claims-cli 0.6.0__tar.gz

claims_cli-0.6.0/.gitignore

@@ -0,0 +1,51 @@
+__pycache__/
+*.pyc
+.venv/
+dist/
+*.egg-info/
+.DS_Store
+
+# adduce writes its report and the repro sidecar here
+.adduce/
+
+# Publisher PDFs for the prior-art dig. Kept locally so `citations verify` can resolve
+# quotations against them; not redistributable, and this repository is public.
+
+# coverage byproducts
+.coverage
+.coverage.*
+htmlcov/
+coverage.xml
+
+# Internal review packets: simulated-referee rounds and self-critique.
+# Local only -- this repository may go public.
+REVIEW_PACKET*.md
+
+# Development corpus: the downloaded articles and shallow clones are rebuildable from
+# frame.json, but the measurements taken from them are not -- ignoring the whole directory
+# lost one 32-article run when the tree was tidied. Subdirectories only, as arxiv_corpus does.
+# documentation site (Docusaurus)
+website/node_modules/
+website/build/
+website/.docusaurus/
+
+# nbconvert output from verifying the demo notebook runs
+notebooks/executed.ipynb
+
+# jupyter lite build state
+.jupyterlite.doit.db
+
+# arXiv LaTeX sources fetched for testing, rebuildable from resolved.json
+
+# CORE-Bench capsules: ~39 MB each, rebuildable from core_train.json
+
+# CORE-Bench capsule metadata: 4.3 MB, fetched from siegelz/core-bench
+
+# Internal reviews are never tracked. They live in reproducible-science-internal-audits.
+docs/reviews/
+
+# Sidecars from provenance_core.locking. The suffixes are deliberately not `.lock` and `.tmp`,
+# so these patterns cannot reach `uv.lock`. They hold nothing: the kernel releases a lock when its
+# holder exits, and a leftover temp file means a write was killed.
+*.provenance-lock
+*.provenance-tmp

claims_cli-0.6.0/PKG-INFO

@@ -0,0 +1,139 @@
+Metadata-Version: 2.5
+Name: claims-cli
+Version: 0.6.0
+Summary: Check what a manuscript states against the files it names: numbers, table cells and quotations
+Project-URL: Changelog, https://github.com/elliottower/reproducible-science/blob/main/CHANGELOG.md
+Project-URL: Homepage, https://github.com/elliottower/reproducible-science
+Project-URL: Issues, https://github.com/elliottower/reproducible-science/issues
+Project-URL: Source, https://github.com/elliottower/reproducible-science/tree/main/packages/claims
+Author-email: Elliot Tower <elliot@elliottower.ai>
+License-Expression: MIT
+Keywords: citations,manuscript,open science,provenance,reproducibility
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
+Classifier: Topic :: Scientific/Engineering
+Requires-Python: >=3.11
+Requires-Dist: citations<0.7,>=0.6
+Requires-Dist: provenance-core<0.7,>=0.6
+Requires-Dist: pydantic>=2.9
+Requires-Dist: pyyaml>=6.0.2
+Provides-Extra: arrays
+Requires-Dist: numpy>=2.2.6; extra == 'arrays'
+Provides-Extra: hdf5
+Requires-Dist: h5py>=3.12; extra == 'hdf5'
+Provides-Extra: netcdf
+Requires-Dist: netcdf4>=1.7.2; extra == 'netcdf'
+Provides-Extra: parquet
+Requires-Dist: pyarrow>=18; extra == 'parquet'
+Provides-Extra: rds
+Requires-Dist: rdata>=1; extra == 'rds'
+Provides-Extra: sheets
+Requires-Dist: openpyxl>=3.1; extra == 'sheets'
+Requires-Dist: xlrd>=2.0.1; extra == 'sheets'
+Provides-Extra: stata-spss
+Requires-Dist: pyreadstat>=1.3.6; extra == 'stata-spss'
+Description-Content-Type: text/markdown
+
+# claims-cli
+
+[![pypi](https://img.shields.io/pypi/v/claims-cli)](https://pypi.org/project/claims-cli/)
+[![python](https://img.shields.io/pypi/pyversions/claims-cli)](https://pypi.org/project/claims-cli/)
+[![license](https://img.shields.io/pypi/l/claims-cli)](https://github.com/elliottower/reproducible-science/blob/main/LICENSE)
+[![docs](https://img.shields.io/badge/docs-live-blue)](https://reproducible.science/)
+
+Check what a manuscript states against the files it names.
+
+The part of [reproducible-science](https://github.com/elliottower/reproducible-science) that reads a manifest of claims and checks each one: a number against a results file, a table cell against a table, a quotation against its source. It installs alone, and it runs no command of yours.
+
+## Install
+
+```bash
+pip install claims-cli
+```
+
+This installs `claims` and one dependency of ours, [`citations`](https://pypi.org/project/citations/), which reads the documents a quotation is checked in. The optional readers are extras, each named for what it reads:
+
+```bash
+pip install "claims-cli[parquet,sheets]"
+```
+
+`arrays`, `hdf5`, `netcdf`, `parquet`, `rds`, `sheets` and `stata-spss`. Without the extra a format needs, a check of that format is `unchecked` and names the extra to install.
+
+## What it reads
+
+`repro.yaml`, a manifest that pins a project's files by sha256 and declares what the manuscript claims about them:
+
+```yaml
+artifacts:
+  - id: results
+    path: analysis/results.json
+    digest: {algorithm: sha256, value: 9f635f9a...}
+
+claims:
+  - id: primary
+    text: The treatment reduced the primary outcome
+    evidence:
+      - kind: metric
+        artifact: results
+        name: p_value
+        pointer: /primary/p
+        reported: "0.031"
+```
+
+The file has the same name and the same format whichever command reads it. The [specification](https://reproducible.science/reference/spec/) defines every kind of evidence and every outcome.
+
+## The commands
+
+```bash
+claims manifest init
+```
+
+Writes a starter `repro.yaml` at the top of the project: each file it finds or is given, pinned, with example claims that verify as written. It never overwrites one.
+
+```bash
+claims verify
+```
+
+Checks every declared assertion against the file it names and prints one line per assertion: `ok`, `MISS` where the file holds another value, `GONE` where the address does not resolve, `--` where the check could not run. It exits 1 where the policy fails, and `--format json` and `--format sarif` print the report as a document.
+
+```bash
+claims stale
+```
+
+Says which claims read a file that is no longer what a recorded re-run produced. It reads `.repro/reproductions.jsonl` and runs nothing.
+
+```bash
+claims why primary
+```
+
+For one claim: each assertion as `claims verify` evaluates it now, each file the claim rests on and the command recorded for it, and then what the record does not establish.
+
+```bash
+claims numbers
+```
+
+Counts every number the manuscript's text prints and says which a claim accounts for.
+
+```bash
+claims export record
+```
+
+Writes one record per checked assertion, for another program to read. `sarif` and `ro-crate` write the same records in those formats.
+
+```bash
+claims origins --repo https://github.com/owner/study=../study
+```
+
+For each artifact copied from another repository, which the manifest says with `origin: {repository, commit, path}`, compares the pinned digest with the bytes of that path at that commit: `same`, `differs`, `absent`, `unknown commit` or `not checked`, each with the reason. A clone given with `--repo` is read with `git cat-file`, so committed bytes are compared and never a working tree. With `--fetch`, a GitHub repository no clone is given for is read from `raw.githubusercontent.com`; without it nothing touches the network. `claims verify` never reads an origin.
+
+## With `repro`
+
+[`reproducible-science`](https://pypi.org/project/reproducible-science/) depends on this package and adds the commands that run things: `repro reproduce`, which runs each declared command again and writes the record `claims stale` reads, `repro check` and `repro audit`, which drive `prereg`, `results` and `citations`, and `repro ci`. Under `repro`, every command above has the same name, does the same thing and prints the same output: `repro verify` and `claims verify` are one function. A project that uses `repro` has no reason to install this by itself. It is separate so that checking a manuscript's claims does not require installing something that executes commands a manifest names.
+
+MIT licensed.

claims_cli-0.6.0/README.md

@@ -0,0 +1,97 @@
+# claims-cli
+
+[![pypi](https://img.shields.io/pypi/v/claims-cli)](https://pypi.org/project/claims-cli/)
+[![python](https://img.shields.io/pypi/pyversions/claims-cli)](https://pypi.org/project/claims-cli/)
+[![license](https://img.shields.io/pypi/l/claims-cli)](https://github.com/elliottower/reproducible-science/blob/main/LICENSE)
+[![docs](https://img.shields.io/badge/docs-live-blue)](https://reproducible.science/)
+
+Check what a manuscript states against the files it names.
+
+The part of [reproducible-science](https://github.com/elliottower/reproducible-science) that reads a manifest of claims and checks each one: a number against a results file, a table cell against a table, a quotation against its source. It installs alone, and it runs no command of yours.
+
+## Install
+
+```bash
+pip install claims-cli
+```
+
+This installs `claims` and one dependency of ours, [`citations`](https://pypi.org/project/citations/), which reads the documents a quotation is checked in. The optional readers are extras, each named for what it reads:
+
+```bash
+pip install "claims-cli[parquet,sheets]"
+```
+
+`arrays`, `hdf5`, `netcdf`, `parquet`, `rds`, `sheets` and `stata-spss`. Without the extra a format needs, a check of that format is `unchecked` and names the extra to install.
+
+## What it reads
+
+`repro.yaml`, a manifest that pins a project's files by sha256 and declares what the manuscript claims about them:
+
+```yaml
+artifacts:
+  - id: results
+    path: analysis/results.json
+    digest: {algorithm: sha256, value: 9f635f9a...}
+
+claims:
+  - id: primary
+    text: The treatment reduced the primary outcome
+    evidence:
+      - kind: metric
+        artifact: results
+        name: p_value
+        pointer: /primary/p
+        reported: "0.031"
+```
+
+The file has the same name and the same format whichever command reads it. The [specification](https://reproducible.science/reference/spec/) defines every kind of evidence and every outcome.
+
+## The commands
+
+```bash
+claims manifest init
+```
+
+Writes a starter `repro.yaml` at the top of the project: each file it finds or is given, pinned, with example claims that verify as written. It never overwrites one.
+
+```bash
+claims verify
+```
+
+Checks every declared assertion against the file it names and prints one line per assertion: `ok`, `MISS` where the file holds another value, `GONE` where the address does not resolve, `--` where the check could not run. It exits 1 where the policy fails, and `--format json` and `--format sarif` print the report as a document.
+
+```bash
+claims stale
+```
+
+Says which claims read a file that is no longer what a recorded re-run produced. It reads `.repro/reproductions.jsonl` and runs nothing.
+
+```bash
+claims why primary
+```
+
+For one claim: each assertion as `claims verify` evaluates it now, each file the claim rests on and the command recorded for it, and then what the record does not establish.
+
+```bash
+claims numbers
+```
+
+Counts every number the manuscript's text prints and says which a claim accounts for.
+
+```bash
+claims export record
+```
+
+Writes one record per checked assertion, for another program to read. `sarif` and `ro-crate` write the same records in those formats.
+
+```bash
+claims origins --repo https://github.com/owner/study=../study
+```
+
+For each artifact copied from another repository, which the manifest says with `origin: {repository, commit, path}`, compares the pinned digest with the bytes of that path at that commit: `same`, `differs`, `absent`, `unknown commit` or `not checked`, each with the reason. A clone given with `--repo` is read with `git cat-file`, so committed bytes are compared and never a working tree. With `--fetch`, a GitHub repository no clone is given for is read from `raw.githubusercontent.com`; without it nothing touches the network. `claims verify` never reads an origin.
+
+## With `repro`
+
+[`reproducible-science`](https://pypi.org/project/reproducible-science/) depends on this package and adds the commands that run things: `repro reproduce`, which runs each declared command again and writes the record `claims stale` reads, `repro check` and `repro audit`, which drive `prereg`, `results` and `citations`, and `repro ci`. Under `repro`, every command above has the same name, does the same thing and prints the same output: `repro verify` and `claims verify` are one function. A project that uses `repro` has no reason to install this by itself. It is separate so that checking a manuscript's claims does not require installing something that executes commands a manifest names.
+
+MIT licensed.

claims_cli-0.6.0/pyproject.toml

@@ -0,0 +1,78 @@
+[build-system]
+build-backend = "hatchling.build"
+requires = [ "hatchling" ]
+
+[project]
+name = "claims-cli"
+version = "0.6.0"
+description = "Check what a manuscript states against the files it names: numbers, table cells and quotations"
+readme = "README.md"
+keywords = [ "citations", "manuscript", "open science", "provenance", "reproducibility" ]
+license = "MIT"
+authors = [ { name = "Elliot Tower", email = "elliot@elliottower.ai" } ]
+requires-python = ">=3.11"
+classifiers = [
+  "Development Status :: 3 - Alpha",
+  "Intended Audience :: Science/Research",
+  "License :: OSI Approved :: MIT License",
+  "Programming Language :: Python :: 3 :: Only",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Programming Language :: Python :: 3.14",
+  "Topic :: Scientific/Engineering",
+]
+dependencies = [
+  # A quotation is checked by `citations`, and a prose locator and the count of a text's
+  # numbers read a document through its extractors, so it is needed to import this package.
+  "citations>=0.6,<0.7",
+  "provenance-core>=0.6,<0.7",
+  # Floors, not bare names: without them a fresh install resolves whatever is oldest on
+  # PyPI, which is not the combination the tests exercise.
+  "pydantic>=2.9",
+  "pyyaml>=6.0.2",
+]
+optional-dependencies.arrays = [
+  "numpy>=2.2.6",
+]
+# Readers for the formats results arrive in besides delimited text and JSON. Each is optional,
+# because most manifests address a CSV or a JSON file and a Parquet or HDF5 library is not
+# something every install should carry. Without one, a check of that format is `unchecked` and
+# names the extra; it never fails. All are MIT, BSD or Apache-2.0. Not `pyreadr`, which is AGPL
+# and would relicense this package; `rdata` parses `.rds` in Python and starts no R process.
+optional-dependencies.hdf5 = [ "h5py>=3.12" ]
+optional-dependencies.netcdf = [ "netcdf4>=1.7.2" ]
+optional-dependencies.parquet = [ "pyarrow>=18" ]
+optional-dependencies.rds = [ "rdata>=1" ]
+optional-dependencies.sheets = [ "openpyxl>=3.1", "xlrd>=2.0.1" ]
+optional-dependencies.stata-spss = [ "pyreadstat>=1.3.6" ]
+urls.Changelog = "https://github.com/elliottower/reproducible-science/blob/main/CHANGELOG.md"
+urls.Homepage = "https://github.com/elliottower/reproducible-science"
+urls.Issues = "https://github.com/elliottower/reproducible-science/issues"
+urls.Source = "https://github.com/elliottower/reproducible-science/tree/main/packages/claims"
+scripts.claims = "claims.cli:main"
+
+[dependency-groups]
+dev = [
+  "numpy>=2.2.6",
+  "pytest>=9.1.1",
+]
+
+[tool.hatch]
+build.targets.wheel.packages = [ "src/claims" ]
+
+[tool.uv]
+# Resolved from the workspace during development and from PyPI once published, so a
+# cross-package change is one commit rather than a release dance.
+sources.citations = { workspace = true }
+
+[tool.deptry]
+# Without this every intra-package import reads as a transitive dependency on itself.
+known_first_party = [ "claims" ]
+
+[tool.pytest]
+ini_options.addopts = "-m 'not corpus'"
+ini_options.markers = [
+  "corpus: reads repositories outside this one, and may fetch them",
+  "integration: exercises a third-party library that is not a runtime dependency",
+]

claims_cli-0.6.0/src/claims/__init__.py

@@ -0,0 +1,132 @@
+"""One contract for verifying declared evidence assertions against pinned artifacts.
+
+A quotation in a manuscript and a number in a results file are the same kind of object: a
+typed value asserted to be extractable from a pinned artifact. This package verifies that
+assertion -- and nothing beyond it. Whether the evidence *supports* the sentence citing it is
+a question about entailment, addressed elsewhere with model-based methods.
+
+    from claims import load, verify
+    from claims.policy import PUBLICATION
+
+    report = verify(load("repro.yaml"))       # facts
+    verdict = PUBLICATION.assess(report)      # judgment
+
+Facts and judgment are separate calls because a missing optional citation is acceptable in a
+draft and not in a submission, and one project's standard should not be everyone's.
+
+Nothing here prints, exits, or reads global state.
+"""
+
+from importlib.metadata import PackageNotFoundError
+from importlib.metadata import version as _version
+
+from claims.exceptions import (
+    ArtifactMissingError,
+    ArtifactUnreadableError,
+    BackendUnavailableError,
+    DigestMismatchError,
+    ManifestError,
+    ReproError,
+    UnknownEvidenceKindError,
+)
+from claims.manifest import DEFAULT_NAME, find, load
+from claims.models import (
+    SCHEMA_VERSION,
+    ArtifactRef,
+    ArtifactState,
+    Availability,
+    Claim,
+    ClaimAssessment,
+    ComparisonMode,
+    ComparisonStatus,
+    Decision,
+    Digest,
+    Evidence,
+    ExecutionStatus,
+    ExtractionStatus,
+    Manifest,
+    MetricEvidence,
+    Outcome,
+    QuoteEvidence,
+    Reason,
+    TableCellEvidence,
+    Validity,
+    VerificationReport,
+    Warning_,
+)
+from claims.policy import EXPLORATORY, PROFILES, PUBLICATION, STRICT, Assessment, Policy, Severity
+from claims.renderers import to_sarif
+from claims.resolve import Resolution, read_table, resolve, resolve_pointer, sniff_delimiter
+from claims.toolchain import ALONE
+from claims.verify import (
+    DEFAULT_BACKENDS,
+    Backend,
+    MetricBackend,
+    QuoteBackend,
+    TableBackend,
+    ValueBackend,
+    verify,
+)
+
+# This package's own version, `claims-cli`'s, whatever else is installed: what `claims
+# --version` prints and what a SARIF log names as its driver's version.
+try:
+    __version__ = _version(ALONE)
+except PackageNotFoundError:  # a source tree with nothing installed
+    __version__ = "0+unknown"
+
+__all__ = [
+    "DEFAULT_BACKENDS",
+    "DEFAULT_NAME",
+    "EXPLORATORY",
+    "PROFILES",
+    "PUBLICATION",
+    "SCHEMA_VERSION",
+    "STRICT",
+    "ArtifactMissingError",
+    "ArtifactRef",
+    "ArtifactState",
+    "ArtifactUnreadableError",
+    "Assessment",
+    "Availability",
+    "Backend",
+    "BackendUnavailableError",
+    "Claim",
+    "ClaimAssessment",
+    "ComparisonMode",
+    "ComparisonStatus",
+    "Decision",
+    "Digest",
+    "DigestMismatchError",
+    "Evidence",
+    "ExecutionStatus",
+    "ExtractionStatus",
+    "Manifest",
+    "ManifestError",
+    "MetricBackend",
+    "MetricEvidence",
+    "Outcome",
+    "Policy",
+    "QuoteBackend",
+    "QuoteEvidence",
+    "Reason",
+    "ReproError",
+    "Resolution",
+    "Severity",
+    "TableBackend",
+    "TableCellEvidence",
+    "UnknownEvidenceKindError",
+    "Validity",
+    "ValueBackend",
+    "VerificationReport",
+    "Warning_",
+    "__version__",
+    "find",
+    "load",
+    "read_table",
+    "resolve",
+    "resolve_pointer",
+    "sniff_delimiter",
+    "to_sarif",
+    "verify",
+]

claims_cli-0.6.0/src/claims/__main__.py

@@ -0,0 +1,5 @@
+"""`python -m claims`, the same as the `claims` command."""
+
+from claims.cli import main
+
+raise SystemExit(main())

claims_cli-0.6.0/src/claims/adapters/__init__.py

@@ -0,0 +1,6 @@
+"""One module per artifact format.
+
+Each adapter translates a locator into the addressing its format already has, and every
+one returns the same `Found` contract. Adding a format is a new module and a registry
+entry, not an edit to a switchboard.
+"""