rejectu 0.0.2 → 0.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 26755280c4039506f58fa4ca6f055128a6cd4986
-  data.tar.gz: 241be0c2c820fc77f611fa66b69dfc842e71f1e1
+  metadata.gz: b1db9ef4eb58f201a481308327961182d1226609
+  data.tar.gz: 3ff4ccab421b53c29bd6ed08219c36218ce761e0
 SHA512:
-  metadata.gz: ff027e3da57a31fa94087f7303d288a61960e93d79d9e09afeb26f58d2b2e93bd5e9549c2d48256fdd82953c9f07c68c9ccd2208cb88057f43fc3ddef5a248c6
-  data.tar.gz: ca9c37d6035dee4d3258023cd3d4af76967c2bd5dc006aa666862be7f9e28905c649bcefaed6bed955d8ed6c6cf8dc8ca498ed989a45483d7618339944d885d6
+  metadata.gz: 4e1f64f8d17d061fb0e521bb63f4923896e00eed84375156dd4b3b3164b3671dce5021112fc08887ddd842381db6814a505cbea131a2d8e090893664294de2a0
+  data.tar.gz: 4a04a170ebb63932dae2c863ebc00777aaccab8b7ec5583aa44065d1f4773208c813afe2782e48dbe2ea79cf9764708951d0e8c181cea12d2f271f90d4c276dd

data/Gemfile.lock

@@ -1,14 +1,17 @@
 PATH
   remote: .
   specs:
-    rejectu (0.0.1)
+    rejectu (0.0.2)
 
 GEM
   remote: https://rubygems.org/
   specs:
+    power_assert (0.2.4)
     rake (10.3.2)
     rake-compiler (0.9.2)
       rake
+    test-unit (3.1.4)
+      power_assert
 
 PLATFORMS
   ruby
@@ -16,3 +19,4 @@ PLATFORMS
 DEPENDENCIES
   rake-compiler (~> 0.9)
   rejectu!
+  test-unit (~> 3.1)

data/README.md

@@ -6,18 +6,25 @@ A simple Ruby extension that verifies that a UTF-8 string does not contain any c
 
 - C extension that uses SSE2 for webscale
 
+### Installation
+
+`gem install rejectu`
+
+or if you're using bundler add the following to your Gemfile
+
+`gem "rejectu"`
+
 ### Usage
 
 ```ruby
 require 'rejectu/rejectu'
 
-Rejectu.valid? "happy! \xf2\xa4\xb7\xa4" # false
-Rejectu.valid? "really happy!" # true
-Rejectu.valid? "this should be good too \xe2\x84\xa2" # true
+Rejectu.valid?("happy! \xf2\xa4\xb7\xa4") # false
+Rejectu.valid?("really happy!") # true
+Rejectu.valid?("this should be good too \xe2\x84\xa2") # true
 
-Rejectu.scrub "happy! \xf2\xa4\xb7\xa4" # => "happy! ?"
-```
-
-### Limitations
+Rejectu.scrub("happy! \xf2\xa4\xb7\xa4") # => "happy! ?"
 
-- The replacement character for `scrub` currently cannot be configured
+# using a custom replacement character
+Rejectu.scrub("happy! \xf2\xa4\xb7\xa4", ".") # => "happy! ."
+```

data/ext/rejectu/rejectu.c

@@ -6,6 +6,7 @@
 
 static VALUE mRejectu = Qnil;
 static VALUE idEncoding, idTo_s;
+static VALUE defaultToken = Qnil;
 
 #ifdef __SSE2__
 static inline int
@@ -125,11 +126,12 @@ is_valid(VALUE self, VALUE str)
 }
 
 static VALUE
-do_scrub(VALUE str)
+do_scrub(VALUE str, VALUE rplToken)
 {
   VALUE out_str;
   unsigned char *p, *end, *out_start, *out;
   long len, out_len;
+  char token = StringValueCStr(rplToken)[0];
 
   validate_utf8_input(str);
 
@@ -151,7 +153,7 @@ do_scrub(VALUE str)
       } else {
         p += 4;
       }
-      *out++ = '?';
+      *out++ = token;
     } else {
       *out++ = *p++;
     }
@@ -165,34 +167,52 @@ do_scrub(VALUE str)
 }
 
 static VALUE
-scrub(VALUE self, VALUE str)
+scrub(int argc, VALUE *argv, VALUE self)
 {
-  if (is_valid(self, str) == Qtrue) {
-    return rb_enc_str_new(RSTRING_PTR(str), RSTRING_LEN(str), rb_utf8_encoding());
+  VALUE input, token;
+  rb_scan_args(argc, argv, "11", &input, &token);
+
+  if (is_valid(self, input) == Qtrue) {
+    return input;
+  }
+
+  if (token == Qnil) {
+    token = defaultToken;
   }
-  return do_scrub(str);
+
+  return do_scrub(input, token);
 }
 
 static VALUE
-scrub_bang(VALUE self, VALUE str)
+scrub_bang(int argc, VALUE *argv, VALUE self)
 {
-  VALUE repl;
-  if (is_valid(self, str) == Qtrue) {
-    return str;
+  VALUE input, token;
+  rb_scan_args(argc, argv, "11", &input, &token);
+
+  if (!is_valid(self, input)) {
+    if (token == Qnil) {
+      token = defaultToken;
+    }
+
+    VALUE repl = do_scrub(input, token);
+    if (!NIL_P(repl)) {
+      rb_str_replace(input, repl);
+    }
   }
-  repl = do_scrub(str);
-  if (!NIL_P(repl)) rb_str_replace(str, repl);
-  return str;
+
+  return input;
 }
 
 void
 Init_rejectu()
 {
   mRejectu = rb_define_module("Rejectu");
+  defaultToken = rb_str_new2("?");
+  rb_global_variable(&defaultToken);
 
   rb_define_singleton_method(mRejectu, "valid?", is_valid, 1);
-  rb_define_singleton_method(mRejectu, "scrub", scrub, 1);
-  rb_define_singleton_method(mRejectu, "scrub!", scrub_bang, 1);
+  rb_define_singleton_method(mRejectu, "scrub", scrub, -1);
+  rb_define_singleton_method(mRejectu, "scrub!", scrub_bang, -1);
 
   idEncoding = rb_intern("encoding");
   idTo_s = rb_intern("to_s");

data/rejectu.gemspec

@@ -1,6 +1,6 @@
 Gem::Specification.new do |s|
   s.name = 'rejectu'
-  s.version = '0.0.2'
+  s.version = '0.1.0'
   s.summary = 'Detects if a UTF-8 string supplementary plane code points'
   s.description = <<-DOC
     This gem detects if a UTF-8 encoded string contains characters from the UTF-8 supplementary
@@ -14,4 +14,5 @@ Gem::Specification.new do |s|
   s.files = `git ls-files`.split("\n")
   s.extensions = ['ext/rejectu/extconf.rb']
   s.add_development_dependency 'rake-compiler', '~> 0.9'
+  s.add_development_dependency 'test-unit', '~> 3.1'
 end

data/test/test_rejectu.rb

@@ -77,10 +77,19 @@ class TestRejectu < Test::Unit::TestCase
     assert_equal "? test string", Rejectu.scrub("\xf2\xa4\xb7\xa4 test string")
   end
 
+  def test_scrub_with_custom_token
+    assert_equal ". test string", Rejectu.scrub("\xf2\xa4\xb7\xa4 test string", ".")
+  end
+
   def test_scrub!
     s = "\xf2\xa4\xb7\xa4 test string"
     assert_equal "? test string", Rejectu.scrub!(s)
     assert_equal "? test string", s
   end
 
+  def test_scrub_with_custom_token!
+    s = "\xf2\xa4\xb7\xa4 test string"
+    assert_equal ". test string", Rejectu.scrub!(s, ".")
+    assert_equal ". test string", s
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rejectu
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.1.0
 platform: ruby
 authors:
 - Scott Francis
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-08-05 00:00:00.000000000 Z
+date: 2015-10-06 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake-compiler
@@ -24,6 +24,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '0.9'
+- !ruby/object:Gem::Dependency
+  name: test-unit
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.1'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.1'
 description: |2
       This gem detects if a UTF-8 encoded string contains characters from the UTF-8 supplementary
       plane (code points >= U+10000).
@@ -63,7 +77,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.2.2
+rubygems_version: 2.2.3
 signing_key: 
 specification_version: 4
 summary: Detects if a UTF-8 string supplementary plane code points