refactor: move parser into own crate

crates/datasources/Cargo.toml

@@ -25,6 +25,7 @@ futures = { workspace = true }
 gcp-bigquery-client = "0.18.0"
 klickhouse = { version = "0.11.2", features = ["tls"] }
 protogen = { path = "../protogen" }
+parser = {path = "../parser"}
 datafusion_ext = { path = "../datafusion_ext" }
 mongodb = "2.8.1"
 mysql_async = { version = "0.33.0", default-features = false, features = [

crates/datasources/src/debug/mod.rs

@@ -37,6 +37,8 @@ use datafusion_ext::errors::ExtensionError;
 use datafusion_ext::functions::VirtualLister;
 use errors::DebugError;
 use futures::Stream;
+use parser::errors::ParserError;
+use parser::options::{OptionValue, ParseOptionValue};
 use protogen::metastore::types::options::TunnelOptions;
 use serde::{Deserialize, Serialize};
 
@@ -48,6 +50,22 @@ pub enum DebugTableType {
     NeverEnding,
 }
 
+impl ParseOptionValue<DebugTableType> for OptionValue {
+    fn parse_opt(self) -> Result<DebugTableType, ParserError> {
+        let opt = match self {
+            Self::QuotedLiteral(s) | Self::UnquotedLiteral(s) => s
+                .parse()
+                .map_err(|e: DebugError| ParserError::ParserError(e.to_string()))?,
+            o => {
+                return Err(ParserError::ParserError(format!(
+                    "Expected a string, got: {}",
+                    o
+                )))
+            }
+        };
+        Ok(opt)
+    }
+}
 impl fmt::Display for DebugTableType {
     fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
         write!(f, "{}", self.as_str())

crates/datasources/src/mongodb/mod.rs

@@ -25,6 +25,9 @@ use mongodb::bson::spec::BinarySubtype;
 use mongodb::bson::{bson, Binary, Bson, Document, RawDocumentBuf};
 use mongodb::options::{ClientOptions, FindOptions};
 use mongodb::{Client, Collection};
+use parser::errors::ParserError;
+use parser::options::{OptionValue, ParseOptionValue};
+use parser::{parser_err, unexpected_type_err};
 use tracing::debug;
 
 use crate::bson::array_to_bson;
@@ -66,6 +69,19 @@ impl FromStr for MongoDbProtocol {
     }
 }
 
+impl ParseOptionValue<MongoDbProtocol> for OptionValue {
+    fn parse_opt(self) -> Result<MongoDbProtocol, ParserError> {
+        let opt = match self {
+            Self::QuotedLiteral(s) | Self::UnquotedLiteral(s) => {
+                s.parse().map_err(|e| parser_err!("{e}"))?
+            }
+            o => return Err(unexpected_type_err!("mongodb protocol", o)),
+        };
+        Ok(opt)
+    }
+}
+
+
 impl Display for MongoDbProtocol {
     fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
         let s = match self {

crates/metastore/Cargo.toml

@@ -1,18 +1,18 @@
 [package]
 name = "metastore"
-version = {workspace = true}
-edition = {workspace = true}
+version = { workspace = true }
+edition = { workspace = true }
 
 [lints]
 workspace = true
 
 [dependencies]
-ioutil = {path = "../ioutil"}
-logutil = {path = "../logutil"}
-protogen = {path = "../protogen"}
+ioutil = { path = "../ioutil" }
+logutil = { path = "../logutil" }
+protogen = { path = "../protogen" }
 sqlbuiltins = { path = "../sqlbuiltins" }
-object_store_util = {path = "../object_store_util"}
-pgrepr = {path = "../pgrepr"}
+object_store_util = { path = "../object_store_util" }
+pgrepr = { path = "../pgrepr" }
 async-trait = { workspace = true }
 datafusion = { workspace = true }
 futures = { workspace = true }
@@ -26,10 +26,6 @@ tracing = { workspace = true }
 uuid = { version = "1.7.0", features = ["v4", "fast-rng", "macro-diagnostics"] }
 bytes = "1.4"
 once_cell = "1.19.0"
-proptest-derive = "0.4"
 tower = "0.4"
 dashmap = "5.5.0"
 catalog = { path = "../catalog" }
-
-[dev-dependencies]
-proptest = "1.4"

crates/parser/Cargo.toml

@@ -0,0 +1,17 @@
+[package]
+name = "parser"
+version.workspace = true
+edition.workspace = true
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+datafusion.workspace = true
+datafusion_ext = { path = "../datafusion_ext" }
+protogen = { path = "../protogen" }
+prql-compiler = "0.11.3"
+thiserror.workspace = true
+
+
+[lints]
+workspace = true

crates/parser/src/errors.rs

@@ -0,0 +1,20 @@
+pub use datafusion::sql::sqlparser::parser::ParserError;
+
+#[derive(Debug, thiserror::Error)]
+pub enum ParseError {
+    #[error("SQL statement currently unsupported: {0}")]
+    UnsupportedSQLStatement(String),
+    #[error(transparent)]
+    SqlParser(#[from] ParserError),
+    #[error("Invalid object name length: {length}, max: {max}")]
+    InvalidNameLength { length: usize, max: usize },
+    #[error("Other error: {0}")]
+    Other(String),
+}
+
+pub type Result<T, E = ParseError> = std::result::Result<T, E>;
+impl From<ParseError> for ParserError {
+    fn from(e: ParseError) -> Self {
+        ParserError::ParserError(e.to_string())
+    }
+}