CSV import · result-copy review
The row count can match while the imported data changes.
A self-created example compares three CSV rows with three records in a detached import-result copy. See missing and extra keys, changed values and changed types, then choose the check you need.
3 source CSV rows3 snapshot records4 findings
Four findings despite equal row counts
| Source position / snapshot alias | Finding |
|---|---|
| CSV 2 / D00001 / F3 | Quantity 0 became NULL |
| CSV 3 / D00002 / F2 | An integer price became text with the same digits |
| CSV 4 / — | A source item is missing from the snapshot |
| — / D00003 | An unexpected record exists only in the snapshot |
CSV positions count logical rows including the header. D aliases follow typed primary-key sorting for that exact snapshot; they are not database rowids. F2 maps price and F3 maps quantity. The two uniquely matched keys include rows with changed fields.
Inspect the actual example reports
The reports keep positions, field aliases and hashes, not original keys, column names or cell values. These are fictional examples, not paid client work. The HTML/TXT explanations are in Korean; JSON includes English finding codes.
MATCH_WITHIN_SCOPE · 0
Open the corrected-copy reportWhat this example checked
Only the selected table, keys and mapped values in a CSV and detached SQLite copy were compared. Zero findings in the fictional corrected copy do not certify an operational import or the entire database. The checker does not fill missing data or modify a production database.
Before importing: decide which columns to send
Build a free explicit mapping plan and check omitted required columns and blank-cell policies. This does not execute an import or inspect your database.
Open the free column planner · English controls availableChoose the check your files need
Find missing or changed values in two CSV copies
If you can export a result copy to CSV and need an exact-string comparison, use the free comparison for up to 20 rows per file. Map one to four key columns and up to five value columns. CSV export alone cannot recover SQLite NULL or integer-versus-text types.
Compare two CSV files freeFind basic product-CSV errors before import
For SKU, product name, nonnegative integer KRW price, stock and image-URL format, use the existing product precheck tool. That product checks individual CSV files; it does not include the CSV-to-SQLite comparison shown above.
14,900 KRW · one-time downloadUp to 1,000 rows, 100 columns and 2MiB per file. Confirm final taxes and totals at checkout. The general checker has Korean and English controls and guides. The same ZIP includes a separate English WooCommerce simple-product companion; its decimal-price scope does not change the general checker’s integer-price rules.
Are both files SQLite snapshots?
The fictional example above compares one CSV with one SQLite snapshot. If both files are detached SQLite snapshots from the same reference time, check the separate Kmong snapshot-review service. It does not include CSV-to-SQLite import verification or the existing CSV tool purchase.
Two detached snapshots, up to 20MiB each, five ordinary tables, 10,000 rows/32 columns per table and declared primary keys without NULL. HTML/JSON/TXT reports link input hashes with record aliases; one explanation/display revision for the same input is included.
Do not send live databases, WAL, SHM or journal files. Confirm current price, supported scope, data rights, redaction and schedule on Kmong before a protected paid order and agreed data intake. This page accepts no files and sends no message. Recovery, original-file changes, migration and production-app verification are excluded. Service communication and explanations are in Korean; English customer support is not promised.
Check the two-snapshot review · Korean service detailsFour decisions before sharing any data
- Preserve production originals and prepare a closed copy you are authorized to read.
- Agree keys that identify the same records. Do not pair duplicate or blank keys arbitrarily.
- Specify compared fields and allowed type conversions. Do not auto-correct blanks, zero or leading zeroes.
- Record limits, unchecked items and input hashes. Do not enter personal or real customer data on this page.