content-hash-cache-pattern

To cache file processing results effectively, compute a SHA-256 content hash of the file to use as the cache key, storing outputs as {hash}.json. This ensures fast reuse on repeated runs by matching file contents rather than file paths.

Path-based file cache keys miss on directory changes because they rely on the file location. A content-hash cache key uses SHA-256 hashing instead, making your cache completely path-independent so moving or renaming files never invalidates valid cache hits.

To auto-invalidate cached data on content changes, use a SHA-256 content hash as the cache key. When the file content changes, the generated hash changes, automatically bypassing the old cache and forcing a fresh processing run for the updated file.

The best way to cache PDF text extraction and image analysis results is using a separate service layer to coordinate SHA-256 content hashing and {hash}.json storage, keeping the actual processing functions pure and separating caching as a distinct concern.

Lazy cache directory creation reduces filesystem churn for large datasets by creating the cache directory only when writing a new cache entry. This avoids unnecessary filesystem operations when processing large batches of files that result in cache hits.

To keep file processing functions pure while adding a caching layer, use a service layer wrapper to coordinate cache lookups, SHA-256 hashing, and {hash}.json storage. This isolates caching as a separate concern, enabling easy testing and predictable behavior.