NexVeridian/wikidata-to-surrealdb
1
0
Fork 0
mirror of https://github.com/NexVeridian/wikidata-to-surrealdb.git synced 2025-09-02 01:49:13 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

ClaimData

Browse source
This commit is contained in:
Elijah McMorris 2023-12-14 03:11:24 -08:00
parent 60cbb47a10
commit e0d2cef391
Signed by: NexVeridian
SSH key fingerprint: SHA256:bsA1SKZxuEcEVHAy3gY1HUeM5ykRJl0U0kQHQn0hMg8
2 changed files with 47 additions and 17 deletions
Download patch file Download diff file Expand all files Collapse all files

2
src/main.rs
View file

@ -21,7 +21,7 @@ async fn main() -> Result<(), Error> {
db.use_ns("wikidata").use_db("wikidata").await?; db.use_ns("wikidata").use_db("wikidata").await?;
let file = File::open("data/ex2.json")?; let file = File::open("data/e.json")?;
let reader = BufReader::new(file); let reader = BufReader::new(file);
for line in reader.lines() { for line in reader.lines() {

62
src/utils.rs
View file

@ -3,9 +3,41 @@ use serde::{Deserialize, Serialize};
use surrealdb::sql::Thing; use surrealdb::sql::Thing;
use wikidata::{ClaimValue, ClaimValueData, Entity, Lang, Pid, WikiId}; use wikidata::{ClaimValue, ClaimValueData, Entity, Lang, Pid, WikiId};
#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub enum ClaimData {
Thing(Thing),
ClaimValueData(ClaimValueData),
}
impl ClaimData {
fn from_cvd(cvd: ClaimValueData) -> Self {
match cvd {
ClaimValueData::Item(qid) => ClaimData::Thing(Thing {
id: qid.0.into(),
tb: "Entity".to_string(),
}),
ClaimValueData::Property(pid) => ClaimData::Thing(Thing {
id: pid.0.into(),
tb: "Property".to_string(),
}),
ClaimValueData::Lexeme(lid) => ClaimData::Thing(Thing {
id: lid.0.into(),
tb: "Lexeme".to_string(),
}),
_ => ClaimData::ClaimValueData(cvd),
}
}
}
#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)] #[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct Claims { pub struct Claims {
pub claims: Vec<(Thing, ClaimValueData)>, pub claims: Vec<Claim>,
}
#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
pub struct Claim {
pub id: Thing,
pub value: ClaimData,
} }
#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)] #[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
@ -18,12 +50,12 @@ pub struct EntityMini {
impl EntityMini { impl EntityMini {
pub fn from_entity(entity: Entity) -> (Thing, (Thing, Claims), Self) { pub fn from_entity(entity: Entity) -> (Thing, (Thing, Claims), Self) {
let thing_claim = Thing { let thing_claim = Thing {
id: get_id(&entity).id, id: get_id_entity(&entity).id,
tb: "Claims".to_string(), tb: "Claims".to_string(),
}; };
( (
get_id(&entity), get_id_entity(&entity),
( (
thing_claim.clone(), thing_claim.clone(),
Self::flatten_claims(entity.claims.clone()), Self::flatten_claims(entity.claims.clone()),
@ -41,23 +73,21 @@ impl EntityMini {
claims: claims claims: claims
.iter() .iter()
.flat_map(|(pid, claim_value)| { .flat_map(|(pid, claim_value)| {
let mut flattened = vec![( let mut flattened = vec![Claim {
Thing { id: Thing {
id: pid.0.into(), id: pid.0.into(),
tb: "Property".to_string(), tb: "Property".to_string(),
}, },
claim_value.data.clone(), value: ClaimData::from_cvd(claim_value.data.clone()),
)]; }];
flattened.extend(claim_value.qualifiers.iter().map( flattened.extend(claim_value.qualifiers.iter().map(
|(qualifier_pid, qualifier_value)| { |(qualifier_pid, qualifier_value)| Claim {
( id: Thing {
Thing { id: qualifier_pid.0.into(),
id: qualifier_pid.0.into(), tb: "Property".to_string(),
tb: "Property".to_string(), },
}, value: ClaimData::from_cvd(qualifier_value.clone()),
qualifier_value.clone(),
)
}, },
)); ));
flattened flattened
@ -67,7 +97,7 @@ impl EntityMini {
} }
} }
fn get_id(entity: &Entity) -> Thing { fn get_id_entity(entity: &Entity) -> Thing {
let (id, tb) = match entity.id { let (id, tb) = match entity.id {
WikiId::EntityId(qid) => (qid.0, "Entity".to_string()), WikiId::EntityId(qid) => (qid.0, "Entity".to_string()),
WikiId::PropertyId(pid) => (pid.0, "Property".to_string()), WikiId::PropertyId(pid) => (pid.0, "Property".to_string()),