sdmxlib 0.41.1__tar.gz → 0.41.2__tar.gz

{sdmxlib-0.41.1 → sdmxlib-0.41.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: sdmxlib
-Version: 0.41.1
+Version: 0.41.2
 Summary: SDMX structural metadata library for Python
 Keywords: sdmx,statistics,metadata,datastructure
 Author: gabrielgellner

{sdmxlib-0.41.1 → sdmxlib-0.41.2}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "sdmxlib"
-version = "0.41.1"
+version = "0.41.2"
 description = "SDMX structural metadata library for Python"
 readme = "README.md"
 license = { text = "Apache-2.0" }

{sdmxlib-0.41.1 → sdmxlib-0.41.2}/src/sdmxlib/formats/sdmx_json/_data_common.py

@@ -294,11 +294,14 @@ class ComponentTable:
         # so position N maps to the Nth-inserted entry. ``to_json``
         # walks this directly.
         self._codes: dict[str, dict[str, dict[str, object]]] = {cid: {} for cid in component_ids}
-        # static label cache (codelist code id → localized name). Read
-        # at code-registration time to fill the ``name`` field; never
-        # iterated for emission, so unobserved codes stay out of the
-        # serialised values array.
-        self._code_labels: dict[str, dict[str, str]] = {cid: {} for cid in component_ids}
+        # Per-component resolved codelists, stored by reference. Labels are
+        # resolved on-demand in ``position()`` via ``ItemList.get`` (O(1));
+        # we deliberately do NOT walk the full codelist here. For census-scale
+        # codelists (CL_GEOGRAPHY: 579k codes) the eager walk dominated wall
+        # time on small data responses — a single-row /data query spent
+        # multiple seconds populating a label cache it would never read more
+        # than once from.
+        self._codelists: dict[str, Any] = {}
 
         for cid in component_ids:
             comp = components.get(cid)
@@ -312,12 +315,10 @@ class ComponentTable:
             self._names[cid] = name
             # TimeDimension carries a Facet representation, not a codelist
             # ref, so it has no ``resolved_codelist`` property — skip the
-            # label-cache step for components without one.
+            # codelist registration for components without one.
             cl = getattr(comp, "resolved_codelist", None)
-            if cl is None:
-                continue
-            for code in cl.codes:
-                self._code_labels[cid][code.id] = code.name.get(lang) or code.id
+            if cl is not None:
+                self._codelists[cid] = cl
 
     def position(self, component_id: str, code_id: str) -> int:
         """Look up ``code_id`` under ``component_id``, assigning on first sight."""
@@ -329,9 +330,16 @@ class ComponentTable:
         table[code_id] = pos
         self._position_str[component_id][code_id] = str(pos)
         # Both SDMX-JSON 1.0 and 2.0 schemas require every value entry
-        # to carry id + name. Pull the localized name from the codelist
-        # label cache if known, otherwise fall back to the id.
-        label = self._code_labels[component_id].get(code_id, code_id)
+        # to carry id + name. Look the localized name up on-demand from
+        # the resolved codelist's id-indexed ItemList (O(1)); fall back
+        # to the id when no codelist is registered or the code is
+        # unknown to it (e.g. observed values outside the enumerated set).
+        label = code_id
+        cl = self._codelists.get(component_id)
+        if cl is not None:
+            code = cl.codes.get(code_id)
+            if code is not None:
+                label = code.name.get(self._lang) or code_id
         entry: dict[str, object] = {"id": code_id, "name": label}
         # TIME_PERIOD start/end is resolved in to_json() once the full set of
         # co-occurring FREQ values is known (see _resolve_time_bounds).