feat: add find_bigdecimals

Cargo.toml

@@ -53,6 +53,7 @@ rayon = { version = "1.5" }
 serde = { version = "1", default-features = false }
 serde_json = { version = "1", default-features = false, features = ["alloc"] }
 snafu = { version = "0.8.4", default-features = false }
+sqlparser = { version = "0.45.0", default-features = false }
 tiny-keccak = { version = "2.0.2", features = [ "keccak" ] }
 tracing = { version = "0.1.36", default-features = false }
 tracing-opentelemetry = { version = "0.22.0" }

crates/proof-of-sql/Cargo.toml

@@ -45,6 +45,7 @@ rayon = { workspace = true, optional = true }
 serde = { workspace = true, features = ["serde_derive"] }
 serde_json = { workspace = true }
 snafu = { workspace = true }
+sqlparser = { workspace = true }
 tiny-keccak = { workspace = true }
 tracing = { workspace = true, features = ["attributes"] }
 zerocopy = { workspace = true }

crates/proof-of-sql/src/lib.rs

@@ -8,6 +8,8 @@ extern crate alloc;
 pub mod base;
 pub mod proof_primitive;
 pub mod sql;
+/// Utilities for working with the library
+pub mod utils;
 
 #[cfg(test)]
 mod tests;

crates/proof-of-sql/src/utils/mod.rs

@@ -0,0 +1,3 @@
+//! This module contains utilities for working with the library
+/// Parse DDLs and find bigdecimal columns
+pub mod parse;

crates/proof-of-sql/src/utils/parse.rs

@@ -0,0 +1,91 @@
+use crate::base::map::IndexMap;
+use alloc::string::ToString;
+use sqlparser::{
+    ast::{DataType, ExactNumberInfo, Statement},
+    dialect::GenericDialect,
+    parser::Parser,
+};
+
+/// Parse a DDL file and return a map of table names to bigdecimal columns
+///
+/// # Panics
+/// Panics if there is an error parsing the SQL
+#[must_use]
+pub fn find_bigdecimals(queries: &str) -> IndexMap<String, Vec<String>> {
+    let dialect = GenericDialect {};
+    let ast = match Parser::parse_sql(&dialect, queries) {
+        Ok(ast) => ast,
+        Err(e) => panic!("Error parsing SQL: {e:?}"),
+    };
+    // Find all `CREATE TABLE` statements
+    ast.iter()
+        .filter_map(|statement| match statement {
+            Statement::CreateTable { name, columns, .. } => {
+                // Find all `DECIMAL` columns where precision > 38
+                // Find the table name
+                // Add the table name and column name to the map
+                let str_name = name.to_string();
+                let big_decimal_cols: Vec<String> = columns
+                    .iter()
+                    .filter_map(|column_def| match column_def.data_type {
+                        DataType::Decimal(ExactNumberInfo::PrecisionAndScale(precision, _))
+                            if precision > 38 =>
+                        {
+                            Some(column_def.name.to_string())
+                        }
+                        _ => None,
+                    })
+                    .collect();
+                Some((str_name, big_decimal_cols))
+            }
+            _ => None,
+        })
+        .collect::<IndexMap<String, Vec<String>>>()
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn test_find_bigdecimals() {
+        let sql = "CREATE TABLE IF NOT EXISTS ETHEREUM.BLOCKS(
+            BLOCK_NUMBER BIGINT NOT NULL,
+            TIME_STAMP TIMESTAMP,
+            BLOCK_HASH VARCHAR,
+            MINER VARCHAR,
+            REWARD DECIMAL(78, 0),
+            SIZE_ INT,
+            GAS_USED INT,
+            GAS_LIMIT INT,
+            BASE_FEE_PER_GAS DECIMAL(78, 0),
+            TRANSACTION_COUNT INT,
+            PARENT_HASH VARCHAR,
+            PRIMARY KEY(BLOCK_NUMBER)
+          );
+          
+        CREATE TABLE IF NOT EXISTS ETHEREUM.BLOCK_DETAILS(
+            BLOCK_NUMBER BIGINT NOT NULL,
+            TIME_STAMP TIMESTAMP,
+            SHA3_UNCLES VARCHAR,
+            STATE_ROOT VARCHAR,
+            TRANSACTIONS_ROOT VARCHAR,
+            RECEIPTS_ROOT VARCHAR,
+            UNCLES_COUNT INT,
+            VERSION VARCHAR,
+            LOGS_BLOOM VARCHAR,
+            NONCE VARCHAR,
+            PRIMARY KEY(BLOCK_NUMBER)
+        );";
+        let bigdecimals = find_bigdecimals(sql);
+        assert_eq!(
+            bigdecimals.get("ETHEREUM.BLOCKS").unwrap(),
+            &["REWARD", "BASE_FEE_PER_GAS"]
+        );
+        let empty_vec: Vec<String> = vec![];
+        assert_eq!(
+            bigdecimals.get("ETHEREUM.BLOCK_DETAILS").unwrap(),
+            &empty_vec
+        );
+    }
+}