"""
This script reads in the given CMSIS device include file (eg stm32f405xx.h),
extracts relevant peripheral constants, and creates qstrs, mpz's and constants
for the stm module.
"""

from __future__ import print_function

import argparse
import re

# Python 2/3 compatibility
import platform

if platform.python_version_tuple()[0] == "2":

    def convert_bytes_to_str(b):
        return b

elif platform.python_version_tuple()[0] == "3":

    def convert_bytes_to_str(b):
        try:
            return str(b, "utf8")
        except ValueError:
            # some files have invalid utf8 bytes, so filter them out
            return "".join(chr(l) for l in b if l <= 126)


# end compatibility code


# given a list of (name,regex) pairs, find the first one that matches the given line
def re_match_first(regexs, line):
    for name, regex in regexs:
        match = re.match(regex, line)
        if match:
            return name, match
    return None, None


class LexerError(Exception):
    def __init__(self, line):
        self.line = line


class Lexer:
    re_io_reg = r"__IO uint(?P<bits>8|16|32)_t +(?P<reg>[A-Z0-9]+)"
    re_comment = r"(?P<comment>[A-Za-z0-9 \-/_()&:\[\]]+)"
    re_addr_offset = r"Address offset: (?P<offset>0x[0-9A-Z]{2,3})"
    regexs = (
        (
            "#define hex",
            re.compile(
                r"#define +(?P<id>[A-Z0-9_]+) +\(?(\(uint32_t\))?(?P<hex>0x[0-9A-F]+)U?L?\)?($| */\*)"
            ),
        ),
        ("#define X", re.compile(r"#define +(?P<id>[A-Z0-9_]+) +(?P<id2>[A-Z0-9_]+)($| +/\*)")),
        (
            "#define X+hex",
            re.compile(
                r"#define +(?P<id>[A-Za-z0-9_]+) +\(?(?P<id2>[A-Z0-9_]+) \+ (?P<hex>0x[0-9A-F]+)U?L?\)?($| +/\*)"
            ),
        ),
        (
            "#define typedef",
            re.compile(
                r"#define +(?P<id>[A-Z0-9_]+(ext)?) +\(\([A-Za-z0-9_]+_(Global)?TypeDef \*\) (?P<id2>[A-Za-z0-9_]+)\)($| +/\*)"
            ),
        ),
        ("typedef struct", re.compile(r"typedef struct$")),
        ("{", re.compile(r"{$")),
        ("}", re.compile(r"}$")),
        (
            "} _t",
            re.compile(r"} *([A-Za-z0-9_]+)_t;$"),
        ),
        (
            "} TypeDef",
            re.compile(r"} *(?P<id>[A-Z][A-Za-z0-9_]*)_(?P<global>([A-Za-z0-9_]+)?)TypeDef;$"),
        ),
        (
            "IO reg",
            re.compile(
                re_io_reg + r" *; */\*!< *" + re_comment + r",? +" + re_addr_offset + r" *\*/"
            ),
        ),
        (
            "IO reg array",
            re.compile(
                re_io_reg
                + r"\[(?P<array>[2-8])\] *; */\*!< *"
                + re_comment
                + r",? +"
                + re_addr_offset
                + r"-(0x[0-9A-Z]{2,3}) *\*/"
            ),
        ),
    )

    def __init__(self, filename):
        self.file = open(filename, "rb")
        self.line_number = 0

    def next_match(self, strictly_next=False):
        while True:
            line = self.file.readline()
            line = convert_bytes_to_str(line)
            self.line_number += 1
            if len(line) == 0:
                return ("EOF", None)
            match = re_match_first(Lexer.regexs, line.strip())
            if strictly_next or match[0] is not None:
                return match

    def must_match(self, kind):
        match = self.next_match(strictly_next=True)
        if match[0] != kind:
            raise LexerError(self.line_number)
        return match


def parse_file(filename):
    lexer = Lexer(filename)

    reg_defs = {}
    consts = {}
    periphs = []
    while True:
        m = lexer.next_match()
        if m[0] == "EOF":
            break
        elif m[0] == "#define hex":
            d = m[1].groupdict()
            consts[d["id"]] = int(d["hex"], base=16)
        elif m[0] == "#define X":
            d = m[1].groupdict()
            if d["id2"] in consts:
                consts[d["id"]] = consts[d["id2"]]
        elif m[0] == "#define X+hex":
            d = m[1].groupdict()
            if d["id2"] in consts:
                consts[d["id"]] = consts[d["id2"]] + int(d["hex"], base=16)
        elif m[0] == "#define typedef":
            d = m[1].groupdict()
            if d["id2"] in consts:
                periphs.append((d["id"], consts[d["id2"]]))
        elif m[0] == "typedef struct":
            lexer.must_match("{")
            m = lexer.next_match()
            regs = []
            while m[0] in ("IO reg", "IO reg array"):
                d = m[1].groupdict()
                reg = d["reg"]
                offset = int(d["offset"], base=16)
                bits = int(d["bits"])
                comment = d["comment"]
                if m[0] == "IO reg":
                    regs.append((reg, offset, bits, comment))
                else:
                    for i in range(int(d["array"])):
                        regs.append((reg + str(i), offset + i * bits // 8, bits, comment))
                m = lexer.next_match()
            if m[0] in ("}", "} _t"):
                pass
            elif m[0] == "} TypeDef":
                d = m[1].groupdict()
                n = d["id"]
                g = d["global"]
                if n not in reg_defs or not g:
                    reg_defs[n] = regs
            else:
                raise LexerError(lexer.line_number)

    return periphs, reg_defs


def print_int_obj(val, needed_mpzs):
    if -0x40000000 <= val < 0x40000000:
        print("MP_ROM_INT(%#x)" % val, end="")
    else:
        print("MP_ROM_PTR(&mpz_%08x)" % val, end="")
        needed_mpzs.add(val)


def print_periph(periph_name, periph_val, needed_mpzs):
    qstr = periph_name.upper()
    print("{ MP_ROM_QSTR(MP_QSTR_%s), " % qstr, end="")
    print_int_obj(periph_val, needed_mpzs)
    print(" },")


def print_regs(reg_name, reg_defs, needed_mpzs):
    reg_name = reg_name.upper()
    for r in reg_defs:
        qstr = reg_name + "_" + r[0]
        print("{ MP_ROM_QSTR(MP_QSTR_%s), " % qstr, end="")
        print_int_obj(r[1], needed_mpzs)
        print(" }, // %s-bits, %s" % (r[2], r[3]))


# This version of print regs groups registers together into submodules (eg GPIO submodule).
# This makes the qstrs shorter, and makes the list of constants more manageable (since
# they are not all in one big module) but it is then harder to compile the constants, and
# is more cumbersome to access.
# As such, we don't use this version.
# And for the number of constants we have, this function seems to use about the same amount
# of ROM as print_regs.
def print_regs_as_submodules(reg_name, reg_defs, modules):
    mod_name_lower = reg_name.lower() + "_"
    mod_name_upper = mod_name_lower.upper()
    modules.append((mod_name_lower, mod_name_upper))

    print(
        """
STATIC const mp_rom_map_elem_t stm_%s_globals_table[] = {
    { MP_ROM_QSTR(MP_QSTR___name__), MP_ROM_QSTR(MP_QSTR_%s) },
"""
        % (mod_name_lower, mod_name_upper)
    )

    for r in reg_defs:
        print(
            "    { MP_ROM_QSTR(MP_QSTR_%s), MP_ROM_INT(%#x) }, // %s-bits, %s"
            % (r[0], r[1], r[2], r[3])
        )

    print(
        """};

STATIC MP_DEFINE_CONST_DICT(stm_%s_globals, stm_%s_globals_table);

const mp_obj_module_t stm_%s_obj = {
    .base = { &mp_type_module },
    .name = MP_QSTR_%s,
    .globals = (mp_obj_dict_t*)&stm_%s_globals,
};
"""
        % (mod_name_lower, mod_name_lower, mod_name_lower, mod_name_upper, mod_name_lower)
    )


def main():
    cmd_parser = argparse.ArgumentParser(description="Extract ST constants from a C header file.")
    cmd_parser.add_argument("file", nargs=1, help="input file")
    cmd_parser.add_argument(
        "--mpz",
        dest="mpz_filename",
        default="build/stmconst_mpz.h",
        help="the destination file of the generated mpz header",
    )
    args = cmd_parser.parse_args()

    periphs, reg_defs = parse_file(args.file[0])

    # add legacy GPIO constants that were removed when upgrading CMSIS
    if "GPIO" in reg_defs and "stm32f4" in args.file[0]:
        reg_defs["GPIO"].append(["BSRRL", 0x18, 16, "legacy register"])
        reg_defs["GPIO"].append(["BSRRH", 0x1A, 16, "legacy register"])

    needed_mpzs = set()

    print("// Automatically generated from %s by make-stmconst.py" % args.file[0])
    print("")

    for periph_name, periph_val in periphs:
        print_periph(periph_name, periph_val, needed_mpzs)

    for reg in (
        "ADC",
        #'ADC_Common',
        #'CAN_TxMailBox',
        #'CAN_FIFOMailBox',
        #'CAN_FilterRegister',
        #'CAN',
        "FDCAN",
        "CRC",
        "DAC",
        "DBGMCU",
        "DMA_Stream",
        "DMA",
        "EXTI",
        "FLASH",
        "GPIO",
        "SYSCFG",
        "I2C",
        "IWDG",
        "PWR",
        "RCC",
        "RTC",
        #'SDIO',
        "SPI",
        "TIM",
        "USART",
        "WWDG",
        "RNG",
        "IPCC",
    ):
        if reg in reg_defs:
            print_regs(reg, reg_defs[reg], needed_mpzs)
        # print_regs_as_submodules(reg, reg_defs[reg], modules)

    # print("#define MOD_STM_CONST_MODULES \\")
    # for mod_lower, mod_upper in modules:
    #    print("    { MP_ROM_QSTR(MP_QSTR_%s), MP_ROM_PTR(&stm_%s_obj) }, \\" % (mod_upper, mod_lower))

    print("")

    with open(args.mpz_filename, "wt") as mpz_file:
        for mpz in sorted(needed_mpzs):
            assert 0 <= mpz <= 0xFFFFFFFF
            print(
                "STATIC const mp_obj_int_t mpz_%08x = {{&mp_type_int}, "
                "{.neg=0, .fixed_dig=1, .alloc=2, .len=2, "
                ".dig=(uint16_t*)(const uint16_t[]){%#x, %#x}}};"
                % (mpz, mpz & 0xFFFF, (mpz >> 16) & 0xFFFF),
                file=mpz_file,
            )


if __name__ == "__main__":
    main()