Optimizing a bit

First make a sort based on column id. This will allow us to only pass through the array of conditions once. Comparing column indecies before making dynamic casts. Early out if column has search index.
realm · Mar 4, 2019 · 1011e96 · 1011e96
1 parent d973a7c
commit 1011e96
Show file tree

Hide file tree

Showing 3 changed files with 39 additions and 28 deletions.
diff --git a/src/realm/query_engine.cpp b/src/realm/query_engine.cpp
@@ -326,30 +326,27 @@ void StringNode<Equal>::_search_index_init()
     }
 }
 
-bool StringNode<Equal>::try_consume_condition(StringNode<Equal>* other)
+void StringNode<Equal>::consume_condition(StringNode<Equal>* other)
 {
     // If a search index is present, don't try to combine conditions since index search is most likely faster.
     // Assuming N elements to search and M conditions to check:
     // 1) search index present:                     O(log(N)*M)
     // 2) no search index, combine conditions:      O(N)
     // 3) no search index, conditions not combined: O(N*M)
     // In practice N is much larger than M, so if we have a search index, choose 1, otherwise if possible choose 2.
-    if (!m_condition_column->has_search_index() && m_condition_column == other->m_condition_column) {
-        REALM_ASSERT(other->m_needles.empty());
-        if (m_needles.empty()) {
-            m_needles.insert(bool(m_value) ? StringData(*m_value) : StringData());
-        }
-        if (bool(other->m_value)) {
-            m_needle_storage.push_back(StringBuffer());
-            m_needle_storage.back().append(*other->m_value);
-            m_needles.insert(StringData(m_needle_storage.back().data(), m_needle_storage.back().size()));
-        }
-        else {
-            m_needles.insert(StringData());
-        }
-        return true;
+    REALM_ASSERT(m_condition_column == other->m_condition_column);
+    REALM_ASSERT(other->m_needles.empty());
+    if (m_needles.empty()) {
+        m_needles.insert(bool(m_value) ? StringData(*m_value) : StringData());
+    }
+    if (bool(other->m_value)) {
+        m_needle_storage.push_back(StringBuffer());
+        m_needle_storage.back().append(*other->m_value);
+        m_needles.insert(StringData(m_needle_storage.back().data(), m_needle_storage.back().size()));
+    }
+    else {
+        m_needles.insert(StringData());
     }
-    return false;
 }
 
 // Requirements of template types:

diff --git a/src/realm/query_engine.hpp b/src/realm/query_engine.hpp
@@ -1186,6 +1186,11 @@ class StringNodeBase : public ParentNode {
         do_verify_column(m_condition_column);
     }
 
+    bool has_search_index() const
+    {
+        return m_condition_column->has_search_index();
+    }
+
     void init() override
     {
         ParentNode::init();
@@ -1549,7 +1554,7 @@ class StringNode<Equal> : public StringNodeEqualBase {
 
     void _search_index_init() override;
 
-    bool try_consume_condition(StringNode<Equal>* other);
+    void consume_condition(StringNode<Equal>* other);
 
     std::unique_ptr<ParentNode> clone(QueryNodeHandoverPatches* patches) const override
     {
@@ -1675,18 +1680,27 @@ class OrNode : public ParentNode {
         m_dD = 10.0;
 
         StringNode<Equal>* first = nullptr;
-        StringNode<Equal>* advance = nullptr;
-        for (auto it = m_conditions.begin(); it != m_conditions.end(); ++it) {
-            if (bool(*it) && (first = dynamic_cast<StringNode<Equal>*>(it->get()))) {
-                for (auto next = it + 1; next != m_conditions.end(); ) {
-                    if ((advance = dynamic_cast<StringNode<Equal>*>(next->get()))) {
-                        if (first->try_consume_condition(advance)) {
-                            next = m_conditions.erase(next);
-                            continue;
-                        }
+        std::sort(m_conditions.begin(), m_conditions.end(),
+                  [](auto& a, auto& b) { return a->m_condition_column_idx < b->m_condition_column_idx; });
+        auto it = m_conditions.begin();
+        while (it != m_conditions.end()) {
+            // Only try to optimize on StringNode<Equal> conditions without search index
+            if (bool(*it) && (first = dynamic_cast<StringNode<Equal>*>(it->get())) && !first->has_search_index()) {
+                auto col_ndx = first->m_condition_column_idx;
+                auto next = it + 1;
+                while (next != m_conditions.end() && (*next)->m_condition_column_idx == col_ndx) {
+                    if (auto advance = dynamic_cast<StringNode<Equal>*>(next->get())) {
+                        first->consume_condition(advance);
+                        next = m_conditions.erase(next);
+                    }
+                    else {
+                        ++next;
                     }
-                    ++next;
                 }
+                it = next;
+            }
+            else {
+                ++it;
             }
         }
 

diff --git a/test/test_parser.cpp b/test/test_parser.cpp
@@ -2770,7 +2770,7 @@ TEST(Parser_ChainedStringEqualQueries)
     rd.shuffle(populated_data.begin(), populated_data.end());
     std::string query;
     bool first = true;
-    size_t column_to_query = 0;
+    char column_to_query = 0;
     for (auto s : populated_data) {
         std::string column_name(1, 'a' + column_to_query);
         query += (first ? "" : " or " ) + column_name + " == '" + s + "'";