recountdiff

recount-diff [OPTIONS] <OLD_RECOUNT_FILE> <NEW_RECOUNT_FILE>

<OLD_RECOUNT_FILE>
    The path to the recount output file representing the older version of the codebase for comparison.

<NEW_RECOUNT_FILE>
    The path to the recount output file representing the newer version of the codebase for comparison.

-h, --help
    Displays a brief help message and exits, detailing command usage and available options.

-V, --version
    Displays the version information for the recount-diff utility and exits.

-q, --quiet
    Suppresses non-error output, showing only significant differences or error messages.

-v, --verbose
    Enables verbose output, providing more detailed information about the comparison process and results.

-b, --blanks
    Compares and reports only the differences in the number of blank lines between the two input files.

-c, --comments
    Compares and reports only the differences in the number of comment lines.

-e, --empty
    Compares and reports only the differences in the number of empty files found.

-f, --files
    Compares and reports only the differences in the total number of files.

-l, --lines
    Compares and reports only the differences in the number of actual code lines (non-blank, non-comment).

-L, --total-lines
    Compares and reports only the differences in the total number of all lines (code + comments + blanks).

-m, --match-paths
    Attempts to match file paths across the two input files even if their internal order differs.

-s, --sort
    Sorts the output by file paths or other relevant criteria for consistent and readable comparison results.

-D, --display-diffs
    Displays granular differences for individual files within the comparison, not just overall summaries.

-x <PATTERN>, --exclude-files=<PATTERN>
    Excludes files whose paths match the specified <PATTERN> from the comparison. Pattern typically supports wildcards.

-i <PATTERN>, --include-files=<PATTERN>
    Includes only files whose paths match the specified <PATTERN> in the comparison. This option usually takes precedence over --exclude-files.

The recount-diff command, typically found as part of the recount suite, is a specialized utility designed to compare and report differences in various code metrics between two recount output files. These input files (often with a .recount extension) contain aggregated statistics such as lines of code (LOC), comment lines, blank lines, and total file counts for a given codebase at a specific point in time.

recount-diff is invaluable for software development teams to track project evolution, quantify the impact of refactoring efforts, or monitor overall codebase growth or shrinkage. Unlike traditional line-by-line textual diffs, this tool focuses on the quantitative changes in source code characteristics, providing a high-level overview. It enables developers and project managers to quickly ascertain how metrics like the total number of files, lines of code, or specific line types have changed between different snapshots or versions of a project, thereby facilitating better project management and code quality assessment.

The recount-diff command is designed to operate on output files generated by the recount utility itself. It cannot directly compare raw source code directories or files. Therefore, prior execution of recount on both versions of the codebase is a prerequisite.

The accuracy of the reported differences is directly dependent on the consistency and correctness of the input .recount files. Inconsistent options used during the initial recount runs (e.g., different exclusion patterns) can lead to misleading comparison results.

As this utility is part of a specific package (recount), its availability and the exact set of options may vary across different Linux distributions or versions of the recount package. Always consult the specific man page or --help output for your installed version.

Each input file for recount-diff is typically a structured text file (e.g., JSON or a custom format) produced by the recount command. These files contain aggregated data per language and per file, providing a detailed breakdown of line types and counts.

A common use case involves running recount on your project at the beginning of a sprint (generating old_project.recount) and again at the end (generating new_project.recount). You can then use recount-diff old_project.recount new_project.recount to get an immediate summary of code added, removed, or modified, helping to assess sprint productivity or codebase changes.

The recount-diff utility emerged as an integral component of the open-source recount project, which was developed to provide a lightweight and efficient way to count lines of code, comments, and blank lines across various programming languages. Recognizing the need to track changes over time, recount-diff was created to complement the core recount tool by enabling quantitative comparisons between different codebase snapshots. Its development reflects the broader trend in software engineering towards metric-driven development and the automation of codebase analysis, allowing for better visibility into project growth and architectural evolution.

recount(1), diff(1), cloc(1)