This is a toydb, based on Chapter 3 : "Designing Data Intensive Applications" by Martin Kleppmann

Supported (v0.1):

• support simple get/put operations
• write the key-value in the file simpletoy.db
• each put appends to the file
• an in memory index stores the offset of the value
• the get operation seeks to the offset and reads the value
• file format : sequence<entry>'\n'
• entry format : \<keysize\>\<keycontents\>\<valuesize\>\<valuecontents\>
• support compaction

TODO:

• delete older key values
• support delete, put tombstone to deleted value
• support multi segments of db file
• support SSTable, LSM tree
• support a B+Tree based implementation, for good sequential read performance