Python使用PyPDF2和ReportLab操作PDF文件的详细指南

pip install PyPDF2 reportlab

from PyPDF2 import PdfReader

# 打开 PDF 文件
reader = PdfReader("example.pdf")

# 获取页面数
num_pages = len(reader.pages)
print(f"Total pages: {num_pages}")

# 读取每一页的内容
for page_num in range(num_pages):
    page = reader.pages[page_num]
    text = page.extract_text()
    print(f"Page {page_num + 1}:\n{text}")

from PyPDF2 import PdfWriter, PdfReader

# 创建 PDF 写入器
writer = PdfWriter()

# 读取两个 PDF 文件并将它们的页面添加到写入器中
pdf_files = ["file1.pdf", "file2.pdf"]
for pdf_file in pdf_files:
    reader = PdfReader(pdf_file)
    for page in reader.pages:
        writer.add_page(page)

# 保存合并后的 PDF 文件
with open("merged_output.pdf", "wb") as output_pdf:
    writer.write(output_pdf)

from PyPDF2 import PdfWriter, PdfReader

reader = PdfReader("example.pdf")
writer = PdfWriter()

# 提取特定页
for i in range(3):  # 这里表示提取第1页到第3页
    writer.add_page(reader.pages[i])

# 保存拆分后的文件
with open("split_output.pdf", "wb") as output_pdf:
    writer.write(output_pdf)

writer = PdfWriter()
reader = PdfReader("example.pdf")

# 添加所有页面
for page in reader.pages:
    writer.add_page(page)

# 加密并设置密码
writer.encrypt("password123")

# 保存加密的文件
with open("encrypted_output.pdf", "wb") as output_pdf:
    writer.write(output_pdf)

from reportlab.lib.pagesizes import A4
from reportlab.pdfgen import canvas

# 创建 PDF 文件
pdf_path = "generated_example.pdf"
pdf_canvas = canvas.Canvas(pdf_path, pagesize=A4)

# 添加文本
pdf_canvas.drawString(100, 750, "Hello, ReportLab!")
pdf_canvas.drawString(100, 730, "This is a simple PDF file created using Python.")

# 保存并关闭 PDF
pdf_canvas.save()
print(f"PDF saved as {pdf_path}")

from reportlab.lib.pagesizes import A4
from reportlab.pdfgen import canvas

# 创建 PDF 文件
pdf_path = "pdf_with_image.pdf"
pdf_canvas = canvas.Canvas(pdf_path, pagesize=A4)

# 添加图片
pdf_canvas.drawImage("example_image.jpg", 100, 500, width=200, height=150)

# 绘制矩形
pdf_canvas.setStrokeColorRGB(0, 0, 1)  # 蓝色边框
pdf_canvas.setFillColorRGB(0.8, 0.8, 1)  # 浅蓝填充
pdf_canvas.rect(100, 450, 200, 100, fill=True)

# 保存 PDF
pdf_canvas.save()
print(f"PDF with image and shapes saved as {pdf_path}")

from reportlab.lib.pagesizes import A4
from reportlab.pdfgen import canvas
from reportlab.platypus import Table, TableStyle
from reportlab.lib import colors

# 创建 PDF 文件
pdf_path = "pdf_with_table.pdf"
pdf_canvas = canvas.Canvas(pdf_path, pagesize=A4)

# 表格数据
data = [
    ["Product", "Price", "Quantity"],
    ["Widget", "$25.00", "10"],
    ["Gadget", "$15.00", "30"],
    ["Doohickey", "$5.00", "50"]
]

# 创建表格
table = Table(data)
table.setStyle(TableStyle([
    ("BACKGROUND", (0, 0), (-1, 0), colors.grey),
    ("TEXTCOLOR", (0, 0), (-1, 0), colors.whitesmoke),
    ("ALIGN", (0, 0), (-1, -1), "CENTER"),
    ("GRID", (0, 0), (-1, -1), 0.5, colors.black),
    ("BACKGROUND", (0, 1), (-1, -1), colors.beige),
]))

# 将表格添加到 PDF
table.wrapOn(pdf_canvas, 400, 300)
table.drawOn(pdf_canvas, 100, 600)

# 保存 PDF
pdf_canvas.save()
print(f"PDF with table saved as {pdf_path}")

from reportlab.lib.pagesizes import A4
from reportlab.pdfgen import canvas
from reportlab.platypus import Table, TableStyle
from reportlab.lib import colors

def create_invoice_template():
    pdf_path = "invoice_template.pdf"
    pdf_canvas = canvas.Canvas(pdf_path, pagesize=A4)

    # 设置页面标题
    pdf_canvas.setFont("Helvetica-Bold", 16)
    pdf_canvas.drawString(220, 800, "Invoice")

    # 公司信息
    pdf_canvas.setFont("Helvetica", 12)
    pdf_canvas.drawString(50, 780, "Company Name: XYZ Ltd.")
    pdf_canvas.drawString(50, 765, "Address: 123 Example St., City")
    pdf_canvas.drawString(50, 750, "Phone: (123) 456-7890")
    pdf_canvas.drawString(50, 735, "Email: contact@xyz.com")

    # 客户信息部分
    pdf_canvas.drawString(50, 700, "Bill To:")
    pdf_canvas.drawString(50, 685, "Customer Name:")
    pdf_canvas.drawString(50, 670, "Customer Address:")

    # 添加表格表头
    data = [["Item", "Description", "Quantity", "Unit Price", "Total"]]
    table = Table(data)
    table.setStyle(TableStyle([
        ("BACKGROUND", (0, 0), (-1, 0), colors.grey),
        ("TEXTCOLOR", (0, 0), (-1, 0), colors.whitesmoke),
        ("ALIGN", (0, 0), (-1, -1), "CENTER"),
        ("FONTNAME", (0, 0), (-1, 0), "Helvetica-Bold"),
        ("FONTSIZE", (0, 0), (-1, 0), 12),
        ("BOTTOMPADDING", (0, 0), (-1, 0), 12),
        ("GRID", (0, 0), (-1, -1), 0.5, colors.black),
    ]))
    table.wrapOn(pdf_canvas, 450, 400)
    table.drawOn(pdf_canvas, 50, 600)

    # 保存模板
    pdf_canvas.save()
    print(f"Invoice template saved as {pdf_path}")

# 生成模板
create_invoice_template()

from PyPDF2 import PdfReader, PdfWriter
from reportlab.lib.pagesizes import A4
from reportlab.pdfgen import canvas
from io import BytesIO

def fill_invoice(customer_name, customer_address, items):
    # 打开模板
    reader = PdfReader("invoice_template.pdf")
    writer = PdfWriter()

    # 创建一个内存缓冲区来绘制覆盖内容
    packet = BytesIO()
    pdf_canvas = canvas.Canvas(packet, pagesize=A4)

    # 填写客户信息
    pdf_canvas.setFont("Helvetica", 12)
    pdf_canvas.drawString(150, 685, customer_name)
    pdf_canvas.drawString(150, 670, customer_address)

    # 填写项目明细
    y = 580
    for item in items:
        pdf_canvas.drawString(50, y, item["item"])
        pdf_canvas.drawString(150, y, item["description"])
        pdf_canvas.drawString(250, y, str(item["quantity"]))
        pdf_canvas.drawString(350, y, f"${item['unit_price']:.2f}")
        pdf_canvas.drawString(450, y, f"${item['quantity'] * item['unit_price']:.2f}")
        y -= 20  # 调整 y 坐标，确保每一项在新行

    # 保存绘制的内容
    pdf_canvas.save()

    # 将覆盖内容作为新页面内容合并
    packet.seek(0)
    overlay = PdfReader(packet)
    for page in reader.pages:
        page.merge_page(overlay.pages[0])
        writer.add_page(page)

    # 保存带内容的发票
    with open("invoice_filled.pdf", "wb") as output_pdf:
        writer.write(output_pdf)
    print("Invoice filled and saved as invoice_filled.pdf")

# 示例数据
customer_name = "John Doe"
customer_address = "456 Example Ave., City"
items = [
    {"item": "Widget", "description": "High-quality widget", "quantity": 5, "unit_price": 20.00},
    {"item": "Gadget", "description": "Advanced gadget", "quantity": 3, "unit_price": 35.00},
    {"item": "Doohickey", "description": "Multi-purpose tool", "quantity": 2, "unit_price": 15.50},
]

# 生成发票
fill_invoice(customer_name, customer_address, items)