Text parse unified patch

text-parse-unified-patch · version 1.0.0 · Text · free, no key needed

Parse a one-file unified patch into path names and 0-based hunks.

Use when you need to: text parse unified patch · parse unified diff · parse unified patch headers and hunks.

Supported

• text parse unified patch
• parse unified diff
• parse unified patch headers and hunks

Not supported

• preserve original CR/LF through unified text
• git diff --git headers
• index/binary markers
• multi-file patches
• offset/fuzz patch apply
• RFC 6902
• filesystem access

Behavior

• Empty unified yields from_name "a", to_name "b", hunks [], and identical true.
• Non-empty patches must start with --- and +++ headers. git diff --git, index, and binary markers are rejected.
• An omitted hunk count is parsed as 1. Output hunk starts are 0-based. Nonzero counts subtract one from the 1-based text start; a zero count keeps the GNU insertion index from the header.
• Body values are {text, eol} records. eol is "" iff a "\ No newline at end of file" line immediately followed that body line; otherwise eol is "\n".
• The unified document uses LF line breaks and does not round-trip original CR/LF from the source file.
• strip is an integer 0..8 (default 0) that removes that many leading slash-separated components from both paths; an empty path after strip is rejected.
• identical is true iff there are no hunks. Unknown input fields are rejected.

Input

• unified (string, required): max length 262144
• strip (integer, optional): min 0; max 8

Output

• from_name (string, required)
• to_name (string, required)
• hunks (array of object, required): max items 1024
• identical (boolean, required)

Limits

• max text bytes: 262144
• max path bytes: 1024
• max line bytes: 16384
• max hunks: 1024
• max strip: 8
• max output bytes: 1048576

Example

Request input:

{
  "unified": "--- a\n+++ b\n@@ -1,1 +1,1 @@\n-a\n+b\n"
}

Response:

{
  "result": {
    "from_name": "a",
    "to_name": "b",
    "hunks": [
      {
        "old_start": 0,
        "old_count": 1,
        "new_start": 0,
        "new_count": 1,
        "ops": [
          {
            "op": "delete",
            "value": {
              "text": "a",
              "eol": "\n"
            }
          },
          {
            "op": "insert",
            "value": {
              "text": "b",
              "eol": "\n"
            }
          }
        ]
      }
    ],
    "identical": false
  }
}

How to call it

MCP

Connect https://computefirst.net/mcp (setup), then call execute with:

{
  "id": "text-parse-unified-patch",
  "version": "1.0.0",
  "input": {
    "unified": "--- a\n+++ b\n@@ -1,1 +1,1 @@\n-a\n+b\n"
  }
}

HTTP (no key)

curl -X POST https://computefirst.net/v1/tools/text-parse-unified-patch/versions/1.0.0/execute \
  -H "Content-Type: application/json" \
  -d '{"unified":"--- a\n+++ b\n@@ -1,1 +1,1 @@\n-a\n+b\n"}'

The machine-readable contract is at /v1/tools/text-parse-unified-patch/versions/1.0.0.

CLI

node cli.mjs run text-parse-unified-patch 1.0.0 --input input.json --base-url https://computefirst.net

Get the client at /clients/cli/.

Related tools

• Text invert unified patch: Invert a one-file unified patch by swapping file names and delete/insert ops.
• Text apply unified patch: Apply a one-file unified patch to text at original line addresses with exact hunk matches.
• Text diff stats: Summarize a line-record diff as keep, insert, delete, and hunk counts.
• Text unified diff: Emit a one-file unified diff of two texts with comma counts and LF document breaks.
• Text conflict parse: Parse merge or diff3 conflict-marker text into ordered text and conflict parts.
• Text extract between: Extract exact substrings between start and end markers, left to right and non-overlapping.