Edit S3 remote file and upload

edit-s3-remote-file.sh

#!/usr/bin/env bash

SUM=$(which sha512sum || which sha || which md5sum || which md5 || true)

# modify the following to match the s3 setup you desire
path="s3://bucket_name/path/$1"

tempfile=$(mktemp ~/.tmpXXXXXX)
trap "rm -f $tempfile" INT TERM HUP EXIT

aws s3 cp $path $tempfile

if [ ! -z "$SUM" ]
then
    FILEHASH=$($SUM $tempfile)
fi

"${EDITOR:-vim}" $tempfile

if [ -z "$FILEHASH" ] || [ "$FILEHASH" != "$($SUM $tempfile)" ]
then
   aws s3 cp --sse aws:kms $tempfile $path
fi

Claude AI improvements

I'll help modify the script to work better with VSCode while maintaining the security aspects. Here's an improved version:

#!/usr/bin/env bash

# Configuration
EDITOR=${EDITOR:-"code -w"} # Default to VSCode in wait mode
S3_BUCKET="s3://bucket_name"
S3_PATH="path"

# Ensure required commands exist
command -v aws >/dev/null 2>&1 || { echo "AWS CLI required but not installed. Aborting." >&2; exit 1; }
command -v code >/dev/null 2>&1 || { echo "VSCode (code) command not found. Ensure VSCode is in your PATH." >&2; exit 1; }

# Determine hash command
if command -v sha512sum >/dev/null 2>&1; then
    SUM="sha512sum"
elif command -v sha256sum >/dev/null 2>&1; then
    SUM="sha256sum"
elif command -v md5sum >/dev/null 2>&1; then
    SUM="md5sum"
else
    echo "No suitable hash command found. Installing sha256sum..."
    exit 1
fi

# Validate input
if [ -z "$1" ]; then
    echo "Usage: $0 <filename>"
    echo "Example: $0 config.json"
    exit 1
fi

# Create temporary directory with meaningful name
TEMP_DIR=$(mktemp -d ~/.s3edit-XXXXXX)
TEMP_FILE="$TEMP_DIR/$1"

# Cleanup function
cleanup() {
    rm -rf "$TEMP_DIR"
}
trap cleanup EXIT INT TERM

# Full S3 path
S3_FULL_PATH="${S3_BUCKET}/${S3_PATH}/$1"

echo "Downloading $S3_FULL_PATH..."

# Download file from S3
if ! aws s3 cp "$S3_FULL_PATH" "$TEMP_FILE"; then
    echo "Error downloading file from S3"
    exit 1
fi

# Calculate initial hash
INITIAL_HASH=$($SUM "$TEMP_FILE")

# Open in VSCode and wait for edit
echo "Opening in VSCode..."
$EDITOR "$TEMP_FILE"

# Calculate new hash
NEW_HASH=$($SUM "$TEMP_FILE")

# Upload if changed
if [ "$INITIAL_HASH" != "$NEW_HASH" ]; then
    echo "File changed, uploading to S3..."
    if aws s3 cp --sse aws:kms "$TEMP_FILE" "$S3_FULL_PATH"; then
        echo "Successfully uploaded to S3"
    else
        echo "Error uploading to S3"
        exit 1
    fi
else
    echo "No changes detected"
fi

Key improvements made:

1. Uses VSCode's wait flag (-w): This makes the script wait until you close the file in VSCode before continuing

2. Better file organization:

   • Creates a temporary directory instead of just a file
   • Preserves the original filename for better VSCode recognition
   • Cleaner cleanup using a trap on a directory

3. Added error handling:

   • Checks for required commands (aws, code)
   • Validates S3 operations
   • Provides meaningful error messages

4. Better user feedback:

   • Progress messages
   • Clear usage instructions
   • Status updates on file changes

5. Configuration section at the top:

   • Easy to modify S3 bucket and path
   • Configurable editor with VSCode default

To use this script:

1. Save it as something like s3edit.sh
2. Make it executable: chmod +x s3edit.sh
3. Run it: ./s3edit.sh myfile.json

You can also set up a VSCode keybinding for this script by adding to your keybindings.json:

{
    "key": "cmd+shift+s",  // or whatever binding you prefer
    "command": "workbench.action.terminal.sendSequence",
    "args": {
        "text": "./s3edit.sh ${file}\n"
    }
}

One additional improvement you might consider is adding a file extension whitelist or specific S3 path restrictions for security. Let me know if you'd like me to add those or explain any part in more detail!