Remove sorting.py (#8000)

Co-authored-by: Przemysław Uznański <[email protected]>
GitOrigin-RevId: 9871913a06a1688ac33bd077377943f8ac71e104

CHANGELOG.md

@@ -10,6 +10,7 @@ This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 
 ### Changed
 - **BREAKING**: `pw.io.deltalake.read` now requires explicit specification of primary key fields.
+- **BREAKING**: `pw.indexing.build_sorted_index`, `pw.indexing.retrieve_prev_next_values`, `pw.indexing.sort_from_index` and `pw.indexing.SortedIndex` are removed. Sorting is now done with `pw.Table.sort`.
 
 ### Fixed
 - `generate_class` method in `Schema` now correctly renders columns of `UnionType` and `None` types.

python/pathway/stdlib/indexing/__init__.py

@@ -19,12 +19,6 @@
     UsearchKnnFactory,
 )
 from .retrievers import AbstractRetrieverFactory
-from .sorting import (
-    SortedIndex,
-    build_sorted_index,
-    retrieve_prev_next_values,
-    sort_from_index,
-)
 from .vector_document_index import (
     default_brute_force_knn_document_index,
     default_lsh_knn_document_index,
@@ -48,13 +42,9 @@
     "TantivyBM25Factory",
     "HybridIndex",
     "HybridIndexFactory",
-    "SortedIndex",
     "default_vector_document_index",
     "default_lsh_knn_document_index",
     "default_usearch_knn_document_index",
     "default_brute_force_knn_document_index",
     "default_full_text_document_index",
-    "retrieve_prev_next_values",
-    "sort_from_index",
-    "build_sorted_index",
 ]

python/pathway/stdlib/statistical/_interpolate.py

@@ -29,6 +29,29 @@ class InterpolateMode(Enum):
     LINEAR = 0
 
 
+def _retrieving_prev_next_value(tab: pw.Table) -> pw.Table:
+    return tab.with_columns(
+        prev_value=pw.coalesce(
+            pw.this.prev_value, tab.ix(pw.this.prev, optional=True).prev_value
+        ),
+        next_value=pw.coalesce(
+            pw.this.next_value, tab.ix(pw.this.next, optional=True).next_value
+        ),
+    )
+
+
+def _retrieve_prev_next_values(ordered_table: pw.Table) -> pw.Table:
+
+    ordered_table = ordered_table[["prev", "next", "value"]]
+    ordered_table = ordered_table.with_columns(
+        prev_value=pw.require(pw.this.id, pw.this.value),
+        next_value=pw.require(pw.this.id, pw.this.value),
+    )
+    return pw.iterate(_retrieving_prev_next_value, tab=ordered_table)[
+        ["prev_value", "next_value"]
+    ]
+
+
 @trace_user_frame
 def interpolate(
     self: pw.Table,
@@ -78,8 +101,6 @@ def interpolate(
     6         | 6.0      | 60.0
     """
 
-    from pathway.stdlib.indexing.sorting import retrieve_prev_next_values
-
     if mode != InterpolateMode.LINEAR:
         raise ValueError(
             """interpolate: Invalid mode. Only Interpolate.LINEAR is currently available."""
@@ -119,7 +140,7 @@ def interpolate(
             timestamp=timestamp, value=value
         )
 
-        table_with_prev_next = retrieve_prev_next_values(sorted_timestamp_value)
+        table_with_prev_next = _retrieve_prev_next_values(sorted_timestamp_value)
 
         interpolated_table = table_with_prev_next + sorted_timestamp_value