secs-gem/tests/test_secs2.cpp

#include <doctest/doctest.h>

#include "secsgem/secs2/codec.hpp"
#include "secsgem/secs2/item.hpp"
#include "secsgem/secs2/message.hpp"

using namespace secsgem::secs2;

TEST_CASE("encode known byte layouts") {
  CHECK(encode(Item::ascii("ABC")) ==
        std::vector<uint8_t>{0x41, 0x03, 'A', 'B', 'C'});
  CHECK(encode(Item::u1(uint8_t{255})) ==
        std::vector<uint8_t>{0xA5, 0x01, 0xFF});
  CHECK(encode(Item::u4(uint32_t{0x01020304})) ==
        std::vector<uint8_t>{0xB1, 0x04, 0x01, 0x02, 0x03, 0x04});
  CHECK(encode(Item::boolean(true)) ==
        std::vector<uint8_t>{0x25, 0x01, 0x01});
  CHECK(encode(Item::binary({0xDE, 0xAD})) ==
        std::vector<uint8_t>{0x21, 0x02, 0xDE, 0xAD});

  // L [2] <A "MD"> <U2 258>
  CHECK(encode(Item::list({Item::ascii("MD"), Item::u2(uint16_t{258})})) ==
        std::vector<uint8_t>{0x01, 0x02, 0x41, 0x02, 'M', 'D', 0xA9, 0x02, 0x01, 0x02});
}

TEST_CASE("round-trip scalar items") {
  auto rt = [](const Item& i) { return decode(encode(i)); };

  CHECK(rt(Item::ascii("hello world")) == Item::ascii("hello world"));
  CHECK(rt(Item::ascii("")) == Item::ascii(""));
  CHECK(rt(Item::u1(uint8_t{0})) == Item::u1(uint8_t{0}));
  CHECK(rt(Item::u2(uint16_t{0xBEEF})) == Item::u2(uint16_t{0xBEEF}));
  CHECK(rt(Item::u4(uint32_t{0xDEADBEEF})) == Item::u4(uint32_t{0xDEADBEEF}));
  CHECK(rt(Item::u8(uint64_t{0x0102030405060708ULL})) == Item::u8(uint64_t{0x0102030405060708ULL}));
  CHECK(rt(Item::i1(int8_t{-5})) == Item::i1(int8_t{-5}));
  CHECK(rt(Item::i2(int16_t{-1234})) == Item::i2(int16_t{-1234}));
  CHECK(rt(Item::i4(int32_t{-123456})) == Item::i4(int32_t{-123456}));
  CHECK(rt(Item::i8(int64_t{-9000000000LL})) == Item::i8(int64_t{-9000000000LL}));
  CHECK(rt(Item::binary({1, 2, 3, 255, 0})) == Item::binary({1, 2, 3, 255, 0}));
  CHECK(rt(Item::boolean({1, 0, 1})) == Item::boolean({1, 0, 1}));
}

TEST_CASE("round-trip floating point preserves bit pattern") {
  CHECK(decode(encode(Item::f4(3.14159f))) == Item::f4(3.14159f));
  CHECK(decode(encode(Item::f8(2.718281828459045))) == Item::f8(2.718281828459045));
  CHECK(decode(encode(Item::f4({1.0f, -2.5f, 0.0f}))) == Item::f4({1.0f, -2.5f, 0.0f}));
}

TEST_CASE("round-trip nested lists") {
  Item nested = Item::list({
      Item::ascii("EQUIP-01"),
      Item::list({Item::u4(uint32_t{12}), Item::u4(uint32_t{34})}),
      Item::list({}),  // empty list
      Item::binary({0xFF}),
  });
  CHECK(decode(encode(nested)) == nested);
}

TEST_CASE("multi-element numeric arrays") {
  auto v = Item::u2({1, 2, 3, 4, 5});
  CHECK(v.size() == 5);
  CHECK(decode(encode(v)) == v);
}

TEST_CASE("decode rejects truncated input") {
  CHECK_THROWS_AS(decode(std::vector<uint8_t>{0x41, 0x03, 'A'}), CodecError);  // ascii len 3, 1 byte
  CHECK_THROWS_AS(decode(std::vector<uint8_t>{0x41}), CodecError);             // missing length byte
  CHECK_THROWS_AS(decode(std::vector<uint8_t>{}), CodecError);                 // empty
}

TEST_CASE("decode rejects trailing bytes") {
  CHECK_THROWS_AS(decode(std::vector<uint8_t>{0xA5, 0x01, 0xFF, 0x00}), CodecError);
}

TEST_CASE("message body round-trip") {
  Item body = Item::list({Item::ascii("MDLN-1"), Item::ascii("1.0.0")});
  Message m(1, 2, false, body);

  auto bytes = m.encode_body();
  Message decoded = Message::from_body(1, 2, false, bytes);

  REQUIRE(decoded.body.has_value());
  CHECK(*decoded.body == body);
  CHECK(decoded.stream == 1);
  CHECK(decoded.function == 2);
}

TEST_CASE("empty message body") {
  Message m(1, 1, true);  // S1F1 W, header-only
  CHECK(m.encode_body().empty());

  Message decoded = Message::from_body(1, 1, true, {});
  CHECK_FALSE(decoded.body.has_value());
}

TEST_CASE("SML rendering") {
  Item body = Item::list({Item::ascii("MDLN"), Item::u4(uint32_t{42})});
  CHECK(to_sml(body) == "<L [2] <A \"MDLN\" > <U4 42 > >");
}

TEST_CASE("JIS-8 encode/decode (E5 §9.5)") {
  // Format byte for JIS-8 = 0x11 << 2 | 0x01 = 0x45 with 1-byte length.
  // 3 bytes payload "abc" (we don't bother with real JIS chars; the wire
  // format is byte-identical to ASCII, only the format code differs).
  Item j = Item::jis8("abc");
  auto bytes = encode(j);
  CHECK(bytes == std::vector<uint8_t>{0x45, 0x03, 'a', 'b', 'c'});

  Item back = decode(bytes);
  CHECK(back.format() == Format::JIS8);
  CHECK(back == j);
}

TEST_CASE("C2 (Unicode 2-byte) encode/decode (E5 §9.5)") {
  // 0x12 << 2 | 0x01 = 0x49 format byte, 1-byte length, then 2 bytes per
  // code point big-endian. Code points: U+00E9 (é), U+4E2D (中).
  Item c = Item::c2({0x00E9, 0x4E2D});
  auto bytes = encode(c);
  CHECK(bytes == std::vector<uint8_t>{0x49, 0x04, 0x00, 0xE9, 0x4E, 0x2D});

  Item back = decode(bytes);
  CHECK(back.format() == Format::C2);
  CHECK(back == c);
}

TEST_CASE("JIS-8 and C2 disambiguate from ASCII / U2 by Format") {
  // Same backing storage, different format code → not equal.
  CHECK(Item::jis8("hi") != Item::ascii("hi"));
  CHECK(Item::c2({0x41, 0x42}) != Item::u2({0x41, 0x42}));
}

TEST_CASE("SML rendering tags JIS-8 with J and C2 with C") {
  CHECK(to_sml(Item::jis8("hi"))         == "<J \"hi\" >");
  CHECK(to_sml(Item::c2({0x41, 0x42}))   == "<C 65 66 >");
}