translate-md-to-en.sh

translate-md-to-en.sh

#!/usr/bin/env bash
#
# translate-md.sh - Translate Japanese Markdown to English via OpenAI API
#
# Usage:
#   chmod +x translate-md-to-en.sh
#   ./translate-md-to-en.sh article.md    # creates article-en.md
#   cat input.md | ./translate-md.sh -    # outputs to stdout
#
# Environment variables:
#   OPENAI_API_KEY  (required) Your OpenAI API key
#   OPENAI_MODEL    (optional) Model to use (default: gpt-4.1-mini)
#

set -euo pipefail

# --- Configuration ---
MODEL="${OPENAI_MODEL:-gpt-4.1-mini}"
TEMPERATURE="0.2"
API_URL="https://api.openai.com/v1/chat/completions"

# --- Check dependencies ---
if ! command -v curl &>/dev/null; then
    echo "Error: curl is required but not installed." >&2
    exit 1
fi

# --- Check API key ---
if [[ -z "${OPENAI_API_KEY:-}" ]]; then
    echo "Error: OPENAI_API_KEY environment variable is not set." >&2
    echo "  export OPENAI_API_KEY='sk-...'" >&2
    exit 1
fi

# --- Parse arguments ---
if [[ $# -lt 1 ]]; then
    echo "Usage: $0 <input.md | ->" >&2
    echo "  Use '-' to read from stdin" >&2
    exit 1
fi

INPUT_FILE="$1"
OUTPUT_FILE=""

# --- Read input ---
if [[ "$INPUT_FILE" == "-" ]]; then
    INPUT_CONTENT=$(cat)
elif [[ -f "$INPUT_FILE" ]]; then
    INPUT_CONTENT=$(cat "$INPUT_FILE")
    # Generate output filename: article.md -> article-en.md
    DIR=$(dirname "$INPUT_FILE")
    BASENAME=$(basename "$INPUT_FILE" .md)
    OUTPUT_FILE="${DIR}/${BASENAME}-en.md"
else
    echo "Error: File not found: $INPUT_FILE" >&2
    exit 1
fi

if [[ -z "$INPUT_CONTENT" ]]; then
    echo "Error: Input is empty." >&2
    exit 1
fi

# --- Build system prompt ---
SYSTEM_PROMPT="You are a professional technical translator specializing in software engineering content."

# --- Build user prompt ---
USER_PROMPT="Translate the following Japanese Markdown document into natural, concise technical English suitable for Medium articles.

Rules:
- Preserve all Markdown structure (headings, lists, links, etc.)
- Do NOT modify content inside code blocks (\`\`\`)
- Keep technical terms, library names, and proper nouns as-is
- Prefer short, direct sentences
- Remove unnecessary Japanese-style prefaces or greetings
- Output ONLY the translated Markdown, no explanations

---

$INPUT_CONTENT"

# --- Escape strings for JSON using Python ---
escape_json() {
    python3 -c "import json,sys; print(json.dumps(sys.stdin.read()))"
}

ESCAPED_SYSTEM=$(echo -n "$SYSTEM_PROMPT" | escape_json)
ESCAPED_USER=$(echo -n "$USER_PROMPT" | escape_json)

# --- Build request body ---
REQUEST_BODY=$(cat <<EOF
{
  "model": "$MODEL",
  "temperature": $TEMPERATURE,
  "messages": [
    {"role": "system", "content": $ESCAPED_SYSTEM},
    {"role": "user", "content": $ESCAPED_USER}
  ]
}
EOF
)

# --- Make API request ---
TMPFILE=$(mktemp)
trap 'rm -f "$TMPFILE"' EXIT

HTTP_STATUS=$(curl -s -w "%{http_code}" -o "$TMPFILE" \
    -X POST "$API_URL" \
    -H "Content-Type: application/json; charset=utf-8" \
    -H "Authorization: Bearer $OPENAI_API_KEY" \
    -d "$REQUEST_BODY")

RESPONSE_BODY=$(cat "$TMPFILE")

# --- Check HTTP status ---
if [[ "$HTTP_STATUS" -ne 200 ]]; then
    echo "Error: API request failed with HTTP status $HTTP_STATUS" >&2
    echo "Response:" >&2
    echo "$RESPONSE_BODY" >&2
    exit 1
fi

# --- Extract translated content ---
extract_content() {
    if command -v jq &>/dev/null; then
        jq -r '.choices[0].message.content // empty'
    else
        python3 -c "
import json, sys
data = json.load(sys.stdin)
content = data.get('choices', [{}])[0].get('message', {}).get('content', '')
print(content)
"
    fi
}

TRANSLATED=$(echo "$RESPONSE_BODY" | extract_content)

if [[ -z "$TRANSLATED" ]]; then
    echo "Error: Failed to extract translated content from response." >&2
    echo "Response:" >&2
    echo "$RESPONSE_BODY" >&2
    exit 1
fi

# --- Output ---
if [[ -n "$OUTPUT_FILE" ]]; then
    echo "$TRANSLATED" > "$OUTPUT_FILE"
    echo "Created: $OUTPUT_FILE" >&2
else
    echo "$TRANSLATED"
fi