# CSV row diff

`csv-row-diff` · version 1.0.0 · CSV & tables · free, no key needed

Compare two CSV documents with identical headers by unique key tuples and report added, removed, and changed rows.

**Use when you need to: csv row diff · diff csv rows by keys · compare keyed csv rows.**

## Supported

- csv row diff
- diff csv rows by keys
- compare keyed csv rows

## Not supported

- outer join
- fuzzy matching
- numeric comparison
- case-insensitive diff
- unkeyed row diff
- read files
- fetch urls

## Behavior

- left and right must share identical headers in the same order.
- keys is a non-empty unique subset of those headers, at most 32 names.
- Duplicate key tuples on either side are rejected because the diff is undefined.
- Rows are classified by exact key-tuple membership and exact remaining-column string tuples.
- added contains right-only keys in right row order; removed and changed follow first-seen left key order.
- changed rows are the right-side (new) values with the original headers; no change column is added.
- No trimming, case folding, type coercion, or numeric comparison is performed.
- Empty documents without headers are rejected; header-only inputs emit the shared header and zero counts.
- The first header name may not begin with U+FEFF because it is ambiguous with a CSV byte-order mark.

## Input

- `left` (string, required): max length 256000
- `right` (string, required): max length 256000
- `keys` (array of string, required): min items 1; max items 32; each min length 1

## Output

- `added` (string, required)
- `removed` (string, required)
- `changed` (string, required)
- `counts` (object, required)

## Limits

- max input bytes: 256000
- max data rows: 5000
- max columns: 256
- max header name bytes: 256
- max keys: 32
- max output bytes: 1048576

## Example

Request input:

```json
{
  "left": "id,name,city\n1,Ada,London\n2,Lin,Paris\n3,Zoe,Berlin\n",
  "right": "id,name,city\n1,Ada,London\n2,Lin,Rome\n4,Bob,Oslo\n",
  "keys": [
    "id"
  ]
}
```

Response:

```json
{
  "result": {
    "added": "id,name,city\n4,Bob,Oslo\n",
    "removed": "id,name,city\n3,Zoe,Berlin\n",
    "changed": "id,name,city\n2,Lin,Rome\n",
    "counts": {
      "added": 1,
      "removed": 1,
      "changed": 1,
      "unchanged": 1
    }
  }
}
```

## How to call it

### MCP

Connect `https://computefirst.net/mcp` ([setup](/docs#connect)), then call `execute` with:

```json
{
  "id": "csv-row-diff",
  "version": "1.0.0",
  "input": {
    "left": "id,name,city\n1,Ada,London\n2,Lin,Paris\n3,Zoe,Berlin\n",
    "right": "id,name,city\n1,Ada,London\n2,Lin,Rome\n4,Bob,Oslo\n",
    "keys": [
      "id"
    ]
  }
}
```

### HTTP (no key)

```sh
curl -X POST https://computefirst.net/v1/tools/csv-row-diff/versions/1.0.0/execute \
  -H "Content-Type: application/json" \
  -d '{"left":"id,name,city\n1,Ada,London\n2,Lin,Paris\n3,Zoe,Berlin\n","right":"id,name,city\n1,Ada,London\n2,Lin,Rome\n4,Bob,Oslo\n","keys":["id"]}'
```

The machine-readable contract is at [/v1/tools/csv-row-diff/versions/1.0.0](/v1/tools/csv-row-diff/versions/1.0.0).

### CLI

```sh
node cli.mjs run csv-row-diff 1.0.0 --input input.json --base-url https://computefirst.net
```

Get the client at [/clients/cli/](/clients/cli/).

## Related tools

- [CSV reconcile](/tools/csv-reconcile): Reconcile two CSV documents by key columns, reporting full population totals and bounded samples for added, removed, changed, invalid-key, and ambiguous rows.
- [CSV add row number](/tools/csv-add-row-number): Prepend a 1-based row_number column of canonical integer strings.
- [CSV anti join](/tools/csv-anti-join): Keep left CSV rows whose exact key tuples do not appear in the right document.
- [CSV dedupe](/tools/csv-dedupe): Remove duplicate CSV data rows while preserving the first occurrence.
- [CSV group count](/tools/csv-group-count): Count CSV rows by exact named-column tuples in first-seen order.
- [CSV slice rows](/tools/csv-slice-rows): Take a contiguous window of CSV data rows by offset and limit, keeping the header.