added GroupBy and Rows calls

CHANGELOG.md

@@ -1,9 +1,12 @@
 ## Change Log
 
+* **master**
+    * Added `field.rows` and `index.group_by` calls.
+
 * **v1.2.0** (2018-12-21)
     * **Compatible with Pilosa 1.2**
     * Added `track_existence` index option.
-    * Added `not_` index method to support `Not` queries. The corresponding index must be created with `track_existence=True`.  
+    * Added `not_` index method to support `Not` queries. The corresponding index must be created with `track_existence=True`.
     * Added mutex and bool fields.
     * Added `field.clear_row`, `field.store` and `index.options` calls.
     * Added support for roaring importing `RowIDColumnID` with timestamp data.

docs/data-model-queries.md

@@ -103,31 +103,33 @@ Please check [Pilosa documentation](https://www.pilosa.com/docs) for PQL details
 
 Index:
 
-* `union(self, *rows)`
-* `intersect(self, *rows)`
-* `difference(self, *rows)`
-* `count(self, row)`
-* `set_column_attrs(self, column_id, attrs)`
-* `xor(self, *rows)`
-* `not_(self, row)`
-* `options(self, row_query, column_attrs=False, exclude_columns=False, exclude_row_attrs=False, shards=[])`
+* `union(*rows)`
+* `intersect(*rows)`
+* `difference(*rows)`
+* `count(row)`
+* `set_column_attrs(column_id, attrs)`
+* `xor(*rows)`
+* `not_(row)`
+* `options(row_query, column_attrs=False, exclude_columns=False, exclude_row_attrs=False, shards=[])`
+* `group_by(*rows_queries, limit=0, filter=None)`
 
 Field:
 
-* `row(self, row_id)`
-* `set(self, row_id, column_id, timestamp=None)`
-* `clear(self, row_id, column_id)`
-* `topn(self, n, row=None, field="", *values)`
-* `range(self, row_id, start, end)`
-* `set_row_attrs(self, row_id, attrs)`
-* `lt(self, n)`
-* `lte(self, n)`
-* `gt(self, n)`
-* `gte(self, n)`
-* `between(self, a, b)`
-* `sum(self, row=None)`
-* `min(self, row=None)`
-* `max(self, row=None)`
-* `setvalue(self, column_id, value)`
-* `store(self, row_query, row)`
-* `clear_row(self, row)`
+* `row(row_id)`
+* `set(row_id, column_id, timestamp=None)`
+* `clear(row_id, column_id)`
+* `topn(n, row=None, field="", *values)`
+* `range(row_id, start, end)`
+* `set_row_attrs(row_id, attrs)`
+* `lt(n)`
+* `lte(n)`
+* `gt(n)`
+* `gte(n)`
+* `between(a, b)`
+* `sum(row=None)`
+* `min(row=None)`
+* `max(row=None)`
+* `setvalue(column_id, value)`
+* `store(row_query, row)`
+* `clear_row(row)`
+* `rows(prev_row=None, limit=0, column=None)`

integration_tests/test_client_it.py

@@ -44,6 +44,7 @@
 from pilosa.client import Client, URI, Cluster, PilosaServerError
 from pilosa.exceptions import PilosaError
 from pilosa.orm import Index, TimeQuantum, Schema, CacheType
+from pilosa.response import GroupCount, FieldRow
 from pilosa.imports import csv_column_reader, csv_field_value_reader, \
     csv_column_id_value, csv_column_key_value, csv_row_key_column_id
 
@@ -579,6 +580,38 @@ def test_range_field(self):
         self.assertEquals(1, len(response.results))
         self.assertEquals(10, response.result.row.columns[0])
 
+    def test_rows(self):
+        client = self.get_client()
+        index = self.index
+        field = index.field("rowsfield")
+        client.ensure_field(field)
+        client.query(index.batch_query(
+            field.set(1, 100),
+            field.set(1, 200),
+            field.set(2, 200),
+        ))
+        resp = client.query(field.rows())
+        target = [1, 2]
+        self.assertEqual(target, resp.result.row_identifiers.ids)
+
+    def test_group_by(self):
+        client = self.get_client()
+        index = self.index
+        field = index.field("groupbyfield")
+        client.ensure_field(field)
+        client.query(index.batch_query(
+            field.set(1, 100),
+            field.set(1, 200),
+            field.set(2, 200),
+        ))
+        resp = client.query(index.group_by(field.rows()))
+        target = [
+            GroupCount([FieldRow("groupbyfield", 1)], 2),
+            GroupCount([FieldRow("groupbyfield", 2)], 1),
+        ]
+        self.assertEqual(target, resp.result.group_counts)
+
+
     def test_exclude_attrs_columns(self):
         client = self.get_client()
         client.query(self.col_index.batch_query(

pilosa/imports.py

@@ -80,8 +80,7 @@ def __init__(self, column_id=0, column_key="", value=0):
         self.value = value
 
     def __hash__(self):
-        return hash("%s:%s:%s" % \
-            (self.column_id, self.column_key, self.value))
+        return hash((self.column_id, self.column_key, self.value))
 
     def __eq__(self, other):
         if id(self) == id(other):

pilosa/internal/public.proto

@@ -1,5 +1,7 @@
 syntax = "proto3";
 
+package internal;
+
 message Row {
 	repeated uint64 Columns = 1;
 	repeated string Keys = 3;
@@ -17,12 +19,13 @@ message Pair {
 	uint64 Count = 2;
 }
 
-message FieldRow {
+message FieldRow{
 	string Field = 1;
 	uint64 RowID = 2;
+	string RowKey = 3;
 }
 
-message GroupCount {
+message GroupCount{
 	repeated FieldRow Group = 1;
 	uint64 Count = 2;
 }
@@ -32,12 +35,6 @@ message ValCount {
 	int64 Count = 2;
 }
 
-message Bit {
-	uint64 RowID = 1;
-	uint64 ColumnID = 2;
-	int64 Timestamp = 3;
-}
-
 message ColumnAttrSet {
 	uint64 ID = 1;
 	string Key = 3;
@@ -78,7 +75,7 @@ message QueryResult {
 	uint64 N = 2;
 	repeated Pair Pairs = 3;
 	bool Changed = 4;
-	ValCount ValCount = 5;
+    ValCount ValCount = 5;
 	repeated uint64 RowIDs = 7;
 	repeated GroupCount GroupCounts = 8;
 	RowIdentifiers RowIdentifiers = 9;
@@ -104,6 +101,16 @@ message ImportValueRequest {
 	repeated int64 Values = 6;
 }
 
+message TranslateKeysRequest {
+	string Index = 1;
+	string Field = 2;
+	repeated string Keys = 3;
+}
+
+message TranslateKeysResponse {
+	repeated uint64 IDs = 3;
+}
+
 message ImportRoaringRequestView {
 	string Name = 1;
 	bytes Data = 2;