MeiliSearch/milli/src/criterion.rs

182 lines
5.6 KiB
Rust
Raw Normal View History

use std::fmt;
use std::str::FromStr;
2020-11-27 12:14:56 +01:00
2021-06-16 18:33:33 +02:00
use serde::{Deserialize, Serialize};
2020-11-27 12:14:56 +01:00
use crate::error::{Error, UserError};
#[derive(Debug, Serialize, Deserialize, Clone, PartialEq, Eq)]
2020-08-12 10:43:02 +02:00
pub enum Criterion {
/// Sorted by decreasing number of matched query terms.
/// Query words at the front of an attribute is considered better than if it was at the back.
2020-08-12 10:43:02 +02:00
Words,
/// Sorted by increasing number of typos.
Typo,
2020-08-12 10:43:02 +02:00
/// Sorted by increasing distance between matched query terms.
Proximity,
/// Documents with quey words contained in more important
2021-08-23 11:37:18 +02:00
/// attributes are considered better.
2020-08-12 10:43:02 +02:00
Attribute,
/// Dynamically sort at query time the documents. None, one or multiple Asc/Desc sortable
/// attributes can be used in place of this criterion at query time.
Sort,
2020-08-12 10:43:02 +02:00
/// Sorted by the similarity of the matched words with the query words.
Exactness,
/// Sorted by the increasing value of the field specified.
Asc(String),
2020-08-12 10:43:02 +02:00
/// Sorted by the decreasing value of the field specified.
Desc(String),
2020-08-12 10:43:02 +02:00
}
impl Criterion {
/// Returns the field name parameter of this criterion.
pub fn field_name(&self) -> Option<&str> {
match self {
Criterion::Asc(name) | Criterion::Desc(name) => Some(name),
_otherwise => None,
}
}
}
impl FromStr for Criterion {
type Err = Error;
2021-08-23 11:37:18 +02:00
fn from_str(text: &str) -> Result<Criterion, Self::Err> {
match text {
2020-12-04 12:02:22 +01:00
"words" => Ok(Criterion::Words),
"typo" => Ok(Criterion::Typo),
2020-12-04 12:02:22 +01:00
"proximity" => Ok(Criterion::Proximity),
"attribute" => Ok(Criterion::Attribute),
"sort" => Ok(Criterion::Sort),
2020-12-04 12:02:22 +01:00
"exactness" => Ok(Criterion::Exactness),
2021-08-23 11:37:18 +02:00
text => match AscDesc::from_str(text) {
Ok(AscDesc::Asc(field)) => Ok(Criterion::Asc(field)),
Ok(AscDesc::Desc(field)) => Ok(Criterion::Desc(field)),
Err(UserError::InvalidAscDescSyntax { name }) => {
Err(UserError::InvalidCriterionName { name }.into())
}
Err(error) => {
Err(UserError::InvalidCriterionName { name: error.to_string() }.into())
}
Ok(AscDesc::Asc(Member::Geo(_))) | Ok(AscDesc::Desc(Member::Geo(_))) => {
Err(UserError::AttributeLimitReached)? // TODO: TAMO: use a real error
}
Err(error) => Err(error.into()),
2021-08-23 11:37:18 +02:00
},
}
}
}
#[derive(Debug, Serialize, Deserialize, Clone, PartialEq)]
pub enum Member {
Field(String),
Geo([f64; 2]),
}
impl FromStr for Member {
type Err = UserError;
fn from_str(text: &str) -> Result<Member, Self::Err> {
if text.starts_with("_geoPoint(") {
let point =
text.strip_prefix("_geoPoint(")
.and_then(|point| point.strip_suffix(")"))
.ok_or_else(|| UserError::InvalidCriterionName { name: text.to_string() })?;
let point = point
.split(',')
2021-09-01 10:56:19 +02:00
.map(|el| el.trim().parse())
.collect::<Result<Vec<f64>, _>>()
.map_err(|_| UserError::InvalidCriterionName { name: text.to_string() })?;
Ok(Member::Geo([point[0], point[1]]))
} else {
Ok(Member::Field(text.to_string()))
}
}
}
impl fmt::Display for Member {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match self {
Member::Field(name) => write!(f, "{}", name),
Member::Geo([lat, lng]) => write!(f, "_geoPoint({}, {})", lat, lng),
}
}
}
impl Member {
pub fn field(&self) -> Option<&str> {
match self {
Member::Field(field) => Some(field),
Member::Geo(_) => None,
}
}
pub fn geo_point(&self) -> Option<&[f64; 2]> {
match self {
Member::Geo(point) => Some(point),
Member::Field(_) => None,
}
}
}
#[derive(Debug, Serialize, Deserialize, Clone, PartialEq)]
2021-08-23 11:37:18 +02:00
pub enum AscDesc {
Asc(Member),
Desc(Member),
2021-08-23 11:37:18 +02:00
}
impl AscDesc {
pub fn member(&self) -> &Member {
2021-08-23 11:37:18 +02:00
match self {
AscDesc::Asc(member) => member,
AscDesc::Desc(member) => member,
2021-08-23 11:37:18 +02:00
}
}
pub fn field(&self) -> Option<&str> {
self.member().field()
}
2021-08-23 11:37:18 +02:00
}
impl FromStr for AscDesc {
type Err = UserError;
/// Since we don't know if this was deserialized for a criterion or a sort we just return a
/// string and let the caller create his own error
2021-08-23 11:37:18 +02:00
fn from_str(text: &str) -> Result<AscDesc, Self::Err> {
match text.rsplit_once(':') {
Some((left, "asc")) => Ok(AscDesc::Asc(left.parse()?)),
Some((left, "desc")) => Ok(AscDesc::Desc(left.parse()?)),
_ => Err(UserError::InvalidCriterionName { name: text.to_string() }),
}
}
}
2020-08-12 10:43:02 +02:00
pub fn default_criteria() -> Vec<Criterion> {
vec![
Criterion::Words,
Criterion::Typo,
2020-08-12 10:43:02 +02:00
Criterion::Proximity,
Criterion::Attribute,
Criterion::Sort,
2020-08-12 10:43:02 +02:00
Criterion::Exactness,
]
}
impl fmt::Display for Criterion {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
use Criterion::*;
match self {
2021-06-16 18:33:33 +02:00
Words => f.write_str("words"),
Typo => f.write_str("typo"),
Proximity => f.write_str("proximity"),
Attribute => f.write_str("attribute"),
Sort => f.write_str("sort"),
2021-06-16 18:33:33 +02:00
Exactness => f.write_str("exactness"),
2021-08-23 11:37:18 +02:00
Asc(attr) => write!(f, "{}:asc", attr),
Desc(attr) => write!(f, "{}:desc", attr),
}
}
}