ExaMeld’s File Operations module unifies file search, storage analysis, and comparison under a single interface. Combining traditional keyword search with AI-powered semantic search, the module enables rapid file location and comprehensive storage analysis across enterprise-scale file systems.
File Search
A dual-engine search architecture combines fast keyword indexing with AI-powered semantic search. The Fast engine excels at keyword matching and pattern-based queries, while the Semantic engine enables conceptual searches — for example, searching for “authentication modules” locates related files even without exact keyword matches. Hybrid scoring blends results from both engines, configurable per search to weight conceptual versus keyword relevance. Cached indexes eliminate re-indexing overhead for repeated searches across the same file corpus.
Storage Analysis
A hierarchical tree view visualizes directory structures with configurable display modes: Size mode highlights storage consumption with color intensity representing file size, Recent mode colors based on file access dates, Type mode groups and colors by file extension, and FileCount mode indicates the density of files within directories. Folder statistics aggregate size and file counts at each tree level. A top-100 largest files report pinpoints storage hogs for efficient capacity management.
File Diff
A two-pane file selection interface enables direct comparison between any two files. Side-by-side comparison displays differences with line-level highlighting, making change identification intuitive even across large files. Beyond standard text comparison, ExaMeld applies structure-aware diffing for JSON, XML, and YAML files — understanding document hierarchy rather than treating content as flat text. Renamed keys, reordered elements, and nested changes are identified accurately rather than flagged as wholesale replacements. A directory browser with breadcrumbs enables rapid navigation and selection of comparison targets without context switching.
Indexing
The indexing system supports incremental updates, scanning only modified portions of the file system on subsequent runs. Filtered reindexing by file size enables strategic indexing of only files within specified ranges, reducing index size when small files are irrelevant to the analysis. The system scales to index up to 10 million files, automatically selecting optimal indexing strategies based on corpus size.