# ----------------------------------------------------------------------
# phplex.py
#
# A lexer for PHP.
# ----------------------------------------------------------------------

import ply.lex as lex
import re

# todo: BAD_CHARACTER
# todo: <script> syntax (does anyone use this?)

states = (
    ('php', 'exclusive'),
    ('quoted', 'exclusive'),
    ('quotedvar', 'exclusive'),
    ('varname', 'exclusive'),
    ('offset', 'exclusive'),
    ('property', 'exclusive'),
    ('heredoc', 'exclusive'),
    ('heredocvar', 'exclusive'),
    ('nowdoc', 'exclusive'),
    ('backticked', 'exclusive'),
    ('backtickedvar', 'exclusive'),
)

# Reserved words
reserved = (
    'ARRAY', 'AS', 'BREAK', 'CASE', 'CLASS', 'CONST', 'CONTINUE', 'DECLARE',
    'DEFAULT', 'DIE', 'DO', 'ECHO', 'ELSE', 'ELSEIF', 'EMPTY', 'ENDDECLARE',
    'ENDFOR', 'ENDFOREACH', 'ENDIF', 'ENDSWITCH', 'ENDWHILE', 'EVAL', 'EXIT',
    'EXTENDS', 'FOR', 'FOREACH', 'FUNCTION', 'GLOBAL', 'IF', 'INCLUDE',
    'INCLUDE_ONCE', 'INSTANCEOF', 'ISSET', 'LIST', 'NEW', 'PRINT', 'REQUIRE',
    'REQUIRE_ONCE', 'RETURN', 'STATIC', 'SWITCH', 'UNSET', 'USE', 'VAR',
    'WHILE', 'FINAL', 'INTERFACE', 'IMPLEMENTS', 'PUBLIC', 'PRIVATE',
    'PROTECTED', 'ABSTRACT', 'CLONE', 'TRY', 'CATCH', 'THROW', 'NAMESPACE',
    'FINALLY', 'TRAIT', 'YIELD',
)

# Not used by parser
unparsed = (
    # Invisible characters
    'WHITESPACE',

    # Open and close tags
    'OPEN_TAG', 'OPEN_TAG_WITH_ECHO', 'CLOSE_TAG',

    # Comments
    'COMMENT', 'DOC_COMMENT',
)

tokens = reserved + unparsed + (
    # Operators
    'PLUS', 'MINUS', 'MUL', 'DIV', 'MOD', 'AND', 'OR', 'NOT', 'XOR', 'SL',
    'SR', 'BOOLEAN_AND', 'BOOLEAN_OR', 'BOOLEAN_NOT', 'IS_SMALLER',
    'IS_GREATER', 'IS_SMALLER_OR_EQUAL', 'IS_GREATER_OR_EQUAL', 'IS_EQUAL',
    'IS_NOT_EQUAL', 'IS_IDENTICAL', 'IS_NOT_IDENTICAL',

    # Assignment operators
    'EQUALS', 'MUL_EQUAL', 'DIV_EQUAL', 'MOD_EQUAL', 'PLUS_EQUAL',
    'MINUS_EQUAL', 'SL_EQUAL', 'SR_EQUAL', 'AND_EQUAL', 'OR_EQUAL',
    'XOR_EQUAL', 'CONCAT_EQUAL',

    # Increment/decrement
    'INC', 'DEC',

    # Arrows
    'OBJECT_OPERATOR', 'DOUBLE_ARROW', 'DOUBLE_COLON',

    # Delimiters
    'LPAREN', 'RPAREN', 'LBRACKET', 'RBRACKET', 'LBRACE', 'RBRACE', 'DOLLAR',
    'COMMA', 'CONCAT', 'QUESTION', 'COLON', 'SEMI', 'AT', 'NS_SEPARATOR',

    # Casts
    'ARRAY_CAST', 'BINARY_CAST', 'BOOL_CAST', 'DOUBLE_CAST', 'INT_CAST',
    'OBJECT_CAST', 'STRING_CAST', 'UNSET_CAST',

    # Escaping from HTML
    'INLINE_HTML',

    # Identifiers and reserved words
    'DIR', 'FILE', 'LINE', 'FUNC_C', 'CLASS_C', 'METHOD_C', 'NS_C',
    'LOGICAL_AND', 'LOGICAL_OR', 'LOGICAL_XOR',
    'HALT_COMPILER',
    'STRING', 'VARIABLE',
    'LNUMBER', 'DNUMBER', 'NUM_STRING',
    'CONSTANT_ENCAPSED_STRING', 'ENCAPSED_AND_WHITESPACE', 'QUOTE',
    'DOLLAR_OPEN_CURLY_BRACES', 'STRING_VARNAME', 'CURLY_OPEN',

    # Heredocs
    'START_HEREDOC', 'END_HEREDOC',

    # Nowdocs
    'START_NOWDOC', 'END_NOWDOC',

    # Backtick
    'BACKTICK',
)

# Newlines
def t_php_WHITESPACE(t):
    r'[ \t\r\n]+'
    t.lexer.lineno += t.value.count("\n")
    return t

# Operators
t_php_PLUS                = r'\+'
t_php_MINUS               = r'-'
t_php_MUL                 = r'\*'
t_php_DIV                 = r'/'
t_php_MOD                 = r'%'
t_php_AND                 = r'&'
t_php_OR                  = r'\|'
t_php_NOT                 = r'~'
t_php_XOR                 = r'\^'
t_php_SL                  = r'<<'
t_php_SR                  = r'>>'
t_php_BOOLEAN_AND         = r'&&'
t_php_BOOLEAN_OR          = r'\|\|'
t_php_BOOLEAN_NOT         = r'!'
t_php_IS_SMALLER          = r'<'
t_php_IS_GREATER          = r'>'
t_php_IS_SMALLER_OR_EQUAL = r'<='
t_php_IS_GREATER_OR_EQUAL = r'>='
t_php_IS_EQUAL            = r'=='
t_php_IS_NOT_EQUAL        = r'(!=(?!=))|(<>)'
t_php_IS_IDENTICAL        = r'==='
t_php_IS_NOT_IDENTICAL    = r'!=='

# Assignment operators
t_php_EQUALS               = r'='
t_php_MUL_EQUAL            = r'\*='
t_php_DIV_EQUAL            = r'/='
t_php_MOD_EQUAL            = r'%='
t_php_PLUS_EQUAL           = r'\+='
t_php_MINUS_EQUAL          = r'-='
t_php_SL_EQUAL             = r'<<='
t_php_SR_EQUAL             = r'>>='
t_php_AND_EQUAL            = r'&='
t_php_OR_EQUAL             = r'\|='
t_php_XOR_EQUAL            = r'\^='
t_php_CONCAT_EQUAL         = r'\.='

# Increment/decrement
t_php_INC                  = r'\+\+'
t_php_DEC                  = r'--'

# Arrows
t_php_DOUBLE_ARROW         = r'=>'
t_php_DOUBLE_COLON         = r'::'

def t_php_OBJECT_OPERATOR(t):
    r'->'
    if re.match(r'[A-Za-z_]', peek(t.lexer)):
        t.lexer.push_state('property')
    return t

# Delimeters
t_php_LPAREN               = r'\('
t_php_RPAREN               = r'\)'
t_php_DOLLAR               = r'\$'
t_php_COMMA                = r','
t_php_CONCAT               = r'\.(?!\d|=)'
t_php_QUESTION             = r'\?'
t_php_COLON                = r':'
t_php_SEMI                 = r';'
t_php_AT                   = r'@'
t_php_NS_SEPARATOR         = r'\\'

def t_php_LBRACKET(t):
    r'\['
    t.lexer.push_state('php')
    return t

def t_php_RBRACKET(t):
    r'\]'
    t.lexer.pop_state()
    return t

def t_php_LBRACE(t):
    r'\{'
    t.lexer.push_state('php')
    return t

def t_php_RBRACE(t):
    r'\}'
    t.lexer.pop_state()
    return t

# Casts
t_php_ARRAY_CAST           = r'\([ \t]*[Aa][Rr][Rr][Aa][Yy][ \t]*\)'
t_php_BINARY_CAST          = r'\([ \t]*[Bb][Ii][Nn][Aa][Rr][Yy][ \t]*\)'
t_php_BOOL_CAST            = r'\([ \t]*[Bb][Oo][Oo][Ll]([Ee][Aa][Nn])?[ \t]*\)'
t_php_DOUBLE_CAST          = r'\([ \t]*([Rr][Ee][Aa][Ll]|[Dd][Oo][Uu][Bb][Ll][Ee]|[Ff][Ll][Oo][Aa][Tt])[ \t]*\)'
t_php_INT_CAST             = r'\([ \t]*[Ii][Nn][Tt]([Ee][Gg][Ee][Rr])?[ \t]*\)'
t_php_OBJECT_CAST          = r'\([ \t]*[Oo][Bb][Jj][Ee][Cc][Tt][ \t]*\)'
t_php_STRING_CAST          = r'\([ \t]*[Ss][Tt][Rr][Ii][Nn][Gg][ \t]*\)'
t_php_UNSET_CAST           = r'\([ \t]*[Uu][Nn][Ss][Ee][Tt][ \t]*\)'

# Comments

def t_php_DOC_COMMENT(t):
    r'/\*\*(.|\n)*?\*/'
    t.lexer.lineno += t.value.count("\n")
    return t

def t_php_COMMENT(t):
    r'/\*(.|\n)*?\*/ | //([^?%\n]|[?%](?!>))*\n? | \#([^?%\n]|[?%](?!>))*\n?'
    t.lexer.lineno += t.value.count("\n")
    return t

# Escaping from HTML

def t_OPEN_TAG(t):
    r'<[?%](([Pp][Hh][Pp][ \t\r\n]?)|=)?'
    if '=' in t.value: t.type = 'OPEN_TAG_WITH_ECHO'
    t.lexer.lineno += t.value.count("\n")
    t.lexer.begin('php')
    return t

def t_php_CLOSE_TAG(t):
    r'[?%]>\r?\n?'
    t.lexer.lineno += t.value.count("\n")
    t.lexer.begin('INITIAL')
    return t

def t_INLINE_HTML(t):
    r'([^<]|<(?![?%]))+'
    t.lexer.lineno += t.value.count("\n")
    return t

# Identifiers and reserved words

reserved_map = {
    '__DIR__':         'DIR',
    '__FILE__':        'FILE',
    '__LINE__':        'LINE',
    '__FUNCTION__':    'FUNC_C',
    '__CLASS__':       'CLASS_C',
    '__METHOD__':      'METHOD_C',
    '__NAMESPACE__':   'NS_C',

    'AND':             'LOGICAL_AND',
    'OR':              'LOGICAL_OR',
    'XOR':             'LOGICAL_XOR',

    '__HALT_COMPILER': 'HALT_COMPILER',
}

for r in reserved:
    reserved_map[r] = r

# Identifier
def t_php_STRING(t):
    r'[A-Za-z_][\w_]*'
    t.type = reserved_map.get(t.value.upper(), 'STRING')
    return t

# Variable
def t_php_VARIABLE(t):
    r'\$[A-Za-z_][\w_]*'
    return t

# Floating literal
def t_php_DNUMBER(t):
    r'(\d*\.\d+|\d+\.\d*)([Ee][+-]?\d+)? | (\d+[Ee][+-]?\d+)'
    return t

# Integer literal
def t_php_LNUMBER(t):
    r'(0b[01]+)|(0x[0-9A-Fa-f]+)|\d+'
    return t

# String literal
def t_php_CONSTANT_ENCAPSED_STRING(t):
    r"'([^\\']|\\(.|\n))*'"
    t.lexer.lineno += t.value.count("\n")
    return t

def t_php_QUOTE(t):
    r'"'
    t.lexer.push_state('quoted')
    return t

def t_quoted_QUOTE(t):
    r'"'
    t.lexer.pop_state()
    return t

def t_quoted_ENCAPSED_AND_WHITESPACE(t):
    r'( [^"\\${] | \\(.|\n) | \$(?![A-Za-z_{]) | \{(?!\$) )+'
    t.lexer.lineno += t.value.count("\n")
    return t

def t_quoted_VARIABLE(t):
    r'\$[A-Za-z_][\w_]*'
    t.lexer.push_state('quotedvar')
    return t

def t_quoted_CURLY_OPEN(t):
    r'\{(?=\$)'
    t.lexer.push_state('php')
    return t

def t_quoted_DOLLAR_OPEN_CURLY_BRACES(t):
    r'\$\{'
    if re.match(r'[A-Za-z_]', peek(t.lexer)):
        t.lexer.push_state('varname')
    else:
        t.lexer.push_state('php')
    return t

def t_quotedvar_QUOTE(t):
    r'"'
    t.lexer.pop_state()
    t.lexer.pop_state()
    return t

def t_quotedvar_LBRACKET(t):
    r'\['
    t.lexer.begin('offset')
    return t

def t_quotedvar_OBJECT_OPERATOR(t):
    r'->(?=[A-Za-z])'
    t.lexer.begin('property')
    return t

def t_quotedvar_ENCAPSED_AND_WHITESPACE(t):
    r'( [^"\\${] | \\(.|\n) | \$(?![A-Za-z_{]) | \{(?!\$) )+'
    t.lexer.lineno += t.value.count("\n")
    t.lexer.pop_state()
    return t

t_quotedvar_VARIABLE = t_php_VARIABLE

def t_quotedvar_CURLY_OPEN(t):
    r'\{(?=\$)'
    t.lexer.begin('php')
    return t

def t_quotedvar_DOLLAR_OPEN_CURLY_BRACES(t):
    r'\$\{'
    if re.match(r'[A-Za-z_]', peek(t.lexer)):
        t.lexer.begin('varname')
    else:
        t.lexer.begin('php')
    return t

def t_varname_STRING_VARNAME(t):
    r'[A-Za-z_][\w_]*'
    return t

t_varname_RBRACE = t_php_RBRACE
t_varname_LBRACKET = t_php_LBRACKET

def t_offset_STRING(t):
    r'[A-Za-z_][\w_]*'
    return t

def t_offset_NUM_STRING(t):
    r'\d+'
    return t

t_offset_VARIABLE = t_php_VARIABLE
t_offset_RBRACKET = t_php_RBRACKET

def t_property_STRING(t):
    r'[A-Za-z_][\w_]*'
    t.lexer.pop_state()
    return t

# Heredocs

def t_php_START_HEREDOC(t):
    r'<<<[ \t]*(?P<label>[A-Za-z_][\w_]*)\r?\n'
    t.lexer.lineno += t.value.count("\n")
    t.lexer.push_state('heredoc')
    t.lexer.heredoc_label = t.lexer.lexmatch.group('label')
    return t

def t_heredoc_END_HEREDOC(t):
    r'(?<=\n)[A-Za-z_][\w_]*'
    if t.value == t.lexer.heredoc_label:
        del t.lexer.heredoc_label
        t.lexer.pop_state()
    else:
        t.type = 'ENCAPSED_AND_WHITESPACE'
    return t

def t_php_START_NOWDOC(t):
    r'''<<<[ \t]*'(?P<label>[A-Za-z_][\w_]*)'\r?\n'''
    t.lexer.lineno += t.value.count("\n")
    t.lexer.push_state('nowdoc')
    t.lexer.nowdoc_label = t.lexer.lexmatch.group('label')
    return t

def t_nowdoc_END_NOWDOC(t):
    r'(?<=\n)[A-Za-z_][\w_]*'
    if t.value == t.lexer.nowdoc_label:
        del t.lexer.nowdoc_label
        t.lexer.pop_state()
    else:
        t.type = 'ENCAPSED_AND_WHITESPACE'
    return t

def t_nowdoc_ENCAPSED_AND_WHITESPACE(t):
    r'[^\n]*\n'
    t.lexer.lineno += t.value.count("\n")
    return t

def t_heredoc_ENCAPSED_AND_WHITESPACE(t):
    r'( [^\n\\${] | \\. | \$(?![A-Za-z_{]) | \{(?!\$) )+\n? | \\?\n'
    t.lexer.lineno += t.value.count("\n")
    return t

def t_heredoc_VARIABLE(t):
    r'\$[A-Za-z_][\w_]*'
    t.lexer.push_state('heredocvar')
    return t

t_heredoc_CURLY_OPEN = t_quoted_CURLY_OPEN
t_heredoc_DOLLAR_OPEN_CURLY_BRACES = t_quoted_DOLLAR_OPEN_CURLY_BRACES

def t_heredocvar_ENCAPSED_AND_WHITESPACE(t):
    r'( [^\n\\${] | \\. | \$(?![A-Za-z_{]) | \{(?!\$) )+\n? | \\?\n'
    t.lexer.lineno += t.value.count("\n")
    t.lexer.pop_state()
    return t

t_heredocvar_LBRACKET = t_quotedvar_LBRACKET
t_heredocvar_OBJECT_OPERATOR = t_quotedvar_OBJECT_OPERATOR
t_heredocvar_VARIABLE = t_quotedvar_VARIABLE
t_heredocvar_CURLY_OPEN = t_quotedvar_CURLY_OPEN
t_heredocvar_DOLLAR_OPEN_CURLY_BRACES = t_quotedvar_DOLLAR_OPEN_CURLY_BRACES

# Backticks
def t_php_BACKTICK(t):
    r"`"
    t.lexer.push_state('backticked')
    return t

def t_backticked_ENCAPSED_AND_WHITESPACE(t):
    r'( [^`\\${] | \\(.|\n) | \$(?![A-Za-z_{]) | \{(?!\$) )+'
    t.lexer.lineno += t.value.count("\n")
    return t

def t_backticked_VARIABLE(t):
    r'\$[A-Za-z_][\w_]*'
    t.lexer.push_state('backtickedvar')
    return t

t_backticked_CURLY_OPEN = t_quoted_CURLY_OPEN
t_backticked_DOLLAR_OPEN_CURLY_BRACES = t_quoted_DOLLAR_OPEN_CURLY_BRACES

def t_backticked_BACKTICK(t):
    r"`"
    t.lexer.pop_state()
    return t

def t_backtickedvar_BACKTICK(t):
    r"`"
    t.lexer.pop_state()
    t.lexer.pop_state()
    return t

t_backtickedvar_LBRACKET = t_quotedvar_LBRACKET
t_backtickedvar_OBJECT_OPERATOR = t_quotedvar_OBJECT_OPERATOR
t_backtickedvar_VARIABLE = t_quotedvar_VARIABLE
t_backtickedvar_CURLY_OPEN = t_quotedvar_CURLY_OPEN
t_backtickedvar_DOLLAR_OPEN_CURLY_BRACES = t_quotedvar_DOLLAR_OPEN_CURLY_BRACES

def t_backtickedvar_ENCAPSED_AND_WHITESPACE(t):
    r'( [^`\\${] | \\(.|\n) | \$(?![A-Za-z_{]) | \{(?!\$) )+'
    t.lexer.lineno += t.value.count("\n")
    t.lexer.pop_state()
    return t

def t_ANY_error(t):
    raise SyntaxError('illegal character', (None, t.lineno, None, t.value))

def peek(lexer):
    try:
        return lexer.lexdata[lexer.lexpos]
    except IndexError:
        return ''

class FilteredLexer(object):
    def __init__(self, lexer):
        self.lexer = lexer
        self.last_token = None

    @property
    def lineno(self):
        return self.lexer.lineno

    @lineno.setter
    def lineno(self, value):
        self.lexer.lineno = value

    @property
    def lexpos(self):
        return self.lexer.lexpos

    @lexpos.setter
    def lexpos(self, value):
        self.lexer.lexpos = value

    def clone(self):
        return FilteredLexer(self.lexer.clone())

    def current_state(self):
        return self.lexer.current_state()

    def input(self, input):
        self.lexer.input(input)

    def next_lexer_token(self):
        """Return next lexer token.

        Can be useful to customize parser behavior without need to touch
        parser code in the token method."""
        return self.lexer.token()

    def token(self):
        t = self.next_lexer_token()

        # Filter out tokens that the parser is not expecting.
        while t and t.type in unparsed:

            # Skip over open tags, but keep track of when we see them.
            if t.type == 'OPEN_TAG':
                if self.last_token and self.last_token.type == 'SEMI':
                    # Rewrite ?><?php as a semicolon.
                    t.type = 'SEMI'
                    t.value = ';'
                    break
                self.last_token = t
                t = self.next_lexer_token()
                continue

            # Rewrite <?= to yield an "echo" statement.
            if t.type == 'OPEN_TAG_WITH_ECHO':
                t.type = 'ECHO'
                break

            # Insert semicolons in place of close tags where necessary.
            if t.type == 'CLOSE_TAG':
                if self.last_token and \
                       self.last_token.type in ('OPEN_TAG', 'SEMI', 'COLON',
                                                'LBRACE', 'RBRACE'):
                    # Dont insert semicolons after these tokens.
                    pass
                else:
                    # Rewrite close tag as a semicolon.
                    t.type = 'SEMI'
                    break

            t = self.next_lexer_token()

        self.last_token = t
        return t

    # Iterator interface
    def __iter__(self):
        return self

    def __next__(self):
        t = self.token()
        if t is None:
            raise StopIteration
        return t

    __next__ = next

full_lexer = lex.lex()
lexer = FilteredLexer(full_lexer)

full_tokens = tokens
tokens = [token for token in tokens if token not in unparsed]

def run_on_argv1():
    lex.runmain(full_lexer)