Updating a file using differences and file format therefor

A non-transitory machine readable storage medium storing instructions, that when executed by a client computer, cause the client computer to perform operations comprising:

• identifying a first sync directory for a first version of a file stored locally on the client computer, wherein the first sync directory includes a plurality of first hash values, each first hash value corresponding to a respective one of a plurality of blocks in the first version of the file, and wherein the first sync directory also includes a first global hash value based on the contents of all the blocks of the first version of the file;
• obtaining, from a remote server computer, a second sync directory for a second version of the file, wherein the second sync directory includes a plurality of second hash values, each second hash value corresponding to a respective one of a plurality of blocks in the second version of the file, and wherein the second sync directory also includes a second global hash \value based on the contents of all the blocks of the second version of the file;
• wherein divisions between the plurality of blocks in the first version of the file and the plurality of blocks in the second version of the file are generated according to an algorithm, the algorithm locating divisions between blocks at positions within a particular file when a value of a rolling checksum reaches a particular value;
• determining whether the first global hash value and the second global hash value match;
• and
• in response to determining that the first global hash value and the second global hash value do not match, generating a new copy of the second version of the file stored locally on the client computer, comprising:
• determining, for each of the plurality of second hash values in the second sync directory for the second version of the file, whether the second hash value has a matching first hash value in the first sync directory for the first version of the file,
• for each second hash value that does not have a matching first hash value:
• downloading, from the remote server, the block in the second version of the file that corresponds to the second hash value, and
• writing the downloaded block to the new copy of the second version of the file, and
• for each second hash value that has a matching first hash value:
• writing the block corresponding to the first hash value matching the second hash value to the new copy of the second version of the file without downloading the block in the second version of the file that corresponds to the second hash value.

A system comprising one or more storage devices storing instructions that, when executed by a client computer, cause the client computer to perform operations comprising:

• identifying a first sync directory for a first version of a file stored locally on the client computer, wherein the first sync directory includes a plurality of first hash values, each first hash value corresponding to a respective one of a plurality of blocks in the first version of the file, and wherein the first sync directory also includes a first global hash value based on the contents of all the blocks of the first version of the file;
• obtaining, from a remote server computer, a second sync directory for a second version of the file, wherein the second sync directory includes a plurality of second hash values, each second hash value corresponding to a respective one of a plurality of blocks in the second version of the file, and wherein the second sync directory also includes a second global hash value based on the contents of all the blocks of the second version of the file;
• wherein divisions between the plurality of blocks in the first version of the file and the plurality of blocks in the second version of the file are generated according to an algorithm, the algorithm locating divisions between blocks at positions within a particular file when a value of a rolling checksum reaches a particular value;
• determining whether the first global hash value and the second global hash value match;
• and
• in response to determining that the first global hash value and the second global hash value do not match, generating a new copy of the second version of the file stored locally on the client computer, comprising:
• determining, for each of the plurality of second hash values in the second sync directory for the second version of the file, whether the second hash value has a matching first hash value in the first sync directory for the first version of the file,
• for each second hash value that does not have a matching first hash value:
• downloading, from the remote server, the block in the second version of the file that corresponds to the second hash value, and
• writing the downloaded block to the new copy of the second version of the file, and
• for each second hash value that has a matching first hash value:
• writing the block corresponding to the first hash value matching the second hash value to the new copy of the second version of the file without downloading the block in the second version of the file that corresponds to the second hash value.