datamaestro-text 2025.9.11__tar.gz → 2026.2.2__tar.gz

{datamaestro_text-2025.9.11/src/datamaestro_text.egg-info → datamaestro_text-2026.2.2}/PKG-INFO

@@ -1,33 +1,31 @@
 Metadata-Version: 2.4
 Name: datamaestro-text
-Version: 2025.9.11
+Version: 2026.2.2
 Summary: Datamaestro module for text-related datasets
+Project-URL: Homepage, https://github.com/experimaestro/datamaestro_text
+Project-URL: Documentation, https://datamaestro-text.readthedocs.io/en/latest/
+Project-URL: Repository, https://github.com/experimaestro/datamaestro_text
+Project-URL: Bug Tracker, https://github.com/experimaestro/datamaestro_text/issues
 Author-email: Benjamin Piwowarski <benjamin@piwowarski.fr>
-License: GPL-3
-Project-URL: homepage, https://github.com/experimaestro/datamaestro_text
-Project-URL: documentation, https://datamaestro-text.readthedocs.io/en/latest/
-Project-URL: repository, https://github.com/experimaestro/datamaestro_text
-Keywords: dataset manager,information retrieval,experiments
+License: GPL-3.0-or-later
+License-File: LICENSE
+Keywords: dataset manager,experiments,information retrieval
 Classifier: Development Status :: 4 - Beta
 Classifier: Intended Audience :: Science/Research
 Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3)
 Classifier: Operating System :: OS Independent
 Classifier: Programming Language :: Python
 Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Requires-Python: >=3.10
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: datamaestro>=1.5.0
-Requires-Dist: ir_datasets>=0.5.8
 Requires-Dist: attrs
-Provides-Extra: dev
-Requires-Dist: pytest; extra == "dev"
-Requires-Dist: docutils; extra == "dev"
-Requires-Dist: sphobjinv; extra == "dev"
-Requires-Dist: flake8; extra == "dev"
-Requires-Dist: sphinx; extra == "dev"
-Dynamic: license-file
+Requires-Dist: datamaestro>=1.6.2
+Requires-Dist: experimaestro
+Requires-Dist: ir-datasets>=0.5.8
+Description-Content-Type: text/markdown
 
  [![pre-commit](https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit&logoColor=white)](https://github.com/pre-commit/pre-commit) [![PyPI version](https://badge.fury.io/py/datamaestro-text.svg)](https://badge.fury.io/py/datamaestro-text)
 

datamaestro_text-2026.2.2/pyproject.toml

@@ -0,0 +1,105 @@
+[build-system]
+requires = ["hatchling", "hatch-vcs"]
+build-backend = "hatchling.build"
+
+[project]
+name = "datamaestro-text"
+authors = [
+    {name = "Benjamin Piwowarski", email = "benjamin@piwowarski.fr"}
+]
+description = "Datamaestro module for text-related datasets"
+readme = "README.md"
+license = {text = "GPL-3.0-or-later"}
+keywords = ["dataset manager", "information retrieval", "experiments"]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Science/Research",
+    "License :: OSI Approved :: GNU General Public License v3 (GPLv3)",
+    "Operating System :: OS Independent",
+    "Programming Language :: Python",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Software Development :: Libraries :: Python Modules",
+]
+requires-python = ">=3.10"
+dynamic = ["version"]
+dependencies = [
+    "datamaestro>=1.6.2",
+    "ir_datasets>=0.5.8",
+    "attrs",
+    "experimaestro",
+]
+
+[project.urls]
+Homepage = "https://github.com/experimaestro/datamaestro_text"
+Documentation = "https://datamaestro-text.readthedocs.io/en/latest/"
+Repository = "https://github.com/experimaestro/datamaestro_text"
+"Bug Tracker" = "https://github.com/experimaestro/datamaestro_text/issues"
+
+[project.entry-points."datamaestro.repositories"]
+text = "datamaestro_text:Repository"
+irds = "datamaestro_text.datasets.irds:Repository"
+
+[tool.hatch.version]
+source = "vcs"
+
+[tool.hatch.version.raw-options]
+local_scheme = "no-local-version"
+
+[tool.hatch.build.hooks.vcs]
+version-file = "src/datamaestro_text/version.py"
+
+[tool.hatch.build.targets.sdist]
+include = [
+    "/src",
+    "/README.md",
+    "/LICENSE",
+    "/pyproject.toml",
+]
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/datamaestro_text"]
+
+[tool.pytest.ini_options]
+junit_family = "xunit2"
+testpaths = ["src/datamaestro_text"]
+norecursedirs = ["node_modules"]
+
+[dependency-groups]
+dev = [
+    "docutils>=0.21.2",
+    "git-cliff>=2.11.0",
+    "pytest>=8.4.1",
+    "ruff>=0.8",
+    "sphinx>=7,<8",
+    "sphobjinv>=2.3.1.3",
+]
+docs = [
+    "datamaestro>=0.8.5",
+    "experimaestro>=2.0.0b29",  # Pre-release needed for EnumType.name() fix
+    "myst-parser>=0.18.0",
+    "sphinx>=6,<8",  # experimaestro.sphinx uses restify removed in Sphinx 8
+    "sphinx-codeautolink>=0.15.0",
+    "sphinx-rtd-theme>=3.1.0",
+    "sphinx-toolbox>=4.1.2",
+]
+
+[tool.ruff]
+line-length = 88
+target-version = "py310"
+
+[tool.ruff.lint]
+select = [
+    "E",      # pycodestyle errors
+    "F",      # pyflakes
+    "W",      # pycodestyle warnings
+    "C90",    # mccabe complexity
+]
+ignore = [
+    "E501",   # line too long (handled by formatter)
+]
+
+[tool.ruff.lint.mccabe]
+max-complexity = 20

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/__init__.py

@@ -1,6 +1,6 @@
 import datamaestro
 
-from .version import version, version_tuple
+from .version import version as version, version_tuple as version_tuple
 
 
 class Repository(datamaestro.Repository):

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/com/github/ikat.py

@@ -1,6 +1,5 @@
 # See documentation on https://datamaestro.readthedocs.io
 
-import bz2
 from datamaestro.download import reference
 from datamaestro.definitions import datatasks, datatags, dataset
 from datamaestro_text.data.conversation.base import ConversationUserTopics

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/com/oscar-corpus.py

@@ -1,4 +1,4 @@
-from datamaestro.definitions import argument, datatasks, datatags, dataset
+from datamaestro.definitions import dataset
 from datamaestro.download.single import filedownloader
 from datamaestro_text.data.text import TextFile
 from datamaestro.utils import HashCheck

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/com/smashwords/bookcorpus.py

@@ -1,6 +1,6 @@
 # See documentation on https://datamaestro.readthedocs.io
 
-from datamaestro.definitions import argument, datatasks, datatags, dataset
+from datamaestro.definitions import datatasks, datatags, dataset
 from datamaestro_text.data.text import TextFolder
 from datamaestro.download.archive import tardownloader
 from datamaestro.utils import HashCheck

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/edu/stanford/aclimdb.py

@@ -1,5 +1,5 @@
 from datamaestro.data.ml import FolderBased, Supervised
-from datamaestro.definitions import datatasks, datatags, dataset
+from datamaestro.definitions import dataset
 from datamaestro.download.archive import tardownloader
 
 

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/edu/stanford/glove.py

@@ -5,7 +5,6 @@ GloVe is an unsupervised learning algorithm for obtaining vector representations
 """
 
 from datamaestro.definitions import dataset
-from datamaestro.data import Base, Generic
 from datamaestro.download import reference
 from datamaestro.download.archive import zipdownloader
 from datamaestro.download.single import filedownloader

datamaestro_text-2026.2.2/src/datamaestro_text/config/fr/granddebat.py

@@ -0,0 +1,186 @@
+# See documentation on https://datamaestro.readthedocs.io
+
+from pathlib import Path
+from datamaestro.definitions import datatags, dataset
+from datamaestro_text.data.debate import GrandDebatFile
+from datamaestro.download.single import filedownloader
+from datamaestro.utils import HashCheck
+from datamaestro.stream import Transform
+import io
+import json
+import ijson
+import os
+import threading
+
+
+class JsonToJsonl(Transform):
+    """Transforms a JSON file with an array into a JSONL file with one line per
+    array element"""
+
+    def __call__(self, fileobj: io.IOBase) -> io.IOBase:
+        # Stream items from the top-level array into a read-end pipe.
+        try:
+            fileobj.seek(0)
+        except Exception:
+            pass
+
+        r_fd, w_fd = os.pipe()
+        r_file = os.fdopen(r_fd, "rb")
+        w_file = os.fdopen(w_fd, "wb")
+
+        def _writer(fin, fout):
+            try:
+                for item in ijson.items(fin, "item"):
+                    line = json.dumps(item, ensure_ascii=False) + "\n"
+                    fout.write(line.encode("utf-8"))
+                fout.close()
+            except Exception:
+                try:
+                    fout.close()
+                except Exception:
+                    pass
+
+        t = threading.Thread(target=_writer, args=(fileobj, w_file), daemon=True)
+        t.start()
+
+        return r_file
+
+
+@filedownloader(
+    "la_transition_ecologique_2019_03_21.jsonl",
+    "http://opendata.auth-6f31f706db6f4a24b55f42a6a79c5086.storage.sbg.cloud.ovh.net/2019-03-21/LA_TRANSITION_ECOLOGIQUE.json",
+    checker=HashCheck("c4ed3a8b8c43d5806d1f090e03f7aa91"),
+    transforms=JsonToJsonl(),
+)
+@datatags("politics", "debate", "french")
+@dataset(
+    GrandDebatFile,
+    url="https://granddebat.fr",
+)
+def transition(la_transition_ecologique_2019_03_21: Path):
+    """Grand Débat National (transition écologique)
+
+    The *Grand Débat National* (GDN) is a country-wide citizen consultation held
+    in France in 2019.
+
+
+    The consultation prompted citizens to express their views across four main
+    themes: *Taxation and public spending*, *Organization of the state and
+    public services*, *Democracy and citizenship*, and *Ecological transition*.
+    A significant portion of this consultation involved online questionnaires,
+    each concluding with a critical open-ended prompt: "Do you have anything to
+    add about [theme]?".
+    """
+    return GrandDebatFile.C(path=la_transition_ecologique_2019_03_21)
+
+
+@filedownloader(
+    "fiscalité_et_dépenses_publiques_2019_03_21.jsonl",
+    "http://opendata.auth-6f31f706db6f4a24b55f42a6a79c5086.storage.sbg.cloud.ovh.net/2019-03-21/LA_FISCALITE_ET_LES_DEPENSES_PUBLIQUES.json",
+    checker=HashCheck("c4ed3a8b8c43d5806d1f090e03f7aa91"),
+    transforms=JsonToJsonl(),
+)
+@datatags("politics", "debate", "french")
+@dataset(
+    GrandDebatFile,
+    url="https://granddebat.fr",
+)
+def fiscalité(fiscalité_et_dépenses_publiques_2019_03_21: Path):
+    """Grand Débat National (fiscalité et dépenses publiques)
+
+    The *Grand Débat National* (GDN) is a country-wide citizen consultation held
+    in France in 2019.
+
+
+    The consultation prompted citizens to express their views across four main
+    themes: *Taxation and public spending*, *Organization of the state and
+    public services*, *Democracy and citizenship*, and *Ecological transition*.
+    A significant portion of this consultation involved online questionnaires,
+    each concluding with a critical open-ended prompt: "Do you have anything to
+    add about [theme]?".
+    """
+    return GrandDebatFile.C(path=fiscalité_et_dépenses_publiques_2019_03_21)
+
+
+@filedownloader(
+    "democratie_et_citoyennete_2019_03_21.jsonl",
+    "http://opendata.auth-6f31f706db6f4a24b55f42a6a79c5086.storage.sbg.cloud.ovh.net/2019-03-21/DEMOCRATIE_ET_CITOYENNETE.json",
+    checker=HashCheck("049aaeca7e51747f2da5d68143c760fa"),
+    transforms=JsonToJsonl(),
+)
+@datatags("politics", "debate", "french")
+@dataset(
+    GrandDebatFile,
+    url="https://granddebat.fr",
+)
+def démocratie(democratie_et_citoyennete_2019_03_21: Path):
+    """Grand Débat National (démocratie et citoyenneté)
+
+    The *Grand Débat National* (GDN) is a country-wide citizen consultation held
+    in France in 2019.
+
+
+    The consultation prompted citizens to express their views across four main
+    themes: *Taxation and public spending*, *Organization of the state and
+    public services*, *Democracy and citizenship*, and *Ecological transition*.
+    A significant portion of this consultation involved online questionnaires,
+    each concluding with a critical open-ended prompt: "Do you have anything to
+    add about [theme]?".
+    """
+    return GrandDebatFile.C(path=democratie_et_citoyennete_2019_03_21)
+
+
+@filedownloader(
+    "organisation_etat_services_publics_2019_03_21.jsonl",
+    "http://opendata.auth-6f31f706db6f4a24b55f42a6a79c5086.storage.sbg.cloud.ovh.net/2019-03-21/ORGANISATION_DE_LETAT_ET_DES_SERVICES_PUBLICS.json",
+    checker=HashCheck("0ccb7c401889f738b73b0caab897a68b"),
+    transforms=JsonToJsonl(),
+)
+@datatags("politics", "debate", "french")
+@dataset(
+    GrandDebatFile,
+    url="https://granddebat.fr",
+)
+def organisation(organisation_etat_services_publics_2019_03_21: Path):
+    """Grand Débat National (organisation de l'État et des services publics)
+
+    The *Grand Débat National* (GDN) is a country-wide citizen consultation held
+    in France in 2019.
+
+
+    The consultation prompted citizens to express their views across four main
+    themes: *Taxation and public spending*, *Organization of the state and
+    public services*, *Democracy and citizenship*, and *Ecological transition*.
+    A significant portion of this consultation involved online questionnaires,
+    each concluding with a critical open-ended prompt: "Do you have anything to
+    add about [theme]?".
+    """
+    return GrandDebatFile.C(path=organisation_etat_services_publics_2019_03_21)
+
+
+@filedownloader(
+    "les_evenements_2019_03_21.jsonl",
+    "http://opendata.auth-6f31f706db6f4a24b55f42a6a79c5086.storage.sbg.cloud.ovh.net/2019-03-21/LES_EVENEMENTS.json",
+    checker=HashCheck("c4ed3a8b8c43d5806d1f090e03f7aa91"),
+    transforms=JsonToJsonl(),
+)
+@datatags("politics", "debate", "french")
+@dataset(
+    GrandDebatFile,
+    url="https://granddebat.fr",
+)
+def evenements(les_evenements_2019_03_21: Path):
+    """Grand Débat National (événements)
+
+    The *Grand Débat National* (GDN) is a country-wide citizen consultation held
+    in France in 2019.
+
+
+    The consultation prompted citizens to express their views across four main
+    themes: *Taxation and public spending*, *Organization of the state and
+    public services*, *Democracy and citizenship*, and *Ecological transition*.
+    A significant portion of this consultation involved online questionnaires,
+    each concluding with a critical open-ended prompt: "Do you have anything to
+    add about [theme]?".
+    """
+    return GrandDebatFile.C(path=les_evenements_2019_03_21)

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/gov/nist/ir/covid.py

@@ -1,5 +1,4 @@
-"""CORD-19 dataset
-"""
+"""CORD-19 dataset"""
 
 from datamaestro.annotations.agreement import useragreement
 from datamaestro.definitions import datatasks, dataset

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/config/io/metamind/research/wikitext.py

@@ -1,4 +1,4 @@
-from datamaestro.data import Base, File
+from datamaestro.data import File
 from datamaestro.definitions import (
     datatasks,
     datatags,

datamaestro_text-2026.2.2/src/datamaestro_text/data/conversation/__init__.py

@@ -0,0 +1,8 @@
+from .base import (
+    AnswerEntry as AnswerEntry,
+    ConversationDataset as ConversationDataset,
+    ConversationHistory as ConversationHistory,
+    ConversationHistoryItem as ConversationHistoryItem,
+    DecontextualizedItem as DecontextualizedItem,
+    EntryType as EntryType,
+)

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/conversation/base.py

@@ -2,13 +2,13 @@ from abc import ABC, abstractmethod
 from enum import Enum
 from datamaestro_text.data.ir.base import IDItem, SimpleTextItem
 from experimaestro import Param
-from typing import Dict, Generic, Iterator, List, Optional, Sequence, Tuple
+from typing import Dict, Iterator, List, Optional, Sequence, Tuple
 from attr import define
 from datamaestro.record import record_type
 from datamaestro.data import Base
 from datamaestro.record import Record, Item
 from datamaestro_text.data.ir import TopicRecord, Topics
-from datamaestro_text.utils.iter import FactoryIterable, LazyList, RangeView
+from datamaestro_text.utils.iter import FactoryIterable, LazyList
 
 # ---- Basic types
 
@@ -267,7 +267,7 @@ class ConversationUserTopics(Topics):
         """Returns an iterator over topics"""
         # Extracts topics from conversations, Each user query is a topic (can perform retrieval on it)
         # TODO: merge with xpmir.learning.DatasetConversationBase -> same logic
-        
+
         records: List[TopicRecord] = []
         for conversation in self.conversations.__iter__():
             nodes = [
@@ -279,4 +279,4 @@ class ConversationUserTopics(Topics):
                 records.append(
                     node.entry.update(ConversationHistoryItem(node.history()))
                 )
-        return iter(records)
+        return iter(records)

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/conversation/canard.py

@@ -11,7 +11,6 @@ from datamaestro_text.data.conversation.base import (
     EntryType,
 )
 from datamaestro_text.data.ir import IDItem, SimpleTextItem
-import logging
 
 
 @define(kw_only=True)
@@ -82,9 +81,9 @@ class CanardDataset(ConversationDataset, File):
                 )
             else:
                 # The utterance before the last is the last user query
-                assert (
-                    entry.history[-2] == history[-1][SimpleTextItem].text
-                ), f"{entry.dialogue_id} {entry.history} / {history[-4:-1]}"
+                assert entry.history[-2] == history[-1][SimpleTextItem].text, (
+                    f"{entry.dialogue_id} {entry.history} / {history[-4:-1]}"
+                )
 
                 # The last utterance is the system side
                 history.append(

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/conversation/ikat.py

@@ -5,7 +5,6 @@ import logging
 from datamaestro.data import File
 from datamaestro.record import Record
 
-from datamaestro_text.data.ir import Topics
 from datamaestro_text.data.ir.base import (
     IDItem,
     SimpleTextItem,

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/conversation/orconvqa.py

@@ -113,9 +113,9 @@ class OrConvQADataset(ConversationDataset, File):
                 if relevance > 0:
                     relevances[rank] = (entry.answer.answer_start, None)
 
-            assert (
-                len(relevances) <= 1
-            ), f"Too many relevance labels ({len(relevances)}) for {entry.query_id}"
+            assert len(relevances) <= 1, (
+                f"Too many relevance labels ({len(relevances)}) for {entry.query_id}"
+            )
 
             history.append(
                 Record(

datamaestro_text-2026.2.2/src/datamaestro_text/data/debate/__init__.py

@@ -0,0 +1,5 @@
+"""Data classes for debate datasets"""
+
+from .granddebat import GrandDebatEntry, GrandDebatFile, GrandDebatResponse
+
+__all__ = ["GrandDebatEntry", "GrandDebatFile", "GrandDebatResponse"]

datamaestro_text-2026.2.2/src/datamaestro_text/data/debate/granddebat.py

@@ -0,0 +1,68 @@
+"""Data classes for the Grand Débat National dataset"""
+
+import json
+from dataclasses import dataclass, field
+from typing import Iterator, List, Optional
+
+from datamaestro.data import File
+
+
+@dataclass
+class GrandDebatResponse:
+    """A response to a question in the Grand Débat National"""
+
+    question_id: str
+    question_title: str
+    value: Optional[str]
+    formatted_value: Optional[str]
+
+
+@dataclass
+class GrandDebatEntry:
+    """An entry (contribution) in the Grand Débat National dataset"""
+
+    id: str
+    reference: str
+    title: str
+    created_at: str
+    published_at: str
+    updated_at: Optional[str]
+    trashed: bool
+    trashed_status: Optional[str]
+    author_id: str
+    author_type: str
+    author_zip_code: str
+    responses: List[GrandDebatResponse] = field(default_factory=list)
+
+
+class GrandDebatFile(File):
+    """A Grand Débat National JSONL file with iteration support"""
+
+    def __iter__(self) -> Iterator[GrandDebatEntry]:
+        """Iterate over entries in the JSONL file"""
+        with self.path.open("r", encoding="utf-8") as f:
+            for line in f:
+                data = json.loads(line)
+                responses = [
+                    GrandDebatResponse(
+                        question_id=r["questionId"],
+                        question_title=r["questionTitle"],
+                        value=r.get("value"),
+                        formatted_value=r.get("formattedValue"),
+                    )
+                    for r in data.get("responses", [])
+                ]
+                yield GrandDebatEntry(
+                    id=data["id"],
+                    reference=data["reference"],
+                    title=data["title"],
+                    created_at=data["createdAt"],
+                    published_at=data["publishedAt"],
+                    updated_at=data.get("updatedAt"),
+                    trashed=data["trashed"],
+                    trashed_status=data.get("trashedStatus"),
+                    author_id=data["authorId"],
+                    author_type=data["authorType"],
+                    author_zip_code=data["authorZipCode"],
+                    responses=responses,
+                )

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/embeddings.py

@@ -20,6 +20,7 @@ class WordEmbeddings(Base):
 
 class WordEmbeddingsText(WordEmbeddings, File):
     """Word embeddings as a text word / values"""
+
     encoding: Meta[str] = "utf-8"
 
     def load(self):

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/ir/__init__.py

@@ -6,7 +6,7 @@ from functools import cached_property
 import logging
 from pathlib import Path
 from attrs import define
-from typing import Callable, Dict, Iterator, List, Optional, Tuple, Type, TYPE_CHECKING
+from typing import Callable, Dict, Iterator, List, Optional, Tuple, Type
 import random
 from experimaestro import Config
 from datamaestro.definitions import datatasks, Param, Meta

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/ir/base.py

@@ -1,7 +1,7 @@
 from abc import ABC, abstractmethod
 from attrs import define
 from typing import List
-from datamaestro.record import Record, Item, record_type
+from datamaestro.record import Record, Item
 
 
 TopicRecord = DocumentRecord = Record

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/ir/csv.py

@@ -1,27 +1,26 @@
 from functools import cached_property
 from pathlib import Path
-from typing import Iterator, Tuple, Type
 
-from experimaestro import Param, Option, Constant, Meta
-from datamaestro.definitions import argument
+from experimaestro import Param, Meta
 from datamaestro.record import Record, RecordType
 import datamaestro_text.data.ir as ir
 from datamaestro_text.data.ir.base import IDItem, SimpleTextItem
 from datamaestro_text.interfaces.plaintext import read_tsv
 
 
-@argument("path", type=Path)
-@argument("separator", type=str, default="\t", ignored=True)
 class AdhocRunWithText(ir.AdhocRun):
     "(qid, doc.id, query, passage)"
-    pass
+
+    path: Meta[Path]
+    separator: Meta[str] = "\t"
 
 
-@argument("path", type=Path)
-@argument("separator", type=str, default="\t", ignored=True)
 class Topics(ir.Topics):
     "Pairs of query id - query using a separator"
 
+    path: Meta[Path]
+    separator: Meta[str] = "\t"
+
     def iter(self):
         return (
             Record(IDItem(qid), SimpleTextItem(title))

datamaestro_text-2026.2.2/src/datamaestro_text/data/ir/data.py

@@ -0,0 +1 @@
+from .base import *  # noqa: F403

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/ir/formats.py

@@ -1,5 +1,5 @@
 from functools import cached_property
-from typing import ClassVar, Tuple, List
+from typing import Tuple, List
 from attrs import define
 from datamaestro.record import record_type
 from ir_datasets.datasets.wapo import WapoDocMedia
@@ -11,9 +11,8 @@ from ir_datasets.datasets.cord19 import Cord19FullTextSection
 class DocumentWithTitle(TextItem):
     """Web document with title and body"""
 
-    body: str
-
     title: str
+    body: str
 
     @cached_property
     def text(self):

{datamaestro_text-2025.9.11 → datamaestro_text-2026.2.2}/src/datamaestro_text/data/ir/stores.py

@@ -82,7 +82,7 @@ class IKatClueWeb22DocumentStore(LZ4DocumentStore):
 
                 file_checksum = hasher.hexdigest()
                 assert file_checksum == checksum, (
-                    f"Expected {checksum}, " f"got {file_checksum} for {filename}"
+                    f"Expected {checksum}, got {file_checksum} for {filename}"
                 )
 
             # Get the MD5 hashes of all the passages