Quote:
Originally Posted by Yil
Once the logic to read/compute over a previous portion of a file is done adding the ability to calculate it for any file would be possible.
|
If I'm not mistaken, CRC32 can be computated "incrementally": if you already have the CRC32 value of the first part of the file, you can use that as the starting value for the algorithm and then feed it only the new bytes. By storing the intermediate CRC32 value (before the resuming), alot of CPU and disk operations can be saved.