Module: Bond::Search

Included in:
Rc
Defined in:
lib/bond/search.rb

Overview

Contains search methods used to filter possible completions given what the user has typed for that completion. For a search method to be used by Bond.complete it must end in ‘_search’ and take two arguments: the Input string and an array of possible completions.

Creating a search method

Say you want to create a custom search which ignores completions containing ‘-’. In a completion file under Rc namespace, define this method:

def ignore_hyphen_search(input, list)
  normal_search(input, list.select {|e| e !~ /-/ })
end

Now you can pass this custom search to any complete() as :search => :ignore_hyphen

Class Attribute Summary collapse

Instance Method Summary collapse

Class Attribute Details

.default_search ⇒ Object

Default search used across missions, set by Bond.config



17
18
19
 
# File 'lib/bond/search.rb', line 17

def default_search
  @default_search
end

Instance Method Details

#anywhere_search(input, list) ⇒ Object

Searches completions anywhere in the string.



26
27
28
 
# File 'lib/bond/search.rb', line 26

def anywhere_search(input, list)
  list.grep(/#{Regexp.escape(input)}/)
end

#default_search(input, list) ⇒ Object

Default search across missions to be invoked by a search that wrap another search i.e. files_search.



49
50
51
 
# File 'lib/bond/search.rb', line 49

def default_search(input, list)
  send("#{Search.default_search}_search", input, list)
end

#files_search(input, list) ⇒ Object

Does default_search on the given paths but only returns ones that match the input’s current directory depth, determined by ‘/’. For example if a user has typed ‘irb/c’, this search returns matching paths that are one directory deep i.e. ‘irb/cmd/ irb/completion.rb irb/context.rb’.



56
57
58
 
# File 'lib/bond/search.rb', line 56

def files_search(input, list)
  incremental_filter(input, list, '/')
end

#ignore_case_search(input, list) ⇒ Object

Searches completions from the beginning and ignores case.



31
32
33
 
# File 'lib/bond/search.rb', line 31

def ignore_case_search(input, list)
  list.grep(/^#{Regexp.escape(input)}/i)
end

#incremental_filter(input, list, delim) ⇒ Object

Used by files_search and modules_search.



66
67
68
69
70
71
72
 
# File 'lib/bond/search.rb', line 66

def incremental_filter(input, list, delim)
  i = 0; input.gsub(delim) {|e| i+= 1 }
  delim_chars = delim.split('').uniq.join('')
  current_matches, future_matches = default_search(input, list).partition {|e|
    e[/^[^#{delim_chars}]*(#{delim}[^#{delim_chars}]+){0,#{i}}$/] }
  (current_matches + future_matches.map {|e| e[/^(([^#{delim_chars}]*#{delim}){0,#{i+1}})/, 1] }).uniq
end

#modules_search(input, list) ⇒ Object

Does the same as files_search but for modules. A module depth is delimited by ‘::’.



61
62
63
 
# File 'lib/bond/search.rb', line 61

def modules_search(input, list)
  incremental_filter(input, list, '::')
end

#normal_search(input, list) ⇒ Object

Searches completions from the beginning of the string.



21
22
23
 
# File 'lib/bond/search.rb', line 21

def normal_search(input, list)
  list.grep(/^#{Regexp.escape(input)}/)
end

#underscore_search(input, list) ⇒ Object

A normal_search which also provides aliasing of underscored words. For example ‘some_dang_long_word’ can be specified as ‘s_d_l_w’. Aliases can be any unique string at the beginning of an underscored word. For example, to choose the first completion between ‘so_long’ and ‘so_larger’, type ‘s_lo’.



39
40
41
42
43
44
45
46
 
# File 'lib/bond/search.rb', line 39

def underscore_search(input, list)
  if input[/_([^_]+)$/]
    regex = input.split('_').map {|e| Regexp.escape(e) }.join("([^_]+)?_")
    list.select {|e| e =~ /^#{regex}/ }
  else
    normal_search(input, list)
  end
end