YAML parsing bottleneck

I have started to notice very significant bottlenecks related to parsing the database file as it becomes exceedingly larger (multiple 100 entries). Some crude profiling shows that the YAML parsing is largely to blame for this.

Since we don't really need to support the full-blown YAML format (because we have a largely flat structure), a home-grown barebones parser could considerably speed things up. This is especially important when working with the CLI and not the TUI when the database needs to be parsed for basically every operation.