c_prog/surnames_validation.sh

#!/bin/bash
# surnames_validation.sh
# Comprehensive validation script for DBLP surname extraction

set -e  # Exit on any error

DBLP_FILE="${1:-dblp.xml.gz}"
MIN_COUNT="${2:-10000}"
TEMP_DIR="/tmp/surnames_validation_$$"

# Colors for output
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
NC='\033[0m' # No Color

echo -e "${BLUE}=== DBLP Surname Extraction Validation ===${NC}"
echo "Input: $DBLP_FILE"
echo "Minimum count: $MIN_COUNT"
echo "Temp directory: $TEMP_DIR"

# Create temp directory
mkdir -p "$TEMP_DIR"

# Function to cleanup
cleanup() {
    echo -e "\n${YELLOW}Cleaning up...${NC}"
    rm -rf "$TEMP_DIR"
}
trap cleanup EXIT

# Check if files exist
if [[ ! -f "surnames" ]]; then
    echo -e "${RED}Error: C program 'surnames' not found!${NC}"
    exit 1
fi

if [[ ! -f "surnames.py" ]]; then
    echo -e "${RED}Error: Python script 'surnames.py' not found!${NC}"
    exit 1
fi

if [[ ! -f "$DBLP_FILE" ]]; then
    echo -e "${RED}Error: DBLP file '$DBLP_FILE' not found!${NC}"
    exit 1
fi

echo -e "\n${BLUE}=== Memory Leak Testing (Valgrind) ===${NC}"
echo "Running valgrind on C program..."
if command -v valgrind &> /dev/null; then
    gunzip -c "$DBLP_FILE" | valgrind --leak-check=full --show-leak-kinds=all \
        --track-origins=yes --verbose --log-file="$TEMP_DIR/valgrind.log" \
        ./surnames > "$TEMP_DIR/c_results_valgrind.txt" 2>&1
    
    echo "Valgrind results:"
    if grep -q "All heap blocks were freed" "$TEMP_DIR/valgrind.log"; then
        echo -e "${GREEN}✓ No memory leaks detected${NC}"
    else
        echo -e "${RED}⚠ Potential memory issues detected${NC}"
        grep -E "(definitely lost|possibly lost|still reachable)" "$TEMP_DIR/valgrind.log" | head -5
    fi
    echo "Full valgrind log: $TEMP_DIR/valgrind.log"
else
    echo -e "${YELLOW}Valgrind not available, skipping memory check${NC}"
fi

echo -e "\n${BLUE}=== Performance Profiling ===${NC}"

# C program performance
echo "Profiling C program..."
if command -v perf &> /dev/null; then
    echo "Using perf for detailed profiling..."
    gunzip -c "$DBLP_FILE" | perf stat -d ./surnames > "$TEMP_DIR/c_results.txt" 2> "$TEMP_DIR/c_perf.log"
    echo "Perf results:"
    grep -E "(task-clock|cycles|instructions|cache-misses)" "$TEMP_DIR/c_perf.log" | head -4
else
    echo "Using time for basic profiling..."
    gunzip -c "$DBLP_FILE" | /usr/bin/time -v ./surnames > "$TEMP_DIR/c_results.txt" 2> "$TEMP_DIR/c_time.log"
    echo "Time results:"
    grep -E "(Elapsed|Maximum resident|Page faults)" "$TEMP_DIR/c_time.log"
fi

# Python program performance
echo -e "\nProfiling Python program..."
gunzip -c "$DBLP_FILE" | /usr/bin/time -v python3 surnames.py > "$TEMP_DIR/python_results.txt" 2> "$TEMP_DIR/python_time.log"
echo "Python time results:"
grep -E "(Elapsed|Maximum resident|Page faults)" "$TEMP_DIR/python_time.log"

echo -e "\n${BLUE}=== Output Validation ===${NC}"

# Compare line counts
c_lines=$(wc -l < "$TEMP_DIR/c_results.txt")
python_lines=$(wc -l < "$TEMP_DIR/python_results.txt")

echo "Result counts:"
echo "  C program: $c_lines surnames"
echo "  Python script: $python_lines surnames"

if [[ $c_lines -eq $python_lines ]]; then
    echo -e "${GREEN}✓ Line counts match${NC}"
else
    echo -e "${YELLOW}⚠ Line counts differ${NC}"
fi

# Compare top 10 results
echo -e "\nTop 10 comparison:"
echo -e "${BLUE}C Program:${NC}"
head -10 "$TEMP_DIR/c_results.txt"
echo -e "${BLUE}Python Script:${NC}"
head -10 "$TEMP_DIR/python_results.txt"

# Detailed difference analysis
echo -e "\n${BLUE}=== Detailed Difference Analysis ===${NC}"
if command -v diff &> /dev/null; then
    echo "Running diff analysis..."
    if diff -q "$TEMP_DIR/c_results.txt" "$TEMP_DIR/python_results.txt" > /dev/null; then
        echo -e "${GREEN}✓ Results are identical!${NC}"
    else
        echo -e "${YELLOW}⚠ Found differences:${NC}"
        diff "$TEMP_DIR/c_results.txt" "$TEMP_DIR/python_results.txt" | head -20
        
        # Count differences
        diff_count=$(diff "$TEMP_DIR/c_results.txt" "$TEMP_DIR/python_results.txt" | wc -l)
        echo "Total difference lines: $diff_count"
    fi
else
    echo "diff not available, using comm..."
    comm -3 <(sort "$TEMP_DIR/c_results.txt") <(sort "$TEMP_DIR/python_results.txt") | head -10
fi

# Specific validation checks
echo -e "\n${BLUE}=== Specific Validation Checks ===${NC}"

# Check Wang count (most frequent)
c_wang=$(grep "^Wang " "$TEMP_DIR/c_results.txt" | awk '{print $2}')
python_wang=$(grep "^Wang " "$TEMP_DIR/python_results.txt" | awk '{print $2}')

echo "Wang frequency check:"
echo "  C: $c_wang"
echo "  Python: $python_wang"
if [[ "$c_wang" == "$python_wang" ]]; then
    echo -e "${GREEN}✓ Wang counts match${NC}"
else
    echo -e "${RED}✗ Wang counts differ${NC}"
fi

# Check for duplicates in C results
echo -e "\nDuplicate check (C program):"
duplicates=$(awk '{print $1}' "$TEMP_DIR/c_results.txt" | sort | uniq -d | wc -l)
if [[ $duplicates -eq 0 ]]; then
    echo -e "${GREEN}✓ No duplicates found${NC}"
else
    echo -e "${RED}✗ Found $duplicates duplicate surnames${NC}"
    awk '{print $1}' "$TEMP_DIR/c_results.txt" | sort | uniq -d | head -5
fi

# Performance comparison
echo -e "\n${BLUE}=== Performance Summary ===${NC}"
c_time=$(grep "Elapsed" "$TEMP_DIR/c_time.log" | awk '{print $8}' | head -1)
python_time=$(grep "Elapsed" "$TEMP_DIR/python_time.log" | awk '{print $8}' | head -1)

echo "Execution times:"
echo "  C program: $c_time"
echo "  Python script: $python_time"

# Save results for later analysis
echo -e "\n${BLUE}=== Results Saved ===${NC}"
echo "Results saved in: $TEMP_DIR"
echo "  C results: $TEMP_DIR/c_results.txt"
echo "  Python results: $TEMP_DIR/python_results.txt"
echo "  Valgrind log: $TEMP_DIR/valgrind.log"
echo "  Performance logs: $TEMP_DIR/*_time.log"

echo -e "\n${GREEN}Validation complete!${NC}"