luxorasap 0.1.18__tar.gz → 0.1.19__tar.gz

{luxorasap-0.1.18 → luxorasap-0.1.19}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: luxorasap
-Version: 0.1.18
+Version: 0.1.19
 Summary: Toolbox da Luxor para ingestão, análise e automação de dados financeiros.
 Author-email: Luxor Group <backoffice@luxor.com.br>
 License: Proprietary – All rights reserved

{luxorasap-0.1.18 → luxorasap-0.1.19}/pyproject.toml

@@ -10,7 +10,7 @@ build-backend = "setuptools.build_meta"
 #############################
 [project]
 name            = "luxorasap"
-version = "0.1.18"
+version = "0.1.19"
 description     = "Toolbox da Luxor para ingestão, análise e automação de dados financeiros."
 readme          = "README.md"
 requires-python = ">=3.9"
@@ -78,7 +78,7 @@ exclude = ["tests*"]
 # bumpver (sem-ver)
 #############################
 [tool.bumpver]
-current_version = "0.1.18"
+current_version = "0.1.19"
 version_pattern = "MAJOR.MINOR.PATCH"
 
 # regex explícito – obrigatório no bumpver 2024+

{luxorasap-0.1.18 → luxorasap-0.1.19}/src/luxorasap/__init__.py

@@ -13,7 +13,7 @@ from types import ModuleType
 try:
     __version__: str = metadata.version(__name__)
 except metadata.PackageNotFoundError:  # editable install
-    __version__ = "0.1.18"
+    __version__ = "0.1.19"
 
 # ─── Lazy loader ─────────────────────────────────────────────────
 def __getattr__(name: str) -> ModuleType:

{luxorasap-0.1.18 → luxorasap-0.1.19}/src/luxorasap/ingest/cloud/__init__.py

@@ -4,7 +4,7 @@ import pandas as pd
 import datetime as dt
 
 from luxorasap.utils.storage import BlobParquetClient
-from luxorasap.utils.dataframe import prep_for_save
+from luxorasap.utils.dataframe import prep_for_save, astype_str_inplace
 from luxorasap.datareader import LuxorQuery
 
 
@@ -37,6 +37,7 @@ def save_table(
     df = prep_for_save(df, index=index, index_name=index_name, normalize=normalize_columns)
     
     #_client.write_df(df.astype(str), f"{directory}/{table_name}.parquet")
+    astype_str_inplace(df)
     _client.write_df(df, f"{directory}/{table_name}.parquet")
 
 

{luxorasap-0.1.18 → luxorasap-0.1.19}/src/luxorasap/utils/dataframe/__init__.py

@@ -1,4 +1,4 @@
 from .transforms import prep_for_save, persist_column_formatting, text_to_lowercase_inplace
 from .reader import read_bytes
 
-__all__ = ["prep_for_save", "persist_column_formatting", "text_to_lowercase_inplace", "read_bytes"]
+__all__ = ["prep_for_save", "persist_column_formatting", "text_to_lowercase_inplace", "read_bytes", "astype_str_inplace"]

{luxorasap-0.1.18 → luxorasap-0.1.19}/src/luxorasap/utils/dataframe/transforms.py

@@ -1,4 +1,5 @@
 import pandas as pd
+import gc
 from pandas.api.types import is_object_dtype, is_string_dtype
 
 
@@ -55,4 +56,21 @@ def prep_for_save(
     if index:
         name = df.index.name or index_name
         df = df.reset_index().rename(columns={"index": name})
-    return persist_column_formatting(df) if normalize else df
+    return persist_column_formatting(df) if normalize else df
+
+
+def astype_str_inplace(df: pd.DataFrame, *, gc_every: int = 3) -> None:
+    """
+    Converte TODAS as colunas para string sem duplicar o DataFrame completo.
+    Em cada passo:
+      • cria uma nova Series (coluna convertida)
+      • substitui a antiga
+      • libera memória da coluna anterior
+    Isso mantém o pico de RAM ≤ tamanho da maior coluna.
+    """
+    for i, col in enumerate(df.columns):
+        df[col] = df[col].astype(str)
+
+        # opcional: força coleta de lixo a cada N colunas
+        if (i + 1) % gc_every == 0:
+            gc.collect()

{luxorasap-0.1.18 → luxorasap-0.1.19}/src/luxorasap/utils/storage/blob.py

@@ -37,7 +37,7 @@ class BlobParquetClient:
     def write_df(self, df, blob_path: str):
         
         blob = self._blob(blob_path)
-        table = pa.Table.from_pandas(df, preserve_index=False)
+        table = pa.Table.from_pandas(df)
         buf = io.BytesIO()
         pq.write_table(table, buf)
         buf.seek(0)

{luxorasap-0.1.18 → luxorasap-0.1.19}/src/luxorasap.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: luxorasap
-Version: 0.1.18
+Version: 0.1.19
 Summary: Toolbox da Luxor para ingestão, análise e automação de dados financeiros.
 Author-email: Luxor Group <backoffice@luxor.com.br>
 License: Proprietary – All rights reserved

{luxorasap-0.1.18 → luxorasap-0.1.19}/tests/test_ingest_cloud.py

@@ -16,7 +16,7 @@ def test_save_table_calls_blob_client(fake_blob, monkeypatch):
     df = pd.DataFrame({"x": [1]})
     cloud.save_table("t1", df, directory="dir")
     assert captured["path"] == "dir/t1.parquet"
-    assert captured["df"].equals(df)
+    assert captured["df"].equals(df.astype(str))
     
 
 def test_incremental_load_merges_correctly(fake_blob, monkeypatch):