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
"""Tasks 专用产物策略:每个输入只发布一个完整的可下载文件。"""
from __future__ import annotations
import json
import zipfile
from pathlib import Path
from ..task_types import TASK_OUTPUT_FORMATS, TaskOutputFormat
from .base import ParseResult
def build_task_output(
result: ParseResult,
formats: list[TaskOutputFormat],
directory: Path,
) -> Path:
"""一次物化并渲染所选格式,全部成功后才返回主产物路径。"""
from docvortex.export import embed_html_images, materialize_middle, validate_materialized_assets
from ..render import render_docx, render_html, render_latex
middle, assets = materialize_middle(result.middle_json)
validate_materialized_assets(middle, assets)
exported = ParseResult(middle_json=middle)
entries: dict[str, bytes] = {}
if "default" in formats:
entries.update(
{
"markdown.md": exported.markdown().encode("utf-8", errors="replace"),
"middle_json.json": exported.to_json().encode("utf-8", errors="replace"),
"structured_content.json": json.dumps(exported.structured_content(), ensure_ascii=False, indent=2).encode(
"utf-8", errors="replace"
),
}
)
if "docx" in formats:
entries["document.docx"] = render_docx(middle, asset_resolver=assets.__getitem__)
if "html" in formats:
entries["document.html"] = embed_html_images(render_html(middle), assets).encode("utf-8", errors="replace")
if "latex" in formats:
entries["document.tex"] = render_latex(middle).encode("utf-8", errors="replace")
directory.mkdir(parents=True, exist_ok=True)
if len(formats) == 1 and formats[0] in {"docx", "html"}:
name, content = next(iter(entries.items()))
target = directory / name
target.write_bytes(content)
return target
target = directory / "result.zip"
with zipfile.ZipFile(target, "w", zipfile.ZIP_DEFLATED, allowZip64=True) as archive:
for name, content in entries.items():
archive.writestr(name, content)
if "default" in formats or "latex" in formats:
archive.writestr("images/", b"")
for name, content in sorted(assets.items()):
archive.writestr(name, content)
return target
__all__ = ["TASK_OUTPUT_FORMATS", "TaskOutputFormat", "build_task_output"]