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
chore: add .zcodeignore for build artifacts and temporary files
myhloli1aa76acf1cdc
diff --git a/.zcodeignore b/.zcodeignorenew file mode 100644index 00000000..466c7449--- /dev/null+++ b/.zcodeignore@@ -0,0 +1,100 @@+*.tar+*.tar.gz+*.zip+venv*/+envs/+slurm_logs/++sync1.sh+data_preprocess_pj1+data-preparation1+__pycache__+*.log+*.pyc+.vscode+debug/+*.ipynb+.idea++# Python build artifacts+*.egg-info/+build/+# PDF.js 的浏览器发行资源必须随源码提交。+!mineru/resources/pdf_preview/vendor/pdfjs/legacy/build/+mineru/resources/pdf_preview/vendor/pdfjs/legacy/build/*+!mineru/resources/pdf_preview/vendor/pdfjs/legacy/build/pdf.min.mjs+!mineru/resources/pdf_preview/vendor/pdfjs/legacy/build/pdf.worker.min.mjs+dist/+*.egg++# vscode history+.history++.DS_Store+.env++bad_words/+bak/++app/tests/*+temp/+tmp/+tmp+.vscode+.vscode/+ocr_demo+.coveragerc+/app/common/__init__.py+/magic_pdf/config/__init__.py+source.dev.env++tmp++projects/web/node_modules+projects/web/dist++projects/web_demo/web_demo/static/+cli_debug/+debug_utils/++# sphinx docs+_build/+++output/+.coverage++# E2E 测试工作目录(.e2e/fixtures-cache 跨 session 复用)+.e2e/++htmlcov/+.zcode/++# ===== ↑ 以上同步自 .gitignore(「从 .gitignore 同步」只重写以上部分)=====+.git/+.hg/+.svn/+node_modules/+bower_components/+jspm_packages/+site-packages/+venv/+coverage/+lcov-report/+cmakefiles/+cmake-build-*/+bazel-*/+pods/+deriveddata/+storybook-static/+playwright-report/+test-results/+allure-results/+allure-report/+cdk.out/+*.dist-info/+eggs/+pip-wheel-metadata/+wheels/+# ----- ↑ 以上为 ZCode 默认排除规则(自定义规则请写在本行下方,不会被同步/恢复改动)-----+# 自定义规则写在下方(本行提示可删除)diff --git a/pyproject.toml b/pyproject.tomlindex bb63c8a6..f17679ef 100644--- a/pyproject.toml+++ b/pyproject.toml@@ -39,7 +39,7 @@ classifiers = ["Programming Language :: Python :: 3.14",]dependencies = [- "docvortex>=0.5.3,<1",+ "docvortex>=0.5.4,<1","gradio>=6.8,<7","mineru[torch] ; sys_platform == 'darwin' and platform_machine == 'arm64'","mineru-vl-utils>=2.0.5,<3",diff --git a/tests/unittest/test_docvortex_import_boundary.py b/tests/unittest/test_docvortex_import_boundary.pyindex 8a680437..ad962051 100644--- a/tests/unittest/test_docvortex_import_boundary.py+++ b/tests/unittest/test_docvortex_import_boundary.py@@ -6,12 +6,9 @@ import astfrom pathlib import Pathimport pytest-from docvortex.document.mhtml import __all__ as MHTML_EXPORTSfrom docvortex.public_api import PUBLIC_API-# DocVortex 0.4.24 已通过模块 __all__ 导出归档接口,公开清单将在后续版本同步。_DECLARED_API = dict(PUBLIC_API)-_DECLARED_API.setdefault("docvortex.document.mhtml", tuple(MHTML_EXPORTS))def _dotted_name(node: ast.AST) -> str:diff --git a/tests/unittest/test_pdf_tight_spacing.py b/tests/unittest/test_pdf_tight_spacing.pyindex a4b9d5f3..8df9c24c 100644--- a/tests/unittest/test_pdf_tight_spacing.py+++ b/tests/unittest/test_pdf_tight_spacing.py@@ -1,12 +1,9 @@-"""验证普通、旋转及 Rust 原生回填共用非 CJK 墨迹词界规则。"""+"""验证普通文本、代码保护与整页/竖排回填共用非 CJK 墨迹词界规则。"""-from copy import deepcopy-from pathlib import Pathfrom typing import Anyimport pytest-from docvortex.document.pdf import PDFDocument-from docvortex.document.pdf.text._contracts import Bbox+from docvortex.document.pdf import Bboxfrom mineru.backend.analysis.pdf.text import nativefrom mineru.backend.analysis.pdf.text.models import _AnalyzeSpanfrom mineru.types import ContentType@@ -106,39 +103,3 @@ def test_vertical_line_fill_uses_tight_spacing(monkeypatch: pytest.MonkeyPatch)detect_scripts=False,)assert target.content == "A B"--[email protected]("enabled", [True, False])-def test_real_snapshot_content_matches_python(enabled: bool) -> None:- """真实论文逐字符验证 Rust 物化与 Python 回退一致,且确实执行 Rust 内容快路径。"""- from docvortex._compute_backend import get_native-- extension = get_native()- if extension is None:- pytest.skip("requires native extension")- path = Path(__file__).resolve().parents[3] / "docvortex/demo/pdfs/中文论文2.pdf"- with PDFDocument(path.read_bytes()) as document:- owner = document[0].get_text_snapshot()- width, height = document[0].size- geometry = owner.materialize_geometry()- results = []- before = extension.text_snapshot_stats()[2]- for current_owner in (None, owner):- span = _AnalyzeSpan(- ContentType.TEXT,- (0.0, 0.0, width, height),- metadata={"chars": [], "height": height, "width": width, "_native_tight_spacing": enabled},- )- native.fill_char_in_spans(- [span],- deepcopy(geometry.chars),- 10.0,- tight_bboxes=geometry.tight_bboxes,- origins=geometry.origins,- detect_scripts=False,- _native_text=current_owner,- )- results.append(span.content)- assert extension.text_snapshot_stats()[2] == before + 1- assert results[0] == results[1]- assert ("Large Language Models" in results[0]) is enabled