Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
454 changes: 418 additions & 36 deletions Cargo.lock

Large diffs are not rendered by default.

1 change: 1 addition & 0 deletions Cargo.toml
Original file line number Diff line number Diff line change
Expand Up @@ -60,6 +60,7 @@ peg = "0.8"
rand = "0.10.1"
oxilangtag = "0.1.5"
fundu = "2.0.1"
hdt = { version = "0.6.0", default-features = false, features = ["nt"] }
memmap2 = "0.9.5"
sprs = "0.11.3"
walkdir = "2.5.0"
Expand Down
10 changes: 10 additions & 0 deletions lib/maplib/src/model.rs
Original file line number Diff line number Diff line change
Expand Up @@ -492,6 +492,16 @@ impl Model {
Ok(())
}

pub fn write_hdt<W: Write>(
&mut self,
buffer: &mut W,
graph: &NamedGraph,
) -> Result<(), MaplibError> {
self.triplestore
.write_hdt(buffer, graph)
.map_err(MaplibError::TriplestoreError)
}

pub fn write_cim_xml<W: Write>(
&mut self,
buffer: &mut W,
Expand Down
1 change: 1 addition & 0 deletions lib/triplestore/Cargo.toml
Original file line number Diff line number Diff line change
Expand Up @@ -29,6 +29,7 @@ uuid.workspace = true
thiserror.workspace = true
oxrdfio.workspace = true
oxttl.workspace = true
hdt.workspace = true
memmap2.workspace = true
sparesults.workspace = true
tracing.workspace = true
Expand Down
2 changes: 2 additions & 0 deletions lib/triplestore/src/errors.rs
Original file line number Diff line number Diff line change
Expand Up @@ -60,6 +60,8 @@ pub enum TriplestoreError {
DecodeError(PolarsError),
#[error("Error collecting lazy triples {0}")]
LazyLoadError(PolarsError),
#[error("HDT error: {0}")]
HDTError(String),
}

impl<T> From<PoisonError<T>> for TriplestoreError {
Expand Down
227 changes: 149 additions & 78 deletions lib/triplestore/src/triples_read.rs
Original file line number Diff line number Diff line change
Expand Up @@ -4,8 +4,9 @@ use crate::{NewTriples, TriplesToAdd};
use std::cmp;

use cimxml_import::{fix_cim_quad, Remapper};
use hdt::Hdt;
use memmap2::MmapOptions;
use oxrdf::{BlankNode, GraphName, NamedNode, NamedOrBlankNode, Quad, Term, Triple};
use oxrdf::{BlankNode, GraphName, Literal, NamedNode, NamedOrBlankNode, Quad, Term, Triple};
use oxrdfio::{
JsonLdProfileSet, LoadedDocument, RdfFormat, RdfParser, RdfSyntaxError, SliceQuadParser,
};
Expand All @@ -25,7 +26,10 @@ use representation::{
use representation::{OBJECT_COL_NAME, SUBJECT_COL_NAME};
use std::collections::HashMap;
use std::fs::File;
use std::io::Cursor;
use std::path::Path;
use std::str::FromStr;
use std::sync::Arc;
use std::time::Instant;
use tracing::{debug, instrument};

Expand All @@ -35,6 +39,7 @@ const UTF8_BOM: [u8; 3] = [0xEF, 0xBB, 0xBF];
pub enum ExtendedRdfFormat {
Normal(RdfFormat),
CIMXML,
HDT,
}

impl Triplestore {
Expand Down Expand Up @@ -70,6 +75,8 @@ impl Triplestore {
ExtendedRdfFormat::Normal(RdfFormat::JsonLd {
profile: JsonLdProfileSet::empty(),
})
} else if path.extension() == Some("hdt".as_ref()) {
ExtendedRdfFormat::HDT
} else {
todo!("Have not implemented file format {:?}", path);
};
Expand Down Expand Up @@ -152,84 +159,23 @@ impl Triplestore {
} else {
matches!(rdf_format, ExtendedRdfFormat::Normal(RdfFormat::NTriples))
};
let mut readers = if matches!(
rdf_format,
ExtendedRdfFormat::Normal(RdfFormat::NTriples)
| ExtendedRdfFormat::Normal(RdfFormat::Turtle)
) && parallel
{
let threads = if let Ok(threads) = std::thread::available_parallelism() {
threads.get()
} else {
1
};

let mut readers = vec![];
if rdf_format == ExtendedRdfFormat::Normal(RdfFormat::Turtle) {
let mut parser = TurtleParser::new();
for (k, v) in prefixes {
parser = parser.with_prefix(k, v.as_str()).unwrap();
}
if !checked {
parser = parser.lenient();
}
if let Some(base_iri) = base_iri {
parser = parser.with_base_iri(base_iri).unwrap();
}
for r in parser.split_slice_for_parallel_parsing(use_slice, threads) {
readers.push(MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::TurtlePar(r),
});
}
} else if rdf_format == ExtendedRdfFormat::Normal(RdfFormat::NTriples) {
let mut parser = NTriplesParser::new();
if !checked {
parser = parser.lenient();
}
for r in parser.split_slice_for_parallel_parsing(use_slice, threads) {
readers.push(MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::NTriplesPar(r),
});
}
}
readers
let hdt;
let mut readers = if rdf_format == ExtendedRdfFormat::HDT {
hdt = Hdt::read(Cursor::new(use_slice))
.map_err(|e| TriplestoreError::HDTError(e.to_string()))?;
vec![hdt_reader(&hdt)]
} else if parallel && rdf_format == ExtendedRdfFormat::Normal(RdfFormat::Turtle) {
parallel_turtle_readers(use_slice, prefixes, checked, base_iri)
} else if parallel && rdf_format == ExtendedRdfFormat::Normal(RdfFormat::NTriples) {
parallel_ntriples_readers(use_slice, checked)
} else {
let use_format = match rdf_format {
ExtendedRdfFormat::Normal(n) => n,
ExtendedRdfFormat::CIMXML => RdfFormat::RdfXml,
};
let mut parser = RdfParser::from(use_format.clone());
if !checked {
parser = parser.lenient();
}
if let Some(base_iri) = &base_iri {
parser = parser.with_base_iri(base_iri).unwrap();
}
let mut for_slice = parser.for_slice(use_slice);
if matches!(use_format, RdfFormat::JsonLd { .. }) {
for_slice = for_slice.with_document_loader(move |url| {
if let Some(doc) = known_contexts.get(url) {
Ok(LoadedDocument {
url: url.to_string(),
content: doc.clone().into_bytes(),
format: RdfFormat::JsonLd {
profile: JsonLdProfileSet::empty(),
},
})
} else {
Err(Box::new(TriplestoreError::MissingContext(url.to_string())))
}
});
}
if matches!(rdf_format, ExtendedRdfFormat::CIMXML) {
vec![MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::CIMXML(for_slice, base_iri.clone()),
}]
} else {
vec![MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::Other(for_slice),
}]
}
vec![rdf_parser_reader(
use_slice,
&rdf_format,
checked,
base_iri,
known_contexts,
)]
};
debug!("Effective parallelization for reading is {}", readers.len());

Expand Down Expand Up @@ -578,6 +524,101 @@ fn create_predicate_map<'a>(
Ok((out_r, graph_predicate_map))
}

fn hdt_reader(hdt: &Hdt) -> MyFromSliceQuadReader<'_> {
MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::HDT(Box::new(
hdt.triples_all().map(hdt_string_triple_to_quad),
)),
}
}

fn parallel_turtle_readers<'a>(
slice: &'a [u8],
prefixes: &HashMap<String, NamedNode>,
checked: bool,
base_iri: Option<String>,
) -> Vec<MyFromSliceQuadReader<'a>> {
let threads = std::thread::available_parallelism()
.map(|t| t.get())
.unwrap_or(1);
let mut parser = TurtleParser::new();
for (k, v) in prefixes {
parser = parser.with_prefix(k, v.as_str()).unwrap();
}
if !checked {
parser = parser.lenient();
}
if let Some(base_iri) = base_iri {
parser = parser.with_base_iri(base_iri).unwrap();
}
parser
.split_slice_for_parallel_parsing(slice, threads)
.into_iter()
.map(|r| MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::TurtlePar(r),
})
.collect()
}

fn parallel_ntriples_readers(slice: &[u8], checked: bool) -> Vec<MyFromSliceQuadReader<'_>> {
let threads = std::thread::available_parallelism()
.map(|t| t.get())
.unwrap_or(1);
let mut parser = NTriplesParser::new();
if !checked {
parser = parser.lenient();
}
parser
.split_slice_for_parallel_parsing(slice, threads)
.into_iter()
.map(|r| MyFromSliceQuadReader {
parser: MyFromSliceQuadReaderKind::NTriplesPar(r),
})
.collect()
}

fn rdf_parser_reader<'a>(
slice: &'a [u8],
rdf_format: &ExtendedRdfFormat,
checked: bool,
base_iri: Option<String>,
known_contexts: HashMap<String, String>,
) -> MyFromSliceQuadReader<'a> {
let use_format = match rdf_format {
ExtendedRdfFormat::Normal(n) => *n,
ExtendedRdfFormat::CIMXML => RdfFormat::RdfXml,
ExtendedRdfFormat::HDT => unreachable!("HDT is handled in read_triples"),
};
let mut parser = RdfParser::from(use_format);
if !checked {
parser = parser.lenient();
}
if let Some(base_iri) = &base_iri {
parser = parser.with_base_iri(base_iri).unwrap();
}
let mut for_slice = parser.for_slice(slice);
if matches!(use_format, RdfFormat::JsonLd { .. }) {
for_slice = for_slice.with_document_loader(move |url| {
let Some(doc) = known_contexts.get(url) else {
return Err(Box::new(TriplestoreError::MissingContext(url.to_string())));
};
Ok(LoadedDocument {
url: url.to_string(),
content: doc.clone().into_bytes(),
format: RdfFormat::JsonLd {
profile: JsonLdProfileSet::empty(),
},
})
});
}
let parser = if matches!(rdf_format, ExtendedRdfFormat::CIMXML) {
MyFromSliceQuadReaderKind::CIMXML(for_slice, base_iri)
} else {
MyFromSliceQuadReaderKind::Other(for_slice)
};
MyFromSliceQuadReader { parser }
}

//Adapted from proposed change to https://github.com/oxigraph/
#[must_use]
pub struct MyFromSliceQuadReader<'a> {
Expand All @@ -589,6 +630,7 @@ pub enum MyFromSliceQuadReaderKind<'a> {
CIMXML(SliceQuadParser<'a>, Option<String>),
TurtlePar(SliceTurtleParser<'a>),
NTriplesPar(SliceNTriplesParser<'a>),
HDT(Box<dyn Iterator<Item = Quad> + Send + 'a>),
}

impl Iterator for MyFromSliceQuadReader<'_> {
Expand All @@ -615,10 +657,39 @@ impl Iterator for MyFromSliceQuadReader<'_> {
Ok(triple) => Ok(triple.in_graph(GraphName::default())),
Err(e) => Err(e.into()),
},
MyFromSliceQuadReaderKind::HDT(iter) => Ok(iter.next()?),
})
}
}

fn hdt_string_triple_to_quad(t: [Arc<str>; 3]) -> Quad {
let [s, p, o] = t;
let subject = if let Some(label) = s.strip_prefix("_:") {
NamedOrBlankNode::BlankNode(BlankNode::new_unchecked(label))
} else {
NamedOrBlankNode::NamedNode(NamedNode::new_unchecked(s.as_ref()))
};
let predicate = NamedNode::new_unchecked(p.as_ref());
Quad::new(
subject,
predicate,
hdt_object_string_to_term(&o),
GraphName::DefaultGraph,
)
}

fn hdt_object_string_to_term(o: &str) -> Term {
if o.starts_with('"') {
return Literal::from_str(o)
.unwrap_or_else(|_| Literal::new_simple_literal(o))
.into();
}
if let Some(label) = o.strip_prefix("_:") {
return BlankNode::new_unchecked(label).into();
}
NamedNode::new_unchecked(o).into()
}

fn get_or_insert_dt(
base_rdfnode_type_ref: BaseRDFNodeTypeRef,
type_map: &mut HashMap<String, BaseRDFNodeType>,
Expand Down
27 changes: 27 additions & 0 deletions lib/triplestore/src/triples_write.rs
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@ use std::collections::HashMap;
use std::io::Write;

mod fast_ntriples;
mod hdt_write;
mod pretty_turtle;
mod serializers;

Expand Down Expand Up @@ -132,6 +133,32 @@ impl Triplestore {
Ok(())
}

pub fn write_hdt<W: Write>(
&mut self,
buf: &mut W,
graph: &NamedGraph,
) -> Result<(), TriplestoreError> {
self.check_graph_exists(graph)?;
let mut builder = hdt_write::HdtBuilder::new();
for (predicate, df_map) in self.graph_triples_map.get(graph).unwrap() {
for ((subject_type, object_type), tt) in df_map {
for (lf, _) in tt.get_lazy_frames(&None, &None)? {
let triples = global_df_as_triples(
lf.collect().unwrap(),
subject_type.clone(),
object_type.clone(),
predicate,
self.global_cats.clone(),
);
for t in &triples {
builder.add_triple(t);
}
}
}
}
builder.finish(buf)
}

pub(crate) fn check_graph_exists(&self, graph: &NamedGraph) -> Result<(), TriplestoreError> {
if !self.graph_triples_map.contains_key(graph) {
Err(TriplestoreError::GraphDoesNotExist(graph.to_string()))
Expand Down
Loading