find_first, find_min, find_max, count added

PDB writing changed for auth_seq_num
version bump

CMakeLists.txt

@@ -25,7 +25,7 @@
 cmake_minimum_required(VERSION 3.16)
 
 # set the project name
-project(cifpp VERSION 5.0.7.1 LANGUAGES CXX)
+project(cifpp VERSION 5.0.8 LANGUAGES CXX)
 
 list(PREPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake")
 

changelog

+Version 5.0.8
+- implemented find_first, find_min, find_max and count in category
+- find1 now throws an exception if condition does not not exactly match one row
+- Change in writing out PDB files, now looking up the original auth_seq_num
+  via the pdbx_xxx_scheme categories based on the atom_site.auth_seq_num ->
+  pdbx_xxx_scheme.pdb_seq_num relationship.
+
 Version 5.0.7.1
 - Use the implementation from zeep for std::experimental::is_detected
 

include/cif++/condition.hpp

@@ -182,6 +182,33 @@ namespace detail
 		uint16_t m_item_ix = 0;
 	};
 
+	struct key_is_not_empty_condition_impl : public condition_impl
+	{
+		key_is_not_empty_condition_impl(const std::string &item_tag)
+			: m_item_tag(item_tag)
+		{
+		}
+
+		condition_impl *prepare(const category &c) override
+		{
+			m_item_ix = get_column_ix(c, m_item_tag);
+			return this;
+		}
+
+		bool test(row_handle r) const override
+		{
+			return not r[m_item_ix].empty();
+		}
+
+		void str(std::ostream &os) const override
+		{
+			os << m_item_tag << " IS NOT NULL";
+		}
+
+		std::string m_item_tag;
+		uint16_t m_item_ix = 0;
+	};
+
 	struct key_equals_condition_impl : public condition_impl
 	{
 		key_equals_condition_impl(item &&i)
@@ -824,6 +851,11 @@ inline condition operator==(const key &key, const empty_type &)
 	return condition(new detail::key_is_empty_condition_impl(key.m_item_tag));
 }
 
+inline condition operator!=(const key &key, const empty_type &)
+{
+	return condition(new detail::key_is_not_empty_condition_impl(key.m_item_tag));
+}
+
 inline condition operator not(condition &&rhs)
 {
 	return condition(new detail::not_condition_impl(std::move(rhs)));

src/model.cpp

@@ -1780,7 +1780,7 @@ std::string structure::insert_compound(const std::string &compoundID, bool is_en
 	{
 		auto &pdbxEntityNonpoly = m_db["pdbx_entity_nonpoly"];
 
-		entity_id = pdbxEntityNonpoly.find1<std::string>("comp_id"_key == compoundID, "entity_id");
+		entity_id = pdbxEntityNonpoly.find_first<std::string>("comp_id"_key == compoundID, "entity_id");
 
 		if (entity_id.empty())
 		{
@@ -1954,7 +1954,7 @@ void structure::change_residue(residue &res, const std::string &newCompound,
 		// create a copy of the entity first
 		auto &entity = m_db["entity"];
 
-		entityID = entity.find1<std::string>("type"_key == "non-polymer" and "pdbx_description"_key == compound->name(), "id");
+		entityID = entity.find_first<std::string>("type"_key == "non-polymer" and "pdbx_description"_key == compound->name(), "id");
 
 		if (entityID.empty())
 		{
@@ -2573,7 +2573,7 @@ std::string structure::create_entity_for_branch(branch &branch)
 
 	auto &entity = m_db["entity"];
 
-	std::string entityID = entity.find1<std::string>("type"_key == "branched" and "pdbx_description"_key == entityName, "id");
+	std::string entityID = entity.find_first<std::string>("type"_key == "branched" and "pdbx_description"_key == entityName, "id");
 
 	if (entityID.empty())
 	{

test/unit-v2-test.cpp

@@ -1774,8 +1774,6 @@ _test.name
 
 BOOST_AUTO_TEST_CASE(c3)
 {
-	cif::VERBOSE = 1;
-
 	auto f = R"(data_TEST
 #
 loop_
@@ -1811,6 +1809,43 @@ _test.name
 	BOOST_CHECK_EQUAL(name, "aap");
 }
 
+BOOST_AUTO_TEST_CASE(c4)
+{
+	auto f = R"(data_TEST
+#
+loop_
+_test.id
+_test.name
+1 aap
+2 noot
+3 mies
+4 .
+5 ?
+    )"_cf;
+
+	auto &db = f.front();
+
+	// query tests
+	BOOST_TEST(db["test"].find_max<int>("id") == 5);
+	BOOST_TEST(db["test"].find_max<int>("id", cif::key("name") != cif::null) == 3);
+
+	BOOST_TEST(db["test"].find_min<int>("id") == 1);
+	BOOST_TEST(db["test"].find_min<int>("id", cif::key("name") == cif::null) == 4);
+
+	// count tests
+	BOOST_TEST(db["test"].count(cif::all()) == 5);
+	BOOST_TEST(db["test"].count(cif::key("name") != cif::null) == 3);
+	BOOST_TEST(db["test"].count(cif::key("name") == cif::null) == 2);
+
+	// find_first tests
+	BOOST_TEST(db["test"].find_first<int>(cif::key("id") == 1, "id") == 1);
+	BOOST_TEST(db["test"].find_first<int>(cif::all(), "id") == 1);
+
+	// find1 tests
+	BOOST_TEST(db["test"].find1<int>(cif::key("id") == 1, "id") == 1);
+	BOOST_CHECK_THROW(db["test"].find1<int>(cif::all(), "id"), cif::multiple_results_error);
+}
+
 // --------------------------------------------------------------------
 // rename test