SOAdvancedDissector/cppprototypeparser.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
#
# Copyright Grégory Soutadé

# This file is part of SOAdvancedDissector

# SOAdvancedDissector 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.
#
# SOAdvancedDissector 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 SOAdvancedDissector.  If not, see <http://www.gnu.org/licenses/>.
#

class CPPPrototypeParser:
    """Class that extract all parts of a cpp method prototype"""

    def __init__(self, fullname=''):
        if fullname:
            self.parse(fullname)
        else:
            self._reset()

    def _reset(self):
        self.fullclassname = ''   # Class name with its namespaces
        self.names = []           # All decoded names : namespaces, class, function
        self.namespaces = []      # Namespaces part
        self.parameters = []      # Parameters part
        self.fullparameters = ''  # Full parameters string
        self.funcname = ''        # Function name
        self.classname = ''       # Class name
        self.fullname = ''        # namespaces + class + function + parameters
        self.has_parameters = False
        self.is_function = False  # Is single function or class method
        
    # Should parse something like :
    # uft::ClassDescriptor<layout::MasterConditionalReference>::queryFunc(uft::StructDescriptor const*, void*, uft::Value const&, void*)
    def parse(self, fullname):
        """Parse CPP method prototype and fill class members
        with right values

        Parameters
        ----------
        fullname : str
            Prototype to parse
        """
        self._reset()
        self.fullname = fullname
        if '(' in fullname:
            self.has_parameters = True
            splitname = fullname.split('(')
            if 'operator()' in fullname:
                namepart = splitname[0] + '()'
                self.fullparameters = '(' + ''.join(splitname[2:])
            else:
                namepart = splitname[0]
                self.fullparameters = '(' + ''.join(splitname[1:])
            self._parseParameters()
        else:
            namepart = fullname
            
        self.names = self._parseName(namepart)

        # Function or class method
        if self.has_parameters:
            self.funcname = self.names[-1]
            # Class method with at least class name + method name
            if len(self.names) >= 2:
                self.classname = self.names[-2]
                self.fullclassname = '::'.join(self.names[:-1])
                # Contains namespaces
                if len(self.names) > 2:
                    self.namespaces = self.names[:-2]
            # Simple function
            else:
                self.is_function = True
                self.funcname = self.names[0]
        else:
            # Class name with or without namespaces
            self.classname = self.names[-1]
            if len(self.names) > 1:
                self.namespaces = self.names[:-1]

        if self.funcname.endswith(' const'):
            self.funcname = self.funcname[:-len(' const')]
            
        if self.classname.endswith(' const'):
            self.classname = self.classname[:-len(' const')]

    def _parseName(self, name):
        """Parse CPP method name and split in different parts
        using '::' as delimiter.
        It supports templates names and function prototypes

        Parameters
        ----------
        name : str
            Name to parse

        Returns
        -------
        list
            Splitted name parts
        """
        nb_parenthesis = 0
        nb_chevrons = 0
        nb_parenthesis = 0
        cur_buf = ''
        names = []
        nb_db_points = 0
        in_template = False
        in_parenthesis = False
        operator = False
        
        for c in name.strip():
            if operator:
                cur_buf += c
                continue
            
            if c == '(':
                cur_buf += c
                nb_parenthesis += 1
                in_parenthesis = True
            elif c == ')':
                cur_buf += c
                nb_parenthesis -= 1
                if nb_parenthesis == 0:
                    in_parenthesis = False
            elif c == ':':
                if in_template or in_parenthesis:
                    cur_buf += c
                    continue
                nb_db_points += 1
                if nb_db_points == 2:
                    names.append(cur_buf)
                    cur_buf = ''
                    nb_db_points = 0
            elif c == '<':
                cur_buf += c
                in_template = True
                nb_chevrons += 1
            elif c == '>':
                cur_buf += c
                nb_chevrons -= 1
                if nb_chevrons == 0:
                    in_template = False
            else:
                cur_buf += c

            # Next characters are part of operator name
            if cur_buf == 'operator':
                operator = True
                
        if cur_buf:
            names.append(cur_buf)

        return names

    def _parseParameters(self):
        """Parse each parameters and do split on them
        using '::' as delimiter.
        Update self.parameters member
        """
        nb_chevrons = 0
        in_template = False
        cur_buf = ''

        # Parse fullparameters except parenthesis
        for c in self.fullparameters[1:-1]:
            if c == ',':
                if in_template: continue
                self.parameters.append(self._parseName(cur_buf))
                cur_buf = ''
            elif c == '<':
                cur_buf += c
                if in_template: continue
                in_template = True
                nb_chevrons += 1
            elif c == '>':
                cur_buf += c
                nb_chevrons -= 1
                if nb_chevrons == 0:
                    in_template = False
            else:
                cur_buf += c

        if cur_buf:
            self.parameters.append(self._parseName(cur_buf))
            
# Some tests
if __name__ == "__main__":
    strings = ['uft::ClassDescriptor<layout::MasterConditionalReference>::queryFunc(uft::StructDescriptor const*, void*, uft::Value const&, void*)',
               'adept::LicenseImpl::getVoucherID()',
               'adept::DRMProcessorImpl::addSignIn()',
               'layout::InlineLayoutEngine::AnnotationGlyphRunCounter::operator()(uft::sref<layout::RunListItem>&)',
               'dp::PointerVector<void>::~PointerVector()',
               'adept::IoCallbackWrapper<adept::DRMProcessorImpl>::IoCallbackWrapper(adept::DRMProcessorImpl*, void (adept::DRMProcessorImpl::*)(dp::Unknown*, bool), void (adept::DRMProcessorImpl::*)(double), void (adept::DRMProcessorImpl::*)(dp::String const&))',
               'tetraphilia::ThreadImpl<T3AppTraits, tetraphilia::PFiber<T3AppTraits>, tetraphilia::NoClientYieldHook<T3AppTraits> >',
               'debugOutputOpen']
    for s in strings:
        p = CPPPrototypeParser(s)
        print('Original {}'.format(s))
        print('Full parameters {}'.format(p.fullparameters))
        print('Names {}'.format(p.names))
        print('Namespaces {}'.format(p.namespaces))
        print('Parameters {}'.format(p.parameters))
        print('Class name {}'.format(p.classname))
        print('Func name {}'.format(p.funcname))
        print('Full class name {}'.format(p.fullclassname))
        print('Full parameters {}'.format(p.fullparameters))
        print('Is function ? {}'.format(p.is_function))
        print('Has parameters ? {}'.format(p.has_parameters))
        print("")