Currently only letters, numbers and a subset of letter-like symbols are searchable. Every other character gets ignored, i.e. treated as whitespace. Furthermore there seems to be no way to search for exact matches e.g. by quoting a string, or to perform a case sensitive search.
I have over 130 private and shared notebooks that I cannot organize in any meaningful way, because only public notebooks can be added to collections. To work around this limitation I would like to start adding tags to each notebook body, using the “#” character as a word boundary delimiter, e.g.:
md`# Notebook title` /* #wip #3d #cleanup */ /* Alternative with "namespace": tag#wip tag#3d tag#cleanup */
The idea here is that the combination of words + # is rare enough to avoid (too many) false positives.
At present this does not work because e.g. “tag#3d” gets indexed/searched as “tag 3d”, performing partial matches for both “tag” and “3d”. It would be great if “#” could instead be treated as a letter, so that “tag#3d” would be considered a single word.