13762570662

Committed 10 Mar 2025 10:35AM UTC coverage: 78.931% (-0.1%) from 79.05%

Build # 13762570662

Build Type

push

github

Committed by

SuperFola

Commit Message

refactor(builtins): use std::numbers instead of <cmath> for M_PI and HUGE_VAL (float inf)

Run Details

5848 of 7409 relevant lines covered (78.93%)

79635.36 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

42.91

/src/arkreactor/Compiler/BytecodeReader.cpp

#include <Ark/Compiler/BytecodeReader.hpp>

#include <Ark/Compiler/Instructions.hpp>
#include <Ark/Builtins/Builtins.hpp>

#include <iomanip>
#include <unordered_map>
#include <picosha2.h>
#include <Ark/Compiler/Serialization/IEEE754Serializer.hpp>
#include <Ark/Compiler/Serialization/IntegerSerializer.hpp>
#include <fmt/core.h>
#include <fmt/color.h>

namespace Ark
{
    using namespace Ark::internal;

    void BytecodeReader::feed(const bytecode_t& bytecode)
    {
        m_bytecode = bytecode;
    }

    void BytecodeReader::feed(const std::string& file)
    {
        std::ifstream ifs(file, std::ios::binary | std::ios::ate);
        if (!ifs.good())
            throw std::runtime_error(fmt::format("[BytecodeReader] Couldn't open file '{}'", file));

        const auto pos = ifs.tellg();
        // reserve appropriate number of bytes
        std::vector<char> temp(static_cast<std::size_t>(pos));
        ifs.seekg(0, std::ios::beg);
        ifs.read(&temp[0], pos);
        ifs.close();

        m_bytecode = bytecode_t(static_cast<std::size_t>(pos));
        for (std::size_t i = 0; i < static_cast<std::size_t>(pos); ++i)
            m_bytecode[i] = static_cast<uint8_t>(temp[i]);
    }

    bool BytecodeReader::checkMagic() const
    {
        return m_bytecode.size() >= bytecode::Magic.size() &&
            m_bytecode[0] == bytecode::Magic[0] &&
            m_bytecode[1] == bytecode::Magic[1] &&
            m_bytecode[2] == bytecode::Magic[2] &&
            m_bytecode[3] == bytecode::Magic[3];
    }

    const bytecode_t& BytecodeReader::bytecode() noexcept
    {
        return m_bytecode;
    }

    Version BytecodeReader::version() const
    {
        if (!checkMagic() || m_bytecode.size() < bytecode::Magic.size() + bytecode::Version.size())
            return Version { 0, 0, 0 };

        return Version {
            .major = static_cast<uint16_t>((m_bytecode[4] << 8) + m_bytecode[5]),
            .minor = static_cast<uint16_t>((m_bytecode[6] << 8) + m_bytecode[7]),
            .patch = static_cast<uint16_t>((m_bytecode[8] << 8) + m_bytecode[9])
        };
    }

    unsigned long long BytecodeReader::timestamp() const
    {
        // 4 (ark\0) + version (2 bytes / number) + timestamp = 18 bytes
        if (!checkMagic() || m_bytecode.size() < bytecode::HeaderSize)
            return 0;

        // reading the timestamp in big endian
        using timestamp_t = unsigned long long;
        return (static_cast<timestamp_t>(m_bytecode[10]) << 56) +
            (static_cast<timestamp_t>(m_bytecode[11]) << 48) +
            (static_cast<timestamp_t>(m_bytecode[12]) << 40) +
            (static_cast<timestamp_t>(m_bytecode[13]) << 32) +
            (static_cast<timestamp_t>(m_bytecode[14]) << 24) +
            (static_cast<timestamp_t>(m_bytecode[15]) << 16) +
            (static_cast<timestamp_t>(m_bytecode[16]) << 8) +
            static_cast<timestamp_t>(m_bytecode[17]);
    }

    std::vector<unsigned char> BytecodeReader::sha256() const
    {
        if (!checkMagic() || m_bytecode.size() < bytecode::HeaderSize + picosha2::k_digest_size)
            return {};

        std::vector<unsigned char> sha(picosha2::k_digest_size);
        for (std::size_t i = 0; i < picosha2::k_digest_size; ++i)
            sha[i] = m_bytecode[bytecode::HeaderSize + i];
        return sha;
    }

    Symbols BytecodeReader::symbols() const
    {
        if (!checkMagic() || m_bytecode.size() < bytecode::HeaderSize + picosha2::k_digest_size ||
            m_bytecode[bytecode::HeaderSize + picosha2::k_digest_size] != SYM_TABLE_START)
            return {};

        std::size_t i = bytecode::HeaderSize + picosha2::k_digest_size + 1;
        const uint16_t size = readNumber(i);
        i++;

        Symbols block;
        block.start = bytecode::HeaderSize + picosha2::k_digest_size;
        block.symbols.reserve(size);

        for (uint16_t j = 0; j < size; ++j)
        {
            std::string content;
            while (m_bytecode[i] != 0)
                content.push_back(static_cast<char>(m_bytecode[i++]));
            i++;

            block.symbols.push_back(content);
        }

        block.end = i;
        return block;
    }

    Values BytecodeReader::values(const Symbols& symbols) const
    {
        if (!checkMagic())
            return {};

        std::size_t i = symbols.end;
        if (m_bytecode[i] != VAL_TABLE_START)
            return {};
        i++;

        const uint16_t size = readNumber(i);
        i++;
        Values block;
        block.start = symbols.end;
        block.values.reserve(size);

        for (uint16_t j = 0; j < size; ++j)
        {
            const uint8_t type = m_bytecode[i];
            i++;

            if (type == NUMBER_TYPE)
            {
                auto exp = deserializeLE<decltype(ieee754::DecomposedDouble::exponent)>(
                    m_bytecode.begin() + static_cast<std::vector<uint8_t>::difference_type>(i), m_bytecode.end());
                i += sizeof(decltype(exp));
                auto mant = deserializeLE<decltype(ieee754::DecomposedDouble::mantissa)>(
                    m_bytecode.begin() + static_cast<std::vector<uint8_t>::difference_type>(i), m_bytecode.end());
                i += sizeof(decltype(mant));

                const ieee754::DecomposedDouble d { exp, mant };
                double val = ieee754::deserialize(d);
                block.values.emplace_back(val);
            }
            else if (type == STRING_TYPE)
            {
                std::string val;
                while (m_bytecode[i] != 0)
                    val.push_back(static_cast<char>(m_bytecode[i++]));
                block.values.emplace_back(val);
            }
            else if (type == FUNC_TYPE)
            {
                const uint16_t addr = readNumber(i);
                i++;
                block.values.emplace_back(addr);
            }
            else
                throw std::runtime_error(fmt::format("Unknown value type: {:x}", type));
            i++;
        }

        block.end = i;
        return block;
    }

    Code BytecodeReader::code(const Values& values) const
    {
        if (!checkMagic())
            return {};

        std::size_t i = values.end;

        Code block;
        block.start = i;

        while (m_bytecode[i] == CODE_SEGMENT_START)
        {
            i++;
            const std::size_t size = readNumber(i) * 4;
            i++;

            block.pages.emplace_back().reserve(size);
            for (std::size_t j = 0; j < size; ++j)
                block.pages.back().push_back(m_bytecode[i++]);

            if (i == m_bytecode.size())
                break;
        }

        return block;
    }

    void BytecodeReader::display(const BytecodeSegment segment,
                                 const std::optional<uint16_t> sStart,
                                 const std::optional<uint16_t> sEnd,
                                 const std::optional<uint16_t> cPage) const
    {
        if (!checkMagic())
        {
            fmt::println("Invalid format");
            return;
        }

        auto [major, minor, patch] = version();
        fmt::println("Version:   {}.{}.{}", major, minor, patch);
        fmt::println("Timestamp: {}", timestamp());
        fmt::print("SHA256:    ");
        for (const auto sha = sha256(); unsigned char h : sha)
            fmt::print("{:02x}", h);
        fmt::print("\n\n");

        // reading the different tables, one after another

        if ((sStart.has_value() && !sEnd.has_value()) || (!sStart.has_value() && sEnd.has_value()))
        {
            fmt::print(fmt::fg(fmt::color::red), "Both start and end parameter need to be provided together\n");
            return;
        }
        if (sStart.has_value() && sEnd.has_value() && sStart.value() >= sEnd.value())
        {
            fmt::print(fmt::fg(fmt::color::red), "Invalid slice start and end arguments\n");
            return;
        }

        const auto syms = symbols();
        const auto vals = values(syms);
        const auto code_block = code(vals);

        // symbols table
        {
            std::size_t size = syms.symbols.size();
            std::size_t sliceSize = size;
            bool showSym = (segment == BytecodeSegment::All || segment == BytecodeSegment::Symbols);

            if (showSym && sStart.has_value() && sEnd.has_value() && (sStart.value() > size || sEnd.value() > size))
                fmt::print(fmt::fg(fmt::color::red), "Slice start or end can't be greater than the segment size: {}\n", size);
            else if (showSym && sStart.has_value() && sEnd.has_value())
                sliceSize = sEnd.value() - sStart.value() + 1;

            if (showSym || segment == BytecodeSegment::HeadersOnly)
                fmt::println("{} (length: {})", fmt::styled("Symbols table", fmt::fg(fmt::color::cyan)), sliceSize);

            for (std::size_t j = 0; j < size; ++j)
            {
                if (auto start = sStart; auto end = sEnd)
                    showSym = showSym && (j >= start.value() && j <= end.value());

                if (showSym)
                    fmt::println("{}) {}", j, syms.symbols[j]);
            }

            if (showSym)
                fmt::print("\n");
            if (segment == BytecodeSegment::Symbols)
                return;
        }

        // values table
        {
            std::size_t size = vals.values.size();
            std::size_t sliceSize = size;

            bool showVal = (segment == BytecodeSegment::All || segment == BytecodeSegment::Values);
            if (showVal && sStart.has_value() && sEnd.has_value() && (sStart.value() > size || sEnd.value() > size))
                fmt::print(fmt::fg(fmt::color::red), "Slice start or end can't be greater than the segment size: {}\n", size);
            else if (showVal && sStart.has_value() && sEnd.has_value())
                sliceSize = sEnd.value() - sStart.value() + 1;

            if (showVal || segment == BytecodeSegment::HeadersOnly)
                fmt::println("{} (length: {})", fmt::styled("Constants table", fmt::fg(fmt::color::cyan)), sliceSize);

            for (std::size_t j = 0; j < size; ++j)
            {
                if (auto start = sStart; auto end = sEnd)
                    showVal = showVal && (j >= start.value() && j <= end.value());

                if (showVal)
                {
                    switch (const auto val = vals.values[j]; val.valueType())
                    {
                        case ValueType::Number:
                            fmt::println("{}) (Number) {}", j, val.number());
                            break;
                        case ValueType::String:
                            fmt::println("{}) (String) {}", j, val.string());
                            break;
                        case ValueType::PageAddr:
                            fmt::println("{}) (PageAddr) {}", j, val.pageAddr());
                            break;
                        default:
                            fmt::print(fmt::fg(fmt::color::red), "Value type not handled: {}\n", types_to_str[static_cast<std::size_t>(val.valueType())]);
                            break;
                    }
                }
            }

            if (showVal)
                fmt::print("\n");
            if (segment == BytecodeSegment::Values)
                return;
        }

        const auto stringify_value = [](const Value& val) -> std::string {
            switch (val.valueType())
            {
                case ValueType::Number:
                    return fmt::format("{} (Number)", val.number());
                case ValueType::String:
                    return fmt::format("{} (String)", val.string());
                case ValueType::PageAddr:
                    return fmt::format("{} (PageAddr)", val.pageAddr());
                default:
                    return "";
            }
        };

        enum class ArgKind
        {
            Symbol,
            Value,
            Builtin,
            Raw
        };

        struct Arg
        {
            ArgKind kind;
            uint16_t arg;
        };

        const std::unordered_map<Instruction, ArgKind> arg_kinds = {
            { LOAD_SYMBOL, ArgKind::Symbol },
            { LOAD_CONST, ArgKind::Value },
            { POP_JUMP_IF_TRUE, ArgKind::Raw },
            { STORE, ArgKind::Symbol },
            { SET_VAL, ArgKind::Symbol },
            { POP_JUMP_IF_FALSE, ArgKind::Raw },
            { JUMP, ArgKind::Raw },
            { CALL, ArgKind::Raw },
            { CALL_BUILTIN, ArgKind::Raw },
            { CAPTURE, ArgKind::Symbol },
            { BUILTIN, ArgKind::Builtin },
            { DEL, ArgKind::Symbol },
            { MAKE_CLOSURE, ArgKind::Value },
            { GET_FIELD, ArgKind::Symbol },
            { PLUGIN, ArgKind::Value },
            { LIST, ArgKind::Raw },
            { APPEND, ArgKind::Raw },
            { CONCAT, ArgKind::Raw },
            { APPEND_IN_PLACE, ArgKind::Raw },
            { CONCAT_IN_PLACE, ArgKind::Raw }
        };

        const auto color_print_inst = [&syms, &vals, &stringify_value](const std::string& name, std::optional<Arg> arg = std::nullopt) {
            fmt::print("{}", fmt::styled(name, fmt::fg(fmt::color::gold)));
            if (arg.has_value())
            {
                switch (auto [kind, idx] = arg.value(); kind)
                {
                    case ArgKind::Symbol:
                        fmt::print(fmt::fg(fmt::color::green), " {}\n", syms.symbols[idx]);
                        break;
                    case ArgKind::Value:
                        fmt::print(fmt::fg(fmt::color::magenta), " {}\n", stringify_value(vals.values[idx]));
                        break;
                    case ArgKind::Builtin:
                        fmt::print(" {}\n", Builtins::builtins[idx].first);
                        break;
                    case ArgKind::Raw:
                        fmt::print(fmt::fg(fmt::color::red), " ({})\n", idx);
                        break;
                }
            }
            else
                fmt::print("\n");
        };

        if (segment == BytecodeSegment::All || segment == BytecodeSegment::Code || segment == BytecodeSegment::HeadersOnly)
        {
            uint16_t pp = 0;

            for (const auto& page : code_block.pages)
            {
                bool displayCode = true;

                if (auto wanted_page = cPage)
                    displayCode = pp == wanted_page.value();

                if (displayCode)
                    fmt::println(
                        "{} {} (length: {})",
                        fmt::styled("Code segment", fmt::fg(fmt::color::magenta)),
                        fmt::styled(pp, fmt::fg(fmt::color::magenta)),
                        page.size());

                if (page.empty())
                {
                    if (displayCode)
                        fmt::print("NOP");
                }
                else
                {
                    if (cPage.value_or(pp) != pp)
                        continue;
                    if (segment == BytecodeSegment::HeadersOnly)
                        continue;
                    if (sStart.has_value() && sEnd.has_value() && ((sStart.value() > page.size()) || (sEnd.value() > page.size())))
                    {
                        fmt::print(fmt::fg(fmt::color::red), "Slice start or end can't be greater than the segment size: {}\n", page.size());
                        return;
                    }

                    for (std::size_t j = sStart.value_or(0), end = sEnd.value_or(page.size()); j < end; j += 4)
                    {
                        const uint8_t inst = page[j];
                        // TEMP
                        const uint8_t padding = page[j + 1];
                        const auto arg = static_cast<uint16_t>((page[j + 2] << 8) + page[j + 3]);

                        // instruction number
                        fmt::print(fmt::fg(fmt::color::cyan), "{:>4}", j / 4);
                        // padding inst arg arg
                        fmt::print(" {:02x} {:02x} {:02x} {:02x} ", inst, padding, page[j + 2], page[j + 3]);

                        if (const auto idx = static_cast<std::size_t>(inst); idx < InstructionNames.size())
                        {
                            const auto inst_name = InstructionNames[idx];
                            if (const auto iinst = static_cast<Instruction>(inst); arg_kinds.contains(iinst))
                                color_print_inst(inst_name, Arg { arg_kinds.at(iinst), arg });
                            else
                                color_print_inst(inst_name);
                        }
                        else
                            fmt::println("Unknown instruction");
                    }
                }
                if (displayCode && segment != BytecodeSegment::HeadersOnly)
                    fmt::print("\n");

                ++pp;
            }
        }
    }

    uint16_t BytecodeReader::readNumber(std::size_t& i) const
    {
        const auto x = static_cast<uint16_t>(m_bytecode[i] << 8);
        const uint16_t y = m_bytecode[++i];
        return x + y;
    }
}

1	#include <Ark/Compiler/BytecodeReader.hpp>
2
3	#include <Ark/Compiler/Instructions.hpp>
4	#include <Ark/Builtins/Builtins.hpp>
5
6	#include <iomanip>
7	#include <unordered_map>
8	#include <picosha2.h>
9	#include <Ark/Compiler/Serialization/IEEE754Serializer.hpp>
10	#include <Ark/Compiler/Serialization/IntegerSerializer.hpp>
11	#include <fmt/core.h>
12	#include <fmt/color.h>
13
14	namespace Ark
15	{
16	using namespace Ark::internal;
17
18	void BytecodeReader::feed(const bytecode_t& bytecode)	254✔
19	{	254✔
20	m_bytecode = bytecode;	254✔
21	}	254✔
22
23	void BytecodeReader::feed(const std::string& file)	×
24	{	×
25	std::ifstream ifs(file, std::ios::binary \| std::ios::ate);	×
26	if (!ifs.good())	×
27	throw std::runtime_error(fmt::format("[BytecodeReader] Couldn't open file '{}'", file));	×
28
29	const auto pos = ifs.tellg();	×
30	// reserve appropriate number of bytes
31	std::vector<char> temp(static_cast<std::size_t>(pos));	×
32	ifs.seekg(0, std::ios::beg);	×
33	ifs.read(&temp[0], pos);	×
34	ifs.close();	×
35
36	m_bytecode = bytecode_t(static_cast<std::size_t>(pos));	×
37	for (std::size_t i = 0; i < static_cast<std::size_t>(pos); ++i)	×
38	m_bytecode[i] = static_cast<uint8_t>(temp[i]);	×
39	}	×
40
41	bool BytecodeReader::checkMagic() const	669✔
42	{	669✔
43	return m_bytecode.size() >= bytecode::Magic.size() &&	1,337✔
44	m_bytecode[0] == bytecode::Magic[0] &&	668✔
45	m_bytecode[1] == bytecode::Magic[1] &&	498✔
46	m_bytecode[2] == bytecode::Magic[2] &&	996✔
47	m_bytecode[3] == bytecode::Magic[3];	498✔
48	}
49
50	const bytecode_t& BytecodeReader::bytecode() noexcept	×
51	{	×
52	return m_bytecode;	×
53	}	×
54
55	Version BytecodeReader::version() const	83✔
56	{	83✔
57	if (!checkMagic() \|\| m_bytecode.size() < bytecode::Magic.size() + bytecode::Version.size())	83✔
58	return Version { 0, 0, 0 };	×
59
60	return Version {	332✔
61	.major = static_cast<uint16_t>((m_bytecode[4] << 8) + m_bytecode[5]),	83✔
62	.minor = static_cast<uint16_t>((m_bytecode[6] << 8) + m_bytecode[7]),	83✔
63	.patch = static_cast<uint16_t>((m_bytecode[8] << 8) + m_bytecode[9])	83✔
64	};
65	}	83✔
66
67	unsigned long long BytecodeReader::timestamp() const	1✔
68	{	1✔
69	// 4 (ark\0) + version (2 bytes / number) + timestamp = 18 bytes
70	if (!checkMagic() \|\| m_bytecode.size() < bytecode::HeaderSize)	1✔
71	return 0;	×
72
73	// reading the timestamp in big endian
74	using timestamp_t = unsigned long long;
75	return (static_cast<timestamp_t>(m_bytecode[10]) << 56) +	3✔
76	(static_cast<timestamp_t>(m_bytecode[11]) << 48) +	2✔
77	(static_cast<timestamp_t>(m_bytecode[12]) << 40) +	2✔
78	(static_cast<timestamp_t>(m_bytecode[13]) << 32) +	2✔
79	(static_cast<timestamp_t>(m_bytecode[14]) << 24) +	2✔
80	(static_cast<timestamp_t>(m_bytecode[15]) << 16) +	2✔
81	(static_cast<timestamp_t>(m_bytecode[16]) << 8) +	2✔
82	static_cast<timestamp_t>(m_bytecode[17]);	1✔
83	}	1✔
84
85	std::vector<unsigned char> BytecodeReader::sha256() const	83✔
86	{	83✔
87	if (!checkMagic() \|\| m_bytecode.size() < bytecode::HeaderSize + picosha2::k_digest_size)	83✔
88	return {};	×
89
90	std::vector<unsigned char> sha(picosha2::k_digest_size);	83✔
91	for (std::size_t i = 0; i < picosha2::k_digest_size; ++i)	2,739✔
92	sha[i] = m_bytecode[bytecode::HeaderSize + i];	2,656✔
93	return sha;	83✔
94	}	166✔
95
96	Symbols BytecodeReader::symbols() const	83✔
97	{	83✔
98	if (!checkMagic() \|\| m_bytecode.size() < bytecode::HeaderSize + picosha2::k_digest_size \|\|	166✔
99	m_bytecode[bytecode::HeaderSize + picosha2::k_digest_size] != SYM_TABLE_START)	83✔
100	return {};	×
101
102	std::size_t i = bytecode::HeaderSize + picosha2::k_digest_size + 1;	83✔
103	const uint16_t size = readNumber(i);	83✔
104	i++;	83✔
105
106	Symbols block;	83✔
107	block.start = bytecode::HeaderSize + picosha2::k_digest_size;	83✔
108	block.symbols.reserve(size);	83✔
109
110	for (uint16_t j = 0; j < size; ++j)	2,463✔
111	{
112	std::string content;	2,380✔
113	while (m_bytecode[i] != 0)	24,125✔
114	content.push_back(static_cast<char>(m_bytecode[i++]));	21,745✔
115	i++;	2,380✔
116
117	block.symbols.push_back(content);	2,380✔
118	}	2,380✔
119
120	block.end = i;	83✔
121	return block;	83✔
122	}	83✔
123
124	Values BytecodeReader::values(const Symbols& symbols) const	83✔
125	{	83✔
126	if (!checkMagic())	83✔
127	return {};	×
128
129	std::size_t i = symbols.end;	83✔
130	if (m_bytecode[i] != VAL_TABLE_START)	83✔
131	return {};	×
132	i++;	83✔
133
134	const uint16_t size = readNumber(i);	83✔
135	i++;	83✔
136	Values block;	83✔
137	block.start = symbols.end;	83✔
138	block.values.reserve(size);	83✔
139
140	for (uint16_t j = 0; j < size; ++j)	2,964✔
141	{
142	const uint8_t type = m_bytecode[i];	2,881✔
143	i++;	2,881✔
144
145	if (type == NUMBER_TYPE)	2,881✔
146	{
147	auto exp = deserializeLE<decltype(ieee754::DecomposedDouble::exponent)>(	1,040✔
148	m_bytecode.begin() + static_cast<std::vector<uint8_t>::difference_type>(i), m_bytecode.end());	520✔
149	i += sizeof(decltype(exp));	520✔
150	auto mant = deserializeLE<decltype(ieee754::DecomposedDouble::mantissa)>(	1,040✔
151	m_bytecode.begin() + static_cast<std::vector<uint8_t>::difference_type>(i), m_bytecode.end());	520✔
152	i += sizeof(decltype(mant));	520✔
153
154	const ieee754::DecomposedDouble d { exp, mant };	520✔
155	double val = ieee754::deserialize(d);	520✔
156	block.values.emplace_back(val);	520✔
157	}	520✔
158	else if (type == STRING_TYPE)	2,361✔
159	{
160	std::string val;	1,150✔
161	while (m_bytecode[i] != 0)	20,939✔
162	val.push_back(static_cast<char>(m_bytecode[i++]));	19,789✔
163	block.values.emplace_back(val);	1,150✔
164	}	1,150✔
165	else if (type == FUNC_TYPE)	1,211✔
166	{
167	const uint16_t addr = readNumber(i);	1,211✔
168	i++;	1,211✔
169	block.values.emplace_back(addr);	1,211✔
170	}	1,211✔
171	else
172	throw std::runtime_error(fmt::format("Unknown value type: {:x}", type));	×
173	i++;	2,881✔
174	}	2,881✔
175
176	block.end = i;	83✔
177	return block;	83✔
178	}	83✔
179
180	Code BytecodeReader::code(const Values& values) const	83✔
181	{	83✔
182	if (!checkMagic())	83✔
183	return {};	×
184
185	std::size_t i = values.end;	83✔
186
187	Code block;	83✔
188	block.start = i;	83✔
189
190	while (m_bytecode[i] == CODE_SEGMENT_START)	1,294✔
191	{
192	i++;	1,294✔
193	const std::size_t size = readNumber(i) * 4;	1,294✔
194	i++;	1,294✔
195
196	block.pages.emplace_back().reserve(size);	1,294✔
197	for (std::size_t j = 0; j < size; ++j)	147,194✔
198	block.pages.back().push_back(m_bytecode[i++]);	145,900✔
199
200	if (i == m_bytecode.size())	1,294✔
201	break;	83✔
202	}	1,294✔
203
204	return block;	83✔
205	}	83✔
206
207	void BytecodeReader::display(const BytecodeSegment segment,	×
208	const std::optional<uint16_t> sStart,
209	const std::optional<uint16_t> sEnd,
210	const std::optional<uint16_t> cPage) const
211	{	×
212	if (!checkMagic())	×
213	{
214	fmt::println("Invalid format");	×
215	return;	×
216	}
217
218	auto [major, minor, patch] = version();	×
219	fmt::println("Version: {}.{}.{}", major, minor, patch);	×
220	fmt::println("Timestamp: {}", timestamp());	×
221	fmt::print("SHA256: ");	×
222	for (const auto sha = sha256(); unsigned char h : sha)	×
223	fmt::print("{:02x}", h);	×
224	fmt::print("\n\n");	×
225
226	// reading the different tables, one after another
227
228	if ((sStart.has_value() && !sEnd.has_value()) \|\| (!sStart.has_value() && sEnd.has_value()))	×
229	{
230	fmt::print(fmt::fg(fmt::color::red), "Both start and end parameter need to be provided together\n");	×
231	return;
232	}	×
233	if (sStart.has_value() && sEnd.has_value() && sStart.value() >= sEnd.value())	×
234	{
235	fmt::print(fmt::fg(fmt::color::red), "Invalid slice start and end arguments\n");	×
236	return;
237	}	×
238
239	const auto syms = symbols();	×
240	const auto vals = values(syms);	×
241	const auto code_block = code(vals);	×
242
243	// symbols table
244	{
245	std::size_t size = syms.symbols.size();	×
246	std::size_t sliceSize = size;	×
247	bool showSym = (segment == BytecodeSegment::All \|\| segment == BytecodeSegment::Symbols);	×
248
249	if (showSym && sStart.has_value() && sEnd.has_value() && (sStart.value() > size \|\| sEnd.value() > size))	×
250	fmt::print(fmt::fg(fmt::color::red), "Slice start or end can't be greater than the segment size: {}\n", size);	×
251	else if (showSym && sStart.has_value() && sEnd.has_value())	×
252	sliceSize = sEnd.value() - sStart.value() + 1;	×
253
254	if (showSym \|\| segment == BytecodeSegment::HeadersOnly)	×
255	fmt::println("{} (length: {})", fmt::styled("Symbols table", fmt::fg(fmt::color::cyan)), sliceSize);	×
256
257	for (std::size_t j = 0; j < size; ++j)	×
258	{
259	if (auto start = sStart; auto end = sEnd)	×
260	showSym = showSym && (j >= start.value() && j <= end.value());	×
261
262	if (showSym)	×
263	fmt::println("{}) {}", j, syms.symbols[j]);	×
264	}	×
265
266	if (showSym)	×
267	fmt::print("\n");	×
268	if (segment == BytecodeSegment::Symbols)	×
269	return;	×
270	}	×
271
272	// values table
273	{
274	std::size_t size = vals.values.size();	×
275	std::size_t sliceSize = size;	×
276
277	bool showVal = (segment == BytecodeSegment::All \|\| segment == BytecodeSegment::Values);	×
278	if (showVal && sStart.has_value() && sEnd.has_value() && (sStart.value() > size \|\| sEnd.value() > size))	×
279	fmt::print(fmt::fg(fmt::color::red), "Slice start or end can't be greater than the segment size: {}\n", size);	×
280	else if (showVal && sStart.has_value() && sEnd.has_value())	×
281	sliceSize = sEnd.value() - sStart.value() + 1;	×
282
283	if (showVal \|\| segment == BytecodeSegment::HeadersOnly)	×
284	fmt::println("{} (length: {})", fmt::styled("Constants table", fmt::fg(fmt::color::cyan)), sliceSize);	×
285
286	for (std::size_t j = 0; j < size; ++j)	×
287	{
288	if (auto start = sStart; auto end = sEnd)	×
289	showVal = showVal && (j >= start.value() && j <= end.value());	×
290
291	if (showVal)	×
292	{
293	switch (const auto val = vals.values[j]; val.valueType())	×
294	{	×
295	case ValueType::Number:
296	fmt::println("{}) (Number) {}", j, val.number());	×
297	break;	×
298	case ValueType::String:
299	fmt::println("{}) (String) {}", j, val.string());	×
300	break;	×
301	case ValueType::PageAddr:
302	fmt::println("{}) (PageAddr) {}", j, val.pageAddr());	×
303	break;	×
304	default:
305	fmt::print(fmt::fg(fmt::color::red), "Value type not handled: {}\n", types_to_str[static_cast<std::size_t>(val.valueType())]);	×
306	break;
307	}	×
308	}	×
309	}	×
310
311	if (showVal)	×
312	fmt::print("\n");	×
313	if (segment == BytecodeSegment::Values)	×
314	return;	×
315	}	×
316
317	const auto stringify_value = [](const Value& val) -> std::string {	×
318	switch (val.valueType())	×
319	{	×
320	case ValueType::Number:
321	return fmt::format("{} (Number)", val.number());	×
322	case ValueType::String:
323	return fmt::format("{} (String)", val.string());	×
324	case ValueType::PageAddr:
325	return fmt::format("{} (PageAddr)", val.pageAddr());	×
326	default:
327	return "";	×
328	}
329	};	×
330
331	enum class ArgKind
332	{
333	Symbol,
334	Value,
335	Builtin,
336	Raw
337	};
338
339	struct Arg
340	{
341	ArgKind kind;
342	uint16_t arg;
343	};
344
345	const std::unordered_map<Instruction, ArgKind> arg_kinds = {	×
346	{ LOAD_SYMBOL, ArgKind::Symbol },
347	{ LOAD_CONST, ArgKind::Value },
348	{ POP_JUMP_IF_TRUE, ArgKind::Raw },
349	{ STORE, ArgKind::Symbol },
350	{ SET_VAL, ArgKind::Symbol },
351	{ POP_JUMP_IF_FALSE, ArgKind::Raw },
352	{ JUMP, ArgKind::Raw },
353	{ CALL, ArgKind::Raw },
354	{ CALL_BUILTIN, ArgKind::Raw },
355	{ CAPTURE, ArgKind::Symbol },
356	{ BUILTIN, ArgKind::Builtin },
357	{ DEL, ArgKind::Symbol },
358	{ MAKE_CLOSURE, ArgKind::Value },
359	{ GET_FIELD, ArgKind::Symbol },
360	{ PLUGIN, ArgKind::Value },
361	{ LIST, ArgKind::Raw },
362	{ APPEND, ArgKind::Raw },
363	{ CONCAT, ArgKind::Raw },
364	{ APPEND_IN_PLACE, ArgKind::Raw },
365	{ CONCAT_IN_PLACE, ArgKind::Raw }
366	};
367
368	const auto color_print_inst = [&syms, &vals, &stringify_value](const std::string& name, std::optional<Arg> arg = std::nullopt) {	×
369	fmt::print("{}", fmt::styled(name, fmt::fg(fmt::color::gold)));	×
370	if (arg.has_value())	×
371	{
372	switch (auto [kind, idx] = arg.value(); kind)	×
373	{	×
374	case ArgKind::Symbol:
375	fmt::print(fmt::fg(fmt::color::green), " {}\n", syms.symbols[idx]);	×
376	break;	×
377	case ArgKind::Value:
378	fmt::print(fmt::fg(fmt::color::magenta), " {}\n", stringify_value(vals.values[idx]));	×
379	break;	×
380	case ArgKind::Builtin:
381	fmt::print(" {}\n", Builtins::builtins[idx].first);	×
382	break;	×
383	case ArgKind::Raw:
384	fmt::print(fmt::fg(fmt::color::red), " ({})\n", idx);	×
385	break;
386	}	×
387	}	×
388	else
389	fmt::print("\n");	×
390	};	×
391
392	if (segment == BytecodeSegment::All \|\| segment == BytecodeSegment::Code \|\| segment == BytecodeSegment::HeadersOnly)	×
393	{
394	uint16_t pp = 0;	×
395
396	for (const auto& page : code_block.pages)	×
397	{
398	bool displayCode = true;	×
399
400	if (auto wanted_page = cPage)	×
401	displayCode = pp == wanted_page.value();	×
402
403	if (displayCode)	×
404	fmt::println(	×
405	"{} {} (length: {})",	×
406	fmt::styled("Code segment", fmt::fg(fmt::color::magenta)),	×
407	fmt::styled(pp, fmt::fg(fmt::color::magenta)),	×
408	page.size());	×
409
410	if (page.empty())	×
411	{
412	if (displayCode)	×
413	fmt::print("NOP");	×
414	}	×
415	else
416	{
417	if (cPage.value_or(pp) != pp)	×
418	continue;	×
419	if (segment == BytecodeSegment::HeadersOnly)	×
420	continue;	×
421	if (sStart.has_value() && sEnd.has_value() && ((sStart.value() > page.size()) \|\| (sEnd.value() > page.size())))	×
422	{
423	fmt::print(fmt::fg(fmt::color::red), "Slice start or end can't be greater than the segment size: {}\n", page.size());	×
424	return;	×
425	}
426
427	for (std::size_t j = sStart.value_or(0), end = sEnd.value_or(page.size()); j < end; j += 4)	×
428	{
429	const uint8_t inst = page[j];	×
430	// TEMP
431	const uint8_t padding = page[j + 1];	×
432	const auto arg = static_cast<uint16_t>((page[j + 2] << 8) + page[j + 3]);	×
433
434	// instruction number
435	fmt::print(fmt::fg(fmt::color::cyan), "{:>4}", j / 4);	×
436	// padding inst arg arg
437	fmt::print(" {:02x} {:02x} {:02x} {:02x} ", inst, padding, page[j + 2], page[j + 3]);	×
438
439	if (const auto idx = static_cast<std::size_t>(inst); idx < InstructionNames.size())	×
440	{
441	const auto inst_name = InstructionNames[idx];	×
442	if (const auto iinst = static_cast<Instruction>(inst); arg_kinds.contains(iinst))	×
443	color_print_inst(inst_name, Arg { arg_kinds.at(iinst), arg });	×
444	else
445	color_print_inst(inst_name);	×
446	}	×
447	else
448	fmt::println("Unknown instruction");	×
449	}	×
450	}
451	if (displayCode && segment != BytecodeSegment::HeadersOnly)	×
452	fmt::print("\n");	×
453
454	++pp;	×
455	}	×
456	}	×
457	}	×
458
459	uint16_t BytecodeReader::readNumber(std::size_t& i) const	2,671✔
460	{	2,671✔
461	const auto x = static_cast<uint16_t>(m_bytecode[i] << 8);	2,671✔
462	const uint16_t y = m_bytecode[++i];	2,671✔
463	return x + y;	5,342✔
464	}	2,671✔
465	}

ArkScript-lang / Ark / 13762570662

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous