PHP array(...) to [...]

convert_array.py

#!/usr/bin/env python3

import os
import sys
import mmap

def deleteFromMmap(f, mmapfile, start, end):
    length = end - start
    size = len(mmapfile)
    newsize = size - length

    mmapfile.move(start, end, size - end)
    mmapfile.flush()
    mmapfile.close()
    f.truncate(newsize)
    return mmap.mmap(f.fileno(), 0)


def insertIntoMmap(f, mmapfile, offset, data):
    length = len(data)
    size = len(mmapfile)
    newsize = size + length

    mmapfile.flush()
    mmapfile.close()
    f.seek(size)
    f.write(data)
    f.flush()
    mmapfile = mmap.mmap(f.fileno(), 0)

    mmapfile.move(offset + length, offset, size - offset)
    mmapfile.seek(offset)
    mmapfile.write(bytes(data, "UTF-8"))
    mmapfile.flush()
    return mmapfile


def convert(path):
    print(path)
    size = os.path.getsize(path)
    with open(path, "r+") as f:
        data = mmap.mmap(f.fileno(), 0)
        orig_index = 0
        comment_str = ""
        while orig_index + 6 < size:

            # Skip strings entirely
            if (chr(data[orig_index]) == "'" or chr(data[orig_index]) == "\"") and chr(data[orig_index - 1]) != "\\" and not comment_str:
                char = data[orig_index]
                orig_index += 1
                while data[orig_index] != char or chr(data[orig_index - 1]) == "\\":
                    orig_index += 1
                orig_index += 1

            # Comments
            comment_substr = data[orig_index:orig_index + 2]
            if not comment_str and (comment_substr == bytes("//", "UTF-8") or comment_substr == bytes("/*", "UTF-8")):
                comment_str = comment_substr
            elif comment_str == bytes("//", "UTF-8") and chr(data[orig_index]) == "\n":
                comment_str = ""
            elif comment_str == bytes("/*", "UTF-8") and comment_substr == bytes("*/", "UTF-8"):
                comment_str = ""

            substr = data[orig_index:orig_index + 6]
            if substr != bytes("array(", "UTF-8"):
                orig_index += 1
                continue

            # Only match array() func, not others
            if chr(data[orig_index - 1]) not in [' ', '\n', '.', '(', '[']:
                orig_index += 1
                continue

            # Closing ')' is at least 6 characters after the start of array
            index = orig_index + 6

            # Match braces
            open_br = 1
            while index < size:
                if chr(data[index]) == ')':
                    if open_br == 1:
                        break
                    else:
                        open_br -= 1
                elif chr(data[index]) == '(':
                    open_br += 1
                index += 1

            print("Found at", orig_index, "Ends at", index)
            data = deleteFromMmap(f, data, orig_index, orig_index + 6)
            data = insertIntoMmap(f, data, orig_index, "[")
            data = deleteFromMmap(f, data, index - 6 + 1, index - 6 + 2)
            data = insertIntoMmap(f, data, index - 6 + 1, "]")
            size -= 5

            orig_index += 1


for root, dirs, files in os.walk(os.getcwd()):
    if "vendor" in root:
        continue
    for f in files:
        if f.endswith('.php'):
            path = root + '/' + f
            convert(path)