public void Process(BlogMLPost postml)
{
Log.DebugFormat("Setting up category references on post {0}", postml.ID);
using (var query = new OxiteReader("SELECT * FROM oxite_Post A, oxite_PostTagRelationship B WHERE A.PostID=B.PostID AND A.PostID='" + postml.ID + "'"))
{
var tagRefs = query.Execute();
foreach (var tag in tagRefs)
{
var commentRef = new BlogMLCategoryReference
{
Ref = tag.TagID.ToString()
};
postml.Categories.Add(commentRef);
}
}
}
/// <summary>
/// Converts the blog export to BlogML
/// </summary>
public void Convert()
{
XDocument blogger = XDocument.Load(_bloggerExportPath);
// extract basic information
string blogTitle = blogger.Element(AtomNamespace + "feed").Element(AtomNamespace + "title").Value;
DateTime blogUpdated = DateTime.Parse(blogger.Element(AtomNamespace + "feed").Element(AtomNamespace + "updated").Value).ToUniversalTime();
string authorName = blogger.Element(AtomNamespace + "feed").Element(AtomNamespace + "author").Element(AtomNamespace + "name").Value;
string authorUri = blogger.Element(AtomNamespace + "feed").Element(AtomNamespace + "author").Element(AtomNamespace + "uri").Value;
string authorEmail = blogger.Element(AtomNamespace + "feed").Element(AtomNamespace + "author").Element(AtomNamespace + "email").Value;
// assume the updated date and then hunt backwards in time
DateTime blogCreated = blogUpdated;
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_MessageBlogTitle,
blogTitle,
blogUpdated));
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_MessageBlogAuthor,
authorName,
authorEmail,
authorUri));
// parse ATOM entries
var query = from entry in blogger.Descendants(AtomNamespace + "entry")
select new
{
Id = entry.Element(AtomNamespace + "id").Value,
Published = DateTime.Parse(entry.Element(AtomNamespace + "published").Value).ToUniversalTime(),
Updated = DateTime.Parse(entry.Element(AtomNamespace + "updated").Value).ToUniversalTime(),
Title = entry.Element(AtomNamespace + "title").Value,
Content = entry.Element(AtomNamespace + "content").Value,
AuthorName = entry.Element(AtomNamespace + "author").Element(AtomNamespace + "name").Value,
AuthorUri = entry.Element(AtomNamespace + "author").Element(AtomNamespace + "uri") == null ? null : entry.Element(AtomNamespace + "author").Element(AtomNamespace + "uri").Value,
AuthorEmail = entry.Element(AtomNamespace + "author").Element(AtomNamespace + "email") == null ? null : entry.Element(AtomNamespace + "author").Element(AtomNamespace + "email").Value,
InReplyTo = entry.Element(ThrNamespace + "in-reply-to") == null ? null : entry.Element(ThrNamespace + "in-reply-to").Attribute("ref").Value,
Categories = (from category in entry.Descendants(AtomNamespace + "category")
select new
{
Scheme = category.Attribute("scheme").Value,
Term = category.Attribute("term").Value,
}),
Links = (from link in entry.Descendants(AtomNamespace + "link")
select new
{
Rel = link.Attribute("rel").Value,
Type = link.Attribute("type").Value,
Href = link.Attribute("href").Value,
}),
};
// separate out the different export categories from the ATOM entries
Dictionary <string, BlogMLPost> posts = new Dictionary <string, BlogMLPost>();
List <CommentWithInReplyTo> comments = new List <CommentWithInReplyTo>();
List <BlogMLAuthor> authors = new List <BlogMLAuthor>();
List <BlogMLCategory> categories = new List <BlogMLCategory>();
NameValueCollection settings = new NameValueCollection();
string template = null;
foreach (var q in query)
{
// update the blog created date as we find earlier entires
if (q.Published < blogCreated)
{
blogCreated = q.Published;
}
// create a content holder
BlogMLContent content = new BlogMLContent();
content.Text = q.Content;
// find categories and the type of entry
List <BlogMLCategoryReference> categoryRefs = new List <BlogMLCategoryReference>();
string entryKind = null;
// get the type of entry and any post categories
foreach (var c in q.Categories)
{
if (c.Scheme == "http://schemas.google.com/g/2005#kind")
{
entryKind = c.Term;
}
else if (c.Scheme == "http://www.blogger.com/atom/ns#")
{
BlogMLCategoryReference categoryRef = new BlogMLCategoryReference();
categoryRef.Ref = UpdateCategoriesGetRef(ref categories, c.Term, q.Published);
categoryRefs.Add(categoryRef);
}
else
{
// we've found a category scheme we don't know about
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_MessageUnexpectedCategoryScheme,
c.Scheme));
}
}
// process entry based on the entry kind
switch (entryKind)
{
case "http://schemas.google.com/blogger/2008/kind#template":
template = q.Content;
break;
case "http://schemas.google.com/blogger/2008/kind#settings":
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_ImportingSettings,
q.Id,
q.Content));
settings.Add(q.Id, q.Content);
break;
case "http://schemas.google.com/blogger/2008/kind#post":
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_ImportingPost,
q.Title));
// get a reference to the author of this entry
BlogMLAuthorReference authorReference = new BlogMLAuthorReference();
authorReference.Ref = UpdateAuthorsGetRef(ref authors, q.AuthorName, q.AuthorEmail, q.Published);
BlogMLPost post = new BlogMLPost();
post.Approved = true;
post.Authors.Add(authorReference);
if (categoryRefs.Count > 0)
{
post.Categories.AddRange(categoryRefs);
}
post.Content = content;
post.DateCreated = q.Published;
post.DateModified = q.Updated;
post.HasExcerpt = false;
post.ID = q.Id;
post.PostType = BlogPostTypes.Normal;
post.Title = q.Title;
posts.Add(q.Id, post);
break;
case "http://schemas.google.com/blogger/2008/kind#comment":
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_ImportingComment,
q.Title));
BlogMLComment comment = new BlogMLComment();
comment.Approved = true;
comment.Content = content;
comment.DateCreated = q.Published;
comment.DateModified = q.Updated;
comment.ID = q.Id;
comment.Title = q.Title;
comment.UserEMail = q.AuthorEmail;
comment.UserName = q.AuthorName;
comment.UserUrl = q.AuthorUri;
comments.Add(new CommentWithInReplyTo(comment, q.InReplyTo));
break;
default:
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_MessageUnexpectedEntryKind,
entryKind));
break;
}
}
// add comments to posts
foreach (CommentWithInReplyTo comment in comments)
{
if (posts.ContainsKey(comment.InReplyTo))
{
BlogMLPost post = posts[comment.InReplyTo];
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_AttachingComment,
comment.Comment.Title,
post.Title));
post.Comments.Add(comment.Comment);
}
else
{
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_OrphanedComment,
comment.Comment.Title));
}
}
// build the blog
LogMessage(Properties.Resources.Converter_BuildingBlogML);
BlogMLBlog blog = new BlogMLBlog();
blog.Authors.AddRange(authors);
blog.Categories.AddRange(categories);
blog.DateCreated = blogCreated;
blog.Posts.AddRange(posts.Values);
blog.Title = blogTitle;
// add blogger settings as extended properties
foreach (string name in settings.Keys)
{
Pair <string, string> pair = new Pair <string, string>();
pair.Key = name;
pair.Value = settings[name];
blog.ExtendedProperties.Add(pair);
}
// output BlogML
LogMessage(string.Format(CultureInfo.CurrentCulture,
Properties.Resources.Converter_WritingBlogML,
_blogmlPath));
XmlWriterSettings writerSettings = new XmlWriterSettings();
writerSettings.CheckCharacters = true;
writerSettings.CloseOutput = true;
writerSettings.ConformanceLevel = ConformanceLevel.Document;
writerSettings.Indent = true;
using (XmlWriter writer = XmlWriter.Create(_blogmlPath, writerSettings))
{
BlogMLSerializer.Serialize(writer, blog);
}
}
