Fuzzy finder only matches from the first letter onward. about wilder.nvim HOT 4 CLOSED

If you're using the python_search_pipeline, there's already an option for that:

call wilder#set_option('pipeline', [
      \   wilder#branch(
      \     wilder#cmdline_pipeline(),
      \     wilder#python_search_pipeline({
      \       'pattern': wilder#python_fuzzy_pattern({
      \         'start_at_boundary': 0,
      \       }),
      \       'engine': 're',
      \     }),
      \   ),
      \ ])

The pattern option can also take a function if you need to customise it further. See :h wilder#python_search_pipeline().

Note 1: start_at_boundary: 0 slows down the regex matching by a lot since it introduces a lot of backtracking.

Note 2: I highly recommend using re2 as the engine for performance reasons. It can be installed with pip install pyre2.

For vim_search_pipeline, I don't intend to add fuzzy matching as the searching is synchronous and Vim's regex engine is not as fast.

from wilder.nvim.

start_at_boundary only applies to the pattern in wilder#python_search_pipeline().

To get the same effect for wilder#cmdline_pipeline(), set 'fuzzy': 2:

call wilder#set_option('pipeline', [
      \   wilder#branch(
      \     wilder#cmdline_pipeline({'fuzzy': 2}),
      \     wilder#python_search_pipeline({
      \       'pattern': wilder#python_fuzzy_pattern({
      \         'start_at_boundary': 0,
      \       }),
      \       'engine': 're',
      \     }),
      \   ),
      \ ])

from wilder.nvim.

I met the same issue. After adding "'start_at_boundary': 0", it seems that it works for '/' and '?', but not for command line search ':'. I wonder if this is due to limitation of nvim.

from wilder.nvim.

Thank you! works as expected.

from wilder.nvim.