Range Indexed Data Access
A big file. One small read.
Website · Python / PyPI · Node.js / npm · Download for Mac
Turn JSON records into a static dataset. Its index finds the stored value your application needs, and the reader retrieves that value from disk or through an ordinary HTTP byte-range request. A field can be a sentence, a complete translated help page, a product description, or a nested JSON object.
| Edition | Install or download | Published version | Requirements |
|---|---|---|---|
| Python library and CLI | pip install rida-format |
2.0.2 | Python 3.10+; encryption included |
| Node.js library and CLI | npm install rida-format |
2.0.1 | Node.js 20+; npm / npx |
| RIDA Studio | RIDA.dmg | 2.0.2 | macOS on Apple Silicon; see requirements |
The package name is rida-format on both registries. Python imports rida; Node.js imports rida-format. npx runs the Node command; it is not a separate format or product. Edition versions are independent.
pip install rida-format
rida --version
rida scan examples/records.json
rida build examples/records.json ExampleData.rida --yes
rida get ExampleData.rida backpack-2812 hindiimport rida
with rida.RidaFile("ExampleData.rida") as data:
print(data.get("backpack-2812", "hindi"))
print(data.get_many("backpack-2812", ["price", "hindi"]))npm install rida-format
npx rida-format --help
npx rida-format scan examples/records.json
npx rida-format build examples/records.json --output ExampleData.rida
npx rida-format get ExampleData.rida backpack-2812 hindiimport { RidaReader } from "rida-format";
const reader = await RidaReader.open("./ExampleData.rida");
try {
console.log(await reader.get("backpack-2812", "hindi"));
console.log(await reader.getMany("backpack-2812", ["price", "hindi"]));
} finally {
await reader.close();
}Run the Python and Node build examples in separate directories, or choose different output names; builders protect existing output files. Download this repository to use its examples/ files.
Open RIDA.dmg, drag RIDA Studio into Applications, then launch it. Choose a JSON file or folder, scan it, configure the output, and build. Auto setup scans and suggests; applying a suggestion does not start a build.
flowchart LR
J[JSON records] --> B[RIDA builder]
B --> D[Data file on disk or CDN]
B --> I[Index and map with reader]
A[Application: record ID + field] --> I
I --> R[Read the stored byte range]
D --> R
R --> V[Decode the requested value]
The reader uses the catalogue to locate a value. HTTP delivery requires 206 Partial Content; a field read does not download the complete dataset. Opening with a bundled/local index avoids an index download. The Node reader can also fetch an index; Python uses its local index/cache workflow. How it works →
| File | Job | Keep it |
|---|---|---|
ExampleData.rida |
Encoded field values | Disk, object storage or CDN |
ExampleIndex.rida |
Catalogue and metadata locating values | With the reader, or an explicitly supported index location |
ExampleMap.json |
Folder labels and field mappings for numbered datasets | With the reader when name resolution is needed |
ExampleKey.rida |
Key generated for an encrypted build | Private; never beside public download files |
Maps and generated key files depend on the build options. RIDA Studio writes a map for its builds. The catalogue lives inside the index; it is not a fifth file. Format and mapping →
- Commerce: retrieve a product's complete Hindi description, then request price and description together when the screen needs both.
- Help centres: organise nested folders of pages, resolve their mapped path, and read the entire Spanish page with its steps and troubleshooting.
- Directories and reference data: deliver known records and fields without starting a database query for each lookup.
- Archived sensor records: use deep folder paths for locations, devices and dates, then read one stored measurement field.
A field can be large. Reading a complete object returns the whole object, including nested content. RIDA selects top-level fields; it does not extract arbitrary nested properties. Detailed examples →
| Guide | What you will find |
|---|---|
| Python manual | Complete installation, CLI, scan/build, record IDs, local/HTTP readers, cache, encryption, API and examples |
| Node.js / npm / npx manual | Complete npm installation, CLI, builder/reader APIs, JSON layouts, nested folders, encryption and examples |
| Mac app | Download, compatibility, installation, Auto popup, separate/shared/all groups and output files |
| How it works | Byte ranges, catalogue, map, numbering, codecs, compression and integrity |
| Build guide | Input shapes, mixed schemas, group strategies, safe builds and large inputs |
| Read guide | Local reads, HTTP, get-many, whole records, map resolution, caching and verification |
| CDN hosting | Bunny, Cloudflare, B2, your own server, HTTP requirements and cache considerations |
| Security | AES-256-GCM, private keys, direct CDN reads, server-only Next.js keys and access-control limits |
| Benchmarks and costs | Reference measurements, modeled provider costs, SQL comparison and limitations |
| API reference | Python and Node method names and CLI differences |
| Hosting-cost examples | Four provider arrangements, calculator assumptions and former-plan comparisons |
| Purpose | The foundation, former paid service and public availability |
| Binary format | Complete byte-level format specification |
| Troubleshooting | Common installation, input, output, HTTP, key and compatibility problems |
| Website pages | Links to every public section, terms, privacy and release information |
| Release history | Edition versions and the Mac release |
Our reference measurements use 27,647 records and 34 fields: approximately 7.69 KB per complete JSON record versus 143 bytes per average stored RIDA field. Reading one average field instead of the complete record models about 98% fewer payload bytes. HTTP headers, encryption and provider accounting add overhead. This is not a promise that every field is 143 bytes or that every bill drops by 98%. Full measurements and assumptions →
RIDA fits collections built ahead of time and read repeatedly: catalogues, translations, documentation, directories and reference archives. Use a database for live writes, transactions, relational joins and dynamic queries. A project can use a database for changing data and RIDA for its static collection.
RIDA is free to use under the RIDA Proprietary License; free availability does not mean an unrestricted open-source license. Use in free or commercial applications is permitted subject to the license and attribution. Standalone redistribution and resale are restricted. See the Python license, Node license, and the license included with your download.
This repository contains public documentation and examples. Install the libraries from PyPI/npm. The Mac installer is provided as a GitHub Release asset rather than a large file in Git history.
Website: https://rida.qasra.org · Publisher: Qasra Foundation · Copyright: Syed Ali Hamza Chishty