Commands

These are the commands that @charlesmike knows. Any comment addressed to the bot, containing the words in the command name should work:

• hello - it will introduce itself;
• index site - indexes the whole website, crawling it as a graph;
  • any comment containing the words “index” and “site” will trigger this. Recommended only for small sites. For bigger sites/blogs, try indexing by sitemap.
• index sitemap - index the site represented by a sitemap.xml file;
  • any comment containing the words “index” and “sitemap”, and also the link to the sitemap.xml file in Markdown.
    E.g.

    @charlemike index [this](link/to/sitemap) sitemap pls

• index page - indexes a single page;
  • the comment has to contain the words “index” and “page”, and also the link to the page in markdown format;
    E.g.

    @charlesmike index [this](link/to/page) page)

• delete - deletes the index;
  • the comment has to contain the word “delete” and the repo’s name (for double-checking);

Configuration

There is an optional config file, which you can use to tune the bot. See here.

Implementation

The bot is written in Java, packaged as a .war file and should work fine on any JavaEE 7 platform. The whole code is available on Github. The architecture is described in detail here.

Works with

The key systems which enable @charlesmike to work are: ElasticSearch, Github API and Amazon Web Services.