TextWatermark Service

An HTTP Service for using the TextWatermark module to watermark text.

Based on Fastapi.

Installation

Modify Dockerfile and docker-compose.yml then run docker-compose -f docker-compose.yml up

Usage

There are 2 steps for using textwatermark-service

1. Create a worker when you want to watermark some text.
2. Do the watermarking job with the worker you created.

Use Swagger UI for testing

Visit http://0.0.0.0:8000/docs and you will see swagger UI.

Use Curl for testing

Create worker

The params JSON string is the same as what you exported by using the TextWatermark package in the console. see https://textwatermark.jd.army/cmdline/#export-parameters-example

curl -X 'POST' \
  'http://0.0.0.0:8000/worker/create?authorize_key=d0f4ac8c14eae3a992aa574a55099e4f' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "params": {
    "tpl_type": "INVISIBLE_CHARS",
    "confusables_chars": [],
    "confusables_chars_key": "",
    "wm_base": 29,
    "method": "INSERT_INTO_POSITION",
    "wm_mode": "REAL_NUMBER",
    "wm_len": 5,
    "wm_flag_bit": true,
    "wm_loop": false,
    "wm_max": "99999",
    "start_at": 1,
    "version": "0.3.2"
  },
  "use_job_id": true,
  "text": "12345678901"
}'

Result:

{
  "worker_id": 65542
}

Do job

curl -X 'POST' \
  'http://0.0.0.0:8000/worker/1/do_job?authorize_key=d0f4ac8c14eae3a992aa574a55099e4f' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/x-www-form-urlencoded' \
  -d 'wm_str='

Result:

{
  "use_job_id": false,
  "job_id": 66433,
  "wm_str": "888888",
  "wm_text": "Ӏ2𝟑𝟺Ƽ𝟔𝟩890123456789012345678901234567890"
}

Get worker record

curl -X 'GET' \
  'http://0.0.0.0:8000/worker/10000?authorize_key=d0f4ac8c14eae3a992aa574a55099e4f' \
  -H 'accept: application/json'

Result:

{
  "use_job_id": false,
  "created": "2023-03-11T15:16:02.791571",
  "text": "1234567890123456789012345678901234567890",
  "params": {
    "tpl_type": "HOMOGRAPH_NUMBERS",
    "confusables_chars": [],
    "confusables_chars_key": "",
    "wm_base": 7,
    "method": 1,
    "wm_mode": 1,
    "wm_len": 8,
    "wm_flag_bit": false,
    "wm_loop": false,
    "wm_max": "999999",
    "start_at": 0,
    "version": "0.2.1"
  },
  "id": 65542
}

Redo job

curl -X 'POST' \
  'http://0.0.0.0:8000/worker/1/job/1/redo?authorize_key=d0f4ac8c14eae3a992aa574a55099e4f' \
  -H 'accept: application/json' \
  -d ''

Result:

{
  "use_job_id": false,
  "job_id": 4904,
  "wm_str": "888888",
  "wm_text": "Ӏ2𝟑𝟺Ƽ𝟔𝟩890123456789012345678901234567890"
}

Get job record

curl -X 'GET' \
  'http://0.0.0.0:8000/worker/1/job/1?authorize_key=d0f4ac8c14eae3a992aa574a55099e4f' \
  -H 'accept: application/json'

Result:

{
  "created": "2023-03-11T15:18:17.792940",
  "id": 4904,
  "worker_id": 65542,
  "wm_str": "888888",
  "worker": {
    "use_job_id": false,
    "created": "2023-03-11T15:16:02.791571",
    "text": "1234567890123456789012345678901234567890",
    "params": {
      "tpl_type": "HOMOGRAPH_NUMBERS",
      "confusables_chars": [],
      "confusables_chars_key": "",
      "wm_base": 7,
      "method": 1,
      "wm_mode": 1,
      "wm_len": 8,
      "wm_flag_bit": false,
      "wm_loop": false,
      "wm_max": "999999",
      "start_at": 0,
      "version": "0.2.1"
    },
    "id": 65542
  }
}

Developing and Testing

Install dependencies

code .

virtualenv venv
. venv/bin/activate

pip install poetry
poetry lock
poetry install

tox

Run uvicorn

Run uvicorn textwatermark_service.main:app --port 8000 --reload

Use wrk for benchmarks

see wrk.md in Benchmarks Directory.