Merge branch 'develop'

README.md

@@ -56,25 +56,25 @@ For example: list of ProductId "in stock" to exclude not available products.
 
 Tests on sets of products with 10 attributes, search with filters by 3 fields.
 
-PHPBench v2.0.3 PHP 8.1.0 + JIT + opcache (no xdebug extension)
+PHPBench v2.1.0 ArrayIndex PHP 8.1.0 + JIT + opcache (no xdebug extension)
 
 | Items count | Memory | Find | Get Filters (aggregates) | Sort by field| Results Found |
 |----------------:|---------:|-----------------:|-------------------------:|-------------:|-----------------:|
 | 10,000 | ~6Mb | ~0.0003 s. | ~0.002 s. | ~0.0001 s. | 907 |
-| 50,000 | ~40Mb | ~0.001 s. | ~0.013 s. | ~0.0006 s. | 4550 |
-| 100,000 | ~80Mb | ~0.003 s. | ~0.029 s. | ~0.001 s. | 8817 |
-| 300,000 | ~189Mb | ~0.011 s. | ~0.108 s. | ~0.005 s. | 26891 |
-| 1,000,000 | ~657Mb | ~0.052 s. | ~0.419 s. | ~0.018 s. | 90520 |
+| 50,000 | ~40Mb | ~0.001 s. | ~0.013 s. | ~0.0005 s. | 4550 |
+| 100,000 | ~80Mb | ~0.003 s. | ~0.030 s. | ~0.001 s. | 8817 |
+| 300,000 | ~189Mb | ~0.011 s. | ~0.101 s. | ~0.005 s. | 26891 |
+| 1,000,000 | ~657Mb | ~0.049 s. | ~0.396 s. | ~0.017 s. | 90520 |
 
-Bench v1.3.3 PHP 8.1.0 + JIT + opcache (no xdebug extension)
+PHPBench v2.1.0 FixedArrayIndex PHP 8.1.0 + JIT + opcache (no xdebug extension)
 
 | Items count | Memory | Find | Get Filters (aggregates) | Sort by field| Results Found |
 |----------------:|---------:|-----------------:|-------------------------:|-------------:|-----------------:|
-| 10,000 | ~7Mb | ~0.0007 s. | ~0.004 s. | ~0.0003 s. | 907 |
-| 50,000 | ~49Mb | ~0.002 s. | ~0.014 s. | ~0.0009 s. | 4550 |
-| 100,000 | ~98Mb | ~0.004 s. | ~0.028 s. | ~0.002 s. | 8817 |
-| 300,000 | ~242Mb | ~0.012 s. | ~0.112 s. | ~0.007 s. | 26891 |
-| 1,000,000 | ~812Mb | ~0.057 s. | ~0.443 s. | ~0.034 s. | 90520 |
+| 10,000 | ~2Mb | ~0.0007 s. | ~0.006 s. | ~0.0002 s. | 907 |
+| 50,000 | ~12Mb | ~0.003 s. | ~0.027 s. | ~0.001 s.  | 4550 |
+| 100,000 | ~23Mb | ~0.006 s. | ~0.057 s. | ~0.002 s. | 8817 |
+| 300,000 | ~70Mb  | ~0.021 s. | ~0.188 s. | ~0.007 s. | 26891 |
+| 1,000,000 | ~233Mb | ~0.080 s. | ~0.674 s. | ~0.032 s. | 90520 |
 
 * Items count - Products in index
 * Memory - RAM used for index
@@ -86,24 +86,24 @@ Bench v1.3.3 PHP 8.1.0 + JIT + opcache (no xdebug extension)
 
 Experimental Golang port bench https:/k-samuel/go-faceted-search
 
-Bench v0.3.1 golang 1.17.3 with parallel aggregates
+Bench v0.3.2 golang 1.17.3 with parallel aggregates
 
 | Items count | Memory | Find | Get Filters (aggregates) | Sort by field| Results Found |
 |----------------:|---------:|-----------------:|-------------------------:|-------------:|-----------------:|
 | 10,000 | ~5Mb | ~0.0004 s. | ~0.001 s. | ~0.0002 s. | 907 |
 | 50,000 | ~15Mb | ~0.002 s. | ~0.010 s. | ~0.001 s. | 4550 |
-| 100,000 | ~21Mb | ~0.006 s. | ~0.028 s. | ~0.002 s. | 8817 |
-| 300,000 | ~47Mb | ~0.020 s. | ~0.091 s. | ~0.010 s. | 26891 |
-| 1,000,000 | ~150Mb | ~0.089 s. | ~0.412 s. | ~0.034 s. | 90520 |
+| 100,000 | ~21Mb | ~0.007 s. | ~0.030 s. | ~0.003 s. | 8817 |
+| 300,000 | ~47Mb | ~0.021 s. | ~0.081 s. | ~0.007 s. | 26891 |
+| 1,000,000 | ~150Mb | ~0.090 s. | ~0.372 s. | ~0.036 s. | 90520 |
 
 ## Examples
 
 Create index using console/crontab etc.
 ```php
 <?php
-use KSamuel\FacetedSearch\Index;
+use KSamuel\FacetedSearch\Index\ArrayIndex;
 
-$searchIndex = new Index();
+$searchIndex = new ArrayIndex();
 /*
  * Getting products data from DB
  * Sort data by $recordId before using Index->addRecord it can improve performance 
@@ -115,7 +115,7 @@ $data = [
 ];
 foreach($data as $item){ 
  $recordId = $item['id'];
- // no ned to add faceted index by id
+ // no need to add faceted index by id
  unset($item['id']);
  $searchIndex->addRecord($recordId, $item);
 }
@@ -129,15 +129,15 @@ Using in application
 
 ```php
 <?php
-use KSamuel\FacetedSearch\Index;
+use KSamuel\FacetedSearch\Index\ArrayIndex;
 use KSamuel\FacetedSearch\Search;
 use KSamuel\FacetedSearch\Filter\ValueFilter;
 use KSamuel\FacetedSearch\Filter\RangeFilter;
 use KSamuel\FacetedSearch\Sorter\ByField;
 
 // load index by product category (use request params)
 $indexData = json_decode(file_get_contents('./first-index.json'), true);
-$searchIndex = new Index();
+$searchIndex = new ArrayIndex();
 $searchIndex->setData($indexData);
 // create search instance
 $search = new Search($searchIndex);
@@ -183,12 +183,12 @@ Note that RangeFilter is slow solution, it is better to avoid facets for highly
 
 ```php
 <?php
-use KSamuel\FacetedSearch\Index;
+use KSamuel\FacetedSearch\Index\ArrayIndex;
 use KSamuel\FacetedSearch\Search;
 use KSamuel\FacetedSearch\Indexer\Number\RangeIndexer;
 use KSamuel\FacetedSearch\Filter\RangeFilter;
 
-$index = new Index();
+$index = new ArrayIndex();
 $rangeIndexer = new RangeIndexer(100);
 $index->addIndexer('price', $rangeIndexer);
 
@@ -210,21 +210,66 @@ $search->find($filters);
 RangeListIndexer allows you to use custom ranges list
 ```php
 <?php
-use KSamuel\FacetedSearch\Index;
+use KSamuel\FacetedSearch\ArrayIndex;
 use KSamuel\FacetedSearch\Indexer\Number\RangeListIndexer;
 
-$index = new Index();
+$index = new ArrayIndex();
 $rangeIndexer = new RangeListIndexer([100,500,1000]); // (0-99)[0],(100-499)[100],(500-999)[500],(1000 & >)[1000] 
 $index->addIndexer('price', $rangeIndexer);
 ```
 Also, you can create your own indexers with range detection method
 
+
+### FixedArrayIndex
+
+FixedArrayIndex is much slower but requires significant less memory.
+Working with an FixedArrayIndex is slightly different from ArrayIndex
+
+The stored index data is compatible, you can transfer it from ArrayIndex to FixedArrayIndex
+
+```php
+<?php
+use KSamuel\FacetedSearch\Index\ArrayIndex;
+use KSamuel\FacetedSearch\Index\FixedArrayIndex;
+
+$searchIndex = new FixedArrayIndex();
+// Switch index into write mode
+$searchIndex->writeMode();
+/*
+ * Getting products data from DB
+ * Sort data by $recordId before using Index->addRecord it can improve performance 
+ */
+$data = [
+ ['id'=>7, 'color'=>'black', 'price'=>100, 'sale'=>true, 'size'=>36], 
+ ['id'=>9, 'color'=>'green', 'price'=>100, 'sale'=>true, 'size'=>40], 
+ // ....
+];
+foreach($data as $item){ 
+ $recordId = $item['id'];
+ // no need to add faceted index by id
+ unset($item['id']);
+ $searchIndex->addRecord($recordId, $item);
+}
+// After the data is added, you need to commit the changes 
+$searchIndex->commitChanges();
+// save index data to some storage 
+$indexData = $searchIndex->getData();
+// We will use file for example
+file_put_contents('./first-index.json', json_encode($indexData));
+
+// Index data is fully compatible. You can create both indexes from the same data 
+$arrayIndex = new ArrayIndex();
+$arrayIndex->setData($indexData);
+
+
+```
+
+
 ### More Examples
 * [Demo](./examples)
 * [Performance test](./tests/performance/readme.md)
 * [Bench](./tests/benchmark/readme.md)
 
-
 ### Tested but discarded concepts
 
 **Bitmap**
@@ -247,4 +292,5 @@ Also, you can create your own indexers with range detection method
 
 # Q&A
 * [Is it possible somehow to implement a full-text filter?](https:/k-samuel/faceted-search/issues/3)
-* [Would that be possible to use a DB as an index instead of a json file?](https:/k-samuel/faceted-search/issues/5)
+* [Would that be possible to use a DB as an index instead of a json file?](https:/k-samuel/faceted-search/issues/5)
+* [Article about project history and base concepts (in Russian)](https://habr.com/ru/post/595765/)

changelog.md

@@ -1,4 +1,35 @@
 # Changelog
+### v2.1.0 (06.01.2022)
+
+### Performance update and FixedArrayIndex
+
+FixedArrayIndex is much slower than ArrayIndex but requires significant less memory.
+
+* FixedArrayIndex added
+* KSamuel\FacetedSearch\Index is deprecated use KSamuel\FacetedSearch\Index\ArrayIndex instead
+* Unit and performance tests for FixedArrayIndex
+* Added sorting of filter fields before processing (performance)
+* Documentation updated
+
+PHPBench v2.1.0 ArrayIndex PHP 8.1.0 + JIT + opcache (no xdebug extension)
+
+| Items count | Memory | Find | Get Filters (aggregates) | Sort by field| Results Found |
+|----------------:|---------:|-----------------:|-------------------------:|-------------:|-----------------:|
+| 10,000 | ~6Mb | ~0.0003 s. | ~0.002 s. | ~0.0001 s. | 907 |
+| 50,000 | ~40Mb | ~0.001 s. | ~0.013 s. | ~0.0005 s. | 4550 |
+| 100,000 | ~80Mb | ~0.003 s. | ~0.030 s. | ~0.001 s. | 8817 |
+| 300,000 | ~189Mb | ~0.011 s. | ~0.101 s. | ~0.005 s. | 26891 |
+| 1,000,000 | ~657Mb | ~0.049 s. | ~0.396 s. | ~0.017 s. | 90520 |
+
+PHPBench v2.1.0 FixedArrayIndex PHP 8.1.0 + JIT + opcache (no xdebug extension)
+
+| Items count | Memory | Find | Get Filters (aggregates) | Sort by field| Results Found |
+|----------------:|---------:|-----------------:|-------------------------:|-------------:|-----------------:|
+| 10,000 | ~2Mb | ~0.0007 s. | ~0.006 s. | ~0.0002 s. | 907 |
+| 50,000 | ~12Mb | ~0.003 s. | ~0.027 s. | ~0.001 s. | 4550 |
+| 100,000 | ~23Mb | ~0.006 s. | ~0.057 s. | ~0.002 s. | 8817 |
+| 300,000 | ~70Mb | ~0.021 s. | ~0.188 s. | ~0.007 s. | 26891 |
+| 1,000,000 | ~233Mb | ~0.080 s. | ~0.674 s. | ~0.032 s. | 90520 |
 
 ### v2.0.3 (30.12.2021)
 Performance update

composer.json

@@ -1,6 +1,6 @@
 {
  "name": "k-samuel/faceted-search",
- "version": "2.0.3",
+ "version": "2.1.0",
  "type": "library",
  "description": "PHP Faceted search",
  "keywords": ["php","faceted search"],

src/Filter/RangeFilter.php

@@ -62,7 +62,14 @@ public function filterResults(array $facetedData, ?array $inputIdKeys = null): a
  continue;
  }
  if (empty($limit)) {
- $limit = $records;
+ /**
+ * @var array<int>|\SplFixedArray<int> $records
+ */
+ if($records instanceof \SplFixedArray){
+ $limit = $records->toArray();
+ }else{
+ $limit = $records;
+ }
  } else {
  // array sum (faster than array_merge here)
  foreach ($records as $item){
@@ -97,6 +104,9 @@ public function filterResults(array $facetedData, ?array $inputIdKeys = null): a
 
  $result = [];
  foreach ($start as $index => $exists) {
+ /**
+ * @var int $index
+ */
  if (isset($compare[$index])) {
  $result[$index] = true;
  }