#include <index_registry.hpp>

Public Member Functions
	IndexRegistry ()=default
	Constructor. More...

	~IndexRegistry ()
	Destructor to clean up temp files. More...

	IndexRegistry (const IndexRegistry &other)=delete

IndexRegistry &	operator= (const IndexRegistry &other)=delete

	IndexRegistry (IndexRegistry &&other)

IndexRegistry &	operator= (IndexRegistry &&other)

void	set_prefix (const string &prefix)
	Prefix for all saved outputs. More...

string	get_prefix () const
	Get the current prefix for saving output files. More...

void	set_intermediate_file_keeping (bool keep_intermediates)

void	register_index (const IndexName &identifier, const string &suffix)
	Register an index containing the given identifier. More...

void	register_index (const IndexName &identifier, const vector< string > &suffixes)

RecipeName	register_recipe (const vector< IndexName > &identifiers, const vector< IndexName > &input_identifiers, const RecipeFunc &exec)

void	register_generalization (const RecipeName &generalizer, const RecipeName &generalizee)

void	provide (const IndexName &identifier, const string &filename, const map< string, string > &scopes={})

void	provide (const IndexName &identifier, const vector< string > &filenames, const map< string, string > &scopes={})

void	reset (const IndexName &identifier)
	Remove a provided index. More...

bool	available (const IndexName &identifier) const

bool	predates (const IndexName &earlier, const IndexName &later) const

vector< string >	get_possible_filenames (const IndexName &identifier) const

vector< string >	require (const IndexName &identifier) const

void	set_target_memory_usage (int64_t bytes)

int64_t	get_target_memory_usage () const
	Get the maximum memory we will try to consume. More...

vector< IndexName >	completed_indexes () const
	Get a list of all indexes that have already been completed or provided. More...

void	make_indexes (const vector< IndexName > &identifiers)

string	to_dot () const
	Returns the recipe graph in dot format. More...

string	to_dot (const vector< IndexName > &targets) const
	Returns the recipe graph in dot format with a plan highlighted. More...

void	reset ()
	Discard any provided or constructed indexes. More...

Static Public Member Functions
static set< string >	get_wildcards (const string &pattern)
	Get the names of all brace-enclosed {wildcards} in the given pattern. More...

static string	substitute_wildcards (const string &pattern, const map< string, string > values)

static int64_t	get_system_memory ()
	Get the amount of free memory. More...

static bool	vcf_is_phased (const string &filepath)
	Determine if a VCF file is phased or not. More...

static bool	gfa_has_haplotypes (const string &filepath)
	Determine if a GFA has haplotypes as W-lines. More...

Public Attributes
bool	check_files = true

Protected Member Functions
vector< IndexGroup >	dependency_order () const
	get a topological ordering of all registered indexes in the dependency DAG More...

IndexingPlan	make_plan (const IndexGroup &end_products) const
	generate a plan to create the indexes More...

bool	has_recipe (const RecipeName &recipe_name) const

const IndexRecipe &	get_recipe (const RecipeName &recipe_name) const
	use a recipe identifier to get the recipe More...

vector< vector< string > >	execute_recipe (const RecipeName &recipe_name, const IndexingPlan *plan, AliasGraph &alias_graph)

IndexFile *	get_index (const IndexName &identifier)
	access index file More...

const IndexFile *	get_index (const IndexName &identifier) const
	access const index file More...

bool	all_finished (const vector< const IndexFile * > &inputs) const

bool	all_finished (const IndexGroup &inputs) const

string	get_work_dir ()
	Function to get and/or initialize the temporary directory in which indexes will live. More...

Protected Attributes
map< IndexName, unique_ptr< IndexFile > >	index_registry
	The storage struct for named indexes. Ordered so it is easier to key on index names. More...

unordered_set< string >	registered_suffixes
	All of the suffixes that have been registered by indexes. More...

map< IndexGroup, vector< IndexRecipe > >	recipe_registry
	The storage struct for recipes, which may make index. More...

map< RecipeName, RecipeName >	generalizations
	Map from generalizees to generalizers. More...

string	work_dir
	Temporary directory in which indexes will live. More...

string	output_prefix = "index"
	filepath that will prefix all saved output More...

bool	keep_intermediates = false
	should intermediate files end up in the scratch or the output directory? More...

int64_t	target_memory_usage = numeric_limits<int64_t>::max()
	the max memory we will attempt to use More...

Friends
class	IndexingPlan

Detailed Description

An object that can record methods to produce indexes and design workflows to create a set of desired indexes.

Constructor & Destructor Documentation

◆ IndexRegistry() [1/3]

vg::IndexRegistry::IndexRegistry ( )

default

Constructor.

◆ ~IndexRegistry()

vg::IndexRegistry::~IndexRegistry ( )

Destructor to clean up temp files.

◆ IndexRegistry() [2/3]

vg::IndexRegistry::IndexRegistry ( const IndexRegistry & other )

delete

◆ IndexRegistry() [3/3]

vg::IndexRegistry::IndexRegistry ( IndexRegistry && other )

Member Function Documentation

◆ all_finished() [1/2]

bool vg::IndexRegistry::all_finished ( const IndexGroup & inputs ) const

protected

◆ all_finished() [2/2]

bool vg::IndexRegistry::all_finished ( const vector< const IndexFile * > & inputs ) const

protected

◆ available()

bool vg::IndexRegistry::available ( const IndexName & identifier ) const

Return true if the given index is available and can be require()'d, and false otherwise.

◆ completed_indexes()

vector< IndexName > vg::IndexRegistry::completed_indexes ( ) const

Get a list of all indexes that have already been completed or provided.

◆ dependency_order()

vector< IndexGroup > vg::IndexRegistry::dependency_order ( ) const

protected

get a topological ordering of all registered indexes in the dependency DAG

◆ execute_recipe()

vector< vector< string > > vg::IndexRegistry::execute_recipe	(	const RecipeName &	recipe_name,
		const IndexingPlan *	plan,
		AliasGraph &	alias_graph
	)

protected

Build the index using the recipe with the provided priority. Expose the plan so that the recipe knows where it is supposed to go.

◆ get_index() [1/2]

IndexFile * vg::IndexRegistry::get_index ( const IndexName & identifier )

protected

access index file

◆ get_index() [2/2]

const IndexFile * vg::IndexRegistry::get_index ( const IndexName & identifier ) const

protected

access const index file

◆ get_possible_filenames()

vector< string > vg::IndexRegistry::get_possible_filenames ( const IndexName & identifier ) const

Get the possible filename(s) associated with the given index with the given prefix. TODO: Get this to account for sample-scoped indexes.

◆ get_prefix()

string vg::IndexRegistry::get_prefix ( ) const

Get the current prefix for saving output files.

◆ get_recipe()

const IndexRecipe & vg::IndexRegistry::get_recipe ( const RecipeName & recipe_name ) const

protected

use a recipe identifier to get the recipe

◆ get_system_memory()

int64_t vg::IndexRegistry::get_system_memory ( )

static

Get the amount of free memory.

◆ get_target_memory_usage()

int64_t vg::IndexRegistry::get_target_memory_usage ( ) const

Get the maximum memory we will try to consume.

◆ get_wildcards()

set< string > vg::IndexRegistry::get_wildcards ( const string & pattern )

static

Get the names of all brace-enclosed {wildcards} in the given pattern.

◆ get_work_dir()

string vg::IndexRegistry::get_work_dir ( )

protected

Function to get and/or initialize the temporary directory in which indexes will live.

◆ gfa_has_haplotypes()

bool vg::IndexRegistry::gfa_has_haplotypes ( const string & filepath )

static

Determine if a GFA has haplotypes as W-lines.

◆ has_recipe()

bool vg::IndexRegistry::has_recipe ( const RecipeName & recipe_name ) const

protected

Check if a recipe identifier correesponds to a recipe.

Recipe identifiers not corresponding to actual recipes are used during planning to represent provided inputs.

TODO: Refactor that with some kind of tagged union or optional.

◆ make_indexes()

void vg::IndexRegistry::make_indexes ( const vector< IndexName > & identifiers )

Create and execute a plan to make the indicated indexes using provided inputs If provided inputs cannot create the desired indexes, throws a InsufficientInputException. When completed, all requested index files will be available via require().

◆ make_plan()

IndexingPlan vg::IndexRegistry::make_plan ( const IndexGroup & end_products ) const

protected

generate a plan to create the indexes

◆ operator=() [1/2]

IndexRegistry& vg::IndexRegistry::operator= ( const IndexRegistry & other )

delete

◆ operator=() [2/2]

IndexRegistry & vg::IndexRegistry::operator= ( IndexRegistry && other )

◆ predates()

bool vg::IndexRegistry::predates	(	const IndexName &	earlier,
		const IndexName &	later
	)		const

For two available indexes, returns true if the modification times on the eariler index are no later than those on the later index.

Useful for enforcing that downstream indexes haven't had their upstream indexes overwritten.

◆ provide() [1/2]

void vg::IndexRegistry::provide	(	const IndexName &	identifier,
		const string &	filename,
		const map< string, string > &	scopes = `{}`
	)

Indicate a serialized file that contains some identified index, optionally with scopes that propagates to descendant files.

TODO: If scopes contain ".", we can run into problems with combinations of different scopes producing the same final string. Right now we only use one kind of scope, which avoids this.

◆ provide() [2/2]

void vg::IndexRegistry::provide	(	const IndexName &	identifier,
		const vector< string > &	filenames,
		const map< string, string > &	scopes = `{}`
	)

Indicate a list of serialized files that contains some identified index, optionally with scopes that propagates to descendant files.

TODO: If scopes contain ".", we can run into problems with combinations of different scopes producing the same final string. Right now we only use one kind of scope, which avoids this.

◆ register_generalization()

void vg::IndexRegistry::register_generalization	(	const RecipeName &	generalizer,
		const RecipeName &	generalizee
	)

Indicate one recipe is a broadened version of another. The indexes consumed and produced by the generalization must be semantically identical to those of the generalizee

◆ register_index() [1/2]

void vg::IndexRegistry::register_index	(	const IndexName &	identifier,
		const string &	suffix
	)

Register an index containing the given identifier.

◆ register_index() [2/2]

void vg::IndexRegistry::register_index	(	const IndexName &	identifier,
		const vector< string > &	suffixes
	)

Register an index containing the given identifier, with multiple possible suffixes. The first suffix where all {wildcards} can be substituted with scopes will be used.

◆ register_recipe()

RecipeName vg::IndexRegistry::register_recipe	(	const vector< IndexName > &	identifiers,
		const vector< IndexName > &	input_identifiers,
		const RecipeFunc &	exec
	)

Register a recipe to produce an index using other indexes or input files. Recipes registered earlier will have higher priority.

◆ require()

vector< string > vg::IndexRegistry::require ( const IndexName & identifier ) const

Get the filename(s) associated with the given index. Aborts if the index is not a known type, or if it is not provided or made.

◆ reset() [1/2]

void vg::IndexRegistry::reset ( )

Discard any provided or constructed indexes.

◆ reset() [2/2]

void vg::IndexRegistry::reset ( const IndexName & identifier )

Remove a provided index.

◆ set_intermediate_file_keeping()

void vg::IndexRegistry::set_intermediate_file_keeping ( bool keep_intermediates )

Should intermediate files be saved to the output directory or the temp directory?

◆ set_prefix()

void vg::IndexRegistry::set_prefix ( const string & prefix )

Prefix for all saved outputs.

◆ set_target_memory_usage()

void vg::IndexRegistry::set_target_memory_usage ( int64_t bytes )

Set the maximum memory that indexing should try to consume (note: this is not strictly adhered to due to difficulties in estimating memory use)

◆ substitute_wildcards()

string vg::IndexRegistry::substitute_wildcards	(	const string &	pattern,
		const map< string, string >	values
	)

static

Substitute wildcards into the given pattern. All wildcards must have values assigned. Extra values not used are allowed.

◆ to_dot() [1/2]

string vg::IndexRegistry::to_dot ( ) const

Returns the recipe graph in dot format.

◆ to_dot() [2/2]

string vg::IndexRegistry::to_dot ( const vector< IndexName > & targets ) const

Returns the recipe graph in dot format with a plan highlighted.

◆ vcf_is_phased()

bool vg::IndexRegistry::vcf_is_phased ( const string & filepath )

static

Determine if a VCF file is phased or not.

Friends And Related Function Documentation

◆ IndexingPlan

friend class IndexingPlan

friend

Member Data Documentation

◆ check_files

bool vg::IndexRegistry::check_files = true

◆ generalizations

map<RecipeName, RecipeName> vg::IndexRegistry::generalizations

protected

Map from generalizees to generalizers.

◆ index_registry

map<IndexName, unique_ptr<IndexFile> > vg::IndexRegistry::index_registry

protected

The storage struct for named indexes. Ordered so it is easier to key on index names.

◆ keep_intermediates

bool vg::IndexRegistry::keep_intermediates = false

protected

should intermediate files end up in the scratch or the output directory?

◆ output_prefix

string vg::IndexRegistry::output_prefix = "index"

protected

filepath that will prefix all saved output

◆ recipe_registry

map<IndexGroup, vector<IndexRecipe> > vg::IndexRegistry::recipe_registry

protected

The storage struct for recipes, which may make index.

◆ registered_suffixes

unordered_set<string> vg::IndexRegistry::registered_suffixes

protected

All of the suffixes that have been registered by indexes.

◆ target_memory_usage

int64_t vg::IndexRegistry::target_memory_usage = numeric_limits<int64_t>::max()

protected

the max memory we will attempt to use

◆ work_dir

string vg::IndexRegistry::work_dir

protected

Temporary directory in which indexes will live.

The documentation for this class was generated from the following files:

src/index_registry.hpp
src/index_registry.cpp

Public Member Functions

Static Public Member Functions

Public Attributes

Protected Member Functions

Protected Attributes

Friends

Detailed Description

Constructor & Destructor Documentation

◆ IndexRegistry() [1/3]

◆ ~IndexRegistry()

◆ IndexRegistry() [2/3]

◆ IndexRegistry() [3/3]

Member Function Documentation

◆ all_finished() [1/2]

◆ all_finished() [2/2]

◆ available()

◆ completed_indexes()

◆ dependency_order()

◆ execute_recipe()

◆ get_index() [1/2]

◆ get_index() [2/2]

◆ get_possible_filenames()

◆ get_prefix()

◆ get_recipe()

◆ get_system_memory()

◆ get_target_memory_usage()

◆ get_wildcards()

◆ get_work_dir()

◆ gfa_has_haplotypes()

◆ has_recipe()

◆ make_indexes()

◆ make_plan()

◆ operator=() [1/2]

◆ operator=() [2/2]

◆ predates()

◆ provide() [1/2]

◆ provide() [2/2]

◆ register_generalization()

◆ register_index() [1/2]

◆ register_index() [2/2]

◆ register_recipe()

◆ require()

◆ reset() [1/2]

◆ reset() [2/2]

◆ set_intermediate_file_keeping()

◆ set_prefix()

◆ set_target_memory_usage()

◆ substitute_wildcards()

◆ to_dot() [1/2]

◆ to_dot() [2/2]

◆ vcf_is_phased()

Friends And Related Function Documentation

◆ IndexingPlan

Member Data Documentation

◆ check_files

◆ generalizations

◆ index_registry

◆ keep_intermediates

◆ output_prefix

◆ recipe_registry

◆ registered_suffixes

◆ target_memory_usage

◆ work_dir