export.sh

#!/bin/bash
# =============================================================================
# Clean Project Export for LLM Analysis (Final Directory Fix)
# =============================================================================

set -e

OUTPUT_DIR="documentation/llm"
OUTPUT_FILE="$OUTPUT_DIR/dump.txt"
PROJECT_PATH="$(pwd)"

# Ensure we are in a git repository
if ! git rev-parse --is-inside-work-tree > /dev/null 2>&1; then
    echo "Error: This script must be run inside a Git repository."
    exit 1
fi

mkdir -p "$OUTPUT_DIR"

echo "=============================================="
echo "  Generating Clean Project Export"
echo "=============================================="

# Start output file with header
{
    echo "==============================================================================="
    echo "PROJECT EXPORT (GIT TRACKED ONLY)"
    echo "Generated: $(date)"
    echo "Project Path: $PROJECT_PATH"
    echo "==============================================================================="
    echo ""
} > "$OUTPUT_FILE"

# 1. Directory Structure (Using Python for a reliable tree)
echo "Generating directory structure..."
{
    echo "DIRECTORY STRUCTURE:"
    echo "==================="
    # This python snippet takes git-tracked files and builds a perfect visual tree
    git ls-files | python3 -c "
import sys
tree = {}
for line in sys.stdin:
    parts = line.strip().split('/')
    curr = tree
    for part in parts:
        curr = curr.setdefault(part, {})
def print_tree(d, indent=''):
    items = sorted(d.items())
    for i, (name, children) in enumerate(items):
        is_last = (i == len(items) - 1)
        print(f'{indent}{\"└── \" if is_last else \"├── \"}{name}')
        print_tree(children, indent + ('    ' if is_last else '│   '))
print_tree(tree)
"
    echo ""
} >> "$OUTPUT_FILE"

# 2. Collect and Process Files
echo "Collecting and cleaning file contents..."
{
    echo "FILE CONTENTS:"
    echo "=============="
    echo ""
} >> "$OUTPUT_FILE"

git ls-files | while read -r FILENAME; do
    # Skip the export script itself and the output file
    if [[ "$FILENAME" == "export.sh" || "$FILENAME" == "$OUTPUT_FILE" ]]; then
        continue
    fi

    # Skip specific binary extensions
    if [[ "$FILENAME" =~ \.(ico|png|jpg|jpeg|gif|dll|exe|pdb|bin|zip|tar|gz|7z|ttf|woff|woff2)$ ]]; then
        continue
    fi

    # Content-based binary check
    if file --mime "$FILENAME" | grep -q "binary"; then
        continue
    fi

    # Null byte check (Crucial for preventing "Unsupported Encoding" in Grok)
    if grep -qP '\x00' "$FILENAME" 2>/dev/null; then
        continue
    fi

    FILESIZE=$(stat -c%s "$FILENAME" 2>/dev/null || stat -f%z "$FILENAME" 2>/dev/null || echo "0")
    
    # Skip large files (>500KB)
    if [ "$FILESIZE" -gt 512000 ]; then
        continue
    fi

    {
        echo "================================================================================"
        echo "FILE: $FILENAME"
        echo "SIZE: $(echo "scale=2; $FILESIZE/1024" | bc 2>/dev/null || echo "0.00") KB"
        echo "================================================================================"
        echo ""
        # tr -d removes non-printable control characters that break LLM parsers
        cat "$FILENAME" | tr -d '\000-\010\013\014\016-\037' 
        echo ""
        echo ""
    } >> "$OUTPUT_FILE"
    
    echo "Processed: $FILENAME"
done

echo ""
echo "Export Complete: $OUTPUT_FILE"