628 lines
24 KiB
Python
628 lines
24 KiB
Python
|
"""
|
|||
|
|
图表解析器 - ChartBuilder 的反向操作
|
|||
|
|
|
|||
|
|
这是 ChartBuilder 的逆向工程：
|
|||
|
|
- ChartBuilder: Config -> XML
|
|||
|
|
- ChartParser: XML -> Config
|
|||
|
|
|
|||
|
|
核心职责：
|
|||
|
|
1. 解析坐标轴映射（主轴/次轴）
|
|||
|
|
2. 解析绘图区中的所有系列
|
|||
|
|
3. 提取嵌入的 Excel 数据
|
|||
|
|
4. 生成可用于 ChartBuilder 的配置
|
|||
|
|
"""
|
|||
|
|
|
|||
|
|
from typing import Dict, List, Tuple, Optional
|
|||
|
|
from lxml import etree
|
|||
|
|
from pptx import Presentation
|
|||
|
|
from pptx.chart.chart import Chart
|
|||
|
|
import pandas as pd
|
|||
|
|
import io
|
|||
|
|
from openpyxl import load_workbook
|
|||
|
|
|
|||
|
|
from .oxml_ns import NAMESPACES
|
|||
|
|
|
|||
|
|
|
|||
|
|
class ChartParser:
|
|||
|
|
"""
|
|||
|
|
图表解析器 - 从现有图表提取配置
|
|||
|
|
|
|||
|
|
使用场景：
|
|||
|
|
1. 从"黄金标准"PPTX 提取图表配置
|
|||
|
|
2. 批量复现图表
|
|||
|
|
3. 数据更新和刷新
|
|||
|
|
"""
|
|||
|
|
|
|||
|
|
def __init__(self, chart: Chart):
|
|||
|
|
"""
|
|||
|
|
初始化解析器
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
chart: python-pptx Chart 对象
|
|||
|
|
"""
|
|||
|
|
self.chart = chart
|
|||
|
|
self.chart_element = chart._element
|
|||
|
|
self.plotArea = self.chart_element.find('.//c:plotArea', namespaces=NAMESPACES)
|
|||
|
|
|
|||
|
|
# 解析坐标轴映射
|
|||
|
|
self.axis_map = self._parse_axes()
|
|||
|
|
|
|||
|
|
print("\n📖 ChartParser 初始化完成")
|
|||
|
|
print(f" - 发现坐标轴: {len(self.axis_map)} 个")
|
|||
|
|
for ax_id, ax_info in self.axis_map.items():
|
|||
|
|
print(f" • {ax_id}: {ax_info['type']} ({ax_info.get('position', 'N/A')})")
|
|||
|
|
|
|||
|
|
def _parse_axes(self) -> Dict[str, Dict]:
|
|||
|
|
"""
|
|||
|
|
解析所有坐标轴，建立 ID 映射表
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
{
|
|||
|
|
"123456": {"type": "category", "position": None},
|
|||
|
|
"789001": {"type": "primary", "position": "l"},
|
|||
|
|
"789002": {"type": "secondary", "position": "r"}
|
|||
|
|
}
|
|||
|
|
"""
|
|||
|
|
axis_map = {}
|
|||
|
|
|
|||
|
|
# 1. 解析所有值轴 (Y轴)
|
|||
|
|
val_axes = self.plotArea.findall('.//c:valAx', namespaces=NAMESPACES)
|
|||
|
|
for val_ax in val_axes:
|
|||
|
|
ax_id_elem = val_ax.find('.//c:axId', namespaces=NAMESPACES)
|
|||
|
|
if ax_id_elem is None:
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
ax_id = ax_id_elem.get('val')
|
|||
|
|
|
|||
|
|
# 查找轴的位置
|
|||
|
|
ax_pos_elem = val_ax.find('.//c:axPos', namespaces=NAMESPACES)
|
|||
|
|
position = ax_pos_elem.get('val') if ax_pos_elem is not None else 'l'
|
|||
|
|
|
|||
|
|
# 判断主轴还是次轴
|
|||
|
|
# 'r' (right) = 次轴, 'l' (left) = 主轴
|
|||
|
|
if position == 'r':
|
|||
|
|
axis_map[ax_id] = {"type": "secondary", "position": position}
|
|||
|
|
else:
|
|||
|
|
axis_map[ax_id] = {"type": "primary", "position": position}
|
|||
|
|
|
|||
|
|
# 2. 解析所有分类轴 (X轴)
|
|||
|
|
cat_axes = self.plotArea.findall('.//c:catAx', namespaces=NAMESPACES)
|
|||
|
|
for cat_ax in cat_axes:
|
|||
|
|
ax_id_elem = cat_ax.find('.//c:axId', namespaces=NAMESPACES)
|
|||
|
|
if ax_id_elem is None:
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
ax_id = ax_id_elem.get('val')
|
|||
|
|
axis_map[ax_id] = {"type": "category", "position": None}
|
|||
|
|
|
|||
|
|
return axis_map
|
|||
|
|
|
|||
|
|
def parse(self) -> Tuple[List[Dict], pd.DataFrame, str, Optional[Dict]]:
|
|||
|
|
"""
|
|||
|
|
解析图表配置和数据
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
(series_config, df, categories_col, layout_info)
|
|||
|
|
|
|||
|
|
series_config: 系列配置列表
|
|||
|
|
df: 数据 DataFrame
|
|||
|
|
categories_col: 分类列名
|
|||
|
|
layout_info: 布局信息（图例、轴配置等）
|
|||
|
|
"""
|
|||
|
|
print("\n" + "=" * 80)
|
|||
|
|
print("🔍 开始解析图表")
|
|||
|
|
print("=" * 80)
|
|||
|
|
|
|||
|
|
# 1. 提取嵌入的 Excel 数据
|
|||
|
|
df, categories_col = self._extract_embedded_data()
|
|||
|
|
|
|||
|
|
print(f"\n📊 数据提取完成:")
|
|||
|
|
print(f" - 行数: {len(df)}")
|
|||
|
|
print(f" - 列数: {len(df.columns)}")
|
|||
|
|
print(f" - 分类列: {categories_col}")
|
|||
|
|
print(f" - 数据列: {list(df.columns[1:])}")
|
|||
|
|
|
|||
|
|
# 2. 解析系列配置
|
|||
|
|
series_config = self._parse_series_config()
|
|||
|
|
|
|||
|
|
print(f"\n📋 系列配置解析完成:")
|
|||
|
|
for i, cfg in enumerate(series_config):
|
|||
|
|
print(f" {i+1}. {cfg['name']}: type={cfg['type']}, axis={cfg['axis']}, key={cfg['key']}")
|
|||
|
|
|
|||
|
|
# 3. ⭐ 新增：解析布局配置
|
|||
|
|
layout_info = self._parse_layout_info()
|
|||
|
|
|
|||
|
|
if layout_info:
|
|||
|
|
print(f"\n⚙️ 布局配置:")
|
|||
|
|
if 'legend' in layout_info:
|
|||
|
|
leg = layout_info['legend']
|
|||
|
|
print(f" - 图例: 位置={leg.get('position')}, 字体={leg.get('font_size_pt')}pt")
|
|||
|
|
if 'category_axis' in layout_info:
|
|||
|
|
cat = layout_info['category_axis']
|
|||
|
|
print(f" - 横轴: 类型={cat.get('type')}, 格式={cat.get('number_format')}")
|
|||
|
|
|
|||
|
|
print("\n" + "=" * 80)
|
|||
|
|
print("✅ 图表解析完成！")
|
|||
|
|
print("=" * 80)
|
|||
|
|
|
|||
|
|
return series_config, df, categories_col, layout_info
|
|||
|
|
|
|||
|
|
def _parse_layout_info(self) -> Dict:
|
|||
|
|
"""
|
|||
|
|
解析布局信息（图例、轴配置、标题、尺寸等）
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
布局信息字典
|
|||
|
|
"""
|
|||
|
|
layout_info = {}
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
# 1. 解析图表标题
|
|||
|
|
try:
|
|||
|
|
if self.chart.has_title:
|
|||
|
|
title_frame = self.chart.chart_title.text_frame
|
|||
|
|
if title_frame and title_frame.text:
|
|||
|
|
layout_info["title"] = {
|
|||
|
|
"text": title_frame.text,
|
|||
|
|
"has_title": True,
|
|||
|
|
}
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f" ⚠️ 标题解析失败: {e}")
|
|||
|
|
|
|||
|
|
# 2. 解析图例
|
|||
|
|
if self.chart.has_legend:
|
|||
|
|
legend = self.chart.legend
|
|||
|
|
layout_info["legend"] = {
|
|||
|
|
"position": legend.position,
|
|||
|
|
"font_size_pt": legend.font.size.pt if legend.font.size else None,
|
|||
|
|
"include_in_layout": legend.include_in_layout,
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
# 3. 解析横轴（分类轴）
|
|||
|
|
try:
|
|||
|
|
category_axis = self.chart.category_axis
|
|||
|
|
layout_info["category_axis"] = {
|
|||
|
|
"type": category_axis.category_type,
|
|||
|
|
"major_unit": category_axis.major_unit if hasattr(category_axis, 'major_unit') else None,
|
|||
|
|
"number_format": category_axis.tick_labels.number_format if hasattr(category_axis.tick_labels, 'number_format') else None,
|
|||
|
|
"font_size_pt": category_axis.tick_labels.font.size.pt if category_axis.tick_labels.font.size else None,
|
|||
|
|
}
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f" ⚠️ 横轴解析失败: {e}")
|
|||
|
|
|
|||
|
|
# 4. 解析纵轴（值轴）
|
|||
|
|
try:
|
|||
|
|
value_axis = self.chart.value_axis
|
|||
|
|
layout_info["value_axis"] = {
|
|||
|
|
"number_format": value_axis.tick_labels.number_format if hasattr(value_axis.tick_labels, 'number_format') else None,
|
|||
|
|
"font_size_pt": value_axis.tick_labels.font.size.pt if value_axis.tick_labels.font.size else None,
|
|||
|
|
"has_major_gridlines": value_axis.has_major_gridlines if hasattr(value_axis, 'has_major_gridlines') else None,
|
|||
|
|
}
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f" ⚠️ 纵轴解析失败: {e}")
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f" ⚠️ 布局信息解析失败: {e}")
|
|||
|
|
|
|||
|
|
return layout_info
|
|||
|
|
|
|||
|
|
def _extract_embedded_data(self) -> Tuple[pd.DataFrame, str]:
|
|||
|
|
"""
|
|||
|
|
提取嵌入在图表中的 Excel 数据
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
(df, categories_col)
|
|||
|
|
"""
|
|||
|
|
try:
|
|||
|
|
# 方法1: 尝试从 chart 的内部部分获取 Excel 数据
|
|||
|
|
# python-pptx 的图表对象包含嵌入的 Excel workbook
|
|||
|
|
chart_part = self.chart.part
|
|||
|
|
|
|||
|
|
# 查找 Excel workbook 关系
|
|||
|
|
for rel in chart_part.rels.values():
|
|||
|
|
if 'package' in rel.reltype or 'oleObject' in rel.reltype or 'embeddings' in rel.reltype:
|
|||
|
|
try:
|
|||
|
|
xlsx_blob = rel.target_part.blob
|
|||
|
|
wb = load_workbook(io.BytesIO(xlsx_blob))
|
|||
|
|
ws = wb.active
|
|||
|
|
|
|||
|
|
# 转换为 DataFrame
|
|||
|
|
data = []
|
|||
|
|
for row in ws.iter_rows(values_only=True):
|
|||
|
|
data.append(row)
|
|||
|
|
|
|||
|
|
if data:
|
|||
|
|
# 第一行是表头
|
|||
|
|
df = pd.DataFrame(data[1:], columns=data[0])
|
|||
|
|
categories_col = df.columns[0]
|
|||
|
|
print(f" ✅ 成功从嵌入的 Excel 提取数据")
|
|||
|
|
return df, categories_col
|
|||
|
|
except Exception as e2:
|
|||
|
|
continue
|
|||
|
|
|
|||
|
|
# 如果上面都失败，尝试从 XML 中提取
|
|||
|
|
print(f" ⚠️ 未找到嵌入的 Excel 数据，从 XML 提取")
|
|||
|
|
return self._extract_data_from_xml()
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f" ⚠️ 提取嵌入数据失败: {e}")
|
|||
|
|
# 如果失败，尝试从 XML 中提取数据点
|
|||
|
|
return self._extract_data_from_xml()
|
|||
|
|
|
|||
|
|
def _extract_data_from_xml(self) -> Tuple[pd.DataFrame, str]:
|
|||
|
|
"""
|
|||
|
|
从 XML 中提取数据点（备用方法）
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
(df, categories_col)
|
|||
|
|
"""
|
|||
|
|
print(" → 尝试从 XML 提取数据...")
|
|||
|
|
from datetime import datetime, timedelta
|
|||
|
|
|
|||
|
|
# 查找所有绘图元素
|
|||
|
|
plot_types = ['barChart', 'lineChart', 'areaChart', 'scatterChart']
|
|||
|
|
|
|||
|
|
data_dict = {}
|
|||
|
|
categories = []
|
|||
|
|
categories_col = "分类"
|
|||
|
|
max_data_points = 0
|
|||
|
|
|
|||
|
|
# 第一遍：找到所有系列和最大数据点数量
|
|||
|
|
for plot_tag in plot_types:
|
|||
|
|
plot_elements = self.plotArea.findall(f'.//c:{plot_tag}', namespaces=NAMESPACES)
|
|||
|
|
|
|||
|
|
for plot_element in plot_elements:
|
|||
|
|
# 提取分类数据（只需提取一次）
|
|||
|
|
if not categories:
|
|||
|
|
cat_elem = plot_element.find('.//c:cat', namespaces=NAMESPACES)
|
|||
|
|
if cat_elem is not None:
|
|||
|
|
# ⭐ 优先尝试 strCache（文本）
|
|||
|
|
cat_cache = cat_elem.find('.//c:strCache', namespaces=NAMESPACES)
|
|||
|
|
if cat_cache is not None:
|
|||
|
|
# ⭐ 修复 A: 强制将字符串转为 datetime
|
|||
|
|
print(" → 发现 strCache，尝试转换为 datetime...")
|
|||
|
|
for pt in cat_cache.findall('.//c:pt', namespaces=NAMESPACES):
|
|||
|
|
v = pt.find('.//c:v', namespaces=NAMESPACES)
|
|||
|
|
if v is not None:
|
|||
|
|
try:
|
|||
|
|
# 尝试用 pandas 智能解析多种日期格式
|
|||
|
|
dt = pd.to_datetime(v.text).to_pydatetime()
|
|||
|
|
categories.append(dt)
|
|||
|
|
print(f" • 转换: {v.text} → {dt}")
|
|||
|
|
except (ValueError, TypeError):
|
|||
|
|
categories.append(v.text) # 回退到字符串
|
|||
|
|
if categories:
|
|||
|
|
print(f" → 成功将 {len(categories)} 个字符串日期转换为 datetime")
|
|||
|
|
else:
|
|||
|
|
# ⭐ 如果没有 strCache，尝试 numCache（日期轴）
|
|||
|
|
num_cache = cat_elem.find('.//c:numCache', namespaces=NAMESPACES)
|
|||
|
|
if num_cache is not None:
|
|||
|
|
# 提取格式代码（如 "yyyy/mm"）
|
|||
|
|
format_code = num_cache.find('.//c:formatCode', namespaces=NAMESPACES)
|
|||
|
|
date_format = format_code.text if format_code is not None else None
|
|||
|
|
|
|||
|
|
# ⭐ 修复 B: 强制将 Excel 序列号转为 datetime
|
|||
|
|
print(f" → 发现 numCache（格式: {date_format}），转换为 datetime...")
|
|||
|
|
for pt in num_cache.findall('.//c:pt', namespaces=NAMESPACES):
|
|||
|
|
v = pt.find('.//c:v', namespaces=NAMESPACES)
|
|||
|
|
if v is not None:
|
|||
|
|
try:
|
|||
|
|
excel_date_num = float(v.text)
|
|||
|
|
# Excel 基准是 1899-12-30
|
|||
|
|
dt = datetime(1899, 12, 30) + timedelta(days=excel_date_num)
|
|||
|
|
categories.append(dt)
|
|||
|
|
except (ValueError, OverflowError):
|
|||
|
|
categories.append(v.text) # 回退到字符串
|
|||
|
|
|
|||
|
|
if categories:
|
|||
|
|
print(f" → 成功将 {len(categories)} 个 Excel 序列号转换为 datetime")
|
|||
|
|
|
|||
|
|
# 提取每个系列的数据
|
|||
|
|
series_elements = plot_element.findall('.//c:ser', namespaces=NAMESPACES)
|
|||
|
|
for ser in series_elements:
|
|||
|
|
# 获取系列名称
|
|||
|
|
name = self._extract_series_name(ser)
|
|||
|
|
|
|||
|
|
# 获取数值
|
|||
|
|
values = self._extract_series_values(ser)
|
|||
|
|
|
|||
|
|
if name and values:
|
|||
|
|
data_dict[name] = values
|
|||
|
|
max_data_points = max(max_data_points, len(values))
|
|||
|
|
|
|||
|
|
# 如果没有分类数据，使用最大数据点数量生成索引
|
|||
|
|
if not categories and max_data_points > 0:
|
|||
|
|
categories = [f"项目{i+1}" for i in range(max_data_points)]
|
|||
|
|
print(f" → 未找到分类数据，生成 {len(categories)} 个默认分类")
|
|||
|
|
elif not categories and not data_dict:
|
|||
|
|
# 如果完全没有数据，返回空 DataFrame
|
|||
|
|
print(" ⚠️ 未找到任何数据，返回空 DataFrame")
|
|||
|
|
return pd.DataFrame({categories_col: []}), categories_col
|
|||
|
|
|
|||
|
|
# 确保所有系列的长度与分类数量一致
|
|||
|
|
target_length = len(categories)
|
|||
|
|
for name, values in data_dict.items():
|
|||
|
|
if len(values) < target_length:
|
|||
|
|
# 补齐 None
|
|||
|
|
data_dict[name] = values + [None] * (target_length - len(values))
|
|||
|
|
elif len(values) > target_length:
|
|||
|
|
# 截断
|
|||
|
|
data_dict[name] = values[:target_length]
|
|||
|
|
print(f" → 系列 '{name}' 数据点从 {len(values)} 截断到 {target_length}")
|
|||
|
|
|
|||
|
|
# 构建 DataFrame
|
|||
|
|
df_data = {categories_col: categories}
|
|||
|
|
df_data.update(data_dict)
|
|||
|
|
df = pd.DataFrame(df_data)
|
|||
|
|
|
|||
|
|
print(f" → 从 XML 提取了 {len(categories)} 行数据，{len(data_dict)} 个系列")
|
|||
|
|
|
|||
|
|
return df, categories_col
|
|||
|
|
|
|||
|
|
def _parse_series_config(self) -> List[Dict]:
|
|||
|
|
"""
|
|||
|
|
解析系列配置
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
系列配置列表
|
|||
|
|
"""
|
|||
|
|
series_config = []
|
|||
|
|
|
|||
|
|
# 图表类型映射
|
|||
|
|
plot_type_map = {
|
|||
|
|
'barChart': 'bar',
|
|||
|
|
'lineChart': 'line',
|
|||
|
|
'areaChart': 'area',
|
|||
|
|
'scatterChart': 'scatter'
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
# 遍历所有图表类型
|
|||
|
|
for plot_tag, chart_type in plot_type_map.items():
|
|||
|
|
plot_elements = self.plotArea.findall(f'.//c:{plot_tag}', namespaces=NAMESPACES)
|
|||
|
|
|
|||
|
|
for plot_element in plot_elements:
|
|||
|
|
# ⭐ 关键修复：在绘图元素级别获取轴引用
|
|||
|
|
plot_axis_type = self._extract_plot_axis(plot_element)
|
|||
|
|
|
|||
|
|
# 遍历该绘图中的所有系列
|
|||
|
|
series_elements = plot_element.findall('.//c:ser', namespaces=NAMESPACES)
|
|||
|
|
|
|||
|
|
for ser in series_elements:
|
|||
|
|
# 解析单个系列
|
|||
|
|
config_entry = self._parse_single_series(ser, chart_type, plot_axis_type)
|
|||
|
|
if config_entry:
|
|||
|
|
series_config.append(config_entry)
|
|||
|
|
|
|||
|
|
return series_config
|
|||
|
|
|
|||
|
|
def _extract_plot_axis(self, plot_element) -> str:
|
|||
|
|
"""
|
|||
|
|
提取绘图元素使用的坐标轴（主轴/次轴）
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
plot_element: 绘图元素 (<c:barChart>, <c:lineChart> 等)
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
'primary' 或 'secondary'
|
|||
|
|
"""
|
|||
|
|
# 查找绘图元素引用的轴 ID
|
|||
|
|
ax_id_refs = plot_element.findall('.//c:axId', namespaces=NAMESPACES)
|
|||
|
|
|
|||
|
|
for ax_id_ref in ax_id_refs:
|
|||
|
|
ax_id = ax_id_ref.get('val')
|
|||
|
|
|
|||
|
|
if ax_id in self.axis_map:
|
|||
|
|
axis_info = self.axis_map[ax_id]
|
|||
|
|
axis_type = axis_info['type']
|
|||
|
|
|
|||
|
|
# 如果是值轴（主轴或次轴），返回它
|
|||
|
|
if axis_type in ('primary', 'secondary'):
|
|||
|
|
return axis_type
|
|||
|
|
|
|||
|
|
# 默认返回主轴
|
|||
|
|
return 'primary'
|
|||
|
|
|
|||
|
|
def _parse_single_series(self, ser, chart_type: str, axis_type: str) -> Optional[Dict]:
|
|||
|
|
"""
|
|||
|
|
解析单个系列
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
ser: 系列元素 (<c:ser>)
|
|||
|
|
chart_type: 图表类型
|
|||
|
|
axis_type: 轴类型（从父绘图元素获取）
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
系列配置字典，或 None
|
|||
|
|
"""
|
|||
|
|
config_entry = {}
|
|||
|
|
|
|||
|
|
# A. 获取系列名称
|
|||
|
|
name = self._extract_series_name(ser)
|
|||
|
|
if not name:
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
config_entry["name"] = name
|
|||
|
|
config_entry["key"] = name # 默认使用 name 作为 key
|
|||
|
|
|
|||
|
|
# B. 设置图表类型
|
|||
|
|
config_entry["type"] = chart_type
|
|||
|
|
|
|||
|
|
# C. 使用传入的轴分配
|
|||
|
|
config_entry["axis"] = axis_type
|
|||
|
|
|
|||
|
|
return config_entry
|
|||
|
|
|
|||
|
|
def _extract_series_name(self, ser) -> Optional[str]:
|
|||
|
|
"""
|
|||
|
|
提取系列名称
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
ser: 系列元素 (<c:ser>)
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
系列名称，或 None
|
|||
|
|
"""
|
|||
|
|
# 方法1: 从 <c:tx> 中提取
|
|||
|
|
tx_elem = ser.find('.//c:tx', namespaces=NAMESPACES)
|
|||
|
|
if tx_elem is not None:
|
|||
|
|
# 尝试从 strRef 中提取
|
|||
|
|
v_elem = tx_elem.find('.//c:v', namespaces=NAMESPACES)
|
|||
|
|
if v_elem is not None and v_elem.text:
|
|||
|
|
return v_elem.text
|
|||
|
|
|
|||
|
|
return None
|
|||
|
|
|
|||
|
|
def _extract_series_values(self, ser) -> List:
|
|||
|
|
"""
|
|||
|
|
提取系列数值
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
ser: 系列元素 (<c:ser>)
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
数值列表
|
|||
|
|
"""
|
|||
|
|
values = []
|
|||
|
|
|
|||
|
|
# 尝试从 <c:val> 提取（柱状图、折线图、面积图）
|
|||
|
|
val_elem = ser.find('.//c:val', namespaces=NAMESPACES)
|
|||
|
|
if val_elem is not None:
|
|||
|
|
num_cache = val_elem.find('.//c:numCache', namespaces=NAMESPACES)
|
|||
|
|
if num_cache is not None:
|
|||
|
|
for pt in num_cache.findall('.//c:pt', namespaces=NAMESPACES):
|
|||
|
|
v = pt.find('.//c:v', namespaces=NAMESPACES)
|
|||
|
|
if v is not None:
|
|||
|
|
try:
|
|||
|
|
values.append(float(v.text))
|
|||
|
|
except (ValueError, TypeError):
|
|||
|
|
values.append(0)
|
|||
|
|
|
|||
|
|
# 尝试从 <c:yVal> 提取（散点图）
|
|||
|
|
if not values:
|
|||
|
|
yVal_elem = ser.find('.//c:yVal', namespaces=NAMESPACES)
|
|||
|
|
if yVal_elem is not None:
|
|||
|
|
num_cache = yVal_elem.find('.//c:numCache', namespaces=NAMESPACES)
|
|||
|
|
if num_cache is not None:
|
|||
|
|
for pt in num_cache.findall('.//c:pt', namespaces=NAMESPACES):
|
|||
|
|
v = pt.find('.//c:v', namespaces=NAMESPACES)
|
|||
|
|
if v is not None:
|
|||
|
|
try:
|
|||
|
|
values.append(float(v.text))
|
|||
|
|
except (ValueError, TypeError):
|
|||
|
|
values.append(0)
|
|||
|
|
|
|||
|
|
return values
|
|||
|
|
|
|||
|
|
|
|||
|
|
def parse_chart_from_pptx(pptx_path: str, slide_idx: int = 0, shape_idx: int = 0) -> Tuple[List[Dict], pd.DataFrame, str]:
|
|||
|
|
"""
|
|||
|
|
从 PPTX 文件解析图表
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
pptx_path: PPTX 文件路径
|
|||
|
|
slide_idx: 幻灯片索引（0-based）
|
|||
|
|
shape_idx: 图表形状索引（0-based）
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
(series_config, df, categories_col)
|
|||
|
|
|
|||
|
|
Example:
|
|||
|
|
>>> series_config, df, categories_col = parse_chart_from_pptx("template.pptx")
|
|||
|
|
>>> # 现在可以用这些配置来复现图表
|
|||
|
|
>>> builder.build(series_config)
|
|||
|
|
"""
|
|||
|
|
# 加载 PPTX
|
|||
|
|
prs = Presentation(pptx_path)
|
|||
|
|
|
|||
|
|
if slide_idx >= len(prs.slides):
|
|||
|
|
raise ValueError(f"幻灯片索引 {slide_idx} 超出范围（共 {len(prs.slides)} 张）")
|
|||
|
|
|
|||
|
|
slide = prs.slides[slide_idx]
|
|||
|
|
|
|||
|
|
# 查找图表
|
|||
|
|
chart = None
|
|||
|
|
chart_shape = None
|
|||
|
|
|
|||
|
|
for shape in slide.shapes:
|
|||
|
|
if hasattr(shape, 'chart'):
|
|||
|
|
if shape_idx == 0:
|
|||
|
|
chart = shape.chart
|
|||
|
|
chart_shape = shape
|
|||
|
|
break
|
|||
|
|
shape_idx -= 1
|
|||
|
|
|
|||
|
|
if chart is None:
|
|||
|
|
raise ValueError(f"在幻灯片 {slide_idx} 上未找到图表")
|
|||
|
|
|
|||
|
|
# 解析图表
|
|||
|
|
parser = ChartParser(chart)
|
|||
|
|
return parser.parse()
|
|||
|
|
|
|||
|
|
|
|||
|
|
def parse_all_charts_from_pptx(pptx_path: str) -> List[Dict]:
|
|||
|
|
"""
|
|||
|
|
解析 PPTX 文件中的所有图表
|
|||
|
|
|
|||
|
|
Args:
|
|||
|
|
pptx_path: PPTX 文件路径
|
|||
|
|
|
|||
|
|
Returns:
|
|||
|
|
图表信息列表
|
|||
|
|
[
|
|||
|
|
{
|
|||
|
|
"slide_idx": 0,
|
|||
|
|
"shape_idx": 0,
|
|||
|
|
"shape_name": "图表 1",
|
|||
|
|
"series_config": [...],
|
|||
|
|
"df": DataFrame,
|
|||
|
|
"categories_col": "日期"
|
|||
|
|
},
|
|||
|
|
...
|
|||
|
|
]
|
|||
|
|
"""
|
|||
|
|
prs = Presentation(pptx_path)
|
|||
|
|
all_charts = []
|
|||
|
|
|
|||
|
|
print(f"\n🔍 扫描 PPTX 文件: {pptx_path}")
|
|||
|
|
print(f" - 共 {len(prs.slides)} 张幻灯片\n")
|
|||
|
|
|
|||
|
|
for slide_idx, slide in enumerate(prs.slides):
|
|||
|
|
chart_count = 0
|
|||
|
|
|
|||
|
|
for shape_idx, shape in enumerate(slide.shapes):
|
|||
|
|
# 使用 has_chart 属性检查，而不是直接访问 chart
|
|||
|
|
if hasattr(shape, 'has_chart') and shape.has_chart:
|
|||
|
|
chart_count += 1
|
|||
|
|
|
|||
|
|
try:
|
|||
|
|
# 解析图表
|
|||
|
|
parser = ChartParser(shape.chart)
|
|||
|
|
series_config, df, categories_col, layout_info = parser.parse()
|
|||
|
|
|
|||
|
|
chart_info = {
|
|||
|
|
"slide_index": slide_idx,
|
|||
|
|
"chart_index": chart_count - 1, # 0-based
|
|||
|
|
"shape_idx": shape_idx,
|
|||
|
|
"shape_name": shape.name if hasattr(shape, 'name') else f"图表 {chart_count}",
|
|||
|
|
"series_config": series_config,
|
|||
|
|
"df": df,
|
|||
|
|
"categories_col": categories_col,
|
|||
|
|
"layout_info": layout_info # 新增：布局信息
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
all_charts.append(chart_info)
|
|||
|
|
|
|||
|
|
print(f"✅ 幻灯片 {slide_idx + 1}, 图表 {chart_count}: {chart_info['shape_name']}")
|
|||
|
|
print(f" - {len(series_config)} 个系列")
|
|||
|
|
|
|||
|
|
except Exception as e:
|
|||
|
|
print(f"⚠️ 幻灯片 {slide_idx + 1}, 图表 {chart_count} 解析失败: {e}")
|
|||
|
|
|
|||
|
|
if chart_count > 0:
|
|||
|
|
print(f" → 幻灯片 {slide_idx + 1}: 发现 {chart_count} 个图表\n")
|
|||
|
|
|
|||
|
|
print(f"\n📊 总计解析 {len(all_charts)} 个图表")
|
|||
|
|
|
|||
|
|
return all_charts
|
|||
|
|
|