feat(transit): 支持Markdown图片转Word图片段落
- 新增 transit/images.py 模块处理图片解析和插入逻辑 - 实现 `<img>` 标签解析为独立图片段落 - 添加图片尺寸检测和自适应缩放功能(超出页面宽度时自动缩放) - 支持图标题居中显示和图片居中对齐 - 优化 body_to_paragraphs 函数,添加图片处理逻辑 - 更新 README.md 使用说明,添加模板文件要求说明 BREAKING CHANGE: 图片处理方式变更,需要包含 sample.docx 模板文件
This commit is contained in:
@@ -1,4 +1,5 @@
|
|||||||
使用方法
|
使用方法
|
||||||
```shell
|
```shell
|
||||||
py .\test.py .\毕业论文初稿.md
|
# 需要 sample.docx 文件 且 该文件有 {{xxx}} 模板引擎的内容
|
||||||
|
python .\test.py .\毕业论文初稿.md
|
||||||
```
|
```
|
||||||
|
|||||||
@@ -7,9 +7,12 @@ Markdown 正文 → Word 段落转换。
|
|||||||
|
|
||||||
import re
|
import re
|
||||||
from copy import deepcopy
|
from copy import deepcopy
|
||||||
|
|
||||||
from docx import Document
|
from docx import Document
|
||||||
from docx.oxml.ns import qn
|
from docx.oxml.ns import qn
|
||||||
|
|
||||||
|
from .images import make_image_paragraph, is_figure_caption, insert_image_paragraphs
|
||||||
|
|
||||||
_PAT_HEADING = re.compile(r"^(#{1,6})\s+(.+)$", re.MULTILINE)
|
_PAT_HEADING = re.compile(r"^(#{1,6})\s+(.+)$", re.MULTILINE)
|
||||||
|
|
||||||
|
|
||||||
@@ -33,17 +36,28 @@ def body_to_paragraphs(
|
|||||||
paragraphs: list[dict] = []
|
paragraphs: list[dict] = []
|
||||||
last_end = 0
|
last_end = 0
|
||||||
|
|
||||||
|
def _add_block(block: str) -> None:
|
||||||
|
block = block.strip()
|
||||||
|
if not block:
|
||||||
|
return
|
||||||
|
# 图片段落
|
||||||
|
img = make_image_paragraph(block)
|
||||||
|
if img:
|
||||||
|
paragraphs.append(img)
|
||||||
|
return
|
||||||
|
# 跳过紧跟在图片后的重复图标题
|
||||||
|
if paragraphs and paragraphs[-1].get("type") == "image" and is_figure_caption(block):
|
||||||
|
return
|
||||||
|
# 普通正文段落
|
||||||
|
paragraphs.append({"text": block, "level": 0, "style": body_style})
|
||||||
|
|
||||||
for m in _PAT_HEADING.finditer(md_text):
|
for m in _PAT_HEADING.finditer(md_text):
|
||||||
# 标题前的普通文本
|
# 标题前的普通文本
|
||||||
if m.start() > last_end:
|
if m.start() > last_end:
|
||||||
pre = md_text[last_end : m.start()].strip()
|
pre = md_text[last_end : m.start()].strip()
|
||||||
if pre:
|
if pre:
|
||||||
for block in re.split(r"\n\s*\n", pre):
|
for block in re.split(r"\n\s*\n", pre):
|
||||||
block = block.strip()
|
_add_block(block)
|
||||||
if block:
|
|
||||||
paragraphs.append(
|
|
||||||
{"text": block, "level": 0, "style": body_style}
|
|
||||||
)
|
|
||||||
|
|
||||||
level = len(m.group(1)) + level_offset
|
level = len(m.group(1)) + level_offset
|
||||||
heading_text = m.group(2).strip()
|
heading_text = m.group(2).strip()
|
||||||
@@ -56,11 +70,7 @@ def body_to_paragraphs(
|
|||||||
tail = md_text[last_end:].strip()
|
tail = md_text[last_end:].strip()
|
||||||
if tail:
|
if tail:
|
||||||
for block in re.split(r"\n\s*\n", tail):
|
for block in re.split(r"\n\s*\n", tail):
|
||||||
block = block.strip()
|
_add_block(block)
|
||||||
if block:
|
|
||||||
paragraphs.append(
|
|
||||||
{"text": block, "level": 0, "style": body_style}
|
|
||||||
)
|
|
||||||
|
|
||||||
return paragraphs
|
return paragraphs
|
||||||
|
|
||||||
@@ -94,6 +104,11 @@ def replace_placeholder(
|
|||||||
parent.remove(para._element)
|
parent.remove(para._element)
|
||||||
|
|
||||||
for pd_data in reversed(paragraphs):
|
for pd_data in reversed(paragraphs):
|
||||||
|
if pd_data.get("type") == "image":
|
||||||
|
insert_image_paragraphs(
|
||||||
|
doc, [pd_data], idx=idx, parent=parent
|
||||||
|
)
|
||||||
|
else:
|
||||||
new_p = doc.add_paragraph(pd_data["text"])
|
new_p = doc.add_paragraph(pd_data["text"])
|
||||||
style_name = pd_data["style"]
|
style_name = pd_data["style"]
|
||||||
|
|
||||||
|
|||||||
151
transit/images.py
Normal file
151
transit/images.py
Normal file
@@ -0,0 +1,151 @@
|
|||||||
|
"""
|
||||||
|
图片处理模块。
|
||||||
|
|
||||||
|
将 Markdown 中的 ``<img>`` 标签解析为独立段落,
|
||||||
|
并在 Word 文档中插入图片及居中图标题。
|
||||||
|
"""
|
||||||
|
|
||||||
|
import struct
|
||||||
|
import re
|
||||||
|
|
||||||
|
from docx import Document
|
||||||
|
from docx.enum.text import WD_ALIGN_PARAGRAPH
|
||||||
|
|
||||||
|
# 匹配 <img src="..." alt="...">
|
||||||
|
_IMG_TAG = re.compile(
|
||||||
|
r'<img\s+([^>]*?)src=["\']([^"\']+)["\']([^>]*?)>', re.IGNORECASE
|
||||||
|
)
|
||||||
|
|
||||||
|
# 从标签属性块中提取名值对
|
||||||
|
_ATTR = re.compile(r'(\w+)\s*=\s*["\']([^"\']*)["\']')
|
||||||
|
|
||||||
|
# 匹配 **图X ...** 重复图标题(注入时跳过)
|
||||||
|
_FIG_CAPTION = re.compile(
|
||||||
|
r'^\*\*(?:图|Table|Figure|Fig\.?)\s*\d+.*\*\*$', re.IGNORECASE
|
||||||
|
)
|
||||||
|
|
||||||
|
|
||||||
|
def make_image_paragraph(block: str) -> dict | None:
|
||||||
|
"""若 *block* 包含 ``<img>`` 标签,返回图片段落字典;否则返回 ``None``。"""
|
||||||
|
m = _IMG_TAG.search(block)
|
||||||
|
if not m:
|
||||||
|
return None
|
||||||
|
attrs = dict(_ATTR.findall(block))
|
||||||
|
return {
|
||||||
|
"type": "image",
|
||||||
|
"src": attrs.get("src", m.group(2)),
|
||||||
|
"alt": attrs.get("alt", ""),
|
||||||
|
}
|
||||||
|
|
||||||
|
|
||||||
|
def is_figure_caption(block: str) -> bool:
|
||||||
|
"""检查 *block* 是否为 ``**图X ...**`` 格式的重复图标题。"""
|
||||||
|
return bool(_FIG_CAPTION.match(block.strip()))
|
||||||
|
|
||||||
|
|
||||||
|
def _get_image_dimensions(image_path: str) -> tuple[int, int] | None:
|
||||||
|
"""读取图片文件头返回 ``(width_px, height_px)``,不支持格式返回 ``None``。
|
||||||
|
|
||||||
|
仅读取文件头,不依赖第三方库。支持 PNG / JPEG / GIF / BMP。
|
||||||
|
"""
|
||||||
|
try:
|
||||||
|
with open(image_path, "rb") as f:
|
||||||
|
header = f.read(32)
|
||||||
|
except Exception:
|
||||||
|
return None
|
||||||
|
|
||||||
|
# PNG: 8-byte signature, then IHDR chunk
|
||||||
|
if header[:8] == b"\x89PNG\r\n\x1a\n":
|
||||||
|
w, h = struct.unpack_from(">II", header, 16)
|
||||||
|
return w, h
|
||||||
|
|
||||||
|
# JPEG: starts with FF D8, scan for SOF marker
|
||||||
|
if header[:2] == b"\xff\xd8":
|
||||||
|
pos = 2
|
||||||
|
while pos < len(header):
|
||||||
|
if header[pos] != 0xFF:
|
||||||
|
return None
|
||||||
|
marker = header[pos + 1]
|
||||||
|
if marker in (0xC0, 0xC1, 0xC2):
|
||||||
|
h, w = struct.unpack_from(">HH", header, pos + 5)
|
||||||
|
return w, h
|
||||||
|
seg_len = struct.unpack_from(">H", header, pos + 2)[0]
|
||||||
|
pos += 2 + seg_len
|
||||||
|
return None
|
||||||
|
|
||||||
|
# GIF: "GIF87a" or "GIF89a"
|
||||||
|
if header[:6] in (b"GIF87a", b"GIF89a"):
|
||||||
|
w, h = struct.unpack_from("<HH", header, 6)
|
||||||
|
return w, h
|
||||||
|
|
||||||
|
# BMP: "BM" signature
|
||||||
|
if header[:2] == b"BM":
|
||||||
|
w, h = struct.unpack_from("<ii", header, 18)
|
||||||
|
return w, abs(h)
|
||||||
|
|
||||||
|
return None
|
||||||
|
|
||||||
|
|
||||||
|
def _get_native_emu(image_path: str) -> int | None:
|
||||||
|
"""读取图片的原生宽度(EMU),失败返回 ``None``。
|
||||||
|
|
||||||
|
Word 默认以 72 DPI 渲染图片,1 px = 914400 / 72 = 12700 EMU。
|
||||||
|
"""
|
||||||
|
dims = _get_image_dimensions(image_path)
|
||||||
|
if dims is None:
|
||||||
|
return None
|
||||||
|
w_px, _ = dims
|
||||||
|
return w_px * 12700
|
||||||
|
|
||||||
|
|
||||||
|
def _constrain_width(image_path: str, page_text_width: int) -> int | None:
|
||||||
|
"""返回图片宽度(EMU),超出页宽时缩至页宽。
|
||||||
|
|
||||||
|
Parameters
|
||||||
|
----------
|
||||||
|
image_path : str
|
||||||
|
图片路径。
|
||||||
|
page_text_width : int
|
||||||
|
页面正文区宽度(EMU),来自 ``section.page_width - margins``。
|
||||||
|
"""
|
||||||
|
native = _get_native_emu(image_path)
|
||||||
|
if native is None:
|
||||||
|
return None
|
||||||
|
return min(native, page_text_width)
|
||||||
|
|
||||||
|
|
||||||
|
def insert_image_paragraphs(
|
||||||
|
doc: Document,
|
||||||
|
paragraphs: list[dict],
|
||||||
|
*,
|
||||||
|
idx: int,
|
||||||
|
parent,
|
||||||
|
):
|
||||||
|
"""在 *doc* 的指定位置插入图片段落序列。
|
||||||
|
|
||||||
|
每条图片段落生成两个 Word 段落:
|
||||||
|
1. 居中图片(超出页宽时自动缩放至页宽,否则保持原尺寸)
|
||||||
|
2. 居中图标题(从 ``alt`` 提取)
|
||||||
|
|
||||||
|
插入顺序保持 ``paragraphs`` 的原有顺序。
|
||||||
|
"""
|
||||||
|
section = doc.sections[0]
|
||||||
|
page_text_width = section.page_width - section.left_margin - section.right_margin
|
||||||
|
|
||||||
|
for pd_data in reversed(paragraphs):
|
||||||
|
# 图标题(在 reversed 中先插入,最终位于图片下方)
|
||||||
|
cap_p = doc.add_paragraph(pd_data.get("alt", ""))
|
||||||
|
cap_p.alignment = WD_ALIGN_PARAGRAPH.CENTER
|
||||||
|
parent.insert(idx, cap_p._element)
|
||||||
|
|
||||||
|
# 图片段落(max-width 行为:超出页宽时压缩,否则原尺寸)
|
||||||
|
img_p = doc.add_paragraph()
|
||||||
|
img_p.alignment = WD_ALIGN_PARAGRAPH.CENTER
|
||||||
|
img_run = img_p.add_run()
|
||||||
|
try:
|
||||||
|
img_width = _constrain_width(pd_data["src"], page_text_width)
|
||||||
|
img_run.add_picture(pd_data["src"], width=img_width)
|
||||||
|
except Exception as exc:
|
||||||
|
print(f"警告:图片加载失败 {pd_data['src']} — {exc}")
|
||||||
|
img_run.add_text(f"[图片加载失败: {pd_data['src']}]")
|
||||||
|
parent.insert(idx, img_p._element)
|
||||||
Reference in New Issue
Block a user