Struct SwhGraphProperties

pub struct SwhGraphProperties<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> { /* private fields */ }

Properties on graph nodes

This structures has many type parameters, to allow loading only some properties, and checking at compile time that only loaded properties are accessed.

Extra properties can be loaded, following the builder pattern on the owning graph. For example, this does not compile:

use swh_graph::graph::SwhGraphWithProperties;
use swh_graph::mph::DynMphf;
use swh_graph::SwhGraphProperties;

swh_graph::graph::SwhUnidirectionalGraph::new(PathBuf::from("./graph"))
    .expect("Could not load graph")
    .init_properties()
    .properties()
    .author_timestamp(42);

but this does:

use swh_graph::graph::SwhGraphWithProperties;
use swh_graph::mph::DynMphf;
use swh_graph::SwhGraphProperties;

swh_graph::graph::SwhUnidirectionalGraph::new(PathBuf::from("./graph"))
    .expect("Could not load graph")
    .init_properties()
    .load_properties(SwhGraphProperties::load_timestamps)
    .expect("Could not load timestamp properties")
    .properties()
    .author_timestamp(42);

Implementations

impl<TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<NoMaps, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

pub fn load_maps<MPHF: LoadableSwhidMphf>( self, ) -> Result<SwhGraphProperties<MappedMaps<MPHF>, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Consumes a SwhGraphProperties and returns a new one with these methods available:

• SwhGraphProperties::node_id_unchecked
• SwhGraphProperties::node_id
• SwhGraphProperties::swhid
• SwhGraphProperties::node_type

pub fn with_maps<MAPS: MaybeMaps>( self, maps: MAPS, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Alternative to load_maps that allows using arbitrary maps implementations

impl<MAPS: Maps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

Functions to map between SWHID and node id.

Only available after calling load_contents or load_all_properties

pub unsafe fn node_id_unchecked(&self, swhid: &SWHID) -> NodeId

Returns the node id of the given SWHID

May return the id of a random node if the SWHID does not exist in the graph.

Safety

Undefined behavior if the swhid does not exist.

pub fn node_id<T: TryInto<SWHID>>( &self, swhid: T, ) -> Result<NodeId, NodeIdFromSwhidError<<T as TryInto<SWHID>>::Error>>

Returns the node id of the given SWHID, or None if it does not exist.

pub fn node_id_from_string_swhid<T: AsRef<str>>( &self, swhid: T, ) -> Result<NodeId, NodeIdFromSwhidError<StrSWHIDDeserializationError>>

Specialized version of node_id when the SWHID is a string

Under the hood, when using GOVMPH, node_id serializes the SWHID to a string, which can be bottleneck. This function skips the serialization by working directly on the string.

pub fn swhid(&self, node_id: NodeId) -> SWHID

Returns the SWHID of a given node

Panics

If the node id does not exist.

pub fn try_swhid(&self, node_id: NodeId) -> Result<SWHID, OutOfBoundError>

Returns the SWHID of a given node, or None if the node id does not exist

pub fn node_type(&self, node_id: NodeId) -> NodeType

Returns the type of a given node

Panics

If the node id does not exist.

pub fn try_node_type( &self, node_id: NodeId, ) -> Result<NodeType, OutOfBoundError>

Returns the type of a given node, or None if the node id does not exist

impl<MAPS: MaybeMaps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, NoTimestamps, PERSONS, CONTENTS, STRINGS, LABELNAMES>

pub fn load_timestamps( self, ) -> Result<SwhGraphProperties<MAPS, MappedTimestamps, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Consumes a SwhGraphProperties and returns a new one with these methods available:

• SwhGraphProperties::author_timestamp
• SwhGraphProperties::author_timestamp_offset
• SwhGraphProperties::committer_timestamp
• SwhGraphProperties::committer_timestamp_offset

pub fn opt_load_timestamps( self, ) -> Result<SwhGraphProperties<MAPS, OptMappedTimestamps, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Equivalent to Self::load_timestamps that does not require all files to be present

pub fn with_timestamps<TIMESTAMPS: MaybeTimestamps>( self, timestamps: TIMESTAMPS, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Alternative to load_timestamps that allows using arbitrary timestamps implementations

impl<MAPS: MaybeMaps, TIMESTAMPS: OptTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

Functions to access timestamps of revision and release nodes

Only available after calling load_timestamps or load_all_properties

pub fn author_timestamp( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<i64>, TIMESTAMPS>

Returns the number of seconds since Epoch that a release or revision was authored at

Panics

If the node id does not exist

pub fn try_author_timestamp( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<i64>, OutOfBoundError>, TIMESTAMPS>

Returns the number of seconds since Epoch that a release or revision was authored at

Returns Err if the node id is unknown, and Ok(None) if the node has no author timestamp

pub fn author_timestamp_offset( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<i16>, TIMESTAMPS>

Returns the UTC offset in minutes of a release or revision’s authorship date

Panics

If the node id does not exist

pub fn try_author_timestamp_offset( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<i16>, OutOfBoundError>, TIMESTAMPS>

Returns the UTC offset in minutes of a release or revision’s authorship date

Returns Err if the node id is unknown, and Ok(None) if the node has no author timestamp

pub fn committer_timestamp( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<i64>, TIMESTAMPS>

Returns the number of seconds since Epoch that a revision was committed at

Panics

If the node id does not exist

pub fn try_committer_timestamp( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<i64>, OutOfBoundError>, TIMESTAMPS>

Returns the number of seconds since Epoch that a revision was committed at

Returns Err if the node id is unknown, and Ok(None) if the node has no committer timestamp

pub fn committer_timestamp_offset( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<i16>, TIMESTAMPS>

Returns the UTC offset in minutes of a revision’s committer date

Panics

If the node id does not exist

pub fn try_committer_timestamp_offset( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<i16>, OutOfBoundError>, TIMESTAMPS>

Returns the UTC offset in minutes of a revision’s committer date

Returns Err if the node id is unknown, and Ok(None) if the node has no committer timestamp

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, NoPersons, CONTENTS, STRINGS, LABELNAMES>

pub fn load_persons( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, MappedPersons, CONTENTS, STRINGS, LABELNAMES>>

Consumes a SwhGraphProperties and returns a new one with these methods available:

• SwhGraphProperties::author_id
• SwhGraphProperties::committer_id

pub fn opt_load_persons( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, OptMappedPersons, CONTENTS, STRINGS, LABELNAMES>>

Equivalent to Self::load_persons that does not require all files to be present

pub fn with_persons<PERSONS: MaybePersons>( self, persons: PERSONS, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Alternative to load_persons that allows using arbitrary persons implementations

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: OptPersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

Functions to access the id of the author or committer of revision/release nodes.

Only available after calling load_persons or load_all_properties

pub fn author_id( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<u32>, PERSONS>

Returns the id of the author of a revision or release, if any

Panics

If the node id does not exist

pub fn try_author_id( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<u32>, OutOfBoundError>, PERSONS>

Returns the id of the author of a revision or release, if any

Returns Err if the node id does not exist, and Ok(None) if the node has no author

pub fn committer_id( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<u32>, PERSONS>

Returns the id of the committer of a revision, if any

Panics

If the node id does not exist

pub fn try_committer_id( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<u32>, OutOfBoundError>, PERSONS>

Returns the id of the committer of a revision, if any

Returns None if the node id does not exist, and Ok(None) if the node has no author

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, NoContents, STRINGS, LABELNAMES>

pub fn load_contents( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, MappedContents, STRINGS, LABELNAMES>>

Consumes a SwhGraphProperties and returns a new one with these methods available:

• SwhGraphProperties::is_skipped_content
• SwhGraphProperties::content_length

pub fn opt_load_contents( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, OptMappedContents, STRINGS, LABELNAMES>>

Equivalent to Self::load_contents that does not require all files to be present

pub fn with_contents<CONTENTS: MaybeContents>( self, contents: CONTENTS, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Alternative to load_contents that allows using arbitrary contents implementations

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: OptContents, STRINGS: MaybeStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

Functions to access properties of content nodes

Only available after calling load_contents or load_all_properties

pub fn is_skipped_content( &self, node_id: NodeId, ) -> PropertiesResult<'_, bool, CONTENTS>

Returns whether the node is a skipped content

Non-content objects get a false value, like non-skipped contents.

Panics

If the node id does not exist.

pub fn try_is_skipped_content( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<bool, OutOfBoundError>, CONTENTS>

Returns whether the node is a skipped content, or Err if the node id does not exist

Non-content objects get a false value, like non-skipped contents.

pub fn content_length( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<u64>, CONTENTS>

Returns the length of the given content.

May be None for skipped contents

Panics

If the node id does not exist.

pub fn try_content_length( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<u64>, OutOfBoundError>, CONTENTS>

Returns the length of the given content, or Err if the node id does not exist

May be Ok(None) for skipped contents

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, NoStrings, LABELNAMES>

pub fn load_strings( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, MappedStrings, LABELNAMES>>

Consumes a SwhGraphProperties and returns a new one with these methods available:

• SwhGraphProperties::message_base64
• SwhGraphProperties::message
• SwhGraphProperties::tag_name_base64
• SwhGraphProperties::tag_name

pub fn opt_load_strings( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, OptMappedStrings, LABELNAMES>>

Equivalent to Self::load_strings that does not require all files to be present

pub fn with_strings<STRINGS: MaybeStrings>( self, strings: STRINGS, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Alternative to load_strings that allows using arbitrary strings implementations

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: OptStrings, LABELNAMES: MaybeLabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

Functions to access message of revision/release nodes, and names of release nodes

Only available after calling load_strings or load_all_properties

pub fn message_base64( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<&[u8]>, STRINGS>

Returns the base64-encoded message of a revision or release, or the base64-encoded URL of an origin

Panics

If the node id does not exist

pub fn try_message_base64( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<&[u8]>, OutOfBoundError>, STRINGS>

Returns the base64-encoded message of a revision or release

Returns Err if the node id is unknown, and Ok(None) if the node has no message.

pub fn message( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<Vec<u8>>, STRINGS>

Returns the message of a revision or release, or the URL of an origin

Panics

If the node id does not exist

pub fn try_message( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<Vec<u8>>, OutOfBoundError>, STRINGS>

Returns the message of a revision or release, or the URL of an origin

Returns Err if the node id is unknown, and Ok(None) if the node has no message.

pub fn tag_name_base64( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<&[u8]>, STRINGS>

Returns the tag name of a release, base64-encoded

Panics

If the node id does not exist

pub fn try_tag_name_base64( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<&[u8]>, OutOfBoundError>, STRINGS>

Returns the tag name of a release, base64-encoded

Returns Err if the node id is unknown, and Ok(None) if the node has no tag name.

pub fn tag_name( &self, node_id: NodeId, ) -> PropertiesResult<'_, Option<Vec<u8>>, STRINGS>

Returns the tag name of a release

Panics

If the node id does not exist

pub fn try_tag_name( &self, node_id: NodeId, ) -> PropertiesResult<'_, Result<Option<Vec<u8>>, OutOfBoundError>, STRINGS>

Returns the tag name of a release

Returns Err if the node id is unknown, and Ok(None) if the node has no tag name.

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, NoLabelNames>

pub fn load_label_names( self, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, MappedLabelNames>>

Consumes a SwhGraphProperties and returns a new one with this methods available:

• SwhGraphProperties::label_name
• SwhGraphProperties::label_name_base64

pub fn with_label_names<LABELNAMES: MaybeLabelNames>( self, label_names: LABELNAMES, ) -> Result<SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>>

Alternative to load_label_names that allows using arbitrary label_names implementations

impl<MAPS: MaybeMaps, TIMESTAMPS: MaybeTimestamps, PERSONS: MaybePersons, CONTENTS: MaybeContents, STRINGS: MaybeStrings, LABELNAMES: LabelNames> SwhGraphProperties<MAPS, TIMESTAMPS, PERSONS, CONTENTS, STRINGS, LABELNAMES>

Functions to access names of arc labels.

Only available after calling load_label_names or load_all_properties.

pub fn num_label_names(&self) -> u64

Returns the total number of labels in the graph

pub fn iter_label_name_ids(&self) -> impl Iterator<Item = LabelNameId>

Returns an iterator on LabelNameIds for this graph

pub fn par_iter_label_name_ids( &self, ) -> impl ParallelIterator<Item = LabelNameId>

Returns an iterator on LabelNameIds for this graph

pub fn label_name_base64(&self, label_name_id: LabelNameId) -> Vec<u8>

Returns the base64-encoded name of an arc label

This is the file name (resp. branch name) of a label on an arc from a directory (resp. snapshot)

Panics

If label_name_id does not exist

pub fn try_label_name_base64( &self, label_name_id: LabelNameId, ) -> Result<Vec<u8>, OutOfBoundError>

Returns the base64-encoded name of an arc label

This is the file name (resp. branch name) of a label on an arc from a directory (resp. snapshot)

pub fn label_name(&self, label_name_id: LabelNameId) -> Vec<u8>

Returns the name of an arc label

This is the file name (resp. branch name) of a label on an arc from a directory (resp. snapshot)

Panics

If label_name_id does not exist

pub fn try_label_name( &self, label_name_id: LabelNameId, ) -> Result<Vec<u8>, OutOfBoundError>

Returns the name of an arc label

This is the file name (resp. branch name) of a label on an arc from a directory (resp. snapshot)

pub fn label_name_id( &self, name: impl AsRef<[u8]>, ) -> Result<LabelNameId, LabelIdFromNameError>

Given a branch/file name, returns the label_name id used by edges with that name, or None if it does not exist.

This is the inverse function of label_name

Unlike in Java where this function is O(1), this implementation is O(log2(num_labels)) because it uses a binary search, as the MPH function can only be read from Java.

impl SwhGraphProperties<NoMaps, NoTimestamps, NoPersons, NoContents, NoStrings, NoLabelNames>

pub fn new(path: impl AsRef<Path>, num_nodes: usize) -> Self

Creates an empty SwhGraphProperties instance, which will load properties from the given path prefix.

pub fn load_all<MPHF: LoadableSwhidMphf>( self, ) -> Result<AllSwhGraphProperties<MPHF>>

Consumes an empty SwhGraphProperties instance and returns a new one with all properties loaded and all methods available.

 use swh_graph::graph::SwhGraphWithProperties;
use swh_graph::mph::DynMphf;
use swh_graph::SwhGraphProperties;

SwhGraphProperties::new(PathBuf::from("./graph"), 123)
    .load_all::<DynMphf>()
    .expect("Could not load properties");

is equivalent to:

use swh_graph::mph::DynMphf;
use swh_graph::SwhGraphProperties;

SwhGraphProperties::new(PathBuf::from("./graph"), 123)
    .load_maps::<DynMphf>()
    .expect("Could not load node2swhid/swhid2node")
    .load_timestamps()
    .expect("Could not load timestamp properties")
    .load_persons()
    .expect("Could not load person properties")
    .load_contents()
    .expect("Could not load content properties")
    .load_strings()
    .expect("Could not load string properties");