Example data
A synthetic school, counted by Leita.
Synthetic example data. Not a school and not a customer. The departments below are not a customer, a deployment, or a testimonial. The figures are this run's totals, not a saving and not a performance claim.
Leita gc9f3469 scanned seed 616-school-1 on a 64 MiB tmpfs. The report calls the volume tmpfs and the filesystem "other", and names the walk engine. The kernel was Linux 7.2.5-3-omarchy x86_64. This run does not measure ext4, btrfs, XFS, NTFS, APFS, a network share, Windows, macOS, or a VPS.
What changed
Four directories stand in for departments. Every counted file is NUL bytes of a fixed length. Between the scans, science/sensors.bin grows from 1 MiB to 3 MiB and arts/posters.bin appears at 1 MiB. Administration and Library are untouched. A short text file at the root carries the disclosure.
| Department | Before (bytes) | After (bytes) | Change |
|---|---|---|---|
| Administration | 2,097,152 | 2,097,152 | unchanged |
| Library | 4,194,304 | 4,194,304 | unchanged |
| Science | 2,097,152 | 4,194,304 | +2,097,152 bytes |
| Arts | 1,048,576 | 2,097,152 | +1,048,576 bytes |
The after scan totals 12,587,008 bytes in 7 files and 4 directories. That is the allocated size of the department files plus the label file (4,096 bytes). Science's directory grew by 2,097,152 bytes and Arts by 1,048,576. The new poster file has no earlier entry, so its change is empty. The label file is under 1 MiB, and this build leaves the change empty for an entry that small; the current scan still counts it.
The comparison scan's duration field is PT0.005S. That is a tmpfs reading of a few megabytes on this machine. It is not a disk benchmark. Leita's summary on this machine began “walk: 24 readers at Full pace”.
What this binary wrote
No licence was loaded. The landing page says a current scan's CSV and JSON are free, and that historical comparisons, before-and-after reports, and HTML and PDF reports are Pro.
This build wrote the comparison JSON and the HTML report without a licence. The JSON is the second scan in one data directory. The HTML is the second scan of the same fixture in another data directory. Both end with: Unlicensed copy made during the 30-day trial. The current-scan CSV and JSON were separate --no-history invocations, and their comparison field is empty. PDF was not published from this run. Scheduled delivery, reusable report profiles, and a total across machines were not run.
Downloads
HTML report of the after scan. Comparison JSON. JSON of the earlier scan. Current-scan CSV. Current-scan JSON. Fixture truth. The runner.
Reproduce
On Linux, with unshare and leita on PATH:
python3 reproduce.pyThe script mounts the tmpfs, writes the fixture, and runs:
leita --top 20 --report before.json --format json /tmp/leita-school-example/mnt/synthetic-example-school
leita --top 20 --report after.json --format json /tmp/leita-school-example/mnt/synthetic-example-school
leita --top 20 --report after.html --format html /tmp/leita-school-example/mnt/synthetic-example-school
leita --top 20 --report current.json --format json --no-history /tmp/leita-school-example/mnt/synthetic-example-school
leita --top 20 --report current.csv --format csv --no-history /tmp/leita-school-example/mnt/synthetic-example-schoolIt exits non-zero if a department total, the aggregate, or the scripted change disagrees with the allocated sizes.