Copyright	(c) Fuzz Leonard 2025
License	MIT
Maintainer	cyborg@bionicfuzz.com
Stability	experimental
Safe Haskell	None
Language	Haskell2010

Clod.FileSystem.Checksums

Contents

Checksum operations
Database operations
Change detection
Database management

Description

This module provides functions for tracking file changes using checksums. It calculates XXH3 (64-bit) hashes of file content and maintains a database of files that have been processed, allowing us to detect new, modified, deleted, and renamed files.

The file checksum database is stored as a Dhall configuration file with the following structure:

{ files =
    { "pathtofile1.txt" =
        { path = "pathtofile1.txt"
        , checksum = "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"
        , lastModified = "2025-01-01T12:00:00Z"
        , optimizedName = "path-to-file1.txt"
        }
    , "pathtofile2.md" =
        { path = "pathtofile2.md"
        , checksum = "7d865e959b2466918c9863afca942d0fb89d7c9ac0c99bafc3749504ded97730"
        , lastModified = "2025-01-02T14:30:00Z"
        , optimizedName = "path-to-file2.md"
        }
    }
, checksums =
    { "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855" = "pathtofile1.txt"
    , "7d865e959b2466918c9863afca942d0fb89d7c9ac0c99bafc3749504ded97730" = "pathtofile2.md"
    }
, lastStagingDir = Some ".staging20250101-120000"
, lastRunTime = "2025-01-01T12:00:00Z"
}

This database allows efficient lookup of files by path or checksum, detection of renamed files (same content with different paths), and tracking of previous staging directories.

Synopsis

calculateChecksum :: ByteString -> Checksum
checksumFile :: FileReadCap -> FilePath -> ClodM Checksum
initializeDatabase :: ClodM ClodDatabase
loadDatabase :: FilePath -> ClodM ClodDatabase
saveDatabase :: FilePath -> ClodDatabase -> ClodM ()
updateDatabase :: ClodDatabase -> FilePath -> Checksum -> UTCTime -> OptimizedName -> ClodDatabase
detectFileChanges :: FileReadCap -> ClodDatabase -> [FilePath] -> FilePath -> ClodM ([(FilePath, FileStatus)], [(FilePath, FilePath)])
findChangedFiles :: ClodDatabase -> [(FilePath, Checksum, UTCTime)] -> ClodM [(FilePath, FileStatus)]
findRenamedFiles :: ClodDatabase -> [(FilePath, FileStatus)] -> [(FilePath, FilePath)]
getFileStatus :: ClodDatabase -> FilePath -> Checksum -> ClodM FileStatus
data FileStatus
- = Unchanged
- | New
- | Modified
- | Deleted
- | Renamed FilePath
cleanupStagingDirectories :: ClodM ()
flushMissingEntries :: FileReadCap -> ClodDatabase -> FilePath -> ClodM ClodDatabase

Checksum operations

calculateChecksum :: ByteString -> Checksum Source #

Calculate XXH3 checksum (64-bit) of a ByteString XXH3 is a fast non-cryptographic hash function with excellent performance

checksumFile :: FileReadCap -> FilePath -> ClodM Checksum Source #

Calculate the checksum of a file Only text files are allowed to be checksummed

Database operations

initializeDatabase :: ClodM ClodDatabase Source #

Initialize a new, empty database

loadDatabase :: FilePath -> ClodM ClodDatabase Source #

Load the database from disk using Dhall

saveDatabase :: FilePath -> ClodDatabase -> ClodM () Source #

Save the database to disk using Dhall serialization

updateDatabase :: ClodDatabase -> FilePath -> Checksum -> UTCTime -> OptimizedName -> ClodDatabase Source #

Update the database with a new file entry

Change detection

detectFileChanges :: FileReadCap -> ClodDatabase -> [FilePath] -> FilePath -> ClodM ([(FilePath, FileStatus)], [(FilePath, FilePath)]) Source #

Detect changes by comparing current files with database

findChangedFiles :: ClodDatabase -> [(FilePath, Checksum, UTCTime)] -> ClodM [(FilePath, FileStatus)] Source #

Find files that need processing (new, modified, renamed)

findRenamedFiles :: ClodDatabase -> [(FilePath, FileStatus)] -> [(FilePath, FilePath)] Source #

Find files that have been renamed

getFileStatus :: ClodDatabase -> FilePath -> Checksum -> ClodM FileStatus Source #

Detect file status by comparing against database

data FileStatus Source #

Data type for tracking file status

Constructors

Unchanged	File has not changed
New	New file
Modified	Existing file with modified content
Deleted	File no longer exists
Renamed FilePath	File was renamed (new path)