rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2026-06-01 08:00:41 +00:00
Code Issues Packages Projects Releases Wiki Activity

expose helpers for json field writer manipulation

Browse Source 
closes #1302
This commit is contained in:
Saroh
2022-05-04 12:27:18 +02:00
parent 7e3c0c5392
commit 9e38343352
3 changed files with 58 additions and 45 deletions
Download Patch File Download Diff File Expand all files Collapse all files

58
src/query/query_parser/query_parser.rs
View File

@@ -7,7 +7,9 @@ use tantivy_query_grammar::{UserInputAst, UserInputBound, UserInputLeaf, UserInp
use super::logical_ast::*;
use crate::core::Index;
use crate::indexer::JsonTermWriter;
use crate::indexer::{
generate_term_from_json_writer, generate_terms_from_json_writer, JsonTermWriter,
};
use crate::query::{
AllQuery, BooleanQuery, BoostQuery, EmptyQuery, Occur, PhraseQuery, Query, RangeQuery,
TermQuery,
@@ -660,26 +662,22 @@ fn generate_literals_for_str(
Ok(Some(LogicalLiteral::Phrase(terms)))
}
enum NumValue {
U64(u64),
I64(i64),
F64(f64),
DateTime(OffsetDateTime),
}
fn infer_type_num(phrase: &str) -> Option<NumValue> {
fn infer_fast_value_term(json_term_writer: &mut JsonTermWriter, phrase: &str) -> Option<Term> {
if let Ok(dt) = OffsetDateTime::parse(phrase, &Rfc3339) {
let dt_utc = dt.to_offset(UtcOffset::UTC);
return Some(NumValue::DateTime(dt_utc));
return Some(generate_term_from_json_writer(
json_term_writer,
DateTime::from_utc(dt_utc),
));
}
if let Ok(u64_val) = str::parse::<u64>(phrase) {
return Some(NumValue::U64(u64_val));
return Some(generate_term_from_json_writer(json_term_writer, u64_val));
}
if let Ok(i64_val) = str::parse::<i64>(phrase) {
return Some(NumValue::I64(i64_val));
return Some(generate_term_from_json_writer(json_term_writer, i64_val));
}
if let Ok(f64_val) = str::parse::<f64>(phrase) {
return Some(NumValue::F64(f64_val));
return Some(generate_term_from_json_writer(json_term_writer, f64_val));
}
None
}
@@ -694,38 +692,12 @@ fn generate_literals_for_json_object(
) -> Result<Vec<LogicalLiteral>, QueryParserError> {
let mut logical_literals = Vec::new();
let mut term = Term::new();
term.set_field(Type::Json, field);
let mut json_term_writer = JsonTermWriter::wrap(&mut term);
for segment in json_path.split('.') {
json_term_writer.push_path_segment(segment);
let mut json_term_writer = JsonTermWriter::initialize(field, json_path, &mut term);
if let Some(term) = infer_fast_value_term(&mut json_term_writer, phrase) {
logical_literals.push(LogicalLiteral::Term(term));
}
if let Some(num_value) = infer_type_num(phrase) {
match num_value {
NumValue::U64(u64_val) => {
json_term_writer.set_fast_value(u64_val);
}
NumValue::I64(i64_val) => {
json_term_writer.set_fast_value(i64_val);
}
NumValue::F64(f64_val) => {
json_term_writer.set_fast_value(f64_val);
}
NumValue::DateTime(dt_val) => {
json_term_writer.set_fast_value(DateTime::from_utc(dt_val));
}
}
logical_literals.push(LogicalLiteral::Term(json_term_writer.term().clone()));
}
json_term_writer.close_path_and_set_type(Type::Str);
let terms = generate_terms_from_json_writer(&mut json_term_writer, phrase, text_analyzer);
drop(json_term_writer);
let term_num_bytes = term.as_slice().len();
let mut token_stream = text_analyzer.token_stream(phrase);
let mut terms: Vec<(usize, Term)> = Vec::new();
token_stream.process(&mut |token| {
term.truncate(term_num_bytes);
term.append_bytes(token.text.as_bytes());
terms.push((token.position, term.clone()));
});
if terms.len() <= 1 {
for (_, term) in terms {
logical_literals.push(LogicalLiteral::Term(term));