Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows.
Python ai4sciencedocument-analysisdocxextract-datalayout-analysisocrparserpdfpdf-converterpdf-extractor-llmpdf-extractor-pretrainpdf-extractor-ragpdf-parserpptxpythonxlsx
from __future__ import annotations
from typing import Any
def inline(content: str, *, styles: list[str] | None = None) -> list[dict[str, Any]]:
"""把测试文字构造成最小 Middle JSON 2.0 TextSpan 列表。"""
span: dict[str, Any] = {"type": "text", "content": content}
if styles:
span["styles"] = styles
return [span]
def equation(latex: str) -> dict[str, str]:
"""构造测试用行内公式 Span。"""
return {"type": "equation_inline", "content": latex}
def hyperlink(url: str, content: str) -> dict[str, Any]:
"""构造测试用超链接 Span。"""
return {"type": "hyperlink", "url": url, "content": inline(content)}
def inline_text(spans: Any) -> str:
"""提取 raw 或 Pydantic InlineSpan 序列的可见文字,供断言复用。"""
if not isinstance(spans, list):
return ""
parts: list[str] = []
for span in spans:
if isinstance(span, dict):
span_type = span.get("type")
content = span.get("content")
else:
span_type = getattr(span, "type", None)
content = getattr(span, "content", None)
if str(span_type) == "hyperlink":
parts.append(inline_text(content))
elif isinstance(content, str):
parts.append(content)
return "".join(parts)
def inline_items(spans: Any) -> list[Any]:
"""深度优先展开 InlineSpan,保留 HyperlinkSpan 本身及其子 Span。"""
if not isinstance(spans, list):
return []
output: list[Any] = []
for span in spans:
output.append(span)
content = span.get("content") if isinstance(span, dict) else getattr(span, "content", None)
span_type = span.get("type") if isinstance(span, dict) else getattr(span, "type", None)
if str(span_type) == "hyperlink":
output.extend(inline_items(content))
return output
def inline_urls(spans: Any) -> list[str]:
"""提取 InlineSpan 序列中的全部超链接目标。"""
urls: list[str] = []
for span in inline_items(spans):
span_type = span.get("type") if isinstance(span, dict) else getattr(span, "type", None)
url = span.get("url") if isinstance(span, dict) else getattr(span, "url", None)
if str(span_type) == "hyperlink" and isinstance(url, str):
urls.append(url)
return urls
def visible_content(content: Any) -> str:
"""统一提取专用字符串内容或 InlineSpan 内容的可见文字。"""
return content if isinstance(content, str) else inline_text(content)