self_encryption 0.6.0

// Copyright 2016 MaidSafe.net limited.
//
// This SAFE Network Software is licensed to you under (1) the MaidSafe.net Commercial License,
// version 1.0 or later, or (2) The General Public License (GPL), version 3, depending on which
// licence you accepted on initial access to the Software (the "Licences").
//
// By contributing code to the SAFE Network Software, or to this project generally, you agree to be
// bound by the terms of the MaidSafe Contributor Agreement, version 1.1.  This, along with the
// Licenses can be found in the root directory of this project at LICENSE, COPYING and CONTRIBUTOR.
//
// Unless required by applicable law or agreed to in writing, the SAFE Network Software distributed
// under the GPL Licence is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
// KIND, either express or implied.
//
// Please review the Licences for the specific language governing permissions and limitations
// relating to use of the SAFE Network Software.

use std::convert::From;
use std::marker::PhantomData;
use std::mem;

use data_map::{ChunkDetails, DataMap};
use sodiumoxide::crypto::hash::sha256;
use super::{MAX_CHUNK_SIZE, MIN_CHUNK_SIZE, SelfEncryptionError, Storage, StorageError, utils};
use super::small_encryptor::SmallEncryptor;

pub const MIN: u64 = 3 * MIN_CHUNK_SIZE as u64;
pub const MAX: u64 = 3 * MAX_CHUNK_SIZE as u64;

// An encryptor for data which will be split into exactly three chunks (i.e. size is between
// `3 * MIN_CHUNK_SIZE` and `3 * MAX_CHUNK_SIZE` inclusive).  Only `close()` will actually cause
// chunks to be stored.  Until then, data is held internally in `buffer`.
pub struct MediumEncryptor<'a, E: StorageError, S: 'a + Storage<E>> {
    pub storage: &'a mut S,
    pub buffer: Vec<u8>,
    original_chunks: Option<Vec<ChunkDetails>>,
    phantom: PhantomData<E>,
}

impl<'a, E: StorageError, S: Storage<E>> MediumEncryptor<'a, E, S> {
    // Constructor for use with pre-existing `DataMap::Chunks` where there are exactly three chunks.
    // Retrieves the chunks from storage and decrypts them to its internal `buffer`.
    pub fn new(storage: &'a mut S,
               chunks: Vec<ChunkDetails>)
               -> Result<MediumEncryptor<'a, E, S>, SelfEncryptionError<E>> {
        debug_assert!(chunks.len() == 3);
        debug_assert!(MIN <= chunks.iter().fold(0, |acc, chunk| acc + chunk.source_size));
        debug_assert!(chunks.iter().fold(0, |acc, chunk| acc + chunk.source_size) <= MAX);

        let mut buffer = Vec::with_capacity(MAX as usize);
        for (index, chunk) in chunks.iter().enumerate() {
            let content = try!(storage.get(&chunk.hash));
            buffer.extend(try!(utils::decrypt_chunk(&content,
                                                    utils::get_pad_key_and_iv(index, &chunks))));
        }
        Ok(MediumEncryptor {
            storage: storage,
            buffer: buffer,
            original_chunks: Some(chunks),
            phantom: PhantomData,
        })
    }

    // Simply appends to internal buffer assuming the size limit is not exceeded.  No chunks are
    // generated by this call.
    pub fn write(&mut self, data: &[u8]) -> Result<(), SelfEncryptionError<E>> {
        debug_assert!(data.len() as u64 + self.len() <= MAX);
        self.original_chunks = None;
        Ok(self.buffer.extend_from_slice(data))
    }

    // This finalises the encryptor - it should not be used again after this call.  Exactly three
    // chunks will be generated and stored by calling this unless the encryptor didn't receive any
    // `write()` calls.
    pub fn close(&mut self) -> Result<DataMap, SelfEncryptionError<E>> {
        if let Some(ref mut chunks) = self.original_chunks {
            let mut swapped_chunks = vec![];
            mem::swap(&mut swapped_chunks, chunks);
            return Ok(DataMap::Chunks(swapped_chunks));
        }
        // Third the contents, with the extra single or two bytes in the last chunk.
        let chunk_contents = vec![&self.buffer[..(self.buffer.len() / 3)],
                                  &self.buffer[(self.buffer.len() / 3)..(2 *
                                                                         (self.buffer.len() / 3))],
                                  &self.buffer[(2 * (self.buffer.len() / 3))..]];
        // Note the pre-encryption hashes and sizes.
        let mut chunk_details = vec![];
        for (index, contents) in chunk_contents.iter().enumerate() {
            chunk_details.push(ChunkDetails {
                chunk_num: index as u32,
                hash: vec![],
                pre_hash: sha256::hash(contents).0.to_vec(),
                source_size: contents.len() as u64,
            });
        }
        // Encrypt the chunks and note the post-encryption hashes
        let partial_details = chunk_details.clone();
        for (index, (contents, mut details)) in chunk_contents.iter()
            .zip(chunk_details.iter_mut())
            .enumerate() {
            let encrypted_contents = try!(utils::encrypt_chunk(contents,
                                          utils::get_pad_key_and_iv(index, &partial_details)));
            let sha256::Digest(hash) = sha256::hash(&encrypted_contents);
            try!(self.storage.put(hash.to_vec(), encrypted_contents));
            details.hash = hash.to_vec();
        }
        Ok(DataMap::Chunks(chunk_details))
    }

    pub fn len(&self) -> u64 {
        self.buffer.len() as u64
    }

    pub fn is_empty(&self) -> bool {
        self.buffer.is_empty()
    }
}

#[cfg_attr(rustfmt, rustfmt_skip)]
impl<'a, E: StorageError, S: Storage<E>> From<SmallEncryptor<'a, E, S>>
        for MediumEncryptor<'a, E, S> {
    fn from(small_encryptor: SmallEncryptor<'a, E, S>) -> MediumEncryptor<'a, E, S> {
        MediumEncryptor {
            storage: small_encryptor.storage,
            buffer: small_encryptor.buffer,
            original_chunks: None,
            phantom: PhantomData,
        }
    }
}



#[cfg(test)]
mod tests {
    use super::*;

    use data_map::DataMap;
    use itertools::Itertools;
    use maidsafe_utilities::SeededRng;
    use rand::Rng;
    use self_encryptor::SelfEncryptor;
    use super::super::{MAX_CHUNK_SIZE, utils};
    use super::super::small_encryptor::{self, SmallEncryptor};
    use test_helpers::SimpleStorage;

    #[test]
    fn consts() {
        assert_eq!(MIN, small_encryptor::MAX + 1);
    }

    // Writes all of `data` to a new encryptor in a single call, then closes and reads back via
    // a `SelfEncryptor`.
    fn basic_write_and_close(data: &[u8]) {
        let mut storage = SimpleStorage::new();
        let data_map;
        {
            let mut encryptor = MediumEncryptor::from(SmallEncryptor::new(&mut storage, vec![]));
            assert_eq!(encryptor.len(), 0);
            assert!(encryptor.is_empty());
            unwrap!(encryptor.write(data));
            assert_eq!(encryptor.len(), data.len() as u64);
            assert!(!encryptor.is_empty());
            data_map = unwrap!(encryptor.close());
        }
        match data_map {
            DataMap::Chunks(ref chunks) => assert_eq!(chunks.len(), 3),
            _ => panic!("Wrong DataMap type returned."),
        }

        let mut self_encryptor = unwrap!(SelfEncryptor::new(&mut storage, data_map));
        let fetched = unwrap!(self_encryptor.read(0, data.len() as u64));
        assert!(fetched == data);
    }

    // Splits `data` into several pieces, then for each piece:
    //  * constructs a new encryptor from existing chunk details (except for the first piece)
    //  * writes the piece
    //  * closes and reads back the full data via a `SelfEncryptor`.
    fn multiple_writes_then_close<T: Rng>(rng: &mut T, data: &[u8]) {
        let mut storage = SimpleStorage::new();
        let mut existing_data = vec![];
        let data_pieces = utils::make_random_pieces(rng, data, MIN as usize);
        let mut current_chunks = vec![];
        for data in data_pieces {
            let data_map;
            {
                let mut encryptor = if current_chunks.is_empty() {
                    SmallEncryptor::new(&mut storage, vec![]).into()
                } else {
                    unwrap!(MediumEncryptor::new(&mut storage, current_chunks))
                };
                unwrap!(encryptor.write(data));
                existing_data.extend_from_slice(data);
                assert_eq!(encryptor.len(), existing_data.len() as u64);
                data_map = unwrap!(encryptor.close());
            }
            match data_map {
                DataMap::Chunks(ref chunks) => {
                    assert_eq!(chunks.len(), 3);
                    current_chunks = chunks.clone()
                }
                _ => panic!("Wrong DataMap type returned."),
            }

            let mut self_encryptor = unwrap!(SelfEncryptor::new(&mut storage, data_map));
            assert_eq!(self_encryptor.len(), existing_data.len() as u64);
            let fetched = unwrap!(self_encryptor.read(0, existing_data.len() as u64));
            assert!(fetched == existing_data);
        }
        assert!(&existing_data[..] == data);
    }

    #[test]
    fn all_unit() {
        let mut rng = SeededRng::new();
        let data = rng.gen_iter().take(MAX as usize).collect_vec();

        basic_write_and_close(&data[..MIN as usize]);
        basic_write_and_close(&data[..MAX_CHUNK_SIZE as usize]);
        basic_write_and_close(&data[..(MAX_CHUNK_SIZE as usize * 2)]);
        basic_write_and_close(&data);

        multiple_writes_then_close(&mut rng, &data[..(MIN as usize * 2)]);
        multiple_writes_then_close(&mut rng, &data[..MAX_CHUNK_SIZE as usize]);
        multiple_writes_then_close(&mut rng, &data[..(MAX_CHUNK_SIZE as usize * 2)]);
        multiple_writes_then_close(&mut rng, &data);
    }
}