Wordpress -> Jekyll converter

_readme.md

Following https://mdavies.net/2014/11/27/migrating-wpjekyll/#migrate-from-wordpress you can convert your posts across to Jekyll format, but it adds superfluous metadata and also doesn't correctly handle Wordpress shorttags. This program helps with that.

Fair warning: it's deliberately dirty / hacky code. I ran it with the posts in a Git repo and inspected file-by-file tweaking as I went.

program.cs

using System.IO;
using System.Linq;
using System.Text.RegularExpressions;
using System.Web;

// <package id = "HtmlAgilityPack" version="1.5.5" targetFramework="net461" />
// <package id = "ReverseMarkdown" version="1.5.0" targetFramework="net461" />

namespace JekyllConverter
{
    class Program
    {
        static void Main(string[] args)
        {
            var converter = new ReverseMarkdown.Converter();
            var dir = args.Length > 0 ? args[0] : @"C:\dev\temp\jekylltest\test";
            var files = Directory.GetFiles(dir);
            var preambleRegex = new Regex(@"(?s)^---.*?^---", RegexOptions.Multiline);
            var codeHtmlRegex = new Regex("<br\\s*/>|<\\/?p>");

            foreach (var file in files)
            {
                var fileToWrite = file;
                var fileContent = File.ReadAllText(file);
                var preamble = preambleRegex.Match(fileContent).Value;
                var content = fileContent.Replace(preamble, "");

                if (file.EndsWith(".html"))
                {
                    var codeBlocks = Regex.Matches(content, @"(?s)\[code(\s+lang(uage)?=""(.+?)"")?\](.*?)\[\/code\]", RegexOptions.Multiline).OfType<Match>().ToList();

                    codeBlocks.ForEach(match => content = content.Replace(match.Value, "<p>|||</p>"));

                    content = Regex.Replace(content, @"(\[caption.+?caption=""(.+?)"".*?](.+?)\[\/caption\])|(\[caption.*?]\s*(<a.+?</a>)(.+?)\[\/caption\])", "$5$3<br /><em>$2$6</em>");

                    var markdown = converter.Convert(content);

                    var index = 0;
                    var markdownWithCode = Regex.Replace(markdown, "\\|\\|\\|", m =>
                    {
                        var match = codeBlocks[index++];
                        return $"```{match.Groups[3].Value}{HttpUtility.HtmlDecode(codeHtmlRegex.Replace(match.Groups[4].Value, ""))}```";
                    });

                    content = markdownWithCode;
                    fileToWrite = Regex.Replace(file, @"\.html$", ".md");
                }

                preamble = Regex.Replace(preamble, @"^parent_id:.+?$\s+", "", RegexOptions.Multiline);
                preamble = Regex.Replace(preamble, @"^published:.+?$\s+", "", RegexOptions.Multiline);
                preamble = Regex.Replace(preamble, @"^password:.+?$\s+", "", RegexOptions.Multiline);
                preamble = Regex.Replace(preamble, @"^status:.+?$\s+", "", RegexOptions.Multiline);
                preamble = Regex.Replace(preamble, @"^meta:[\s\S]+?(?=(^\S))+", "", RegexOptions.Multiline);
                preamble = Regex.Replace(preamble, @"^author:[\s\S]+?display_name:\s+(.+?)$[\s\S]+?(?=(^\S))+", "author: $1\n", RegexOptions.Multiline);

                File.WriteAllText(fileToWrite, preamble + content);
                if (fileToWrite != file)
                    File.Delete(file);
            }
        }
    }
}