Somehow I managed to get the script function properly.
Here is my changes (with a little help from a Claude Code
)
>>mongo.rb
def posts(offset = 0, page_size = 1000)
post_keys = mongo.find(_key: "posts:pid").skip(offset).limit(page_size).pluck(:value)
post_keys
.map { |pid| post(pid) }
.compact # <-- drops any nil results (orphaned pids)
post_keys.map { |post_key| post(post_key) }
end
def post(id)
post = mongo.find(_key: "post:#{id}").first
return nil if post.nil? # <-- null checking
post["timestamp"] = timestamp_to_date(post["timestamp"])
if post["upvoted_by"] = mongo.find(_key: "pid:#{id}:upvote").first
post["upvoted_by"] = mongo.find(_key: "pid:#{id}:upvote").first[:members]
else
post["upvoted_by"] = []
end
post["pid"] = post["pid"].to_s
post["deleted"] = post["deleted"].to_s
post
end
>>nodebb.rb
create_posts(posts, total: post_count, offset: offset) do |post|
# skip if post is null
# skip if it's merged_post
next if post.nil?
next if @merged_posts_map[post["pid"]]
# skip if it's deleted
next if post["deleted"] == "1"
raw = post["content"]
post_id = "p#{post["pid"]}"
next if raw.blank?
topic = topic_lookup_from_imported_post_id("t#{post["tid"]}")
unless topic
puts "Topic with id #{post["tid"]} not found, skipping"
next
end
It seems working the right way now.
Although I don’t know how correct this is from the point of view of Discourse’s internal architecture, at first glance it seems to work.
Any suggestions for improvement and optimization are most welcome.