Survey Parser (C)

A high-performance C-based data parser that processes raw survey data and converts it into structured, analysis-ready datasets using Likert scale normalization.

Overview

This project was built to simulate real-world data ingestion challenges, where input data is often inconsistent, incomplete, or unstructured.

The Survey Parser reads raw survey files, validates and cleans the data, and transforms responses into a standardized format suitable for downstream analysis.

It emphasizes:

• Data integrity
• Efficient memory usage
• Robust error handling
• Structured data transformation

Key Features

• Custom File Parser
  Processes raw text-based survey files using efficient file I/O in C.

• Likert Scale Normalization
  Converts qualitative survey responses (e.g., "Strongly Agree") into consistent numerical representations.

• Data Validation & Cleaning
  Handles:

  • Missing values
  • Invalid inputs
  • Inconsistent formatting

• Error Handling
  Detects and safely handles malformed data without crashing.

• Memory-Efficient Design
  Uses manual memory management to ensure performance and reliability.

How It Works

1. Input
   Raw survey data is read from a text file.

2. Parsing
   The program tokenizes each line and extracts structured fields.

3. Validation
   Each response is checked for correctness and completeness.

4. Transformation
   Responses are mapped to a normalized Likert scale (e.g., 1–5).

5. Output
   Cleaned data is written to a structured format (e.g., CSV).

Installation & Usage

Compile: make

Run: ./test_scripts.sh

Skills Demonstrated

• Low-level programming in C
• File I/O and string parsing
• Data cleaning and normalization
• Memory management (malloc/free)
• Error handling and edge case coverage
• Data pipeline thinking (raw → processed → structured)

Real-World Relevance

This project mirrors real-world data workflows where:

• Raw inputs are inconsistent and require cleaning
• Data must be normalized before analysis
• Reliability and correctness are critical

These concepts directly apply to:

• Data analysis pipelines
• Risk analysis systems
• Business intelligence workflows

Future Improvements

• Add support for JSON/CSV input formats
• Implement multi-threaded parsing for large datasets
• Export results to database systems (SQLite/PostgreSQL)
• Add statistical summaries (mean, distribution, etc.)
• Build a simple visualization layer

Author

Samuel Echefu
GitHub: https://github.com/samnne
LinkedIn: https://linkedin.com/in/samuel-echefu-251a21336