/* vim: set ts=4 sw=4 tw=99 et:

 *

 * Copyright (C) 2012 David Anderson

 *

 * This file is part of JITCraft.

 *

 * JITCraft is free software: you can redistribute it and/or modify it under

 * the terms of the GNU General Public License as published by the Free

 * Software Foundation, either version 3 of the License, or (at your option)

 * any later version.

 * 

 * Foobar is distributed in the hope that it will be useful, but WITHOUT ANY

 * WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS

 * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

 *

 * You should have received a copy of the GNU General Public License along with

 * JITCraft. If not, see http://www.gnu.org/licenses/.

 */

#include <assert.h>

#include <ctype.h>

#include <math.h>

#include <limits.h>

#include <string.h>

#include "../Utility.h"

#include "../Heap-inl.h"

#include "Scanner.h"

using namespace ke;

const char *ke::TokenNames[] =

{

#define _(name, str) str,

    TOKENMAP(_)

#undef _

    NULL

};

Scanner::Scanner(CompileContext &cc, const char *stream, size_t length)

  : cc_(cc),

    stream_(stream),

    length_(length),

    pushed_(false),

    allowTags_(true),

    saved_('\0'),

    pos_(stream_)

{

    cur_.line = 1;

    cur_.col = 1;

    last_ = cur_;

}

char

Scanner::getChar()

{

    if (saved_ != '\0') {

        char temp = saved_;

        saved_ = '\0';

        return temp;

    }

    if (pos_ >= stream_ + length_)

        return '\0';

    last_.col = cur_.col++;

    return *pos_++;

}

void

Scanner::putChar(char c)

{

    assert(saved_ == '\0');

    saved_ = c;

    cur_.col = last_.col;

}

void

Scanner::calcNewLine(char c)

{

    cur_.line++;

    cur_.col = 1;

    last_ = cur_;

    char o;

    if (c == '\r' && ((o = getChar()) != '\n'))

        putChar(o);

}

void

Scanner::eatSingleLine()

{

    char c;

    while ((c = getChar()) != '\0') {

        if (c == '\n' || c == '\r') {

            calcNewLine(c);

            break;

        }

    }

}

bool

Scanner::eatComment()

{

    char c;

    while ((c = getChar()) != '\0') {

        while (c == '*') {

            c = getChar();

            if (c == '\0')

                return false;

            if (c == '/')

                return true;

        }

        if (c == '\r' || c == '\n')

            calcNewLine(c);

    }

    return false;

}

static inline bool

IsDigit(char c)

{

    return c >= '0' && c <= '9';

}

// Based off the logic in sc2.c's ftoi()...

double

Scanner::ParseDouble(const char *string)

{

    const char *ptr = string;

    double number = 0.0;

    while (IsDigit(*ptr)) {

        number = (number * 10) + (*ptr - '0');

        ptr++;

    }

    assert(*ptr == '.');

    ptr++;

    double fraction = 0.0;

    double multiplier = 1.0;

    while (IsDigit(*ptr)) {

        fraction = (fraction * 10) + (*ptr - '0');

        multiplier = multiplier / 10.0;

        ptr++;

    }

    number += fraction * multiplier;

    if (*ptr++ == 'e') {

        int sign = 1;

        if (*ptr == '-') {

            sign = -1;

            ptr++;

        }

        int exponent = 0;

        while (IsDigit(*ptr)) {

            exponent = (exponent * 10) + (*ptr - '0');

            ptr++;

        }

        multiplier = pow(10.0, exponent * sign);

        number *= multiplier;

    }

    return number;

}

static inline bool

IsHexDigit(char c)

{

    return IsDigit(c) ||

           (c >= 'a' && c <= 'f') ||

           (c >= 'A' && c <= 'F');

}

// Based on the logic for litchar() in sc2.c.

int

Scanner::readEscapeCode()

{

    char c = getChar();

    if (c == '\\')

        return c;

    switch (c) {

      case 'a':

        return '\a';

      case 'b':

        return '\b';

      case 'e':

        return 27;          // Apparently \e is non-standard.

      case 'f':

        return '\f';

      case 'n':

        return '\n';

      case 'r':

        return '\r';

      case 't':

        return '\t';

      case '\v':

        return '\v';

      case 'x':

      {

          unsigned digits = 0;

          char r = 0;

          c = getChar();

          while (IsHexDigit(c) && digits < 2) {

              if (IsDigit(c))

                  c = (c << 4) + (c - '0');

              else

                  c = (c << 4) + (tolower(c) - 'a' + 10);

              digits++;

              c = getChar();

           }

          // Swallow a trailing ';'

          if (c != ';')

              putChar(c);

          return r;

      }

      case '\'':

      case '\"':

      case '%':

        return c;

      default:

      {

        if (IsDigit(c)) {

            // \ddd

            char r = 0;

            while (IsDigit(c)) {

                r = r * 10 + (c - '0');

                c = getChar();

            }

            // Swallow a trailing ;

            if (c != ';')

                putChar(c);

            return r;

        }

        cc_.reportError(begin(), Message_UnknownEscapeCode, c);

        return INT_MAX;

      }

    }

}

TokenKind

Scanner::stringLiteral()

{

    literal_.clear();

    for (;;) {

        char c = getChar();

        if (c == '\"')

            break;

        if (c == '\r' || c == '\n' || c == '\0') {

            cc_.reportError(begin(), Message_UnterminatedString);

            return TOK_ERROR;

        }

        if (c == '\\') {

            int code = readEscapeCode();

            if (code == INT_MAX)

                return TOK_ERROR;

            c = char(code);

        }

        if (!literal_.append(c))

            return TOK_ERROR;

    }

    if (!literal_.append('\0'))

        return TOK_ERROR;

    return TOK_STRING_LITERAL;

}

TokenKind

Scanner::hexLiteral()

{

    literal_.clear();

    char c;

    for (;;) {

        c = getChar();

        if (!IsHexDigit(c))

            break;

        if (!literal_.append(c))

            return TOK_ERROR;

    }

    putChar(c);

    if (!literal_.append('\0'))

        return TOK_ERROR;

    return TOK_HEX_LITERAL;

}

TokenKind

Scanner::numberLiteral(char first)

{

    literal_.clear();

    if (!literal_.append(first))

        return TOK_ERROR;

    char c;

    for (;;) {

        c = getChar();

        if (!IsDigit(c))

            break;

        if (!literal_.append(c))

            return TOK_ERROR;

    }

    // Detect a hexadecimal string.

    if (literal_.length() == 1 &&

        literal_[0] == '0' &&

        (c == 'x' || c == 'X'))

    {

        return hexLiteral();

    }

    if (c != '.') {

        putChar(c);

        if (!literal_.append('\0'))

            return TOK_ERROR;

        return TOK_INTEGER_LITERAL;

    }

    if (!literal_.append(c))

        return TOK_ERROR;

    c = getChar();

    if (!IsDigit(c)) {

        cc_.reportError(begin(), Message_ExpectedDigitForFloat, c);

        return TOK_ERROR;

    }

    if (!literal_.append(c))

        return TOK_ERROR;

    for (;;) {

        c = getChar();

        if (!IsDigit(c))

            break;

        if (!literal_.append(c))

            return TOK_ERROR;

    }

    if (c != 'e') {

        putChar(c);

    } else {

        if (!literal_.append(c))

            return TOK_ERROR;

        c = getChar();

        if (c == '-') {

            if (!literal_.append(c))

                return TOK_ERROR;

            c = getChar();

        }

        if (!IsDigit(c)) {

            cc_.reportError(begin(), Message_ExpectedDigitForFloat, c);

            return TOK_ERROR;

        }

        if (!literal_.append(c))

            return TOK_ERROR;

        for (;;) {

            c = getChar();

            if (!IsDigit(c)) {

                putChar(c);

                break;

            }

            if (!literal_.append(c))

                return TOK_ERROR;

        }

    }

    if (!literal_.append('\0'))

        return TOK_ERROR;

    return TOK_FLOAT_LITERAL;

}

static inline TokenKind

FindIdentifier(const char *str)

{

    if (!strcmp(str,  "new"))

        return TOK_NEW;

    if (!strcmp(str,  "for"))

        return TOK_FOR;

    if (!strcmp(str,  "return"))

        return TOK_RETURN;

    if (!strcmp(str,  "if"))

        return TOK_IF;

    if (!strcmp(str,  "else"))

        return TOK_ELSE;

    if (!strcmp(str,  "enum"))

        return TOK_ENUM;

    if (!strcmp(str,  "while"))

        return TOK_WHILE;

    if (!strcmp(str,  "do"))

        return TOK_DO;

    if (!strcmp(str,  "break"))

        return TOK_BREAK;

    if (!strcmp(str,  "continue"))

        return TOK_CONTINUE;

    if (!strcmp(str,  "forward"))

        return TOK_FORWARD;

    if (!strcmp(str,  "public"))

        return TOK_PUBLIC;

    if (!strcmp(str,  "const"))

        return TOK_CONST;

    if (!strcmp(str,  "native"))

        return TOK_NATIVE;

    if (!strcmp(str,  "stock"))

        return TOK_STOCK;

    if (!strcmp(str,  "true"))

        return TOK_TRUE;

    if (!strcmp(str,  "false"))

        return TOK_FALSE;

    if (!strcmp(str,  "struct"))

        return TOK_STRUCT;

    if (!strcmp(str,  "switch"))

        return TOK_SWITCH;

    if (!strcmp(str,  "case"))

        return TOK_CASE;

    if (!strcmp(str,  "default"))

        return TOK_DEFAULT;

    if (!strcmp(str,  "typedef"))

        return TOK_TYPEDEF;

    if (!strcmp(str,  "void"))

        return TOK_VOID;

    if (!strcmp(str,  "int"))

        return TOK_INT;

    if (!strcmp(str,  "sizeof"))

        return TOK_SIZEOF;

    if (!strcmp(str,  "import"))

        return TOK_IMPORT;

    return TOK_NONE;

}

TokenKind

Scanner::identifier(char first)

{

    literal_.clear();

    if (!literal_.append(first))

        return TOK_ERROR;

    char c;

    for (;;) {

        c = getChar();

        if (!isascii(c) || (!isalnum(c) && c != '_'))

            break;

        if (!literal_.append(c))

            return TOK_ERROR;

    }

    if (!literal_.append('\0'))

        return TOK_ERROR;

    // Most keywords do not act as labels.

    TokenKind kind = FindIdentifier(literal());

    if (kind != TOK_NONE) {

        putChar(c);

        return kind;

    }

    if (c == ':' && allowTags_)

        return TOK_LABEL;

    putChar(c);

    // _ is a special case which acts as a tag and a keyword.

    if (!strcmp(literal(), "_"))

        return TOK_UNDERSCORE;

    return TOK_NAME;

}

TokenKind

Scanner::next()

{

    assert(!pushed_);

    for (;;) {

        begin_ = cur_;

        char c = getChar();

        switch (c) {

          case '/':

          {

            c = getChar();

            if (c == '/') {

                eatSingleLine();

                break;

            }

            if (c == '*') {

                if (!eatComment()) {

                    cc_.reportError(begin(), Message_UnterminatedComment);

                    return TOK_ERROR;

                }

                break;

            }

            putChar(c);

            return TOK_SLASH;

          }

          case '\n':

          case '\r':

            calcNewLine(c);

            break;

          case ';': return TOK_SEMICOLON;

          case '{': return TOK_LBRACE;

          case '}': return TOK_RBRACE;

          case '(': return TOK_LPAREN;

          case ')': return TOK_RPAREN;

          case '[': return TOK_LBRACKET;

          case ']': return TOK_RBRACKET;

          case '~': return TOK_TILDE;

          case '?': return TOK_QMARK;

          case ':': return TOK_COLON;

          case '.':

            c = getChar();

            if (c == '.') {

                c = getChar();

                if (c == '.')

                    return TOK_ELLIPSES;

                cc_.reportError(begin_, Message_UnexpectedCharacter, c, uint8(c));

                return TOK_ERROR;

            }

            putChar(c);

            return TOK_DOT;

          case '*':

            c = getChar();

            if (c == '=')

                return TOK_ASSIGN_MUL;

            putChar(c);

            return TOK_STAR;

          case '+':

            c = getChar();

            if (c == '=')

                return TOK_ASSIGN_ADD;

            if (c == '+')

                return TOK_INCREMENT;

            putChar(c);

            return TOK_PLUS;

          case '&':

            c = getChar();

            if (c == '=')

                return TOK_ASSIGN_BITAND;

            if (c == '&')

                return TOK_AND;

            putChar(c);

            return TOK_BITAND;

          case '|':

            c = getChar();

            if (c == '=')

                return TOK_ASSIGN_BITOR;

            if (c == '|')

                return TOK_OR;

            putChar(c);

            return TOK_BITOR;

          case '^':

            c = getChar();

            if (c == '=')

                return TOK_ASSIGN_BITXOR;

            putChar(c);

            return TOK_BITXOR;

          case '%':

            c = getChar();

            if (c == '=')

                return TOK_ASSIGN_MOD;