comment_strip-ruby 0.2.0 → 0.2.1

data/test/unit/tc_strip_c_defects.rb

@@ -0,0 +1,179 @@
+#! /usr/bin/env ruby
+
+$:.unshift File.join(__dir__, '../..', 'lib')
+
+
+require 'comment_strip'
+
+require 'xqsr3/extensions/test/unit'
+
+require 'test/unit'
+
+# Contractual / line-preserving expectations that currently fail against
+# LanguageFamilies::C. These cases document known defects; do not weaken the
+# expectations to match the buggy output — fix the scanner instead.
+class Test_C_strip_defects < Test::Unit::TestCase
+
+  include ::CommentStrip
+
+
+  # --- End-of-input / unterminated constructs --------------------------------
+
+  def test_trailing_slash_at_eof_is_preserved_after_identifier
+
+    # A lone '/' that never becomes '//' or '/*' must remain in the output.
+    assert_equal 'a/', strip('a/', 'C')
+  end
+
+  def test_trailing_slash_at_eof_alone_is_preserved
+
+    assert_equal '/', strip('/', 'C')
+  end
+
+  def test_trailing_slash_at_eof_after_spaces_is_preserved
+
+    assert_equal 'x = y /', strip('x = y /', 'C')
+  end
+
+  def test_unterminated_block_comment_preserves_embedded_newlines
+
+    # Line-preserving policy: newlines inside an open /* … */ that never
+    # closes should still appear in the output (as blank lines), matching the
+    # behaviour used when the comment *does* close.
+    input = "before/*\nline2\nline3"
+    expected = "before\n\n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+
+  # --- CRLF / line-ending fidelity ------------------------------------------
+
+  def test_crlf_inside_block_comment_counts_as_one_line_break
+
+    # One Windows line break inside a block comment must contribute one
+    # preserved newline, not one per CR and LF byte.
+    input = "a/*\r\nb*/c"
+    expected = "a\nc"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_crlf_only_block_comment_body_preserves_single_blank_line
+
+    input = "/*\r\n*/\r\nint x;\r\n"
+    expected = "\n\r\nint x;\r\n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_multiple_crlf_lines_inside_block_comment
+
+    input = "start/*\r\none\r\ntwo\r\n*/end"
+    expected = "start\n\n\nend"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+
+  # --- C++ digit separators (apostrophe is not a character literal) ---------
+
+  def test_cxx14_digit_separator_does_not_swallow_following_line_comment
+
+    # 1'000 uses '\'' as a digit separator between digits — there is no
+    # closing quote. The trailing // comment must still be stripped.
+    input = "int x = 1'000; // one thousand\n"
+    expected = "int x = 1'000; \n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_cxx14_digit_separator_does_not_swallow_following_block_comment
+
+    input = "long y = 1'000; /* million */\n"
+    expected = "long y = 1'000; \n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+
+  # --- Translation-phase / line-splice fidelity (C family claim) ------------
+
+  def test_backslash_newline_splices_before_line_comment_recognition
+
+    # In C/C++, backslash-newline is removed in translation phase 2, before
+    # comments are recognised. So the following is one // comment spanning
+    # both physical lines; `int x;` must not survive as code.
+    input = "// comment continues \\\nint x;\n"
+    expected = "\n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_backslash_newline_inside_block_comment_does_not_invent_extra_lines
+
+    # After splicing, this is a single-line block comment with no newline.
+    input = "a/* foo \\\n bar */b"
+    expected = "ab"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+
+  # --- Documented family breadth (rdoc / README list C#, Go, Java, Rust) ----
+
+  def test_go_raw_string_backticks_protect_comment_markers
+
+    # Go raw strings use `…`; // and /* inside them are not comments.
+    input = "s := `// not a comment /* nor this */`; // real\n"
+    expected = "s := `// not a comment /* nor this */`; \n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_csharp_verbatim_string_backslash_is_not_an_escape
+
+    # In C# @"…", backslash is literal. @"\"; ends the string after one
+    # backslash character; the following // is a real comment.
+    input = "s = @\"\\\"; // real\n"
+    expected = "s = @\"\\\"; \n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_rust_lifetime_then_line_comment
+
+    # A lone Rust lifetime `'a` is not a C character literal. Stuck in the
+    # one-character-literal state, // must still be recognised afterwards.
+    input = "fn f<'a>() {} // lifetime\n"
+    expected = "fn f<'a>() {} \n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+  def test_cxx_raw_string_embedded_double_quote_does_not_end_string
+
+    # C++ R"(…)" may contain " without ending the literal.
+    input = "s = R\"(\")\"; // real\n"
+    expected = "s = R\"(\")\"; \n"
+
+    assert_equal expected, strip(input, 'C')
+  end
+
+
+  # --- Dispatch / diagnostics -----------------------------------------------
+
+  def test_unrecognised_family_message_has_no_typo
+
+    begin
+      strip('', 'NotAFamily')
+      flunk 'expected RuntimeError'
+    rescue RuntimeError => x
+
+      assert_match(/unrecognised or not supported\z/, x.message)
+      assert_no_match(/supported1/, x.message)
+    end
+  end
+end
+
+# ############################## end of file ############################# #

data/test/unit/tc_strip_hash_line_defects.rb

@@ -0,0 +1,154 @@
+#! /usr/bin/env ruby
+
+$:.unshift File.join(__dir__, '../..', 'lib')
+
+
+require 'comment_strip'
+
+require 'xqsr3/extensions/test/unit'
+
+require 'test/unit'
+
+# Contractual expectations that currently fail against
+# LanguageFamilies::HashLine. These cases document known defects for the
+# README-claimed Ruby / Python / shell family; do not weaken the expectations
+# to match buggy output — fix the scanner instead.
+class Test_HashLine_strip_defects < Test::Unit::TestCase
+
+  include ::CommentStrip
+
+
+  # --- Empty and escaped single-quoted strings ------------------------------
+  #
+  # The scanner reuses the C one-character-literal machine for `'…'`, which
+  # is wrong for Ruby / Python / shell string literals.
+
+  def test_empty_single_quoted_string_then_hash_comment
+
+    # '' must open and immediately close a string; the following # is a
+    # comment.
+    input = "x='' # comment\n"
+    expected = "x='' \n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_empty_single_quoted_string_mid_statement_then_hash_comment
+
+    input = "a=''; b=2 # comment\n"
+    expected = "a=''; b=2 \n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_escaped_quote_inside_single_quoted_string_then_hash_comment
+
+    # Ruby / Python-style escaped quote inside '...'; # after the closing
+    # quote is a comment.
+    input = "s='it\\'s' # comment\n"
+    expected = "s='it\\'s' \n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_adjacent_empty_single_quoted_strings_then_hash_comment
+
+    # ''+'' must be two empty strings; # after them is a comment. The
+    # one-character-literal machine treats the second quote of each '' as
+    # the character body and never returns to :text for the #.
+    input = "x = ''+'' # comment\n"
+    expected = "x = ''+'' \n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+
+  # --- Language-specific quoting claimed by README (Ruby / Python / shell) --
+
+  def test_python_triple_single_quotes_protect_hash
+
+    input = "s = '''a # b'''\n"
+    expected = "s = '''a # b'''\n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_python_triple_single_quotes_then_real_comment
+
+    input = "s = '''a # b''' # real\n"
+    expected = "s = '''a # b''' \n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_ruby_percent_q_braces_protect_hash
+
+    input = "s = %q{a # b}\n"
+    expected = "s = %q{a # b}\n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_ruby_percent_q_braces_then_real_comment
+
+    input = "s = %q{a # b} # real\n"
+    expected = "s = %q{a # b} \n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_ruby_percent_w_brackets_protect_hash
+
+    input = "a = %w[one # two]\n"
+    expected = "a = %w[one # two]\n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_ruby_percent_q_parentheses_protect_hash
+
+    input = "s = %q(a # b)\n"
+    expected = "s = %q(a # b)\n"
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_ruby_heredoc_protects_hash
+
+    input = <<-EOF
+s = <<~TXT
+a # not comment
+TXT
+# real comment
+EOF
+    expected = <<-EOF
+s = <<~TXT
+a # not comment
+TXT
+
+EOF
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+
+  def test_ruby_heredoc_single_quoted_protects_hash
+
+    input = <<-EOF
+s = <<~'TXT'
+a # not comment
+TXT
+x = 1 # real
+EOF
+    expected = [
+      "s = <<~'TXT'\n",
+      "a # not comment\n",
+      "TXT\n",
+      "x = 1 ",
+      "\n",
+    ].join
+
+    assert_equal expected, strip(input, :Hash_Line)
+  end
+end
+
+# ############################## end of file ############################# #

metadata

@@ -1,39 +1,58 @@
 --- !ruby/object:Gem::Specification
 name: comment_strip-ruby
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - Matt Wilson
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-03-31 00:00:00.000000000 Z
+date: 2026-08-30 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: xqsr3
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '0.38'
+        version: 0.39.5
+    - - "<"
+      - !ruby/object:Gem::Version
+        version: '1.0'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.39.5
+    - - "<"
       - !ruby/object:Gem::Version
-        version: '0.38'
+        version: '1.0'
 description: 'Source code comment stripping library
 
   '
-email: matthew@synesis.com.au
+email:
+- matthew@synesis.com.au
 executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- AUTHORS.md
+- CHANGES.md
+- CONTRIBUTING.md
+- EXAMPLES.md
+- FAQ.md
+- INSTALL.md
 - LICENSE
+- NEWS.md
 - README.md
+- Rakefile
+- SECURITY.md
+- TODO.md
+- examples/read_c_family_source.rb
+- examples/read_c_family_source_from_stdin.rb
 - lib/comment_strip.rb
 - lib/comment_strip/language_families/c.rb
 - lib/comment_strip/language_families/hash_line.rb
@@ -41,13 +60,19 @@ files:
 - lib/comment_strip/version.rb
 - test/scratch/ex1.rb
 - test/unit/tc_strip_c.rb
+- test/unit/tc_strip_c_defects.rb
 - test/unit/tc_strip_hash_line.rb
+- test/unit/tc_strip_hash_line_defects.rb
 - test/unit/tc_version.rb
 - test/unit/ts_all.rb
 homepage: https://github.com/synesissoftware/comment_strip.r
 licenses:
 - BSD-3-Clause
-metadata: {}
+metadata:
+  bug_tracker_uri: https://github.com/synesissoftware/comment_strip.r/issues
+  changelog_uri: https://github.com/synesissoftware/comment_strip.r/blob/master/CHANGES.md
+  homepage_uri: https://github.com/synesissoftware/comment_strip.r
+  source_code_uri: https://github.com/synesissoftware/comment_strip.r
 post_install_message:
 rdoc_options: []
 require_paths:
@@ -57,9 +82,6 @@ required_ruby_version: !ruby/object:Gem::Requirement
   - - ">="
     - !ruby/object:Gem::Version
       version: '2.0'
-  - - "<"
-    - !ruby/object:Gem::Version
-      version: '4'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
@@ -69,5 +91,5 @@ requirements: []
 rubygems_version: 3.2.15
 signing_key:
 specification_version: 4
-summary: comment_strip.r
+summary: Comment Strip, for Ruby
 test_files: []