File Repair via Server-Side Characteristic Matching

The patent segments the file into multiple blocks and identifies specific infected blocks containing virus characteristics. Only these specific blocks are repaired or replaced, while the rest of the file remains intact. This selective approach removes the virus while preserving the overall file structure and content.

The patent extracts and removes only the virus-infected portions (specific blocks containing virus characteristics) from the file, rather than removing or corrupting the entire file. The clean blocks are retained and reassembled to form the repaired file.

The patent uses characteristic information (fingerprints) of virus-infected blocks to identify and locate virus instances in files. Instead of developing custom cleaning codes for each virus type, the system copies and compares characteristic patterns to detect and repair infections efficiently.

The patent changes the approach from virus-specific cleaning codes to a parameter-based identification system using characteristic information of infected blocks. This allows the same repair mechanism to handle multiple virus types by matching their characteristic patterns rather than requiring specialized codes for each.

The patent applies local quality by treating different blocks of the file differently based on their infection status. Clean blocks are preserved in their original form, while only infected blocks are repaired or replaced. This maintains the original file content where possible while removing viruses locally.

The patent divides the file into segments (blocks) and processes each segment independently based on its infection status. This segmentation allows selective repair of only the necessary portions while preserving the integrity and content of uninfected portions.