Class: Rake::FileList

Relationships & Source Files
Inherits:	Object
Defined in:	lib/rake/file_list.rb

Overview

A FileList is essentially an array with a few helper methods defined to make file manipulation a bit easier.

FileLists are lazy. When given a list of glob patterns for possible files to be included in the file list, instead of searching the file structures to find the files, a FileList holds the pattern for latter use.

This allows us to define a number of FileList to match any number of files, but only search out the actual files when then FileList itself is actually used. The key is that the first time an element of the FileList/Array is requested, the pending patterns are resolved into a real list of file names.

Constant Summary

ARRAY_METHODS =

List of array methods (that are not in Object) that need to be delegated.

# File 'lib/rake/file_list.rb', line 43
```
(Array.instance_methods - Object.instance_methods).map(&:to_s)
```

DEFAULT_IGNORE_PATTERNS =
# File 'lib/rake/file_list.rb', line 380

[
  /(^|[\/\\])CVS([\/\\]|$)/,
  /(^|[\/\\])\.svn([\/\\]|$)/,
  /\.bak$/,
  /~$/
]

DEFAULT_IGNORE_PROCS =
# File 'lib/rake/file_list.rb', line 386

[
  proc { |fn| fn =~ /(^|[\/\\])core$/ && ! File.directory?(fn) }
]

DELEGATING_METHODS =
# File 'lib/rake/file_list.rb', line 60

(ARRAY_METHODS + MUST_DEFINE - MUST_NOT_DEFINE).map(&:to_s).sort.uniq

GLOB_PATTERN =
# File 'lib/rake/file_list.rb', line 85
```
%r{[*?\[\{]}
```
MUST_DEFINE =

List of additional methods that must be delegated.

# File 'lib/rake/file_list.rb', line 46
```
%w[inspect <=>]
```
MUST_NOT_DEFINE =

List of methods that should not be delegated here (we define special versions of them explicitly below).

# File 'lib/rake/file_list.rb', line 50
```
%w[to_a to_ary partition * <<]
```
SPECIAL_RETURN =

List of delegated methods that return new array values which need wrapping.

# File 'lib/rake/file_list.rb', line 54
```
%w[
  map collect sort sort_by select find_all reject grep
  compact flatten uniq values_at
  + - & |
]
```

Class Method Summary

.[](*args)

Create a new file list including the files listed.
.glob(pattern, *args)

Get a sorted list of files matching the pattern.
.new(*patterns) {|_self| ... } ⇒ FileList constructor

Create a file list from the globbable patterns given.

Instance Method Summary

#*(other)

Redefine * to return either a string or a new file list.
#<<(obj)
#==(array)

A FileList is equal through array equality.
#add(*filenames)

Alias for #include.
#clear_exclude

Clear all the exclude patterns so that we exclude nothing.
#egrep(pattern, *options)

Grep each of the files in the filelist using the given pattern.
#exclude(*patterns, &block)

Register a list of file name patterns that should be excluded from the list.
#excluded_from_list?(fn) ⇒ Boolean

Should the given file name be excluded from the list?
#existing

Return a new file list that only contains file names from the current file list that exist on the file system.
#existing!

Modify the current file list so that it contains only file name that exist on the file system.
#ext(newext = "")

Return a new FileList with String#ext method applied to each member of the array.
#gsub(pat, rep)

Return a new FileList with the results of running gsub against each element of the original list.
#gsub!(pat, rep)

Same as #gsub except that the original file list is modified.
#include(*filenames) (also: #add)

Add file names defined by glob patterns to the file list.
#is_a?(klass) ⇒ Boolean (also: #kind_of?)

Lie about our class.
#kind_of?(klass)

Alias for #is_a?.
#pathmap(spec = nil, &block)

Apply the pathmap spec to each of the included file names, returning a new file list with the modified paths.
#resolve

Resolve all the pending adds now.
#sub(pat, rep)

Return a new FileList with the results of running sub against each element of the original list.
#sub!(pat, rep)

Same as #sub except that the original file list is modified.
#to_a

Return the internal array object.
#to_ary

Return the internal array object.
#to_s

Convert a FileList to a string by joining all elements with a space.
#add_matching(pattern) private

Add matching glob patterns.

Constructor Details

.new(*patterns) {|_self| ... } ⇒ `FileList`

Create a file list from the globbable patterns given. If you wish to perform multiple includes or excludes at object build time, use the “yield self” pattern.

Example:

file_list = FileList.new('lib/**/*.rb', 'test/test*.rb')

pkg_files = FileList.new('lib/**/*') do |fl|
  fl.exclude(/\bCVS\b/)
end

Yields:

(_self)

Yield Parameters:

_self (FileList) —

the object that the method was called on

[ GitHub ]

# File 'lib/rake/file_list.rb', line 98


def initialize(*patterns)
  @pending_add = []
  @pending = false
  @exclude_patterns = DEFAULT_IGNORE_PATTERNS.dup
  @exclude_procs = DEFAULT_IGNORE_PROCS.dup
  @items = []
  patterns.each { |pattern| include(pattern) }
  yield self if block_given?
end

Class Method Details

.[](*args)

Create a new file list including the files listed. Similar to:

FileList.new(*args)

[ GitHub ]

# File 'lib/rake/file_list.rb', line 399


def [](*args)
  new(*args)
end

.glob(pattern, *args)

Get a sorted list of files matching the pattern. This method should be preferred to Dir and Dir.glob(pattern) because the files returned are guaranteed to be sorted.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 406


def glob(pattern, *args)
  Dir.glob(pattern, *args).sort
end

Instance Method Details

#*(other)

Redefine * to return either a string or a new file list.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 192


def *(other)
  result = @items * other
  case result
  when Array
    self.class.new.import(result)
  else
    result
  end
end

#<<(obj)

[ GitHub ]

# File 'lib/rake/file_list.rb', line 202


def <<(obj)
  resolve
  @items << Rake.from_pathname(obj)
  self
end

#==(array)

A FileList is equal through array equality.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 170


def ==(array)
  to_ary == array
end

#add(*filenames)

Alias for #include.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 127


alias :add :include

#add_matching(pattern) (private)

Add matching glob patterns.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 349


def add_matching(pattern)
  self.class.glob(pattern).each do |fn|
    self << fn unless excluded_from_list?(fn)
  end
end

#clear_exclude

Clear all the exclude patterns so that we exclude nothing.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 163


def clear_exclude
  @exclude_patterns = []
  @exclude_procs = []
  self
end

#egrep(pattern, *options)

Grep each of the files in the filelist using the given pattern. If a block is given, call the block on each matching line, passing the file name, line number, and the matching line of text. If no block is given, a standard emacs style file:linenumber:line message will be printed to standard out. Returns the number of matched items.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 292


def egrep(pattern, *options)
  matched = 0
  each do |fn|
    begin
      open(fn, "r", *options) do |inf|
        count = 0
        inf.each do |line|
          count += 1
          if pattern.match(line)
            matched += 1
            if block_given?
              yield fn, count, line
            else
              puts "#{fn}:#{count}:#{line}"
            end
          end
        end
      end
    rescue StandardError => ex
      $stderr.puts "Error while processing '#{fn}': #{ex}"
    end
  end
  matched
end

#exclude(*patterns, &block)

Register a list of file name patterns that should be excluded from the list. Patterns may be regular expressions, glob patterns or regular strings. In addition, a block given to exclude will remove entries that return true when given to the block.

Note that glob patterns are expanded against the file system. If a file is explicitly added to a file list, but does not exist in the file system, then an glob pattern in the exclude list will not exclude the file.

Examples:

FileList['a.c', 'b.c'].exclude("a.c") => ['b.c']
FileList['a.c', 'b.c'].exclude(/^a/)  => ['b.c']

If “a.c” is a file, then …

FileList['a.c', 'b.c'].exclude("a.*") => ['b.c']

If “a.c” is not a file, then …

FileList['a.c', 'b.c'].exclude("a.*") => ['a.c', 'b.c']

[ GitHub ]

# File 'lib/rake/file_list.rb', line 149


def exclude(*patterns, &block)
  patterns.each do |pat|
    if pat.respond_to? :to_ary
      exclude(*pat.to_ary)
    else
      @exclude_patterns << Rake.from_pathname(pat)
    end
  end
  @exclude_procs << block if block_given?
  resolve_exclude unless @pending
  self
end

#excluded_from_list?(fn) ⇒ `Boolean`

Should the given file name be excluded from the list?

NOTE: This method was formerly named “exclude?”, but Rails introduced an exclude? method as an array method and setup a conflict with file list. We renamed the method to avoid confusion. If you were using “FileList#exclude?” in your user code, you will need to update.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 363


def excluded_from_list?(fn)
  return true if @exclude_patterns.any? do |pat|
    case pat
    when Regexp
      fn =~ pat
    when GLOB_PATTERN
      flags = File::FNM_PATHNAME
      # Ruby <= 1.9.3 does not support File::FNM_EXTGLOB
      flags |= File::FNM_EXTGLOB if defined? File::FNM_EXTGLOB
      File.fnmatch?(pat, fn, flags)
    else
      fn == pat
    end
  end
  @exclude_procs.any? { |p| p.call(fn) }
end

#existing

Return a new file list that only contains file names from the current file list that exist on the file system.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 319


def existing
  select { |fn| File.exist?(fn) }
end

#existing!

Modify the current file list so that it contains only file name that exist on the file system.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 325


def existing!
  resolve
  @items = @items.select { |fn| File.exist?(fn) }
  self
end

#ext(newext = "")

Return a new FileList with String#ext method applied to each member of the array.

This method is a shortcut for:

array.collect { |item| item.ext(newext) }

ext is a user added method for the Array class.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 283


def ext(newext="")
  collect { |fn| fn.ext(newext) }
end

#gsub(pat, rep)

Return a new FileList with the results of running gsub against each element of the original list.

Example:

FileList['lib/test/file', 'x/y'].gsub(/\//, "\\")
   #=> ['lib\\test\\file', 'x\\y']

[ GitHub ]

# File 'lib/rake/file_list.rb', line 252


def gsub(pat, rep)
  inject(self.class.new) { |res, fn| res << fn.gsub(pat, rep) }
end

#gsub!(pat, rep)

Same as #gsub except that the original file list is modified.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 263


def gsub!(pat, rep)
  each_with_index { |fn, i| self[i] = fn.gsub(pat, rep) }
  self
end

#include(*filenames) Also known as: #add

Add file names defined by glob patterns to the file list. If an array is given, add each element of the array.

Example:

file_list.include("*.java", "*.cfg")
file_list.include %w( math.c lib.h *.o )

[ GitHub ]

# File 'lib/rake/file_list.rb', line 115


def include(*filenames)
  # TODO: check for pending
  filenames.each do |fn|
    if fn.respond_to? :to_ary
      include(*fn.to_ary)
    else
      @pending_add << Rake.from_pathname(fn)
    end
  end
  @pending = true
  self
end

#is_a?(klass) ⇒ `Boolean` Also known as: #kind_of?

Lie about our class.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 186


def is_a?(klass)
  klass == Array || super(klass)
end

#kind_of?(klass)

Alias for #is_a?.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 189


alias kind_of? is_a?

#pathmap(spec = nil, &block)

Apply the pathmap spec to each of the included file names, returning a new file list with the modified paths. (See String#pathmap for details.)

[ GitHub ]

# File 'lib/rake/file_list.rb', line 271


def pathmap(spec=nil, &block)
  collect { |fn| fn.pathmap(spec, &block) }
end

#resolve

Resolve all the pending adds now.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 209


def resolve
  if @pending
    @pending = false
    @pending_add.each do |fn| resolve_add(fn) end
    @pending_add = []
    resolve_exclude
  end
  self
end

#sub(pat, rep)

Return a new FileList with the results of running sub against each element of the original list.

Example:

FileList['a.c', 'b.c'].sub(/\.c$/, '.o')  => ['a.o', 'b.o']

[ GitHub ]

# File 'lib/rake/file_list.rb', line 241


def sub(pat, rep)
  inject(self.class.new) { |res, fn| res << fn.sub(pat, rep) }
end

#sub!(pat, rep)

Same as #sub except that the original file list is modified.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 257


def sub!(pat, rep)
  each_with_index { |fn, i| self[i] = fn.sub(pat, rep) }
  self
end

#to_a

Return the internal array object.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 175


def to_a
  resolve
  @items
end

#to_ary

Return the internal array object.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 181


def to_ary
  to_a
end

#to_s

Convert a FileList to a string by joining all elements with a space.

[ GitHub ]

# File 'lib/rake/file_list.rb', line 343


def to_s
  resolve
  self.join(" ")
end

Class: Rake::FileList

Overview

Constant Summary

Class Method Summary

Instance Method Summary

Constructor Details

.new(*patterns) {|_self| ... } ⇒ FileList

Class Method Details

.[](*args)

.glob(pattern, *args)

Instance Method Details

#*(other)

#<<(obj)

#==(array)

#add(*filenames)

#add_matching(pattern) (private)

#clear_exclude

#egrep(pattern, *options)

#exclude(*patterns, &block)

#excluded_from_list?(fn) ⇒ Boolean

#existing

#existing!

#ext(newext = "")

#gsub(pat, rep)

#gsub!(pat, rep)

#include(*filenames) Also known as: #add

#is_a?(klass) ⇒ Boolean Also known as: #kind_of?

#kind_of?(klass)

#pathmap(spec = nil, &block)

#resolve

#sub(pat, rep)

#sub!(pat, rep)

#to_a

#to_ary

#to_s

.new(*patterns) {|_self| ... } ⇒ `FileList`

#excluded_from_list?(fn) ⇒ `Boolean`

#is_a?(klass) ⇒ `Boolean` Also known as: #kind_of?