Update README

Reviewed-on: #1

.gitmodules

@@ -1,3 +0,0 @@
-[submodule "tests/lib/catch2"]
-	path = tests/lib/catch2
-	url = https://github.com/catchorg/Catch2.git

ArffFiles.hpp

@@ -7,10 +7,10 @@
 #include <sstream>
 #include <fstream>
 #include <cctype> // std::isdigit
-#include <algorithm> // std::all_of
+#include <algorithm> // std::all_of std::transform
 
 class ArffFiles {
-    const std::string VERSION = "1.0.0";
+    const std::string VERSION = "1.1.0";
 public:
     ArffFiles() = default;
     void load(const std::string& fileName, bool classLast = true)
@@ -28,8 +28,9 @@ public:
             attributes.erase(attributes.begin());
             labelIndex = 0;
         }
+        preprocessDataset(labelIndex);
         generateDataset(labelIndex);
-    };
+    }
     void load(const std::string& fileName, const std::string& name)
     {
         int labelIndex;
@@ -48,27 +49,67 @@ public:
         if (!found) {
             throw std::invalid_argument("Class name not found");
         }
+        preprocessDataset(labelIndex);
         generateDataset(labelIndex);
-    };
-    std::vector<std::string> getLines() const { return lines; };
-    unsigned long int getSize() const { return lines.size(); };
-    std::string getClassName() const { return className; };
-    std::string getClassType() const { return classType; };
-    std::vector<std::string> getLabels() const { return labels; }
+    }
+    std::vector<std::string> getLines() const { return lines; }
+    unsigned long int getSize() const { return lines.size(); }
+    std::string getClassName() const { return className; }
+    std::string getClassType() const { return classType; }
+    std::map<std::string, std::vector<std::string>> getStates() const { return states; }
+    std::vector<std::string> getLabels() const { return states.at(className); }
     static std::string trim(const std::string& source)
     {
         std::string s(source);
         s.erase(0, s.find_first_not_of(" '\n\r\t"));
         s.erase(s.find_last_not_of(" '\n\r\t") + 1);
         return s;
-    };
-    std::vector<std::vector<float>>& getX() { return X; };
+    }
+    std::vector<std::vector<float>>& getX() { return X; }
     std::vector<int>& getY() { return y; }
+    std::map<std::string, bool> getNumericAttributes() const { return numeric_features; }
     std::vector<std::pair<std::string, std::string>> getAttributes() const { return attributes; };
-    std::vector<int> factorize(const std::vector<std::string>& labels_t)
+    std::vector<std::string> split(const std::string& text, char delimiter)
+    {
+        std::vector<std::string> result;
+        std::stringstream ss(text);
+        std::string token;
+        while (std::getline(ss, token, delimiter)) {
+            result.push_back(trim(token));
+        }
+        return result;
+    }
+    std::string version() const { return VERSION; }
+protected:
+    std::vector<std::string> lines;
+    std::map<std::string, bool> numeric_features;
+    std::vector<std::pair<std::string, std::string>> attributes;
+    std::string className;
+    std::string classType;
+    std::vector<std::vector<float>> X;
+    std::vector<std::vector<std::string>> Xs;
+    std::vector<int> y;
+    std::map<std::string, std::vector<std::string>> states;
+private:
+    void preprocessDataset(int labelIndex)
+    {
+        //
+        // Learn the numeric features
+        //
+        numeric_features.clear();
+        for (const auto& attribute : attributes) {
+            auto feature = attribute.first;
+            if (feature == className)
+                continue;
+            auto values = attribute.second;
+            std::transform(values.begin(), values.end(), values.begin(), ::toupper);
+            numeric_features[feature] = values == "REAL" || values == "INTEGER" || values == "NUMERIC";
+        }
+    }
+    std::vector<int> factorize(const std::string feature, const std::vector<std::string>& labels_t)
     {
         std::vector<int> yy;
-        labels.clear();
+        states.at(feature).clear();
         yy.reserve(labels_t.size());
         std::map<std::string, int> labelMap;
         int i = 0;
@@ -77,54 +118,46 @@ public:
                 labelMap[label] = i++;
                 bool allDigits = std::all_of(label.begin(), label.end(), ::isdigit);
                 if (allDigits)
-                    labels.push_back("Class " + label);
+                    states[feature].push_back("Class " + label);
                 else
-                    labels.push_back(label);
+                    states[feature].push_back(label);
             }
             yy.push_back(labelMap[label]);
         }
         return yy;
-    };
-    std::string version() const { return VERSION; };
-protected:
-    std::vector<std::string> lines;
-    std::vector<std::pair<std::string, std::string>> attributes;
-    std::string className;
-    std::string classType;
-    std::vector<std::vector<float>> X;
-    std::vector<int> y;
-    std::vector<std::string> labels;
-private:
+    }
     void generateDataset(int labelIndex)
     {
         X = std::vector<std::vector<float>>(attributes.size(), std::vector<float>(lines.size()));
+        Xs = std::vector<std::vector<std::string>>(attributes.size(), std::vector<std::string>(lines.size()));
         auto yy = std::vector<std::string>(lines.size(), "");
-        auto removeLines = std::vector<int>(); // Lines with missing values
         for (size_t i = 0; i < lines.size(); i++) {
             std::stringstream ss(lines[i]);
             std::string value;
             int pos = 0;
             int xIndex = 0;
-            while (getline(ss, value, ',')) {
+            auto tokens = split(lines[i], ',');
+            for (const auto& token : tokens) {
                 if (pos++ == labelIndex) {
-                    yy[i] = value;
+                    yy[i] = token;
                 } else {
-                    if (value == "?") {
-                        X[xIndex++][i] = -1;
-                        removeLines.push_back(i);
-                    } else
-                        X[xIndex++][i] = stof(value);
+                    if (numeric_features[attributes[xIndex].first]) {
+                        X[xIndex][i] = stof(token);
+                    } else {
+                        Xs[xIndex][i] = token;
+                    }
+                    xIndex++;
                 }
             }
         }
-        for (auto i : removeLines) {
-            yy.erase(yy.begin() + i);
-            for (auto& x : X) {
-                x.erase(x.begin() + i);
+        for (size_t i = 0; i < attributes.size(); i++) {
+            if (!numeric_features[attributes[i].first]) {
+                auto data = factorize(attributes[i].first, Xs[i]);
+                std::transform(data.begin(), data.end(), X[i].begin(), [](int x) { return float(x);});
             }
         }
-        y = factorize(yy);
-    };
+        y = factorize(className, yy);
+    }
     void loadCommon(std::string fileName)
     {
         std::ifstream file(fileName);
@@ -152,12 +185,19 @@ private:
             if (line[0] == '@') {
                 continue;
             }
+            if (line.find("?", 0) != std::string::npos) {
+                // ignore lines with missing values
+                continue;
+            }
             lines.push_back(line);
         }
         file.close();
+        for (const auto& attribute : attributes) {
+            states[attribute.first] = std::vector<std::string>();
+        }
         if (attributes.empty())
             throw std::invalid_argument("No attributes found");
-    };
+    }
 };
 
-#endif
+#endif

CHANGELOG.md

@@ -5,6 +5,34 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [1.2.0] 2025-06-27 Refactoring and Improvements
+
+### Added
+
+- Refactored code to improve readability and maintainability
+- Improved error handling with exceptions
+- Claude TECHNICAL_REPORT.md for detailed analysis
+- Claude CLAUDE.md for AI engine usage
+- Actions to build and upload the conan package to Cimmeria
+
+
+
+## [1.1.0] 2024-07-24 String Values in Features
+
+### Added
+
+- Allow string values in features
+- Library logo
+
+### Fixed
+
+- Fixed bug in numeric attributes states
+
+### Removed
+
+- Catch2 git submodule
+- iostream include
+
 ## [1.0.0] 2024-05-21 Initial Release
 
 ### Added

CLAUDE.md

@@ -0,0 +1,83 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+ArffFiles is a header-only C++ library for reading ARFF (Attribute-Relation File Format) files and converting them into STL vectors. The library handles both numeric and categorical features, automatically factorizing categorical attributes.
+
+## Build System
+
+This project uses CMake with Conan for package management:
+- **CMake**: Primary build system (requires CMake 3.20+)
+- **Conan**: Package management for dependencies
+- **Makefile**: Convenience wrapper for common tasks
+
+## Common Development Commands
+
+### Building and Testing
+```bash
+# Build and run tests (recommended)
+make build && make test
+
+# Alternative manual build process
+mkdir build_debug
+cmake -S . -B build_debug -D CMAKE_BUILD_TYPE=Debug -D ENABLE_TESTING=ON -D CODE_COVERAGE=ON
+cmake --build build_debug -t unit_tests_arffFiles -j 16
+cd build_debug/tests && ./unit_tests_arffFiles
+```
+
+### Testing Options
+```bash
+# Run tests with verbose output
+make test opt="-s"
+
+# Clean test artifacts
+make clean
+```
+
+### Code Coverage
+Code coverage is enabled when building with `-D CODE_COVERAGE=ON` and `-D ENABLE_TESTING=ON`. Coverage reports are generated during test runs.
+
+## Architecture
+
+### Core Components
+
+**Single Header Library**: `ArffFiles.hpp` contains the complete implementation.
+
+**Main Class**: `ArffFiles`
+- Header-only design for easy integration
+- Handles ARFF file parsing and data conversion
+- Automatically determines numeric vs categorical features
+- Supports flexible class attribute positioning
+
+### Key Methods
+- `load(fileName, classLast=true)`: Load with class attribute at end/beginning
+- `load(fileName, className)`: Load with specific named class attribute
+- `getX()`: Returns feature vectors as `std::vector<std::vector<float>>`
+- `getY()`: Returns labels as `std::vector<int>`
+- `getNumericAttributes()`: Returns feature type mapping
+
+### Data Processing Pipeline
+1. **File Parsing**: Reads ARFF format, extracts attributes and data
+2. **Feature Detection**: Automatically identifies numeric vs categorical attributes
+3. **Preprocessing**: Handles missing values (lines with '?' are skipped)
+4. **Factorization**: Converts categorical features to numeric codes
+5. **Dataset Generation**: Creates final X (features) and y (labels) vectors
+
+### Dependencies
+- **Catch2**: Testing framework (fetched via CMake FetchContent)
+- **Standard Library**: Uses STL containers (vector, map, string)
+- **C++17**: Minimum required standard
+
+### Test Structure
+- Tests located in `tests/` directory
+- Sample ARFF files in `tests/data/`
+- Single test executable: `unit_tests_arffFiles`
+- Uses Catch2 v3.3.2 for test framework
+
+### Conan Integration
+The project includes a `conanfile.py` that:
+- Automatically extracts version from CMakeLists.txt
+- Packages as a header-only library
+- Exports only the main header file

CMakeLists.txt

@@ -41,7 +41,12 @@ add_subdirectory(config)
 # -------
 if (ENABLE_TESTING)
   MESSAGE("Testing enabled")
-  add_git_submodule("tests/lib/catch2")  
+  Include(FetchContent)
+    FetchContent_Declare(Catch2
+        GIT_REPOSITORY https://github.com/catchorg/Catch2.git
+        GIT_TAG        v3.3.2
+    )
+    FetchContent_MakeAvailable(Catch2)
   include(CTest)
   add_subdirectory(tests)
 endif (ENABLE_TESTING)

Makefile

@@ -1,6 +1,6 @@
 SHELL := /bin/bash
 .DEFAULT_GOAL := help
-.PHONY: help build test clean
+.PHONY: help build test clean conan-build conan-upload
 
 f_debug = build_debug
 test_targets = unit_tests_arffFiles
@@ -44,6 +44,16 @@ test: ## Run tests (opt="-s") to verbose output the tests
 	done
 	@echo ">>> Done";
 
+conan-build: ## Build Conan package locally
+	@echo ">>> Building Conan package...";
+	@conan create . --profile default
+	@echo ">>> Done";
+
+conan-upload: ## Upload package to Cimmeria JFrog Artifactory
+	@echo ">>> Uploading to Cimmeria JFrog Artifactory...";
+	@conan upload arff-files --all -r Cimmeria --confirm
+	@echo ">>> Done";
+
 help: ## Show help message
 	@IFS=$$'\n' ; \
 	help_lines=(`fgrep -h "##" $(MAKEFILE_LIST) | fgrep -v fgrep | sed -e 's/\\$$//' | sed -e 's/##/:/'`); \

README.md

@@ -1,14 +1,211 @@
-# ArffFiles
+# <img src="logo.png" alt="logo" width="50"/> ArffFiles
 
 ![C++](https://img.shields.io/badge/c++-%2300599C.svg?style=flat&logo=c%2B%2B&logoColor=white)
 [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](<https://opensource.org/licenses/MIT>)
-![Gitea Release](https://img.shields.io/gitea/v/release/rmontanana/arfffiles?gitea_url=https://gitea.rmontanana.es:3000)
-![Gitea Last Commit](https://img.shields.io/gitea/last-commit/rmontanana/arfffiles?gitea_url=https://gitea.rmontanana.es:3000&logo=gitea)
+![Gitea Release](https://img.shields.io/gitea/v/release/rmontanana/arfffiles?gitea_url=https://gitea.rmontanana.es)
+![Gitea Last Commit](https://img.shields.io/gitea/last-commit/rmontanana/arfffiles?gitea_url=https://gitea.rmontanana.es&logo=gitea)
 
-Header-only library to read Arff Files and return STL vectors with the data read.
+A modern C++17 header-only library to read **ARFF (Attribute-Relation File Format)** files and convert them into STL vectors for machine learning and data analysis applications.
 
-### Tests
+## Features
+
+- 🔧 **Header-only**: Simply include `ArffFiles.hpp` - no compilation required
+- 🚀 **Modern C++17**: Clean, efficient implementation using modern C++ standards
+- 🔄 **Automatic Type Detection**: Distinguishes between numeric and categorical attributes
+- 📊 **Flexible Class Positioning**: Support for class attributes at any position
+- 🎯 **STL Integration**: Returns standard `std::vector` containers for seamless integration
+- 🧹 **Data Cleaning**: Automatically handles missing values (lines with '?' are skipped)
+- 🏷️ **Label Encoding**: Automatic factorization of categorical features into numeric codes
+
+## Requirements
+
+- **C++17** compatible compiler
+- **Standard Library**: Uses STL containers (no external dependencies)
+
+## Installation
+
+### Using Conan
 
 ```bash
-make build && make test
+# Add the package to your conanfile.txt
+[requires]
+arff-files/1.0.1
+
+# Or install directly
+conan install arff-files/1.0.1@
 ```
+
+### Manual Installation
+
+Simply download `ArffFiles.hpp` and include it in your project:
+
+```cpp
+#include "ArffFiles.hpp"
+```
+
+## Quick Start
+
+```cpp
+#include "ArffFiles.hpp"
+#include <iostream>
+
+int main() {
+    ArffFiles arff;
+    
+    // Load ARFF file (class attribute at the end by default)
+    arff.load("dataset.arff");
+    
+    // Get feature matrix and labels
+    auto& X = arff.getX();  // std::vector<std::vector<float>>
+    auto& y = arff.getY();  // std::vector<int>
+    
+    std::cout << "Dataset size: " << arff.getSize() << " samples" << std::endl;
+    std::cout << "Features: " << X.size() << std::endl;
+    std::cout << "Classes: " << arff.getLabels().size() << std::endl;
+    
+    return 0;
+}
+```
+
+## API Reference
+
+### Loading Data
+
+```cpp
+// Load with class attribute at the end (default)
+arff.load("dataset.arff");
+
+// Load with class attribute at the beginning
+arff.load("dataset.arff", false);
+
+// Load with specific named class attribute
+arff.load("dataset.arff", "class_name");
+```
+
+### Accessing Data
+
+```cpp
+// Get feature matrix (each inner vector is a feature, not a sample)
+std::vector<std::vector<float>>& X = arff.getX();
+
+// Get labels (encoded as integers)
+std::vector<int>& y = arff.getY();
+
+// Get dataset information
+std::string className = arff.getClassName();
+std::vector<std::string> labels = arff.getLabels();
+unsigned long size = arff.getSize();
+
+// Get attribute information
+auto attributes = arff.getAttributes();  // std::vector<std::pair<std::string, std::string>>
+auto numericFeatures = arff.getNumericAttributes();  // std::map<std::string, bool>
+```
+
+### Utility Methods
+
+```cpp
+// Get library version
+std::string version = arff.version();
+
+// Access raw lines (after preprocessing)
+std::vector<std::string> lines = arff.getLines();
+
+// Get label states mapping
+auto states = arff.getStates();  // std::map<std::string, std::vector<std::string>>
+```
+
+## Data Processing Pipeline
+
+1. **File Parsing**: Reads ARFF format, extracts `@attribute` declarations and data
+2. **Missing Value Handling**: Skips lines containing `?` (missing values)
+3. **Feature Type Detection**: Automatically identifies `REAL`, `INTEGER`, `NUMERIC` vs categorical
+4. **Label Positioning**: Handles class attributes at any position in the data
+5. **Factorization**: Converts categorical features and labels to numeric codes
+6. **Data Organization**: Creates feature matrix `X` and label vector `y`
+
+## Example: Complete Workflow
+
+```cpp
+#include "ArffFiles.hpp"
+#include <iostream>
+
+int main() {
+    try {
+        ArffFiles arff;
+        arff.load("iris.arff");
+        
+        // Display dataset information
+        std::cout << "Dataset: " << arff.getClassName() << std::endl;
+        std::cout << "Samples: " << arff.getSize() << std::endl;
+        std::cout << "Features: " << arff.getX().size() << std::endl;
+        
+        // Show class labels
+        auto labels = arff.getLabels();
+        std::cout << "Classes: ";
+        for (const auto& label : labels) {
+            std::cout << label << " ";
+        }
+        std::cout << std::endl;
+        
+        // Show which features are numeric
+        auto numericFeatures = arff.getNumericAttributes();
+        for (const auto& [feature, isNumeric] : numericFeatures) {
+            std::cout << feature << ": " << (isNumeric ? "numeric" : "categorical") << std::endl;
+        }
+        
+    } catch (const std::exception& e) {
+        std::cerr << "Error: " << e.what() << std::endl;
+        return 1;
+    }
+    
+    return 0;
+}
+```
+
+## Supported ARFF Features
+
+- ✅ Numeric attributes (`@attribute feature REAL/INTEGER/NUMERIC`)
+- ✅ Categorical attributes (`@attribute feature {value1,value2,...}`)
+- ✅ Comments (lines starting with `%`)
+- ✅ Missing values (automatic skipping of lines with `?`)
+- ✅ Flexible class attribute positioning
+- ✅ Case-insensitive attribute declarations
+
+## Error Handling
+
+The library throws `std::invalid_argument` exceptions for:
+- Unable to open file
+- No attributes found in file
+- Specified class name not found
+
+## Development
+
+### Building and Testing
+
+```bash
+# Build and run tests
+make build && make test
+
+# Run tests with verbose output
+make test opt="-s"
+
+# Clean test artifacts
+make clean
+```
+
+### Using CMake Directly
+
+```bash
+mkdir build_debug
+cmake -S . -B build_debug -D CMAKE_BUILD_TYPE=Debug -D ENABLE_TESTING=ON
+cmake --build build_debug -t unit_tests_arffFiles
+cd build_debug/tests && ./unit_tests_arffFiles
+```
+
+## License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
+
+## Contributing
+
+Contributions are welcome! Please feel free to submit a Pull Request.

TECHNICAL_REPORT.md

@@ -0,0 +1,242 @@
+# ArffFiles Library - Technical Analysis Report
+
+**Generated**: 2025-06-27  
+**Version Analyzed**: 1.1.0  
+**Library Type**: Header-only C++17 ARFF File Parser  
+
+## Executive Summary
+
+The ArffFiles library is a functional header-only C++17 implementation for parsing ARFF (Attribute-Relation File Format) files. While it successfully accomplishes its core purpose, several significant weaknesses in design, performance, and robustness have been identified that could impact production use.
+
+**Overall Assessment**: ⚠️ **MODERATE RISK** - Functional but requires improvements for production use.
+
+---
+
+## 🟢 Strengths
+
+### 1. **Architectural Design**
+- ✅ **Header-only**: Easy integration, no compilation dependencies
+- ✅ **Modern C++17**: Uses appropriate standard library features
+- ✅ **Clear separation**: Public/protected/private access levels well-defined
+- ✅ **STL Integration**: Returns standard containers for seamless integration
+
+### 2. **Functionality**
+- ✅ **Flexible class positioning**: Supports class attributes at any position
+- ✅ **Automatic type detection**: Distinguishes numeric vs categorical attributes
+- ✅ **Missing value handling**: Skips lines with '?' characters
+- ✅ **Label encoding**: Automatic factorization of categorical features
+- ✅ **Case-insensitive parsing**: Handles @ATTRIBUTE/@attribute variations
+
+### 3. **API Usability**
+- ✅ **Multiple load methods**: Three different loading strategies
+- ✅ **Comprehensive getters**: Good access to internal data structures
+- ✅ **Utility functions**: Includes trim() and split() helpers
+
+### 4. **Testing Coverage**
+- ✅ **Real datasets**: Tests with iris, glass, adult, and Japanese vowels datasets
+- ✅ **Edge cases**: Tests different class positioning scenarios
+- ✅ **Data validation**: Verifies parsing accuracy with expected values
+
+---
+
+## 🔴 Critical Weaknesses
+
+### 1. **Memory Management & Performance Issues**
+
+#### **Inefficient Data Layout** (HIGH SEVERITY)
+```cpp
+// Line 131: Inefficient memory allocation
+X = std::vector<std::vector<float>>(attributes.size(), std::vector<float>(lines.size()));
+```
+- **Problem**: Feature-major layout instead of sample-major
+- **Impact**: Poor cache locality, inefficient for ML algorithms
+- **Memory overhead**: Double allocation for `X` and `Xs` vectors
+- **Performance**: Suboptimal for large datasets
+
+#### **Redundant Memory Usage** (MEDIUM SEVERITY)
+```cpp
+std::vector<std::vector<float>> X;           // Line 89
+std::vector<std::vector<std::string>> Xs;    // Line 90
+```
+- **Problem**: Maintains both numeric and string representations
+- **Impact**: 2x memory usage for categorical features
+- **Memory waste**: `Xs` could be deallocated after factorization
+
+#### **No Memory Pre-allocation** (MEDIUM SEVERITY)
+- **Problem**: Multiple vector resizing during parsing
+- **Impact**: Memory fragmentation and performance degradation
+
+### 2. **Error Handling & Robustness**
+
+#### **Unsafe Type Conversions** (HIGH SEVERITY)
+```cpp
+// Line 145: No exception handling
+X[xIndex][i] = stof(token);
+```
+- **Problem**: `stof()` can throw `std::invalid_argument` or `std::out_of_range`
+- **Impact**: Program termination on malformed numeric data
+- **Missing validation**: No checks for valid numeric format
+
+#### **Insufficient Input Validation** (HIGH SEVERITY)
+```cpp
+// Line 39: Unsafe comparison without bounds checking
+for (int i = 0; i < attributes.size(); ++i)
+```
+- **Problem**: No validation of file structure integrity
+- **Missing checks**: 
+  - Empty attribute names
+  - Duplicate attribute names
+  - Malformed attribute declarations
+  - Inconsistent number of tokens per line
+
+#### **Resource Management** (MEDIUM SEVERITY)
+```cpp
+// Line 163-194: No RAII for file handling
+std::ifstream file(fileName);
+// ... processing ...
+file.close(); // Manual close
+```
+- **Problem**: Manual file closing (though acceptable here)
+- **Potential issue**: No exception safety guarantee
+
+### 3. **Algorithm & Design Issues**
+
+#### **Inefficient String Processing** (MEDIUM SEVERITY)
+```cpp
+// Line 176-182: Inefficient attribute parsing
+std::stringstream ss(line);
+ss >> keyword >> attribute;
+type = "";
+while (ss >> type_w)
+    type += type_w + " ";  // String concatenation in loop
+```
+- **Problem**: Repeated string concatenation is O(n²)
+- **Impact**: Performance degradation on large files
+- **Solution needed**: Use string reserve or stringstream
+
+#### **Suboptimal Lookup Performance** (LOW SEVERITY)
+```cpp
+// Line 144: Map lookup in hot path
+if (numeric_features[attributes[xIndex].first])
+```
+- **Problem**: Hash map lookup for every data point
+- **Impact**: Unnecessary overhead during dataset generation
+
+### 4. **API Design Limitations**
+
+#### **Return by Value Issues** (MEDIUM SEVERITY)
+```cpp
+// Line 55-60: Expensive copies
+std::vector<std::string> getLines() const { return lines; }
+std::map<std::string, std::vector<std::string>> getStates() const { return states; }
+```
+- **Problem**: Large object copies instead of const references
+- **Impact**: Unnecessary memory allocation and copying
+- **Performance**: O(n) copy cost for large datasets
+
+#### **Non-const Correctness** (MEDIUM SEVERITY)
+```cpp
+// Line 68-69: Mutable references without const alternatives
+std::vector<std::vector<float>>& getX() { return X; }
+std::vector<int>& getY() { return y; }
+```
+- **Problem**: No const versions for read-only access
+- **Impact**: API design inconsistency, potential accidental modification
+
+#### **Type Inconsistency** (LOW SEVERITY)
+```cpp
+// Line 56: Mixed return types
+unsigned long int getSize() const { return lines.size(); }
+```
+- **Problem**: Should use `size_t` or `std::size_t`
+- **Impact**: Type conversion warnings on some platforms
+
+### 5. **Thread Safety**
+
+#### **Not Thread-Safe** (MEDIUM SEVERITY)
+- **Problem**: No synchronization mechanisms
+- **Impact**: Unsafe for concurrent access
+- **Missing**: Thread-safe accessors or documentation warning
+
+### 6. **Security Considerations**
+
+#### **Path Traversal Vulnerability** (LOW SEVERITY)
+```cpp
+// Line 161: No path validation
+void loadCommon(std::string fileName)
+```
+- **Problem**: No validation of file path
+- **Impact**: Potential directory traversal if user input not sanitized
+- **Mitigation**: Application-level validation needed
+
+#### **Resource Exhaustion** (MEDIUM SEVERITY)
+- **Problem**: No limits on file size or memory usage
+- **Impact**: Potential DoS with extremely large files
+- **Missing**: File size validation and memory limits
+
+### 7. **ARFF Format Compliance**
+
+#### **Limited Format Support** (MEDIUM SEVERITY)
+- **Missing features**:
+  - Date attributes (`@attribute date "yyyy-MM-dd HH:mm:ss"`)
+  - String attributes (`@attribute text string`)
+  - Relational attributes (nested ARFF)
+  - Sparse data format (`{0 X, 3 Y, 5 Z}`)
+
+#### **Parsing Edge Cases** (LOW SEVERITY)
+```cpp
+// Line 188: Simplistic missing value detection
+if (line.find("?", 0) != std::string::npos)
+```
+- **Problem**: Doesn't handle quoted '?' characters
+- **Impact**: May incorrectly skip valid data containing '?' in strings
+
+---
+
+## 🔧 Recommended Improvements
+
+### High Priority
+1. **Add exception handling** around `stof()` calls
+2. **Implement proper input validation** for malformed data
+3. **Fix memory layout** to sample-major organization
+4. **Add const-correct API methods**
+5. **Optimize string concatenation** in parsing
+
+### Medium Priority
+1. **Implement RAII** patterns consistently
+2. **Add memory usage limits** and validation
+3. **Provide const reference getters** for large objects
+4. **Document thread safety** requirements
+5. **Add comprehensive error reporting**
+
+### Low Priority
+1. **Extend ARFF format support** (dates, strings, sparse)
+2. **Optimize lookup performance** with cached indices
+3. **Add file path validation**
+4. **Implement move semantics** for performance
+
+---
+
+## 📊 Performance Metrics (Estimated)
+
+| Dataset Size | Memory Overhead | Performance Impact |
+|--------------|-----------------|-------------------|
+| Small (< 1MB) | ~200% | Negligible |
+| Medium (10MB) | ~300% | Moderate |
+| Large (100MB+) | ~400% | Significant |
+
+**Note**: Overhead includes duplicate storage and inefficient layout.
+
+---
+
+## 🎯 Conclusion
+
+The ArffFiles library successfully implements core ARFF parsing functionality but suffers from several design and implementation issues that limit its suitability for production environments. The most critical concerns are:
+
+1. **Lack of robust error handling** leading to potential crashes
+2. **Inefficient memory usage** limiting scalability
+3. **Performance issues** with large datasets
+
+While functional for small to medium datasets in controlled environments, significant refactoring would be required for production use with large datasets or untrusted input.
+
+**Recommendation**: Consider this library suitable for prototyping and small-scale applications, but plan for refactoring before production deployment.

cmake/modules/CodeCoverage.cmake

@@ -137,7 +137,7 @@
 
 include(CMakeParseArguments)
 
-option(CODE_COVERAGE_VERBOSE "Verbose information" FALSE)
+option(CODE_COVERAGE_VERBOSE "Verbose information" TRUE)
 
 # Check prereqs
 find_program( GCOV_PATH gcov )
@@ -160,7 +160,11 @@ foreach(LANG ${LANGUAGES})
     endif()
   elseif(NOT "${CMAKE_${LANG}_COMPILER_ID}" MATCHES "GNU"
          AND NOT "${CMAKE_${LANG}_COMPILER_ID}" MATCHES "(LLVM)?[Ff]lang")
-    message(FATAL_ERROR "Compiler is not GNU or Flang! Aborting...")
+         if ("${LANG}" MATCHES "CUDA")
+             message(STATUS "Ignoring CUDA")
+        else()
+            message(FATAL_ERROR "Compiler is not GNU or Flang! Aborting...")
+        endif()
   endif()
 endforeach()
 

conanfile.py

@@ -0,0 +1,43 @@
+import re
+from conan import ConanFile
+from conan.tools.files import copy
+
+
+class ArffFilesConan(ConanFile):
+    name = "arff-files"
+    version = "X.X.X"
+    description = (
+        "Header-only library to read ARFF (Attribute-Relation File Format) files and return STL vectors with the data read."
+    )
+    url = "https://github.com/rmontanana/ArffFiles"
+    license = "MIT"
+    homepage = "https://github.com/rmontanana/ArffFiles"
+    topics = ("arff", "data-processing", "file-parsing", "header-only", "cpp17")
+    no_copy_source = True
+    exports_sources = "ArffFiles.hpp", "LICENSE", "README.md"
+    package_type = "header-library"
+
+    def init(self):
+        # Read the CMakeLists.txt file to get the version
+        with open("CMakeLists.txt", "r") as f:
+            lines = f.readlines()
+        for line in lines:
+            if "VERSION" in line:
+                # Extract the version number using regex
+                match = re.search(r"VERSION\s+(\d+\.\d+\.\d+)", line)
+                if match:
+                    self.version = match.group(1)
+
+    def package(self):
+        # Copy header file to include directory
+        copy(self, "*.hpp", src=self.source_folder, dst=self.package_folder, keep_path=False)
+        # Copy license and readme for package documentation
+        copy(self, "LICENSE", src=self.source_folder, dst=self.package_folder, keep_path=False)
+        copy(self, "README.md", src=self.source_folder, dst=self.package_folder, keep_path=False)
+
+    def package_info(self):
+        # Header-only library configuration
+        self.cpp_info.bindirs = []
+        self.cpp_info.libdirs = []
+        # Set include directory (header will be in package root)
+        self.cpp_info.includedirs = ["."]

tests/TestArffFiles.cc

@@ -18,7 +18,7 @@ public:
 TEST_CASE("Version Test", "[ArffFiles]")
 {
     ArffFiles arff;
-    REQUIRE(arff.version() == "1.0.0");
+    REQUIRE(arff.version() == "1.1.0");
 }
 TEST_CASE("Load Test", "[ArffFiles]")
 {
@@ -65,14 +65,14 @@ TEST_CASE("Load Test", "[ArffFiles]")
 TEST_CASE("Load with class name", "[ArffFiles]")
 {
     ArffFiles arff;
-    arff.load(Paths::datasets("glass"), "Type");
+    arff.load(Paths::datasets("glass"), std::string("Type"));
     REQUIRE(arff.getClassName() == "Type");
     REQUIRE(arff.getClassType() == "{ 'build wind float', 'build wind non-float', 'vehic wind float', 'vehic wind non-float', containers, tableware, headlamps}");
     REQUIRE(arff.getLabels().size() == 6);
-    REQUIRE(arff.getLabels()[0] == "'build wind float'");
-    REQUIRE(arff.getLabels()[1] == "'vehic wind float'");
+    REQUIRE(arff.getLabels()[0] == "build wind float");
+    REQUIRE(arff.getLabels()[1] == "vehic wind float");
     REQUIRE(arff.getLabels()[2] == "tableware");
-    REQUIRE(arff.getLabels()[3] == "'build wind non-float'");
+    REQUIRE(arff.getLabels()[3] == "build wind non-float");
     REQUIRE(arff.getLabels()[4] == "headlamps");
     REQUIRE(arff.getLabels()[5] == "containers");
     REQUIRE(arff.getSize() == 214);
@@ -108,12 +108,41 @@ TEST_CASE("Load with class name as first attribute", "[ArffFiles]")
         {1.86094, 1.89165, 1.93921, 1.71752},
         {-0.207383, -0.193249, -0.239664, -0.218572} }
     };
+    auto X = arff.getX();
     for (int i = 0; i < 4; ++i) {
         for (int j = 0; j < 4; ++j)
-            REQUIRE(arff.getX()[i][j] == Catch::Approx(expected[i][j]));
+            REQUIRE(X[i][j] == Catch::Approx(expected[i][j]));
     }
     auto expected_y = std::vector<int>{ 0, 0, 0, 0 };
     for (int i = 120; i < 124; ++i)
         REQUIRE(arff.getY()[i] == expected_y[i - 120]);
 }
+TEST_CASE("Adult dataset", "[ArffFiles]")
+{
+    ArffFiles arff;
+    arff.load(Paths::datasets("adult"), std::string("class"));
+    REQUIRE(arff.getClassName() == "class");
+    REQUIRE(arff.getClassType() == "{ >50K, <=50K }");
+    REQUIRE(arff.getLabels().size() == 2);
+    REQUIRE(arff.getLabels()[0] == "<=50K");
+    REQUIRE(arff.getLabels()[1] == ">50K");
+    REQUIRE(arff.getSize() == 45222);
+    REQUIRE(arff.getLines().size() == 45222);
+    REQUIRE(arff.getLines()[0] == "25, Private, 226802, 11th, 7, Never-married, Machine-op-inspct, Own-child, Black, Male, 0, 0, 40, United-States, <=50K");
+    auto X = arff.getX();
+    REQUIRE(X[0][0] == 25);
+    REQUIRE(X[1][0] == 0);
+    REQUIRE(X[2][0] == 226802);
+    REQUIRE(X[3][0] == 0);
+    REQUIRE(X[4][0] == 7);
+    REQUIRE(X[5][0] == 0);
+    REQUIRE(X[6][0] == 0);
+    REQUIRE(X[7][0] == 0);
+    REQUIRE(X[8][0] == 0);
+    REQUIRE(X[9][0] == 0);
+    REQUIRE(X[10][0] == 0);
+    REQUIRE(X[11][0] == 0);
+    REQUIRE(X[12][0] == 40);
+    REQUIRE(X[13][0] == 0);
+}