An example of the way how undocumented option of MSVC compiler, /d1reportallclasslayout, can be used to generate static reflection information for C++ sources

refgen.py

# an example of the way how undocumented option of MSVC compiler, /d1reportallclasslayout,
# can be used to generate static reflection information for C++ sources

import sys
import re
import subprocess

cl_exe = 'C:\\Program Files (x86)\\Microsoft Visual Studio\\2017\\Community\\VC\\Tools\\MSVC\\14.16.27023\\bin\Hostx64\\x64\\cl.exe'

file = 'test.cc'


#
class Cxx:

    class Attribute:

        def __init__(self, offset, name):
            self.offset = offset
            self.name = name

    class Class:

        def __init__(self, name, size):
            self.name = name
            self.size = size
            self.attributes = list()

    class Module:

        def __init__(self):
            self.classes = list()


def log_v(str):
    print('v: {}'.format(str))


def log_e(str):
    print('e: {}'.format(str))


if __name__ == '__main__':

    classesInformation = Cxx.Module()

    cxx = subprocess.run([cl_exe, file, '/c', '/nologo', '/d1reportAllClassLayout'], capture_output=True)

    if cxx.returncode != 0:
        log_e('compiler process failed!')

    classLineRegex = re.compile(r'class\s+([a-zA-Z_:][a-zA-Z0-9_:]*)\s+size\(([1-9][0-9]*)\):')
    delimiterRegex = re.compile(r'\s*\+---\s*')
    attributeRegex = re.compile(r'\s*([0-9]+)\s*\|\s*([a-zA-Z_][a-zA-Z0-9_]*)\s*|([a-zA-Z_][a-zA-Z0-9_]*)\s*')

    class State:
        class Undefined:
            pass
        class Class:
            pass
        class Children:
            pass

    state = State.Undefined
    classInfo = None

    for line in cxx.stdout.decode('utf-8').splitlines()[1:]:

        line = line.replace('\t', ' ').strip()
        if not line:
            continue

        if state == State.Undefined:

            regexMatch = classLineRegex.fullmatch(line)
            if regexMatch:
                name = regexMatch.group(1)
                size = int(regexMatch.group(2))
                log_v('class found: {}, {}'.format(name, size))

                state = State.Class
                classInfo = Cxx.Class(name, size)
                classesInformation.classes.append(classInfo)
            else:
                log_e('unexpected: {}'.format(line))

            continue

        if state == State.Class:

            if delimiterRegex.fullmatch(line):
                state = State.Children
            else:
                log_e('unexpected: {}'.format(line))

            continue

        if state == State.Children:

            if delimiterRegex.fullmatch(line):
                state = State.Undefined
                classInfo = None
                continue

            attributeMatch = attributeRegex.match(line)
            if attributeMatch:
                # size, (prefix), name
                name = attributeMatch.group(3) if attributeMatch.group(3) else attributeMatch.group(2)
                offset = int(attributeMatch.group(1))
                log_v('    attribute {} at {}'.format(name, offset))
                classInfo.attributes.append(Cxx.Attribute(offset, name))
                continue

            if '<alignment member>' in line:
                continue

            log_e('unexpected: {}'.format(line))

        log_e('unexpected situation!')
        sys.exit(2)

    log_v('gracefully finished')