Carving is the practice of searching an input for files or other kinds of objects based on content, rather than on metadata. File carving is a powerful tool for recovering files and fragments of files when directory entries are corrupt or missing, as may be the case with old files that have been deleted or when performing an analysis on damaged media. Memory carving is a useful tool for analyzing physical and virtual memory dumps when the memory structures are unknown or have been overwritten.
based on the input's content. but it's possible (and sometimes practical) to carve individual files or physical memory.
Most file carvers operate by looking for file headers and/or footers, and then "carving out" the blocks between these two boundaries. Semantic Carving performs carving based on an analysis of the contents of the proposed files.
File carving should be done on a disk image, rather than on the original disk.
File carving tools are listed on the Tools:Data_Recovery wiki page.
Many carving programs have an option to only look at or near sector boundaries where headers are found. However, searching the entire input can find files that have been embedded into other files, such as JPEGs being embedded into Microsoft Word documents. This may be considered an advantage or a disadvantage, depending on the circumstances.
Today most file carving programs will only recover files that are contiguous on the media.
DFRWS2006 featured a file carving challenge. As a condition of entering the challenge, all tools and techniques developed to solve the challenge had to be open sourced.