如何在AWS Lambda中通过Python实现XLSX转PDF？能否无需EC2效仿同类服务？

Absolutely—you don’t need to spin up an EC2 instance to handle XLSX to PDF conversions with Python in AWS Lambda. The key is leveraging cross-platform tools that work with Lambda’s Linux runtime. Here are the most practical approaches:

1. Use LibreOffice (Most Reliable for Complex Spreadsheets) #

LibreOffice is a free, open-source office suite that can convert XLSX to PDF via command line, and it works perfectly on Linux (Lambda’s default runtime). You can package LibreOffice into an AWS Lambda Layer to reuse it across functions.

Steps to implement: #

• Build a Lambda Layer with LibreOffice:
  1. Spin up an Amazon Linux 2 EC2 instance (temporarily, just to build the layer—you can terminate it afterward).
  2. Install LibreOffice: sudo amazon-linux-extras install libreoffice
  3. Zip the LibreOffice binaries (usually located at /usr/lib64/libreoffice and any required dependencies).
  4. Upload this zip as a Lambda Layer.
• Python code in Lambda:
  Use subprocess to call LibreOffice’s conversion command. Example:

  import subprocess
  import os
  import tempfile
  import boto3
  
  s3 = boto3.client('s3')
  
  def lambda_handler(event, context):
      # Download XLSX from S3 to temp directory
      bucket = event['bucket']
      xlsx_key = event['xlsx_key']
      xlsx_path = "/tmp/input.xlsx"
      pdf_path = "/tmp/output.pdf"
      pdf_key = xlsx_key.replace('.xlsx', '.pdf')
  
      s3.download_file(bucket, xlsx_key, xlsx_path)
  
      # Call LibreOffice to convert
      cmd = [
          "/opt/libreoffice/program/soffice",
          "--headless",
          "--convert-to", "pdf",
          "--outdir", os.path.dirname(pdf_path),
          xlsx_path
      ]
      subprocess.run(cmd, check=True)
  
      # Upload PDF back to S3
      s3.upload_file(pdf_path, bucket, pdf_key)
  
      return {"statusCode": 200, "message": "Conversion complete", "pdf_key": pdf_key}
  

  Note: Make sure your Lambda function has enough memory (at least 512MB) and timeout (10-15 seconds) to handle LibreOffice’s startup overhead.

2. Pure Python Libraries (For Simple Spreadsheets) #

If your XLSX files have basic formatting (no complex charts, pivot tables, or conditional formatting), you can use pure Python libraries to read the spreadsheet and generate a PDF directly.

Example with openpyxl + reportlab: #

• openpyxl reads the XLSX content.
• reportlab generates a PDF with the extracted data.

from openpyxl import load_workbook
from reportlab.pdfgen import canvas
from reportlab.lib.pagesizes import letter
import boto3

s3 = boto3.client('s3')

def lambda_handler(event, context):
    bucket = event['bucket']
    xlsx_key = event['xlsx_key']
    xlsx_path = "/tmp/input.xlsx"
    pdf_path = "/tmp/output.pdf"
    pdf_key = xlsx_key.replace('.xlsx', '.pdf')

    s3.download_file(bucket, xlsx_key, xlsx_path)
    wb = load_workbook(xlsx_path)
    ws = wb.active

    pdf = canvas.Canvas(pdf_path, pagesize=letter)
    y = 750  # Start position from top of page
    col_width = 150  # Adjust based on your content

    for row in ws.iter_rows(values_only=True):
        for idx, cell_value in enumerate(row):
            pdf.drawString(50 + (idx * col_width), y, str(cell_value) if cell_value else "")
        y -= 20
        if y < 50:
            pdf.showPage()
            y = 750

    pdf.save()
    s3.upload_file(pdf_path, bucket, pdf_key)

    return {"statusCode": 200, "pdf_key": pdf_key}

Limitations: This won’t replicate complex Excel formatting (colors, cell merging, formulas) as accurately as LibreOffice.

3. Serverless Container Images #

If you find packaging LibreOffice into a layer too tricky, you can build a custom Docker image with Python and LibreOffice pre-installed, then deploy it as a Lambda container. This gives you full control over the runtime environment.

Basic Dockerfile example: #

FROM public.ecr.aws/lambda/python:3.11

# Install LibreOffice and dependencies
RUN yum install -y libreoffice && yum clean all

# Copy Python code
COPY app.py ${LAMBDA_TASK_ROOT}

# Install any required Python packages
COPY requirements.txt .
RUN pip install -r requirements.txt --target "${LAMBDA_TASK_ROOT}"

CMD ["app.lambda_handler"]

All these approaches avoid the need for a persistent EC2 instance—they run entirely within Lambda’s serverless environment. For most real-world use cases (especially with complex spreadsheets), the LibreOffice layer method is the most robust choice.

内容的提问来源于stack exchange，提问作者Utopion