paratext-cli 0.1.0__tar.gz

paratext_cli-0.1.0/.github/workflows/ci.yml

@@ -0,0 +1,20 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Install uv
+        uses: astral-sh/setup-uv@v5
+      - name: Sync (dev extra)
+        run: uv sync --extra dev
+      - name: Lint
+        run: uv run ruff check
+      - name: Test
+        run: uv run pytest -q

paratext_cli-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,65 @@
+name: Publish
+
+# Publishes paratext-cli to PyPI via Trusted Publishing (OIDC) — no API token is
+# stored anywhere. PyPI must have a publisher registered for this repo and this
+# workflow filename; see docs/publishing.md.
+#
+# Triggered by publishing a GitHub Release, so a publish is always a deliberate
+# act with a changelog attached. A PyPI version can never be reused, so the job
+# checks the tag against pyproject.toml and runs the tests before uploading.
+
+on:
+  release:
+    types: [published]
+  workflow_dispatch:
+    inputs:
+      dry-run:
+        description: "Build and check, but don't upload"
+        type: boolean
+        default: true
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    environment: pypi
+    permissions:
+      # Mints the short-lived OIDC token uv exchanges for a PyPI credential.
+      # Nothing else here needs write access.
+      id-token: write
+    steps:
+      - uses: actions/checkout@v4
+      - name: Install uv
+        uses: astral-sh/setup-uv@v5
+
+      - name: Check the tag matches the packaged version
+        if: github.event_name == 'release'
+        run: |
+          tag="${GITHUB_REF_NAME#v}"
+          pkg=$(uv version --short)
+          if [ "$tag" != "$pkg" ]; then
+            echo "::error::release tag $tag != pyproject version $pkg." \
+                 "Bump the version or retag — a published version can't be replaced."
+            exit 1
+          fi
+          echo "publishing version $pkg"
+
+      # The release could be cut from any commit; CI having passed on main is not
+      # the same as it passing here. Cheap insurance against an unfixable upload.
+      - name: Sync (dev extra)
+        run: uv sync --extra dev
+      - name: Lint
+        run: uv run ruff check
+      - name: Test
+        run: uv run pytest -q
+
+      - name: Build
+        run: uv build
+
+      - name: Check metadata
+        run: uvx twine check dist/*
+
+      - name: Publish to PyPI
+        if: github.event_name == 'release' || !inputs.dry-run
+        # `always` so a misconfigured publisher fails loudly instead of falling
+        # back to looking for a token that doesn't exist.
+        run: uv publish --trusted-publishing always

paratext_cli-0.1.0/.gitignore

@@ -0,0 +1,28 @@
+# Python
+__pycache__/
+*.pyc
+.venv/
+*.egg-info/
+dist/
+build/
+
+# Editor
+.vscode/
+.idea/
+
+# Environment
+.env
+
+# Outputs (anchored to repo root so they don't shadow src/paratext/review/)
+/runs/
+/output/
+/review/
+/logs/
+
+# Working notes are kept outside this repo (it is public); these rules are a
+# guard so an internal note dropped in here can't be committed by accident.
+HANDOFF.md
+WORKING-NOTES.md
+
+# Deployed service venv (built from a wheel; not source)
+/deploy/

paratext_cli-0.1.0/AGENTS.md

@@ -0,0 +1,254 @@
+# paratext — agent guide
+
+Modular metadata-extraction pipeline for digitised library/archive collections,
+driven by a multimodal model. This orients AI coding agents; `README.md` is the
+human guide.
+
+paratext is **built to be adapted**. Most users need a project of their own, and
+some need a new input format or a new export format. Those are extension points,
+not forks — the recipes below are the supported way in.
+
+## Rules
+
+- **Read a file in full before editing it.** These modules are short by design;
+  there is no excuse for patching blind.
+- **Run `uv run pytest -q` and `uv run ruff check` after every change.** Both must
+  be clean. Line length 100, rules E/F/I/W.
+- **Never edit `prompt.md` and `schema.py` casually.** A prompt edit rolls a new
+  review round and invalidates comparisons; a schema change needs a
+  `schema_version` bump. Both cost a model run to re-evaluate.
+- **Keep a project's fields in step across all three places** (schema, prompt,
+  view) and call `audit_project()` from its tests.
+- **Add a notice, not a silent fallback.** If you write a degraded path, append to
+  `Source.notices` so `extract` reports it. Preprocessing that silently does
+  nothing looks like a clean run and quietly costs accuracy — it is the worst
+  failure mode in this codebase.
+- **Don't hardcode institution-specific values.** Detector weights, verso
+  thresholds and grid regions are all config.
+- **Ask before removing functionality** that looks deliberate.
+- Say "multimodal model" or "the model" — never "VLM".
+
+## Setup, build, run
+
+```bash
+uv sync --extra dev          # add --extra detector for the torch card detector
+uv run paratext …            # run the CLI against local source
+uv run pytest -q
+uv run ruff check
+```
+
+Python 3.11–3.13 via `uv`. The model endpoint is any OpenAI-compatible server;
+set `base-url` in `paratext.toml` or `PARATEXT_BASE_URL`.
+
+**The installed-vs-source trap:** a non-editable install means source edits do
+nothing until reinstall. `paratext inspect` reports the *installed* project — when
+behaviour doesn't match the source you're reading, check there first.
+
+## Architecture
+
+One CLI runs the loop: `extract` (model → JSONL) → `package` (JSONL → review
+dataset) → `review` (web UI) → `export` (publish). `run` does extract+package.
+
+| Module | Responsibility |
+| --- | --- |
+| `cli.py` | Argparse wiring, round resolution, command bodies |
+| `config.py` | `paratext.toml` + `PARATEXT_*` resolution |
+| `projects/` | The `Project` plug-in contract, `View`, `audit_project` |
+| `sources.py` | Input adapters (`image_source`, `pdf_source`) |
+| `extract.py` / `runner.py` | Sample loop; model call, retries, image encoding |
+| `packaging.py` | JSONL → `samples.json` + `images/` + `view.json` |
+| `store.py` | SQLite annotations + gold labels — **not** web code |
+| `datasets.py` | Discovering packaged rounds, resolving `view.json` |
+| `records.py` | Format-neutral gold selection, shared by all exports |
+| `catalogue.py` / `hf_export.py` | MARC/DC and Hugging Face exports |
+| `review/server.py` | HTTP layer only; `review/static/` is the vanilla-JS frontend |
+| `cards.py` | Optional scanned-card tools (verso, crop, show-through) |
+| `carbon.py` | Grid-intensity providers for `--green` |
+
+`store.py` and `datasets.py` are deliberately outside `review/` so exporters can
+read a reviewed round without starting a web server.
+
+Output is JSONL with a `_provenance` header (git commit, prompt hash, model,
+schema version, timestamp), and is resumable: re-running with the same `--output`
+skips ids already present.
+
+---
+
+# Extending paratext
+
+## Recipe: a new project
+
+The common case — a user's own collection. `paratext new <name>` scaffolds it;
+these are the files it makes and what to edit.
+
+```
+my_cards/
+    prompt.md     # behaviour lives here
+    schema.py     # Pydantic model = the fields, also sent as response_format
+    __init__.py   # wires them into PROJECT
+```
+
+```python
+from paratext.projects import Project, load_prompt
+from paratext.sources import image_source
+
+from .schema import Record
+
+PROJECT = Project(
+    name="my-cards",
+    schema_version="v1",
+    prompt=load_prompt(__file__),
+    schema=Record,
+    source=image_source(),
+)
+```
+
+Register it in the user's `pyproject.toml`:
+
+```toml
+[project.entry-points."paratext.projects"]
+my-cards = "my_cards:PROJECT"
+```
+
+Then `uv sync` — entry points are re-scanned per request, so the review server
+picks up a new project without a restart, but **framework code changes still need
+one**.
+
+Optional, in order of how often you'll need them:
+
+- `view=View(...)` — curate the review display. Defaults to every schema field.
+- `curate(rec) -> Curation` — `keep` / `drop` / `quarantine` per record.
+- `build_record(rec, images)` — extra keys in `samples.json`.
+- `ground_truth(rec)` — attach existing catalogue data for side-by-side review.
+- `disable_thinking` — some models need `enable_thinking=False`.
+
+**Field discipline:** the field names appear in schema, prompt and view with no
+compile-time link. `audit_project(PROJECT)` checks the view's fields exist in the
+schema and that every model-output field is named in the prompt. Call it from a
+test — `paratext new` generates one.
+
+Put behaviour in `prompt.md` (prose, examples, versioned by the prompt hash).
+Keep `Field(description=...)` short and structural: it is *also* sent to the
+model and shouldn't restate the prompt in a second voice.
+
+## Recipe: a new source adapter
+
+For an input shape neither `image_source` nor `pdf_source` covers — a IIIF
+manifest, a METS/ALTO tree, a CSV of URLs, a database query.
+
+A `Source` is two functions that must agree on the metadata they pass between
+them, plus optional notices and a descriptive config:
+
+```python
+from paratext.sources import Source
+from paratext.projects import Sample
+from paratext.packaging import save_image
+
+def iiif_source(*, max_items: int | None = None) -> Source:
+    notices: list[str] = []
+
+    def _iter(source: Path, limit: int | None) -> Iterator[Sample]:
+        # yield one Sample per unit of work
+        yield Sample(
+            id="unique-stable-id",           # becomes the JSONL/review key
+            images=[pil_image],              # what the model sees
+            metadata={"iiif_url": url},      # anything materialise() will need
+        )
+
+    def _materialise(rec: dict, out: Path, max_size: int) -> list[str]:
+        # write review images under `out`, return paths relative to it
+        rel = f"images/{rec['id']}/image.jpg"
+        save_image(local_path, out / rel, max_size)
+        return [rel]
+
+    return Source(
+        iter_samples=_iter,
+        materialise=_materialise,
+        notices=notices,
+        config={"kind": "iiif", "max_items": max_items},   # shown by `paratext inspect`
+    )
+```
+
+Rules specific to sources:
+
+- **`id` must be stable across runs** — resume and round-updating both key on it.
+- **`metadata` is the contract between the two halves.** `_materialise` runs at
+  packaging time, long after iteration, and gets only the JSONL record.
+- **Degrade loudly.** Falling back? `notices.append(...)`, and say what to do
+  about it.
+- **`config` is descriptive only** — nothing reads it back to make decisions; it
+  exists so `paratext inspect` can show what preprocessing was applied.
+- If your images need no special handling, reuse `packaging.default_materialise`
+  rather than writing the same six lines again.
+
+A source can also pre-classify a sample to skip the model entirely: set
+`metadata["preclassified"] = {...}` and `extract` writes it straight through
+(this is how the verso filter avoids paying for blank card backs).
+
+## Recipe: a new export format
+
+For a shape the built-in `hf` / `marc` / `dc` don't cover — EAD, MODS, a local ILS
+format, a CSV for a spreadsheet workflow.
+
+**Never re-derive which records are gold.** `records.select_records()` is the
+single source of truth and every format shares it:
+
+```python
+from paratext.records import select_records
+
+sel = select_records(dataset_dir, project, db_path=db_path)
+for rec in sel.records:
+    rec.label            # {field: value} — the gold label
+    rec.status           # verified | corrected | rejected
+    rec.verdict          # the model's original verdict, kept for corrected rows
+    rec.document_id      # stable id, falls back to rec.sid
+    rec.images           # resolved paths (may be empty)
+sel.schema_fields        # field order from the Pydantic schema
+sel.provenance           # model, prompt_hash, schema_version
+```
+
+Then follow `catalogue.py`'s shape — a pure `build_*(records, mapping)` that
+returns bytes or a tree, and a `run()` that selects, builds and writes:
+
+1. Write `build_ead(records, mapping) -> ET.ElementTree` with no I/O, so it's
+   unit-testable without a dataset on disk.
+2. Add a `run(dataset_dir, project, fmt)` that calls `select_records`, builds,
+   and writes to `EXPORT_ROOT`.
+3. Plumb it into the CLI in `cli.py`: add the name to `--format`'s `choices`, to
+   `_FORMAT_MENU`, and to the `_cmd_export` dispatch.
+4. If fields need mapping to a target vocabulary, reuse `resolve_mapping()` —
+   config under `[project.<name>.export.<fmt>]` wins, then canonical inference by
+   field name, and anything left goes to the wizard. Unmapped fields are dropped
+   with a warning, **never** a hard error.
+
+To expose it in the review UI's export modal as well, add a tab in
+`review/static/app.js` (`openExportModal`) and an endpoint in `review/server.py`
+alongside `_api_export_catalogue`.
+
+---
+
+## Conventions
+
+- **Rounds.** `run` writes `review/<project>-r<N>`, keyed on the **prompt hash**:
+  editing `prompt.md` rolls a new round (the UI diffs the two latest); re-running
+  the same prompt updates the current round in place, keeping its annotations.
+  Rounds are linear — reverting a prompt starts a new round, it doesn't return to
+  the old one. `--round N` forces; `--fresh` rebuilds and discards annotations.
+- **Reading feedback back into the prompt.** `annotations.db` holds verdicts and
+  notes (`annotations`) and human-corrected answers (`gold_labels`); read it via
+  `store.py`. With the server running, `GET /api/stats` gives accuracy and
+  eval-gold size. Tighten the fields reviewers corrected most.
+  The `annotations.corrections` column is unrelated — handwritten corrections on
+  the *card*, not reviewer edits.
+- **Export gold** is `good_enough` rows (`_label_status: verified`) plus corrected
+  `gold_labels` rows (`corrected`). `_verdict` preserves the model's original
+  verdict, so accuracy still measures the model, not the human who fixed it.
+- **`paratext.cards`** (verso filter, RetinaNet crop, show-through suppression) is
+  **off by default** and calibrated on one library's scans. Never assume it
+  transfers to another collection unchanged.
+- **`paratext carbon` / `run --green`** is opt-in carbon-aware scheduling; the grid
+  region is declared in `[carbon]`, never auto-detected. Stdlib `urllib` only.
+- **Config keys are kebab-case** (`base-url`), matching the CLI flag that sets
+  them. Snake_case still parses so old configs work, but generate kebab-case.
+- **Bump `schema_version`** when a schema changes.
+- **Commits:** short messages, no Co-Authored-By trailer.

paratext_cli-0.1.0/LICENSE

@@ -0,0 +1,202 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

paratext_cli-0.1.0/NOTICE

@@ -0,0 +1,7 @@
+paratext
+Copyright 2026 National Library of Scotland
+
+This product includes software developed at the
+National Library of Scotland (https://www.nls.uk/).
+
+Licensed under the Apache License, Version 2.0. See LICENSE for the full text.