oda-data 2.1.0__tar.gz → 2.1.2__tar.gz

{oda_data-2.1.0 → oda_data-2.1.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: oda_data
-Version: 2.1.0
+Version: 2.1.2
 Summary: A python package to work with Official Development Assistance data from the OECD DAC.
 License: MIT
 Author: Jorge Rivera

{oda_data-2.1.0 → oda_data-2.1.2}/oda_data/__init__.py

@@ -29,6 +29,9 @@ def set_data_path(path):
     global ODAPaths
 
     ODAPaths.raw_data = Path(path).resolve()
+    # Create on explicit user request; avoids import-time side effects
+    ODAPaths.raw_data.mkdir(parents=True, exist_ok=True)
+    ODAPaths.pydeflate = ODAPaths.raw_data / ".pydeflate"
     set_pydeflate_path(ODAPaths.raw_data)
     set_cache_dir(ODAPaths.raw_data)
 

{oda_data-2.1.0 → oda_data-2.1.2}/oda_data/api/sources.py

@@ -1,6 +1,7 @@
 from abc import abstractmethod
 from pathlib import Path
 from typing import Optional
+import threading
 
 import pandas as pd
 from cachetools import TTLCache
@@ -15,7 +16,6 @@ from oda_reader import (
 )
 from oda_reader._cache import memory
 
-from oda_data import config
 from oda_data.clean_data.common import (
     clean_raw_df,
     convert_dot_stat_to_data_explorer_codes,
@@ -52,7 +52,10 @@ class Source:
     """
 
     memory_cache = TTLCache(maxsize=20, ttl=6000)
-    disk_cache = OnDiskCache(ODAPaths.raw_data, ttl_seconds=86400)
+
+    # Shared, process-wide disk cache and lock
+    _shared_disk_cache: Optional[OnDiskCache] = None
+    _shared_disk_cache_lock = threading.Lock()
 
     def __init__(self):
         self.de_providers = None
@@ -63,6 +66,24 @@ class Source:
         self.schema = None
         self._param_hash = None
 
+    @property
+    def disk_cache(self) -> OnDiskCache:
+        # Snapshot path to avoid inconsistency if it changes mid-check
+        current_raw = ODAPaths.raw_data
+        # Fast-path without lock (shared, process-wide cache)
+        dc = Source._shared_disk_cache
+        if dc is not None and dc.base_dir == current_raw:
+            return dc
+
+        # Slow-path with lock to ensure single initialization across threads
+        with Source._shared_disk_cache_lock:
+            dc = Source._shared_disk_cache
+            if dc is None or dc.base_dir != current_raw:
+                Source._shared_disk_cache = OnDiskCache(
+                    current_raw, ttl_seconds=86400
+                )
+            return Source._shared_disk_cache
+
     def _add_filter(self, column: str, predicate: str, value: str | int | list) -> None:
         """Adds a filter to the dataset, ensuring no duplicate columns.
 
@@ -122,7 +143,6 @@ class DACSource(Source):
     """
 
     memory_cache = TTLCache(maxsize=20, ttl=6000)
-    disk_cache = OnDiskCache(ODAPaths.raw_data, ttl_seconds=86400)
 
     def __init__(self):
         super().__init__()
@@ -522,7 +542,6 @@ class AidDataSource(Source):
     """
 
     memory_cache = TTLCache(maxsize=20, ttl=6000)
-    disk_cache = OnDiskCache(ODAPaths.raw_data, ttl_seconds=86400)
 
     def __init__(self):
         super().__init__()
@@ -565,7 +584,6 @@ class AidDataData(AidDataSource):
     """Class to handle the AidData data."""
 
     memory_cache = TTLCache(maxsize=20, ttl=6000)
-    disk_cache = OnDiskCache(config.ODAPaths.raw_data, ttl_seconds=86400)
 
     def __init__(
         self,

{oda_data-2.1.0 → oda_data-2.1.2}/oda_data/clean_data/common.py

@@ -12,8 +12,6 @@ from oda_data.clean_data.validation import validate_currency
 from oda_data.config import ODAPaths
 from oda_data.logger import logger
 
-set_pydeflate_path(ODAPaths.raw_data)
-
 
 def clean_column_name(column_name: str) -> str:
     """Clean a column name by removing spaces, special characters, and lowercasing.
@@ -140,10 +138,20 @@ def convert_units(
     currency: str = "USD",
     base_year: Optional[int] = None,
 ):
+    # Ensure pydeflate has the current data path and directory exists (lazy init)
+    try:
+        ODAPaths.raw_data.mkdir(parents=True, exist_ok=True)
+    except Exception:
+        # In read-only environments, let downstream functions surface meaningful errors
+        pass
+    set_pydeflate_path(ODAPaths.raw_data)
+
     if indicator is None:
         indicator = ""
 
-    if ".40." in indicator or (currency == "USD" and base_year is None):
+    if ((".40." in indicator) and ("DAC1.40.1" != indicator)) or (
+        currency == "USD" and base_year is None
+    ):
         return data.assign(currency=currency, prices="current")
 
     elif base_year is None:

{oda_data-2.1.0 → oda_data-2.1.2}/oda_data/config.py

@@ -6,7 +6,7 @@ class ODAPaths:
 
     project = Path(__file__).resolve().parent.parent
     scripts = project / "oda_data"
-    raw_data = scripts / ".raw_data"
+    raw_data = (Path.cwd() / ".raw_data").resolve()
     pydeflate = raw_data / ".pydeflate"
     indicators = scripts / "indicators"
     names = scripts / "tools" / "names"

{oda_data-2.1.0 → oda_data-2.1.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "oda_data"
-version = "2.1.0"
+version = "2.1.2"
 description = "A python package to work with Official Development Assistance data from the OECD DAC."
 readme = "README.md"
 authors = [