications()

std::vector< domain::publication > get_recent_publications

(

std::uint32_t

limit = 100

)

Gets recent publication history across all datasets.

Parameters

limit

Maximum number of records to return.

Returns

Recent publication records, newest first.