fix: replace awk-based .env parsing with read loop to handle values containing spaces

[amathxbt]

Bug

progress.sh loads .env using awk field splitting:

export $(cat .env | grep -v '#' | sed 's/\r$//' | awk '/=/ {print $1}' ) || error_exit "Failed to load .env file"

awk splits on whitespace by default. This means any .env value that contains a space is silently truncated:

# .env
ETH_RPC_URL=http://localhost:8545  # works fine
SOME_KEY=value with spaces         # ❌ awk prints only "SOME_KEY=value"

Additionally:

• grep -v '#' strips entire lines containing # anywhere, not just comment lines — it would incorrectly strip SOME_KEY=value#tag
• The || error_exit guard never fires on parsing errors because export of an empty string list succeeds silently
• export $(...) with word-split output is fragile: if a value accidentally contains shell metacharacters, it can cause unexpected behaviour

Fix

Replace the one-liner with a while read loop that processes the file line-by-line, correctly skipping blank lines and comment lines while exporting each assignment verbatim:

if [ -f .env ]; then
    while IFS= read -r line || [ -n "$line" ]; do
        # Skip blank lines and comment lines
        [[ -z "$line" || "$line" =~ ^[[:space:]]*# ]] && continue
        # Only export lines that contain an assignment
        if [[ "$line" =~ ^[A-Za-z_][A-Za-z0-9_]*= ]]; then
            export "$line"
        fi
    done < .env
fi

This correctly handles:

• Values with spaces (KEY=value with spaces)
• Inline comments being preserved as part of values (consistent with dotenv spec)
• Files that do not end with a newline (the || [ -n "$line" ] condition)