feat(transit): 改进正文段落到Word文档的转换功能

支持自定义标题级别偏移量和正文样式,增强样式应用的灵活性。
- 新增 level_offset 参数用于调整标题级别
- 新增 body_style 参数用于设置正文段落样式
- 改进样式应用逻辑,支持多种样式的降级机制
- 更新配置文件以支持新的样式配置选项
- 修改解析器使致谢、参考文献和附录部分只提取正文内容
This commit is contained in:
zzy
2026-05-08 21:44:09 +08:00
parent ae70d05672
commit c29a3e6af0
4 changed files with 75 additions and 28 deletions

View File

@@ -11,12 +11,22 @@ from docx import Document
_PAT_HEADING = re.compile(r"^(#{1,6})\s+(.+)$", re.MULTILINE) _PAT_HEADING = re.compile(r"^(#{1,6})\s+(.+)$", re.MULTILINE)
def body_to_paragraphs(md_text: str) -> list[dict]: def body_to_paragraphs(
md_text: str,
*,
level_offset: int = 0,
body_style: str = "Body Text Indent",
) -> list[dict]:
"""将 Markdown 正文按标题和段落拆分为结构化列表。 """将 Markdown 正文按标题和段落拆分为结构化列表。
返回的每个元素:: Parameters
{"text": str, "level": int, "style": str} ----------
其中 ``style`` 为 ``Heading N`` 或 ``Normal``。 md_text : str
正文 Markdown。
level_offset : int
标题级别偏移量(正文从 ``##`` 开始时传 ``-1``,使其输出为 ``Heading 1``)。
body_style : str
正文段落的 Word 样式名。
""" """
paragraphs: list[dict] = [] paragraphs: list[dict] = []
last_end = 0 last_end = 0
@@ -30,10 +40,10 @@ def body_to_paragraphs(md_text: str) -> list[dict]:
block = block.strip() block = block.strip()
if block: if block:
paragraphs.append( paragraphs.append(
{"text": block, "level": 0, "style": "Normal"} {"text": block, "level": 0, "style": body_style}
) )
level = len(m.group(1)) level = len(m.group(1)) + level_offset
heading_text = m.group(2).strip() heading_text = m.group(2).strip()
paragraphs.append( paragraphs.append(
{"text": heading_text, "level": level, "style": f"Heading {level}"} {"text": heading_text, "level": level, "style": f"Heading {level}"}
@@ -47,21 +57,31 @@ def body_to_paragraphs(md_text: str) -> list[dict]:
block = block.strip() block = block.strip()
if block: if block:
paragraphs.append( paragraphs.append(
{"text": block, "level": 0, "style": "Normal"} {"text": block, "level": 0, "style": body_style}
) )
return paragraphs return paragraphs
def replace_placeholder(doc: Document, placeholder: str, paragraphs: list[dict]): def replace_placeholder(
doc: Document,
placeholder: str,
paragraphs: list[dict],
*,
default_body_style: str | None = None,
):
"""在 *doc* 中找到包含 *placeholder* 的段落,替换为 *paragraphs* 列表。 """在 *doc* 中找到包含 *placeholder* 的段落,替换为 *paragraphs* 列表。
每个段落的 ``style`` 字段会从文档样式中查找并应用。 正文段落的样式优先级:
1. ``style`` 字段指定的样式名(来自 ``body_to_paragraphs`` 的 ``body_style``
2. 占位符段落自身的样式(模板中已设好的样式)
3. ``Normal``
""" """
placeholder_found = False placeholder_found = False
for para in doc.paragraphs: for para in doc.paragraphs:
if placeholder in para.text: if placeholder in para.text:
placeholder_found = True placeholder_found = True
placeholder_style = para.style.name if para.style else None
parent = para._element.getparent() parent = para._element.getparent()
idx = list(parent).index(para._element) idx = list(parent).index(para._element)
parent.remove(para._element) parent.remove(para._element)
@@ -69,19 +89,36 @@ def replace_placeholder(doc: Document, placeholder: str, paragraphs: list[dict])
for pd_data in reversed(paragraphs): for pd_data in reversed(paragraphs):
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"]
try:
new_p.style = doc.styles[style_name] # 尝试应用样式,逐步降级
except KeyError: applied = _apply_style(new_p, doc, style_name)
matched = False if not applied and style_name.startswith("Heading"):
for s in doc.styles: # 标题样式找不到
if s.name.lower() == style_name.lower():
new_p.style = s
matched = True
break
if not matched:
new_p.style = doc.styles["Normal"] new_p.style = doc.styles["Normal"]
elif not applied:
# 正文样式找不到 → 尝试占位符自身的样式
if placeholder_style:
_apply_style(new_p, doc, placeholder_style)
if new_p.style.name == "Normal" and placeholder_style:
new_p.style = doc.styles[placeholder_style]
parent.insert(idx, new_p._element) parent.insert(idx, new_p._element)
break break
if not placeholder_found: if not placeholder_found:
print(f"警告:未找到占位符 '{placeholder}',正文段落未注入。") print(f"警告:未找到占位符 '{placeholder}',正文段落未注入。")
def _apply_style(paragraph, doc, style_name: str) -> bool:
"""尝试给段落应用样式,成功返回 ``True``。"""
try:
paragraph.style = doc.styles[style_name]
return True
except KeyError:
pass
# 大小写不敏感匹配
for s in doc.styles:
if s.name.lower() == style_name.lower():
paragraph.style = s
return True
return False

View File

@@ -22,6 +22,8 @@ class ThesisConfig:
body_end_keywords: list[str] = field( body_end_keywords: list[str] = field(
default_factory=lambda: ["致谢", "参考文献", "附录"] default_factory=lambda: ["致谢", "参考文献", "附录"]
) )
body_style: str = "Body Text Indent"
level_offset: int = -1
def to_dict(self) -> dict: def to_dict(self) -> dict:
"""转成模板渲染用的扁平字典,排除 options 命名空间。""" """转成模板渲染用的扁平字典,排除 options 命名空间。"""
@@ -60,4 +62,6 @@ def load_config(path: str | Path) -> ThesisConfig:
body_end_keywords=opts.get( body_end_keywords=opts.get(
"body_end_keywords", ["致谢", "参考文献", "附录"] "body_end_keywords", ["致谢", "参考文献", "附录"]
), ),
body_style=opts.get("body_style", "Body Text Indent"),
level_offset=opts.get("level_offset", -1),
) )

View File

@@ -113,19 +113,19 @@ def parse_markdown(
else: else:
data["body_md"] = "" data["body_md"] = ""
# ── 致谢 ── # ── 致谢(仅正文,不含标题行) ──
ack = _find_section(content, ["致谢"]) ack = _find_section(content, ["致谢"])
if ack: if ack:
data["acknowledgement"] = content[ack[0] : ack[1]].strip() data["acknowledgement"] = _get_section_body(content, ack)
# ── 参考文献 ── # ── 参考文献(仅正文,不含标题行) ──
ref = _find_section(content, ["参考文献"]) ref = _find_section(content, ["参考文献"])
if ref: if ref:
data["reference"] = content[ref[0] : ref[1]].strip() data["reference"] = _get_section_body(content, ref)
# ── 附录 ── # ── 附录(仅正文,不含标题行) ──
app = _find_section(content, ["附录"]) app = _find_section(content, ["附录"])
if app: if app:
data["appendix"] = content[app[0] : app[1]].strip() data["appendix"] = _get_section_body(content, app)
return data return data

View File

@@ -92,7 +92,10 @@ def generate_thesis(
# 5. 解析正文为段落列表 # 5. 解析正文为段落列表
body_md = ctx.get("body_md", "") body_md = ctx.get("body_md", "")
body_paragraphs = body_to_paragraphs(body_md) if body_md else [] body_paragraphs = (
body_to_paragraphs(body_md, level_offset=config.level_offset, body_style=config.body_style)
if body_md else []
)
# 6. 占位符 # 6. 占位符
ctx["body_placeholder"] = "__CONTEXT_PLACEHOLDER__" ctx["body_placeholder"] = "__CONTEXT_PLACEHOLDER__"
@@ -107,7 +110,10 @@ def generate_thesis(
# 9. 正文注入 # 9. 正文注入
final_doc = Document(str(temp_path)) final_doc = Document(str(temp_path))
replace_placeholder(final_doc, "__CONTEXT_PLACEHOLDER__", body_paragraphs) replace_placeholder(
final_doc, "__CONTEXT_PLACEHOLDER__", body_paragraphs,
default_body_style=config.body_style,
)
final_doc.save(str(output_path)) final_doc.save(str(output_path))
temp_path.unlink(missing_ok=True) temp_path.unlink(missing_ok=True)