wordpress-xml-splitter.py

#!/usr/bin/env python3
'''
Created on May 8, 2010 by @anasimtiaz
Original script URL: http://anasimtiaz.com/?p=51
That link is dead, I suppose this is current: https://anasimtiaz.com/2014/03/08/wordpress-xml-splitter-again/

Forked on May 28, 2016 by @danielwrobert
https://gist.github.com/danielwrobert/6c9ca8de8199d5430621f481673d4baa

Changed into a command-line script by @Patrick-1994- on Jan 24th 2022

VERSION 3.0
Added the -s  to change the split size by @Patrick-1994- in 2023
(also --help and --version,so now I have to give this a version number^^)
'''

import os
import sys
import re

DEFAULT_SPLIT_SIZE = 500000

VERSION = "3.0"

#procInfo={};
def writeHeader(currentFile):
    header = '''<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
    xmlns:excerpt="http://wordpress.org/export/1.2/excerpt/"
    xmlns:content="http://purl.org/rss/1.0/modules/content/"
    xmlns:wfw="http://wellformedweb.org/CommentAPI/"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:wp="http://wordpress.org/export/1.2/"
>
<channel>
<wp:wxr_version>1.2</wp:wxr_version>
'''
    
    currentFile.write(header);

def writeFooter(currentFile):
    footer = '''
</channel>
</rss>'''
    
    currentFile.write(footer);

def errExit():
    sys.exit(1);

def err_print(s):
    print(s, file=sys.stderr)
    
def startProc(input_file_path, outDir, split_size):
    filePath,fileName = os.path.split(input_file_path)
    fileNameTxt = os.path.splitext(fileName)[0];
    fileNameExt = os.path.splitext(fileName)[1];

    xmlFileObj = open(os.path.join(filePath,fileName), "r");
    xmlFile = xmlFileObj.read();
    totalCount = len(xmlFile);
    iteration = 0;
    currentCount = 0;
    
    EOF = False;
    while(EOF==False):
        currentFileName = fileNameTxt + "_" + str(iteration) + fileNameExt;
        currentFile = open(os.path.join(outDir,currentFileName), 'w');
        print( 'Writing file ' + currentFileName);
        if iteration != 0:
            writeHeader(currentFile);
            
        if (currentCount+split_size) < totalCount:
            # find index (?) of last </item> before split_size bytes are reached
            xFile_i = xmlFile[currentCount:currentCount+split_size];
            incrFile = xFile_i.rfind('</item>') + len('</item>');

            currentFile.write(xFile_i[:incrFile]);
            currentCount += incrFile;
        else:
            xFile_i = xmlFile[currentCount:];
            currentFile.write(xFile_i);
            print('Finished processing \n');
            EOF = True;
            
        if EOF != True:
            writeFooter(currentFile);
            
        iteration += 1;
       
            
            
def print_usage():
    err_print(f"usage: [ -s SPLIT_SIZE ] {ARGV0} input_file_path [outDir]\noutDir is the current working directory by default")

def print_version():
    print(VERSION)
            
if __name__ == '__main__':
    split_size = DEFAULT_SPLIT_SIZE
    ARGV0 = sys.argv[0]
    command_line_args = sys.argv[1:]
    if len(sys.argv) == 1: # no args
        print_usage()
        errExit()

    if command_line_args[0] == "--help":
        print_usage()
        sys.exit(0)
    if command_line_args[0] == "--version":
        print_version()
        sys.exit(0)
    if command_line_args[0] == "-s":
        if not command_line_args[1].isnumeric():
            err_print( "-s given but the arg is not numeric" )
            errExit()
        split_size = int(command_line_args[1])
        command_line_args = command_line_args[2:]

    input_file_path = command_line_args[0]

    if( len(command_line_args) == 2 ):
        outDir          = command_line_args[1]
    else:
        outDir          = os.path.abspath( os.path.realpath(".") )
    
    if not os.path.isfile(input_file_path):
        err_print(f"""input file "{input_file_path}" does not exist""")
        print_usage()
        errExit()

    if not os.path.isdir(outDir):
        err_print(f"""output directory "{input_file_path}" does not exist""")
        print_usage()
        errExit()

    
    startProc(input_file_path, outDir, split_size)