tty-file 0.7.1 → 0.8.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 5d22dc89aced7b8ab2593a332739545706562d2683f93e9a88ea7fdec5a06453
-  data.tar.gz: c94f59cbaaccf16ce290e08ca44ad9c4ebfb7f105d953db9c6ba0dc00de369da
+  metadata.gz: ff0231767c10a14f731dcf153855ef3c652f0db9996e0a9588ad17b8c6bb4bdc
+  data.tar.gz: a441339b4d10b0ab58e6775d1df7471c36a475700ffdd385ad18d75c0b28459b
 SHA512:
-  metadata.gz: 2a583724c8f93ae02a69cea8c171076f53305fea40b7e6cf04d9de0ec3a0d275847a1d5a3da87ae42453b40835d15165ca7b32156c47b6d5e280b05cd07b3d2c
-  data.tar.gz: 0e737e0619ef7db2d1cc18507b768ca4d62d88cef2dbf12501f8b855598c5aaaed11ecc88b87505a9e4cfd759fc1bfb83afc2fc1b293f6a9e56405482823668e
+  metadata.gz: 322c2d767842c0a315c98eb01175cb0b2ca1b32ce6a5d961ef274d67db9428c5c6b32a710c81345a6383ca92cd67862f2a3354a853f587ed00b47196eb75f664
+  data.tar.gz: 340eb4e7b495e9ed91224fd6eec7df4e4990c4ef5dd5024e08110285560bf3e1bf5dd6c33433493f145f8b5dc3f76aa60052fbca5933816dbca0569c60e2ef77

data/CHANGELOG.md

@@ -1,5 +1,13 @@
 # Change log
 
+## [v0.8.0] - 2019-07-25
+
+### Added
+* Add #read_to_char for reading file content limited by bytes until a valid char
+
+### Fixed
+* Fix #binary? to correctly identify large UTF-8 files as non-binary
+
 ## [v0.7.1] - 2019-05-06
 
 ### Changed
@@ -88,6 +96,7 @@
 
 * Initial implementation and release
 
+[v0.8.0]: https://github.com/piotrmurach/tty-file/compare/v0.7.1...v0.8.0
 [v0.7.1]: https://github.com/piotrmurach/tty-file/compare/v0.7.0...v0.7.1
 [v0.7.0]: https://github.com/piotrmurach/tty-file/compare/v0.6.0...v0.7.0
 [v0.6.0]: https://github.com/piotrmurach/tty-file/compare/v0.5.0...v0.6.0

data/lib/tty/file.rb

@@ -53,8 +53,8 @@ module TTY
     def binary?(relative_path)
       bytes = ::File.new(relative_path).size
       bytes = 2**12 if bytes > 2**12
-      buffer = ::File.read(relative_path, bytes, 0) || ''
-      buffer = buffer.force_encoding(Encoding.default_external)
+      buffer = read_to_char(relative_path, bytes, 0)
+
       begin
         return buffer !~ /\A[\s[[:print:]]]*\z/m
       rescue ArgumentError => error
@@ -64,6 +64,35 @@ module TTY
     end
     module_function :binary?
 
+    # Read bytes from a file up to valid character
+    #
+    # @param [String, Pathname] relative_path
+    #   the path to file
+    #
+    # @param [Integer] bytes
+    #
+    # @example
+    #   TTY::File.read_to_char()
+    #
+    # @return [String]
+    #
+    # @api public
+    def read_to_char(relative_path, bytes = nil, offset = nil)
+      buffer = ""
+      ::File.open(relative_path) do |file|
+        buffer = file.read(bytes) || ""
+        buffer = buffer.dup.force_encoding(Encoding.default_external)
+
+        while !file.eof? && !buffer.valid_encoding? &&
+              (buffer.bytesize < bytes + 10)
+
+          buffer += file.read(1).force_encoding(Encoding.default_external)
+        end
+      end
+      buffer
+    end
+    module_function :read_to_char
+
     # Create checksum for a file, io or string objects
     #
     # @param [File, IO, String, Pathname] source

data/lib/tty/file/version.rb

@@ -2,6 +2,6 @@
 
 module TTY
   module File
-    VERSION = "0.7.1"
+    VERSION = "0.8.0"
   end # File
 end # TTY

data/spec/spec_helper.rb

@@ -30,16 +30,16 @@ module Helpers
     ::File.realpath(path)
   end
 
-  def fixtures_path(filename = nil)
-    ::File.join(dir_path('spec', 'fixtures'), filename.to_s)
+  def fixtures_path(*args)
+    ::File.join(dir_path('spec', 'fixtures'), *args)
   end
 
-  def tmp_path(filename = nil)
-    ::File.join(dir_path('tmp'), filename.to_s)
+  def tmp_path(*args)
+    ::File.join(dir_path('tmp'), *args)
   end
 
-  def tmp_pathname(filename = nil)
-    Pathname.new(tmp_path(filename))
+  def tmp_pathname(*args)
+    Pathname.new(tmp_path(*args))
   end
 
   def exists_and_identical?(source, dest)

data/spec/unit/binary_spec.rb

@@ -54,6 +54,12 @@ RSpec.describe TTY::File, '#binary?' do
     expect(TTY::File.binary?(file)).to eq(false)
   end
 
+  it "indentifies a file as non-binary when greater than 4096 bytes with unicode chars", unless: RSpec::Support::OS.windows? do
+    file = tmp_pathname("binary", "unicode.txt")
+
+    expect(TTY::File.binary?(file)).to eq(false)
+  end
+
   it "indetifies a null-terminated string file as binary" do
     Tempfile.open('tty-file-binary-spec') do |file|
       file.write("Binary content.\0")

data/spec/unit/read_to_char_spec.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+RSpec.describe TTY::File, "#read_to_char" do
+  before do
+    @saved_verbosity =  $VERBOSE
+    @saved_encoding = Encoding.default_external
+    $VERBOSE = nil
+    Encoding.default_external = Encoding::UTF_8
+  end
+
+  after do
+    Encoding.default_external = @saved_encoding
+    $VERBOSE = @saved_verbosity
+  end
+
+  it "reads file up to valid char", unless: RSpec::Support::OS.windows? do
+    file = tmp_pathname("binary", "unicode.txt")
+    bytes = 4096
+
+    content = TTY::File.read_to_char(file, bytes)
+
+    expect(content.bytesize).to eq(bytes + 2)
+  end
+end

data/tasks/console.rake

@@ -8,3 +8,4 @@ task :console do
   ARGV.clear
   IRB.start
 end
+task c: %w[ console ]

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: tty-file
 version: !ruby/object:Gem::Version
-  version: 0.7.1
+  version: 0.8.0
 platform: ruby
 authors:
 - Piotr Murach
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2019-05-06 00:00:00.000000000 Z
+date: 2019-07-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: pastel
@@ -149,6 +149,7 @@ files:
 - spec/unit/escape_glob_path_spec.rb
 - spec/unit/inject_into_file_spec.rb
 - spec/unit/prepend_to_file_spec.rb
+- spec/unit/read_to_char_spec.rb
 - spec/unit/remove_file_spec.rb
 - spec/unit/replace_in_file_spec.rb
 - spec/unit/tail_file_spec.rb