substrate: Add basic file utility to read all the lines from a file

substrate/file_utils

@@ -0,0 +1,66 @@
+// SPDX-License-Identifier: BSD-3-Clause
+
+#ifndef SUBSTRATE_FILE_UTILS
+#define SUBSTRATE_FILE_UTILS
+
+#include <algorithm>
+#include <iterator>
+#include <string>
+#include <vector>
+#include <substrate/fd>
+#include <substrate/mmap>
+
+namespace substrate
+{
+	template<typename Str = std::string>
+		std::vector<Str> read_lines(fd_t &file, typename Str::value_type separator = '\n') noexcept
+	{
+		using char_t = typename Str::value_type;
+
+		if (!file.valid())
+			return {};
+
+		const auto map{file.map(PROT_READ)};
+		if (!map.valid())
+			return {};
+
+		std::vector<Str> result;
+
+		const auto *const cbegin{map.address<char_t>()};
+
+		const auto *const cend
+		{
+			[&]()
+			{
+				const auto *begin{cbegin};
+				std::advance(begin, map.length() / sizeof(char_t));
+				return begin;
+			}()
+		};
+
+		for (const auto *begin{cbegin}; begin < cend;)
+		{
+			const auto *boundary
+			{
+				std::find_if
+				(
+					begin,
+					cend,
+					[&](const auto chr) { return chr == separator || (separator == '\n' && chr == '\r'); }
+				)
+			};
+
+			result.emplace_back(begin, boundary);
+			if (separator == '\n' && *boundary == '\r' && (boundary + 1) != cend && *(boundary + 1) == '\n')
+				std::advance(boundary, 2);
+			else
+				std::advance(boundary, 1);
+
+			begin = boundary;
+		}
+
+		return result;
+	}
+} // namespace substrate
+
+#endif // SUBSTRATE_FILE_UTILS

test/file_utils.cxx

@@ -0,0 +1,110 @@
+// SPDX-License-Identifier: BSD-3-Clause
+#include <array>
+#include <substrate/file_utils>
+#include <substrate/indexed_iterator>
+#include <catch2/catch_test_macros.hpp>
+
+using substrate::fd_t;
+using substrate::indexedIterator_t;
+using substrate::read_lines;
+
+TEST_CASE("lines reader", "[boundedIterator_t]")
+{
+	const std::array<std::string, 6> lines
+	{
+		{
+			"## Enforcement Responsibilities 1",
+			"",
+			"Community leaders are responsible for clarifying and enforcing our standards of",
+			"acceptable behavior and will take appropriate and fair corrective action in",
+			"response to any behavior that they deem inappropriate, threatening, offensive,",
+			"or harmful."
+		}
+	};
+
+	{
+		fd_t text{"info.txt", O_WRONLY | O_CREAT | O_EXCL | O_TEXT, substrate::normalMode};
+		REQUIRE(text.valid());
+		for (const auto &i : lines)
+		{
+			REQUIRE(text.write(i));
+			REQUIRE(text.write('\n'));
+		}
+	}
+
+	{
+		fd_t text{"info.txt", O_RDONLY | O_TEXT};
+		REQUIRE(text.valid());
+		// Memory mapping here will cause the conversion to be skipped
+		const auto input {read_lines(text)};
+		REQUIRE(lines.size() == input.size());
+		for (const auto &line : indexedIterator_t<decltype(lines)>{lines})
+		{
+			const auto &contents {line.second};
+			const auto &input_line {input[line.first]};
+			REQUIRE(input_line.find_first_of('\n', input_line.size() - 1) == std::string::npos);
+			REQUIRE(input_line.find_first_of('\r', input_line.size() - 1) == std::string::npos);
+			REQUIRE(input_line == contents);
+		}
+	}
+
+	{
+		const auto result{unlink("info.txt")};
+		if (result != 0)
+			REQUIRE(errno == 0);
+		else
+			SUCCEED();
+	}
+}
+
+TEST_CASE("lines reader with std::u16string", "[boundedIterator_t]")
+{
+	const std::array<std::u16string, 6> lines
+	{{
+		u"## Enforcement Responsibilities 2",
+		u"",
+		u"Community leaders are responsible for clarifying and enforcing our standards of",
+		u"acceptable behavior and will take appropriate and fair corrective action in",
+		u"response to any behavior that they deem inappropriate, threatening, offensive,",
+		u"or harmful."
+	}};
+
+	{
+		fd_t text{"info.txt", O_WRONLY | O_CREAT | O_EXCL | O_BINARY, substrate::normalMode};
+		REQUIRE(text.valid());
+		for (const auto &line : lines)
+		{
+			REQUIRE(text.write(line.data(), line.size() * sizeof(std::u16string::value_type)));
+			REQUIRE(text.write(u'\n'));
+		}
+	}
+
+	{
+		fd_t text{"info.txt", O_RDONLY | O_BINARY};
+		REQUIRE(text.valid());
+		const auto input{substrate::read_lines<std::u16string>(text)};
+		REQUIRE(lines.size() == input.size());
+		for (const auto &line : indexedIterator_t<decltype(lines)>{lines})
+		{
+			const auto &contents {line.second};
+			const auto &input_line {input[line.first]};
+			REQUIRE(input_line.find_first_of(u'\n', input_line.size() - 1) == std::u16string::npos);
+			REQUIRE(input_line.find_first_of(u'\r', input_line.size() - 1) == std::u16string::npos);
+			REQUIRE(input_line == contents);
+		}
+	}
+
+	{
+		const auto result{unlink("info.txt")};
+		if (result != 0)
+			REQUIRE(errno == 0);
+		else
+			SUCCEED();
+	}
+}
+
+TEST_CASE()
+{
+	unlink("info.txt");
+	SUCCEED();
+}

test/meson.build

@@ -6,7 +6,7 @@ testSrcs = [
 	'buffer_utils.cxx', 'pointer_utils.cxx',
 	'crypto/twofish.cxx', 'crypto/sha256.cxx', 'crypto/sha512.cxx',
 	'zip_container.cxx', 'affinity.cxx', 'threaded_queue.cxx', 'thread_pool.cxx',
-	'mmap.cxx'
+	'mmap.cxx', 'file_utils.cxx'
 ]
 
 if target_machine.system() == 'linux'