From 206bf7e9f0d088a130fe89fbb76498ae7cd084d5 Mon Sep 17 00:00:00 2001
From: Andre-Philippe Paquet <appaquet@gmail.com>
Date: Sun, 25 Feb 2024 21:28:32 -0500
Subject: [PATCH] increase value size

---
 CHANGELOG.md |  6 ++++--
 Cargo.toml   |  3 +--
 src/data.rs  | 12 +++++++-----
 3 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index ae6f9f7..bf85778 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -5,9 +5,11 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
-## [0.6.1] - 2024-02-25
+## [0.7.0] - 2024-02-25
 
-- Fix: delete tmp directory when external sorted is used
+- Breaking: increase value size to 24bits (from 16bits), which means that the
+  maximum value size is now 16MB (from 64KB).
+- Fix: delete tmp directory when external sorted is used.
 
 ## [0.6.0] - 2024-02-19
 
diff --git a/Cargo.toml b/Cargo.toml
index 3e3fcfb..5cbaf67 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -7,7 +7,7 @@ license = "Apache-2.0"
 name = "extindex"
 readme = "README.md"
 repository = "https://github.com/appaquet/extindex-rs"
-version = "0.6.1"
+version = "0.7.0"
 
 [lib]
 bench = false
@@ -26,7 +26,6 @@ log = "0.4"
 memmap2 = "0.9"
 serde = { version = "1.0", optional = true }
 smallvec = "1.13.1"
-integer-encoding = "4.0.0"
 
 [dev-dependencies]
 criterion = "0.5"
diff --git a/src/data.rs b/src/data.rs
index 086f53e..bfefbb7 100644
--- a/src/data.rs
+++ b/src/data.rs
@@ -21,7 +21,7 @@ use crate::{Entry as CrateEntry, Serializable};
 
 const INDEX_FILE_MAGIC_HEADER_SIZE: usize = 2;
 const INDEX_FILE_MAGIC_HEADER: [u8; INDEX_FILE_MAGIC_HEADER_SIZE] = [40, 12];
-const INDEX_FILE_VERSION: u8 = 0;
+const INDEX_FILE_VERSION: u8 = 1;
 
 pub const OBJECT_ID_ENTRY: u8 = 0;
 pub const OBJECT_ID_CHECKPOINT: u8 = 1;
@@ -212,7 +212,7 @@ where
 
         output.write_u8(OBJECT_ID_ENTRY)?;
         output.write_u16::<LittleEndian>(key_size as u16)?;
-        output.write_u16::<LittleEndian>(value_size as u16)?;
+        output.write_u24::<LittleEndian>(value_size as u32)?;
 
         if let Some(key_data) = key_data.as_ref() {
             output.write_all(key_data)?;
@@ -241,11 +241,13 @@ where
         }
 
         let key_size = data_cursor.read_u16::<LittleEndian>()? as usize;
-        let data_size = data_cursor.read_u16::<LittleEndian>()? as usize;
+        let data_size = data_cursor.read_u24::<LittleEndian>()? as usize;
         let key = <K as Serializable>::deserialize(data_cursor, key_size)?;
         let value = <V as Serializable>::deserialize(data_cursor, data_size)?;
 
-        let entry_file_size = 1 + 2 + 2 + key_size + data_size;
+        let entry_file_size = 1 + // obj id
+            2 + 3 +  // key and value size bytes
+            key_size + data_size;
         let entry = CrateEntry { key, value };
         Ok((Entry { entry }, entry_file_size))
     }
@@ -259,7 +261,7 @@ where
         }
 
         let key_size = data_cursor.read_u16::<LittleEndian>()? as usize;
-        let _data_size = data_cursor.read_u16::<LittleEndian>()? as usize;
+        let _data_size = data_cursor.read_u24::<LittleEndian>()? as usize;
 
         let key = <K as Serializable>::deserialize(&mut data_cursor, key_size)?;
         Ok(key)