File size: 4,196 Bytes
ad477a1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 |
---
license: mit
tags:
- nodejs
pretty_name: nodejs-all.json
---
# Node.js API Dataset
> **Source**: [Node.js](https://nodejs.org/) official documentation (JSON variant)
> **Processing Type**: Extractive, Hierarchical Flattening.
## Overview
This dataset contains a structured representation of the Node.js API, derived from the official `nodejs.json` distribution. Unlike raw documentation dumps, this dataset has been processed into two distinct formats to serve different machine learning and analytical purposes: **Macro-level (Documents)** and **Micro-level (Granular Items)**.
This "Dataset-as-a-Repo" approach ensures that the data is not just a transient output of a pipeline but a versioned, maintained artifact suitable for training high-quality code models.
## Methodology & Design Choices
### 1. The "Abstract-to-Concrete" Philosophy
The core design philosophy here is that "code intelligence" requires understanding both the *forest* (modules, high-level concepts) and the *trees* (individual function signatures, property types).
- **Raw Input**: The `nodejs.all.json` is a massive, nested structure that can be overwhelming for simple sequential models.
- **Transformation**: We "pulled apart" the JSON to create focused training examples.
### 2. Dual-Format Output
We intentionally avoided a "one-size-fits-all" schema.
- **Documents (`nodejs_documents.jsonl`)**: Preserves the cohesiveness of a module. Good for teaching a model "concept association" (e.g., that `fs.readFile` belongs with `fs.writeFile`).
- **Granular (`nodejs_granular.jsonl`)**: Isolates every single function and property. Good for "instruction tuning" (e.g., "Write a function signature for `http.createServer`").
### 3. File Formats
- **JSONL**: Chosen for its streaming capabilities and human readability. Perfect for NLP pipelines.
- **Parquet**: Chosen for the "Granular" dataset to allow fast columnar access, filtering, and analysis (e.g., "Find all methods with > 3 arguments").
## Dataset Structure
### Output Location
All processed files are located in `output/`:
```text
output/
├── nodejs_documents.jsonl # High-level module data
├── nodejs_granular.jsonl # Individual API items
└── nodejs_granular.parquet # Parquet version of granular data
```
### Schema: Documents (`nodejs_documents.jsonl`)
Representing a whole Module (e.g., `Buffer`, `http`).
| Field | Type | Description |
|-------|------|-------------|
| `module_name` | string | Name of the module (e.g., `fs`). |
| `type` | string | Usually `module` or `global`. |
| `description` | string | Raw HTML/Markdown description of the module. |
| `content` | json-string | Full nested JSON blob of the module's contents (methods, props). |
### Schema: Granular (`nodejs_granular.jsonl`)
Representing a single API item (Function, Property, Event).
| Field | Type | Description |
|-------|------|-------------|
| `id` | string | Unique namespaced ID (e.g., `fs.readFile`). |
| `parent` | string | Parent module (e.g., `fs`). |
| `type` | string | `method`, `property`, `event`, etc. |
| `name` | string | Short name (e.g., `readFile`). |
| `description` | string | Description of *just* this item. |
| `metadata` | json-string | Detailed signatures, params, stability indices. |
## Use Cases
### 1. Pre-Training Code Models
Feed `nodejs_documents.jsonl` into a language model to teach it the general structure and API surface of Node.js. The large context windows of modern LLMs can easily ingest entire modules.
### 2. Instruction Tuning / RAG
Use `nodejs_granular.jsonl` to build a Retrieval Augmented Generation (RAG) system.
- **Query**: "How do I read a file in Node?"
- **Retrieval**: Search against the `description` field in the granular dataset.
- **Context**: Retrieve the exact `metadata` (signature) for `fs.readFile`.
### 3. API Analysis
Use `nodejs_granular.parquet` with Pandas/DuckDB to answer meta-questions:
- *Which Node.js APIs are marked as Experimental?*
- *What is the average number of arguments for `fs` methods vs `http` methods?*
## Provenance
- **Original File**: `datasets/raw/nodejs.all.json`
- **Script**: `src/process_nodejs.py`
- **Maintainer**: Antigravity (Agent) / User |