Phase 1: foundation, bug-fix upstream, source router (deepXIV + biomed)

.github/workflows/test.yml

@@ -0,0 +1,27 @@
+name: test
+on:
+  pull_request:
+  push:
+    branches: [main]
+permissions:
+  contents: read
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+jobs:
+  pytest:
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+    strategy:
+      fail-fast: false
+      matrix:
+        python: ["3.11", "3.12"]
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python }}
+          cache: pip
+          cache-dependency-path: pyproject.toml
+      - run: pip install -e ".[dev]"
+      - run: pytest -q

.gitignore

@@ -3,6 +3,9 @@ __pycache__/
 *.pyc
 *.pyo
 *.pyd
+*.egg-info/
+build/
+dist/
 
 # macOS
 .DS_Store
@@ -20,3 +23,4 @@ __pycache__/
 # Local runtime environment
 /.venv311/
 /.runtime/
+.venv/

README.md

@@ -132,6 +132,25 @@ Core artifacts:
 Global index:
 - `reports/run_index.tsv`
 
+## Sources & backends
+
+OpenRevise's source router can be configured to retrieve evidence from multiple categories. Pipeline code dispatches by `source_type` to the registered backend, with optional dependencies allowing graceful degradation.
+
+- **Preprints (arXiv / medRxiv / bioRxiv):** retrieval is provided via [DeepXiv-SDK](https://github.com/DeepXiv/deepxiv_sdk) (optional, install with `pip install "openrevise[preprint-deepxiv]"`) or a built-in arXiv API client (`pip install "openrevise[preprint-arxiv]"`). If neither is installed, preprint retrieval is disabled and pipelines using non-preprint sources continue to work.
+- **Biomedical literature (PubMed / PMC / Europe PMC):** built-in clients using NCBI E-utilities and the EuropePMC REST API. No optional dependencies required beyond `requests` (a base dep).
+- **Scholarly indexes (Semantic Scholar / OpenAlex):** retained from the existing `ideaclaw.sources.scholar` module.
+- **Regulatory & guidelines (FDA / EMA / NCCN / ESMO / ASCO / CSCO):** generic HTTP fetchers driven by `config/source_registry.yaml`.
+- **Local files (PDF / DOCX / PPTX / images):** evidence extraction via `openrevise.sources.evidence_extractors`.
+
+## Quick install
+
+```bash
+pip install openrevise                          # core (no preprint search)
+pip install "openrevise[preprint-arxiv]"       # + native arXiv client
+pip install "openrevise[preprint-deepxiv]"     # + DeepXiv-SDK preprint backend
+pip install "openrevise[all]"                  # everything
+```
+
 ## Repository Structure
 | Path | Purpose |
 |---|---|

config/revision_patch_spec_template.json

@@ -1,4 +1,9 @@
 {
+  "meta": {
+    "notes": [
+      "Demo: when replacement contains subgroup label tuples (e.g. ITT/mITT), pipeline label-value gate checks prefix/value binding before revise."
+    ]
+  },
   "footnote_sources": {
     "src_reg_notice": "Source: Regulatory authority notice URL ... verified on YYYY-MM-DD.",
     "src_journal": "Source: Journal paper citation ... verified on YYYY-MM-DD."

config/source_registry.yaml

@@ -235,3 +235,24 @@ sources:
   alias_keywords:
   - sun yat-sen
   - sysu
+- source_id: arxiv
+  source_type: preprint
+  tier: B
+  country: INTL
+  institution_tier: ''
+  reliability_rule: preprint_unreviewed
+  alias_keywords: [arxiv]
+- source_id: medrxiv
+  source_type: preprint
+  tier: B
+  country: INTL
+  institution_tier: ''
+  reliability_rule: preprint_unreviewed
+  alias_keywords: [medrxiv]
+- source_id: biorxiv
+  source_type: preprint
+  tier: B
+  country: INTL
+  institution_tier: ''
+  reliability_rule: preprint_unreviewed
+  alias_keywords: [biorxiv]

pyproject.toml

@@ -0,0 +1,54 @@
+[build-system]
+requires = ["setuptools>=68", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "openrevise"
+version = "0.1.0.dev0"
+description = "Evidence-gated revision infrastructure for high-stakes documents"
+readme = "README.md"
+license = { text = "Apache-2.0" }
+requires-python = ">=3.11"
+authors = [{ name = "StartripAI" }]
+classifiers = [
+  "Development Status :: 3 - Alpha",
+  "License :: OSI Approved :: Apache Software License",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+]
+dependencies = [
+  "lxml>=4.9",
+  "PyYAML>=6.0",
+  "requests>=2.31",
+]
+
+[project.optional-dependencies]
+preprint-deepxiv = ["deepxiv-sdk>=0.1"]
+preprint-arxiv = ["sentence-transformers>=2.2", "feedparser>=6.0"]
+biomed = ["biopython>=1.83"]
+mcp = ["mcp>=0.9"]
+docx = ["python-docx>=1.1"]
+pdf = ["pypdf>=4.0", "pdfminer.six>=20231228"]
+xlsx = ["openpyxl>=3.1"]
+all = ["openrevise[preprint-deepxiv,preprint-arxiv,biomed,mcp,docx,pdf,xlsx]"]
+dev = ["pytest>=8.0", "pytest-cov>=4.1", "ruff>=0.5", "requests-mock>=1.11"]
+
+[project.scripts]
+openrevise = "openrevise.cli:main"
+openrevise-mcp = "openrevise.mcp.server:main"
+
+[project.urls]
+Homepage = "https://github.com/StartripAI/OpenRevise"
+Issues = "https://github.com/StartripAI/OpenRevise/issues"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+include = ["openrevise*"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+addopts = "-q --strict-markers"
+
+[tool.ruff]
+line-length = 100
+target-version = "py311"

src/openrevise/__init__.py

@@ -0,0 +1,10 @@
+"""OpenRevise: Evidence-gated revision infrastructure for high-stakes documents."""
+
+from importlib.metadata import PackageNotFoundError, version
+
+try:
+    __version__ = version("openrevise")
+except PackageNotFoundError:  # pragma: no cover - only hit when package isn't installed
+    __version__ = "0.0.0+unknown"
+
+__all__ = ["__version__"]

scripts/build_q_source_map.py → src/openrevise/artifacts/build_q_source_map.py

@@ -14,7 +14,7 @@
 import xml.etree.ElementTree as ET
 from pathlib import Path
 from typing import Dict, List, Tuple
-from run_artifact_utils import is_valid_run_id
+from openrevise.artifacts.run_artifact_utils import is_valid_run_id
 
 W = "{http://schemas.openxmlformats.org/wordprocessingml/2006/main}"
 QUESTION_PREFIX_RE = re.compile(

scripts/update_run_index.py → src/openrevise/artifacts/update_run_index.py

@@ -11,7 +11,7 @@
 from pathlib import Path
 from typing import Dict, List
 
-from run_artifact_utils import read_tsv, write_tsv
+from openrevise.artifacts.run_artifact_utils import read_tsv, write_tsv
 
 
 RUN_INDEX_FIELDS: List[str] = [
@@ -85,7 +85,7 @@ def main() -> int:
     parser.add_argument(
         "--index",
         type=Path,
-        default=Path(__file__).resolve().parents[1] / "reports" / "run_index.tsv",
+        default=Path(__file__).resolve().parents[3] / "reports" / "run_index.tsv",
     )
     parser.add_argument("--marker", required=True)
     parser.add_argument("--run-id", required=True)