Functions & Generator Reference
OpenMimic provides a comprehensive suite of declarative function generators. Each field in mimic.json specifies one of the following generator types.
1. uuid
Generates standard RFC 4122 v4 UUIDs deterministically based on the generator seed.
Schema
{
"uuid": {
"uppercase": false
}
}
Options
uppercase(boolean, default:false): Iftrue, returns the UUID in uppercase (e.g.3F982A7E-...).
2. distribution
Outputs mathematical distributions (Gaussian / Normal or Uniform).
Normal (Gaussian) Distribution
Generates values around a mean with standard deviation:
{
"distribution": {
"normal": {
"mean": 100.0,
"std_dev": 15.0
}
}
}
Uniform Distribution
Generates values evenly spread between min and max:
{
"distribution": {
"uniform": {
"min": 10.0,
"max": 50.0
}
}
}
3. now
Generates deterministic timestamps that advance with the generator logical clock.
Schema
{
"now": {
"format": "RFC3339"
}
}
Options
format(string):"RFC3339":2026-06-15T10:00:00Z"epoch_ms": Unix timestamp in milliseconds (1750000000000)"epoch_s": Unix timestamp in seconds (1750000000)- Custom Go layout string (e.g.
"2006-01-02 15:04:05").
4. sequential_integer
Generates monotonically incrementing or stepping integers.
Schema
{
"sequential_integer": {
"start": 1,
"step": 1
}
}
Output Formula
\(\text{value} = \text{start} + (\text{event\_index} \times \text{step})\)
5. string_faker
Synthesizes context-aware strings and pattern-based identifiers.
Semantic Categories
{
"string_faker": {
"category": "email"
}
}
Available categories:
"name"/"full_name": Generates realistic human names (e.g."Alice Smith")."first_name","last_name"."email": Generates formatted email addresses."city": Major world cities."country": Global country names.
Custom Mask / Pattern
You can specify a pattern using # for random digits:
{
"string_faker": {
"pattern": "ORDER_####"
}
}
Output: ORDER_4821, ORDER_0945.
6. weighted_choice
Simulates probabilistic distributions, failure rates, or categorical states.
Schema
{
"weighted_choice": {
"options": [
{ "value": "SUCCESS", "weight": 95.0 },
{ "value": "FAILURE", "weight": 5.0 }
]
}
}
In this example, "FAILURE" is produced with an exact 5% probability across the event stream.
7. constant
Outputs a fixed literal value.
{
"constant": {
"string_val": "ACTIVE"
}
}
Supports string_val, int_val, double_val, and bool_val.
8. binary_media (Synthetic Audio & Images)
Generates valid binary media containers directly into streaming payloads (useful for multimodal pipelines, computer vision benchmarks, and media processing ingestion).
Schema
{
"binary_media": {
"format": "wav",
"sample_rate": 44100,
"duration_seconds": 1.0,
"frequency_hz": 440.0
}
}
Supported Media Formats:
wav: Generates RFC PCM WAV audio streams containing a pure sinusoidal tone (e.g. 440 Hz standard A4) with validRIFFandWAVEfmtheaders.png: Generates valid PNG images with official magic header0x89 0x50 0x4E 0x47 0x0D 0x0A 0x1A 0x0A,IHDR, andIENDchunks.jpeg: Generates valid JPEG images with SOI marker0xFF 0xD8, JFIF application header, and EOI marker0xFF 0xD9.
9. aggregation (Streaming CEP Window Rollups)
Calculates streaming rollups (sum, avg, min, max, count) over sliding or tumbling event windows from correlated upstream streams.
Schema
{
"aggregation": {
"source_stream_id": "stream-orders",
"window_type": "sliding",
"window_size_events": 50,
"aggregation_op": "avg",
"target_field": "order_amount"
}
}
10. markov_chain & entity_pool
entity_pool: Selects foreign keys from a pre-allocated bounded pool of entities (e.g.cust_00001throughcust_00500) to maintain relational referential integrity across topics.markov_chain: Simulates realistic state progressions (e.g.CART -> CHECKOUT -> PAYMENT_PENDING -> ORDER_COMPLETED) according to transition probability matrices.
11. video_stream (Real-Time Synthetic Camera Broadcaster)
Simulates continuous real-time camera video streams (CCTV, drone, traffic cams, and optical sensors) encoded with Motion JPEG (MJPEG) frames.
Schema
{
"video_stream": {
"codec": "mjpeg",
"width": 640,
"height": 480,
"fps": 30,
"pattern": "bouncing_ball",
"render_timestamp": true
}
}
Motion Pattern Options:
bouncing_ball: Smooth sinusoidal harmonic target motion ideal for testing computer vision object detection and tracking algorithms.color_bars: Standard SMPTE-style 7-color optical calibration bars.noise: Realistic high-frequency optical sensor static noise.