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
"""Primitive output functions for MinerU CLI runtimes."""
from __future__ import annotations
import json
import sys
from dataclasses import asdict, is_dataclass
from typing import Any
from pydantic import BaseModel
from pydantic_core import to_json
from rich.console import Console
from rich.text import Text
from ..utils.i18n import t
from .contracts import RenderableObject
console = Console(markup=False)
stderr_console = Console(stderr=True, markup=False)
def print_rich(*objects: RenderableObject) -> None:
console.print(*objects)
def print_text(value: object) -> None:
sys.stdout.write(f"{value}\n")
def print_success(msg: str) -> None:
if console:
console.print(msg, style="green")
else:
print_text(msg)
def print_info(msg: str) -> None:
if console:
console.print(msg, style="dim")
else:
print_text(msg)
def print_notice(msg: str) -> None:
stderr_console.print(msg, style="dim")
def print_error(msg: str) -> None:
text = Text(t("Error:"), style="red")
text.append(f" {msg}")
stderr_console.print(text)
def print_json(data: Any) -> None:
if isinstance(data, BaseModel):
print_text(to_json(data, indent=2).decode("utf-8"))
return
if is_dataclass(data) and not isinstance(data, type):
print_text(json.dumps(asdict(data), ensure_ascii=False, indent=2))
return
print_text(json.dumps(data, ensure_ascii=False, indent=2))