searchengines-web/MeiliSearch
searchengines-web/MeiliSearch
1
0
Fork 0
mirror of https://github.com/meilisearch/MeiliSearch synced 2025-07-04 04:17:10 +02:00
Code Issues Releases Wiki Activity

add boolean support for csv documents

Browse source
This commit is contained in:
Tamo 2023-03-09 11:12:49 +01:00
parent df3986cd83
commit c5f22be6e1
3 changed files with 151 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

28
milli/src/documents/builder.rs
View file

@ -116,12 +116,13 @@ impl<W: Write> DocumentsBatchBuilder<W> {
let value = &record[*i];
match type_ {
AllowedType::Number => {
if value.trim().is_empty() {
let trimmed_value = value.trim();
if trimmed_value.is_empty() {
to_writer(&mut self.value_buffer, &Value::Null)?;
} else if let Ok(integer) = value.trim().parse::<i64>() {
} else if let Ok(integer) = trimmed_value.parse::<i64>() {
to_writer(&mut self.value_buffer, &integer)?;
} else {
match value.trim().parse::<f64>() {
match trimmed_value.parse::<f64>() {
Ok(float) => {
to_writer(&mut self.value_buffer, &float)?;
}
@ -135,6 +136,25 @@ impl<W: Write> DocumentsBatchBuilder<W> {
}
}
}
AllowedType::Boolean => {
let trimmed_value = value.trim();
if trimmed_value.is_empty() {
to_writer(&mut self.value_buffer, &Value::Null)?;
} else {
match trimmed_value.parse::<bool>() {
Ok(bool) => {
to_writer(&mut self.value_buffer, &bool)?;
}
Err(error) => {
return Err(Error::ParseBool {
error,
line,
value: value.to_string(),
});
}
}
}
}
AllowedType::String => {
if value.is_empty() {
to_writer(&mut self.value_buffer, &Value::Null)?;
@ -173,6 +193,7 @@ impl<W: Write> DocumentsBatchBuilder<W> {
#[derive(Debug)]
enum AllowedType {
String,
Boolean,
Number,
}
@ -181,6 +202,7 @@ fn parse_csv_header(header: &str) -> (&str, AllowedType) {
match header.rsplit_once(':') {
Some((field_name, field_type)) => match field_type {
"string" => (field_name, AllowedType::String),
"boolean" => (field_name, AllowedType::Boolean),
"number" => (field_name, AllowedType::Number),
// if the pattern isn't reconized, we keep the whole field.
_otherwise => (header, AllowedType::String),

17
milli/src/documents/mod.rs
View file

@ -90,6 +90,7 @@ impl DocumentsBatchIndex {
#[derive(Debug)]
pub enum Error {
ParseFloat { error: std::num::ParseFloatError, line: usize, value: String },
ParseBool { error: std::str::ParseBoolError, line: usize, value: String },
InvalidDocumentFormat,
InvalidEnrichedData,
InvalidUtf8(Utf8Error),
@ -136,6 +137,9 @@ impl fmt::Display for Error {
Error::ParseFloat { error, line, value } => {
write!(f, "Error parsing number {:?} at line {}: {}", value, line, error)
}
Error::ParseBool { error, line, value } => {
write!(f, "Error parsing boolean {:?} at line {}: {}", value, line, error)
}
Error::InvalidDocumentFormat => {
f.write_str("Invalid document addition format, missing the documents batch index.")
}
@ -274,6 +278,19 @@ mod test {
]);
}
#[test]
fn csv_types_dont_panic() {
let csv1_content =
"id:number,b:boolean,c,d:number\n1,,,\n2,true,doggo,2\n3,false,the best doggo,-2\n4,,\"Hello, World!\",2.5";
let csv1 = csv::Reader::from_reader(Cursor::new(csv1_content));
let mut builder = DocumentsBatchBuilder::new(Vec::new());
builder.append_csv(csv1).unwrap();
let vector = builder.into_inner().unwrap();
DocumentsBatchReader::from_reader(Cursor::new(vector)).unwrap();
}
#[test]
fn out_of_order_csv_fields() {
let csv1_content = "id:number,b\n1,0";