# Find CSV rows with no matching key

Return rows in CSV A whose key has no match in CSV B, or the reverse, with composite keys, alternate delimiters, and browser-local processing.

## Run it

- **CLI:** `gizza tool csv-anti-join "id,name
1,Alice
2,Bob
3,Carol" 'b=id,city
2,Berlin
3,Cairo
4,Delhi' 'key=id'`
- **Web:** https://gizza.ai/tools/csv-anti-join/
- **Agents:** machine-readable descriptor (parameters JSON Schema) at https://gizza.ai/tools/csv-anti-join/tool.json

## Inputs

- `a` — CSV A (left table) _(field)_
- `b` — CSV B (right table) _(field)_
- `key` — Key column(s) in A _(field)_
- `key_b` — Key column(s) in B (blank = same as A) _(field)_
- `direction` — Rows to return _(field)_
- `delimiter` — Delimiter _(field)_
- `case_sensitive` — Case-sensitive key matching _(field)_
- `trim_keys` — Trim whitespace around key values _(field)_

## Output

- Anti-join result CSV (text)

## Query parameters

Open the tool pre-filled and auto-run via URL:

- `a` — CSV A (left table)
- `b` — CSV B (right table)
- `key` — Key column(s) in A
- `key_b` — Key column(s) in B (blank = same as A)
- `direction` — Rows to return
- `delimiter` — Delimiter
- `case_sensitive` — Case-sensitive key matching
- `trim_keys` — Trim whitespace around key values

Example: `https://gizza.ai/tools/csv-anti-join/?a=id%2Cname%0A1%2CAlice%0A2%2CBob%0A3%2CCarol&b=id%2Ccity%0A2%2CBerlin%0A3%2CCairo%0A4%2CDelhi&key=id&key_b=id&direction=a-only&delimiter=%2C&case_sensitive=true&trim_keys=true`

---

## About this tool

CSV anti-join is the spreadsheet version of "show me what is missing." Paste two CSV tables,
choose the key column, and the tool returns rows from CSV A whose key does **not** appear in CSV B.
That is useful for reconciling exports, finding customers not present in another system, locating
inventory codes missing from a reference list, or preparing a cleanup list before a migration.

Use **A only** for the standard left anti-join. Switch to **B only** for the reverse check, or
**Both sides** to emit a symmetric difference with a leading `_source` column that marks whether
each unmatched row came from A or B. Keys can be header names (`id`) or 1-based column indexes
(`1`), and comma-separated keys (`first,last`) form a composite match. If the key columns have
different names in the two CSVs, set the B key separately — matching is by value, not by header.

Duplicate keys in the lookup table do not multiply rows. If A contains two unmatched rows with the
same key, both A rows are kept, because they are real rows missing from B. You can choose comma,
tab, semicolon, pipe, or any single-character delimiter, and optionally make key matching
case-insensitive or trim whitespace around key values.

## Worked example

CSV A:

```csv
id,name
1,Alice
2,Bob
3,Carol
```

CSV B:

```csv
id,city
2,Berlin
3,Cairo
4,Delhi
```

Key: `id`, direction: `A only`.

Output:

```csv
id,name
1,Alice
```

## FAQ

<details>
<summary>What is an anti-join?</summary>

An anti-join returns rows from one table when their key does not exist in another table. It is the
opposite of an inner join: instead of keeping matches, it keeps non-matches. In SQL terms, this is
similar to `LEFT JOIN ... WHERE right.key IS NULL` or a `NOT EXISTS` query.

</details>

<details>
<summary>Can the key column have a different name in CSV B?</summary>

Yes. Put the A key in **Key column(s) in A** and the B key in **Key column(s) in B**. For example,
A might use `customer_id` while B uses `id`. The tool compares the key values, so the header names
do not need to match. Leave the B key blank when both files use the same key name or index.

</details>

<details>
<summary>How are duplicate keys handled?</summary>

The comparison side is treated as a key set, so duplicate keys in B do not multiply or otherwise
affect A's output. On the output side, every unmatched row is kept. If CSV A has two rows with key
`x` and `x` is absent from B, both rows appear in the result.

</details>

<details>
<summary>Can I compare on more than one column?</summary>

Yes. Use a comma-separated list such as `first,last` or `1,3`. The same number of key columns must
be supplied for CSV B. A row matches only when all key parts match after the case-sensitivity and
trim settings are applied.

</details>

<details>
<summary>Does the tool upload my CSV data?</summary>

No. The anti-join runs in WebAssembly inside your browser tab. Your pasted CSV text is processed
locally and is not sent to a server.

</details>

## Related tools

- [Avro to JSON Converter](https://gizza.ai/tools/avro-to-json/): Decode Apache Avro Object Container Files (.avro / OCF) to JSON, NDJSON, or the embedded schema — no .avsc needed, free and private in your browser.
- [Bencode decoder](https://gizza.ai/tools/bencode-decoder/): Decode bencode (the BitTorrent / .torrent serialization format) into readable JSON, and re-encode JSON back into canonical bencode — in your browser. Nothing is uploaded.
- [CAMT.053 Statement Parser](https://gizza.ai/tools/camt053-parse/): Parse a camt.053 (or camt.052/054) ISO 20022 bank statement XML into JSON or CSV in your browser: balances, dates, counterparties, references. No upload.
- [CSV Cell Diff](https://gizza.ai/tools/csv-cell-diff/): Compare two CSVs column-by-column and highlight every individual cell that changed, plus added and removed rows and columns.
- [Change CSV Delimiter](https://gizza.ai/tools/csv-change-delimiter/): Re-save CSV or delimited data with a different separator (comma, tab, semicolon, pipe) with correct requoting — in your browser. Free, private, no upload.