From be89667eb8844c47716ed6ca311a484d136d4c9d Mon Sep 17 00:00:00 2001
From: Connor Tsui <connor.tsui20@gmail.com>
Date: Sat, 17 Feb 2024 22:13:13 -0500
Subject: [PATCH] remove substrait

---
 .gitignore                                    |  2 +
 .../src/execution/substrait/deserialize.rs    | 20 -------
 eggstrain/src/execution/substrait/example.rs  | 52 -------------------
 eggstrain/src/execution/substrait/mod.rs      |  3 --
 .../execution/substrait/substrait_arrow.rs    | 20 -------
 5 files changed, 2 insertions(+), 95 deletions(-)
 delete mode 100644 eggstrain/src/execution/substrait/deserialize.rs
 delete mode 100644 eggstrain/src/execution/substrait/example.rs
 delete mode 100644 eggstrain/src/execution/substrait/mod.rs
 delete mode 100644 eggstrain/src/execution/substrait/substrait_arrow.rs

diff --git a/.gitignore b/.gitignore
index 6985cf1..1c0f3e8 100644
--- a/.gitignore
+++ b/.gitignore
@@ -12,3 +12,5 @@ Cargo.lock
 
 # MSVC Windows builds of rustc generate these, which store debugging information
 *.pdb
+
+data/
diff --git a/eggstrain/src/execution/substrait/deserialize.rs b/eggstrain/src/execution/substrait/deserialize.rs
deleted file mode 100644
index 5123102..0000000
--- a/eggstrain/src/execution/substrait/deserialize.rs
+++ /dev/null
@@ -1,20 +0,0 @@
-use std::fs;
-use substrait::proto::{Plan, ReadRel};
-
-pub fn read_str(path: &str) -> String {
-    fs::read_to_string(path).unwrap_or_else(|_| panic!("Unable to read file {}", path))
-}
-
-pub fn get_plan(path: &str) -> Plan {
-    let plan = serde_json::from_str::<Plan>(&read_str(path))
-        .unwrap_or_else(|_| panic!("Could not parse json {:?} into Plan", path));
-    println!("{}", serde_json::to_string_pretty(&plan).unwrap());
-    plan
-}
-
-pub fn get_read(path: &str) -> ReadRel {
-    let read = serde_json::from_str::<ReadRel>(&read_str(path))
-        .unwrap_or_else(|_| panic!("Could not parse json {:?} into Plan", path));
-    println!("{}", serde_json::to_string_pretty(&read).unwrap());
-    read
-}
diff --git a/eggstrain/src/execution/substrait/example.rs b/eggstrain/src/execution/substrait/example.rs
deleted file mode 100644
index 011e238..0000000
--- a/eggstrain/src/execution/substrait/example.rs
+++ /dev/null
@@ -1,52 +0,0 @@
-//! Want to be able to parse substrait into Arrow schema
-
-use arrow::datatypes::{DataType, Field, Schema};
-use substrait::proto::rel::RelType;
-use substrait::proto::ReadRel;
-
-/// We want the plan to look like this:
-///
-/// ```text
-/// |-+ Aggregate({sales = sum(quantity_price)}, group_by=(product_name, product_id))
-///  |-+ InnerJoin(on=orders.product_id = products.product_id)
-///   |- ReadTable(orders)
-///     |-+ Filter(INDEX_IN("Computers", categories) IS NULL)
-///      |- ReadTable(products)
-/// ```
-pub fn tpch_query1() -> RelType {
-    // Go bottom up
-    let read = ReadRel {
-        common: None,
-        base_schema: todo!(),
-        filter: todo!(),
-        best_effort_filter: todo!(),
-        projection: todo!(),
-        advanced_extension: todo!(),
-        read_type: todo!(),
-    };
-
-    todo!()
-}
-
-pub fn tpch_lineitem_schema() -> Schema {
-    let fields = vec![
-        Field::new("L_ORDERKEY", DataType::UInt64, false),
-        Field::new("L_PARTKEY", DataType::UInt64, false),
-        Field::new("L_SUPPKEY", DataType::UInt64, false),
-        Field::new("L_LINENUMBER", DataType::Int64, false),
-        Field::new("L_QUANTITY", DataType::Decimal128(16, 16), false),
-        Field::new("L_EXTENDEDPRICE", DataType::Decimal128(16, 16), false),
-        Field::new("L_DISCOUNT", DataType::Decimal128(16, 16), false),
-        Field::new("L_TAX", DataType::Decimal128(16, 16), false),
-        Field::new("L_RETURNFLAG", DataType::Boolean, false),
-        Field::new("L_LINESTATUS", DataType::Boolean, false),
-        Field::new("L_SHIPDATE", DataType::Date32, false),
-        Field::new("L_COMMITDATE", DataType::Date32, false),
-        Field::new("L_RECEIPTDATE", DataType::Date32, false),
-        Field::new("L_SHIPINSTRUCT", DataType::FixedSizeBinary(25), false),
-        Field::new("L_SHIPMODE", DataType::FixedSizeBinary(10), false),
-        Field::new("L_COMMENT", DataType::FixedSizeBinary(44), false),
-    ];
-
-    Schema::new(fields)
-}
diff --git a/eggstrain/src/execution/substrait/mod.rs b/eggstrain/src/execution/substrait/mod.rs
deleted file mode 100644
index 0fb207d..0000000
--- a/eggstrain/src/execution/substrait/mod.rs
+++ /dev/null
@@ -1,3 +0,0 @@
-pub mod deserialize;
-pub mod example;
-pub mod substrait_arrow;
diff --git a/eggstrain/src/execution/substrait/substrait_arrow.rs b/eggstrain/src/execution/substrait/substrait_arrow.rs
deleted file mode 100644
index 42b7c20..0000000
--- a/eggstrain/src/execution/substrait/substrait_arrow.rs
+++ /dev/null
@@ -1,20 +0,0 @@
-use arrow::datatypes::{DataType, Schema};
-use std::sync::Arc;
-use substrait::proto::{Type, NamedStruct};
-
-/// https://docs.rs/substrait/latest/substrait/proto/struct.NamedStruct.html
-/// https://docs.rs/substrait/latest/substrait/proto/type/struct.Struct.html
-pub fn schema_translate(substrait_schema: NamedStruct) -> Arc<Schema> {
-    for (i, name) in substrait_schema.names.iter().enumerate() {
-        todo!()
-    }
-
-    todo!()
-}
-
-/// https://docs.rs/substrait/latest/substrait/proto/struct.Type.html
-/// https://docs.rs/arrow/latest/arrow/datatypes/enum.DataType.html
-pub fn type_translate(typ: Type) -> DataType {
-    todo!()
-}
-