v0.3.2

.bumpversion.cfg

@@ -1,5 +1,5 @@
 [bumpversion]
-current_version = 0.3.1
+current_version = 0.3.2
 commit = True
 tag = True
 message = 🔖 Bump version: {current_version} → {new_version}

.github/workflows/build-docker.yml

@@ -34,10 +34,10 @@ jobs:
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
       - name: Build and push release
-        uses: docker/build-push-action@v5
+        uses: docker/build-push-action@v6
         with:
           context: .
-          # platforms: linux/amd64,linux/arm64
+          platforms: linux/amd64,linux/arm64
           push: true
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}

.github/workflows/python.yml

@@ -25,8 +25,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        # python-version: ["3.11", "3.12"]
-        python-version: ["3.11"]
+        python-version: ["3.11", "3.12"]
 
     steps:
     - uses: actions/checkout@v4
@@ -79,6 +78,5 @@ jobs:
       uses: coverallsapp/github-action@v2
       with:
         parallel-finished: true
-        # carryforward: "run-3.11,run-3.12"
-        carryforward: "run-3.11"
+        carryforward: "run-3.11,run-3.12"
         github-token: ${{ secrets.GITHUB_TOKEN }}

.pre-commit-config.yaml

@@ -51,7 +51,7 @@ repos:
         exclude: (test_[\w]+\.py)$
 
   - repo: https://github.com/codespell-project/codespell
-    rev: v2.2.6
+    rev: v2.3.0
     hooks:
       - id: codespell
         exclude: (test_[\w]+\.py|\.csv|\.i?jsonl?|\.lock)$

Dockerfile

@@ -1,6 +1,6 @@
-FROM ghcr.io/investigativedata/ftm-docker:main
+FROM ghcr.io/investigativedata/ftmq:main
 
-RUN apt-get update && apt-get -y upgrade && && apt-get -y autoremove && apt-get clean
+RUN apt-get update && apt-get -y upgrade && apt-get -y autoremove && apt-get clean
 
 COPY ftm_columnstore /app/ftm_columnstore
 COPY setup.py /app/setup.py
@@ -11,5 +11,3 @@ COPY VERSION /app/VERSION
 WORKDIR /app
 RUN pip install --no-cache-dir -U pip setuptools
 RUN pip install --no-cache-dir .
-
-ENTRYPOINT ["ftmcs"]

VERSION

@@ -1 +1 @@
-0.3.1
+0.3.2

ftm_columnstore/__init__.py

@@ -2,20 +2,22 @@
 import warnings
 
 import nomenklatura.settings
-import numpy as np
 
 from ftm_columnstore.engine import get_engine
 from ftm_columnstore.store import get_store
 
+# import numpy as np
+
+
 # shut up
 logging.getLogger("clickhouse_driver.columns.service").setLevel(logging.ERROR)
 logging.getLogger("numpy.core.fromnumeric").setLevel(logging.ERROR)
 warnings.filterwarnings("ignore", category=DeprecationWarning)
-warnings.filterwarnings("ignore", category=np.VisibleDeprecationWarning)
+# warnings.filterwarnings("ignore", category=np.VisibleDeprecationWarning)
 
 # FIXME sqlalchemy monkey patch not working
 nomenklatura.settings.DB_URL = "sqlite:///:memory:"
 
-__version__ = "0.3.1"
+__version__ = "0.3.2"
 
 __all__ = ["get_engine", "get_store"]

ftm_columnstore/cli.py

@@ -2,10 +2,9 @@
 from typing import Annotated, Optional
 
 import typer
-from ftmq.io import smart_read_proxies, smart_write_proxies
 from rich import print
 
-from ftm_columnstore import get_engine, get_store, settings
+from ftm_columnstore import get_engine, settings
 
 log = logging.getLogger(__name__)
 
@@ -34,52 +33,6 @@ def cli_init(
     engine.ensure(recreate=recreate, exists_ok=True)
 
 
-@cli.command("iterate", help="Read entities from the store.")
-def cli_iterate(
-    dataset: Annotated[
-        Optional[str], typer.Option("-d", help="Dataset to read from")
-    ] = None,
-    out_uri: Annotated[
-        str, typer.Option("-o", help="Entities uri (as interpreted by `ftmq`)")
-    ] = "-",
-):
-    """
-    Read entities from the store and write json to `-o` (default: stdout)
-    """
-    store = get_store(dataset=dataset)
-    smart_write_proxies(out_uri, store.iterate(dataset), serialize=True)
-
-
-@cli.command("write", help="Write entity fragments to the store.")
-def cli_write(
-    in_uri: Annotated[
-        str, typer.Option("-i", help="Entities uri (as read by `ftmq`)")
-    ] = "-",
-    dataset: Annotated[
-        Optional[str], typer.Option("-d", help="Dataset to write to")
-    ] = None,
-    optimize: Annotated[
-        Optional[bool],
-        typer.Option(
-            ..., help="Optimize after import (don't do that when writing parallel)"
-        ),
-    ] = False,
-):
-    """
-    Write json entities from `infile` to store.
-    """
-    store = get_store(dataset=dataset)
-    with store.writer() as bulk:
-        for ix, proxy in enumerate(smart_read_proxies(in_uri)):
-            bulk.add_entity(proxy)
-            if ix and ix % 10_000 == 0:
-                print("Write entity `%d` ..." % ix)
-
-    if optimize:
-        engine = get_engine()
-        engine.optimize()
-
-
 @cli.command("optimize")
 def cli_optimize(
     full: Annotated[

ftm_columnstore/settings.py

@@ -1,14 +1,13 @@
 import os
 
-VERSION = "0.3.1"
+VERSION = "0.3.2"
 
 
-def get_env(name, default=None):
+def get_env(name, default):
     value = os.environ.get(name)
     if value is not None:
         return str(value)
-    if default is not None:
-        return str(default)
+    return str(default)
 
 
 DATABASE_URI = get_env("DATABASE_URI", "clickhouse://localhost/default")

ftm_columnstore/store.py

@@ -22,10 +22,10 @@ class BaseClickhouseStore(nk.SQLStore):
     def __init__(
         self,
         dataset: DS,
-        resolver: Resolver[CE],
+        linker: Resolver[CE],
         uri: str | None = None,
     ):
-        super().__init__(dataset, resolver)
+        super().__init__(dataset, linker)
         self.metadata = MetaData()
         self.table = make_statement_table(self.metadata)
         self.engine = get_engine(uri)
@@ -73,12 +73,12 @@ def pop(self, entity_id: str) -> list[Statement]:
 
 @cache
 def get_store(
-    uri: str | None = None,
     catalog: C | None = None,
     dataset: Dataset | str | None = None,
-    resolver: Resolver | str | None = None,
+    uri: str | None = None,
+    linker: Resolver | str | None = None,
 ) -> ClickhouseStore:
     get_metadata.cache_clear()
     if isinstance(dataset, str):
         dataset = Dataset(name=dataset)
-    return ClickhouseStore(catalog, dataset, uri=uri, resolver=resolver)
+    return ClickhouseStore(catalog, dataset, uri=uri, linker=linker)