- ruby-dev - ml.ruby-lang.org

[ruby-dev:52159] [Ruby Feature#21393] One-line pattern matching syntax as a method argument
by koic (Koichi ITO) 02 Jun '25

02 Jun '25

Issue #21393 has been reported by koic (Koichi ITO). ---------------------------------------- Feature #21393: One-line pattern matching syntax as a method argument https://bugs.ruby-lang.org/issues/21393 * Author: koic (Koichi ITO) * Status: Open ---------------------------------------- ## Context Using one-line `in` pattern matching syntax as a method argument causes a syntax error. ```console $ ruby -vce 'do_something(expression in pattern)' ruby 3.4.4 (2025-05-14 revision a38531fd3f) +PRISM [x86_64-darwin24] ruby: -e:1: syntax errors found (SyntaxError) > 1 | do_something(expression in pattern) | ^~ unexpected 'in'; expected a `)` to close the arguments | ^ unexpected ')', ignoring it | ^ unexpected ')', expecting end-of-input ``` Wrapping the argument in parentheses avoids the syntax error: ```console $ ruby -vce 'do_something((expression in pattern))' ruby 3.4.4 (2025-05-14 revision a38531fd3f) +PRISM [x86_64-darwin24] Syntax OK ``` This was unexpected, as I assumed `do_something(expression in pattern)` would work without parentheses. For reference, `do_something(expression => pattern)`, which results in a void context, is valid syntax. ```console $ ruby -vce 'do_something(expression => pattern)' ruby 3.4.4 (2025-05-14 revision a38531fd3f) +PRISM [x86_64-darwin24] Syntax OK ``` ## Proposal It seems more appropriate that `do_something(expression in pattern)` should also be valid syntax. ## Additional Information This behavior is consistent not only in Prism, but also in `parse.y`. ```console $ ruby --parser=parse.y -vce 'do_something(expression in pattern)' ruby 3.4.4 (2025-05-14 revision a38531fd3f) [x86_64-darwin24] -e:1: syntax error, unexpected 'in', expecting ')' do_something(expression in pattern) ruby: compile error (SyntaxError) ``` I encountered this while working on a RuboCop issue: https://github.com/rubocop/rubocop/issues/14217 -- https://bugs.ruby-lang.org/

2 2

[ruby-dev:52158] [Ruby Feature#14916] Proposal to add Array#===
by matheusrich (Matheus Richard) 16 May '25

16 May '25

Issue #14916 has been updated by matheusrich (Matheus Richard). This is probably unnecessary now that pattern matching exists. ---------------------------------------- Feature #14916: Proposal to add Array#=== https://bugs.ruby-lang.org/issues/14916#change-113306 * Author: osyo (manga osyo) * Status: Open ---------------------------------------- ## 概要 `Array#===` を追加する提案になります。基本的な動作は『`Array#==` の `===` で比較する版』になります。 ## 仕様配列の各要素をそれぞれ順に `===` で比較し、全要素が `true` の場合に `true` を返す。そうでない場合は `false` を返す。 ### 動作例 ```ruby # 配列の各要素を #=== を使用して比較する [ String, /\w/ ] === [ "a", "c", 7 ] # => false [ String, /\w/, (1..10) ] === [ "a", "c", 7 ] # => true [ String, /\w/, (1..10) ] === [ "a", "!", 42 ] # => false ``` ### 真を返すケース * レシーバと引数が配列で同じサイズかつ、各要素を `===` で比較したときに全て `true` になる場合 * レシーバと引数が同じオブジェクトの場合 ### 偽を返すケース * レシーバと引数が配列で同じサイズかつ、各要素を `===` で比較したときに `false` がある場合 * レシーバと引数の配列のサイズが異なる場合 * 引数が配列以外の場合 ## ユースケース ### 引数の値によって処理を変える可変長引数で引数を受け取り、そのまま case-when で値を精査する ```ruby def plus *args case args # 数値の場合 when [Integer, Integer] args[0] + args[1] # 数字の場合 when [/^\d+$/, /^\d+$/] args[0].to_i + args[1].to_i # それ以外はエラー else raise "Error" end end p plus 1, 2 # => 3 p plus "3", "4" # => 7 p plus "homu", "mado" # Error (RuntimeError) ``` ### FizzBuzz を用いた例任意の処理の結果を複数回参照したい場合、配列でまとめて case-when で利用する ```ruby def fizzbuzz n _ = proc { true } case [n % 3, n % 5] # n % 3 === 0 && n % 5 === 0 when [0, 0] "FizzBuzz" # n % 3 === 0 when [0, _] "Fizz" # n % 5 === 0 when [_, 0] "Buzz" else n end end p (1..20).map &method(:fizzbuzz) # => [1, 2, "Fizz", 4, "Buzz", "Fizz", 7, 8, "Fizz", "Buzz", 11, "Fizz", 13, 14, "FizzBuzz", 16, 17, "Fizz", 19, "Buzz"] ``` ## 関連しそうなチケット * [Feature #14869: Proposal to add Hash#=== - Ruby trunk - Ruby Issue Tracking System](https://bugs.ruby-lang.org/issues/14869?next_issue_id=14866&prev_is… * [Feature #14913: Extend case to match several values at once - Ruby trunk - Ruby Issue Tracking System](https://bugs.ruby-lang.org/issues/14913) 以上、 `Array#===` に関する提案になります。挙動に関して疑問点や意見などございましたらコメント頂けると助かります。 ---Files-------------------------------- array_eqq.patch (3.06 KB) array_eqq.patch (3.2 KB) array_eqq.patch (3.41 KB) -- https://bugs.ruby-lang.org/

1 0

[ruby-dev:52155] [Ruby Bug#21278] configureスクリプトが検出するruby_versionの値に'+0'が付く
by yasuhirokimura (Yasuhiro KIMURA) 22 Apr '25

22 Apr '25

Issue #21278 has been reported by yasuhirokimura (Yasuhiro KIMURA). ---------------------------------------- Bug #21278: configureスクリプトが検出するruby_versionの値に'+0'が付く https://bugs.ruby-lang.org/issues/21278 * Author: yasuhirokimura (Yasuhiro KIMURA) * Status: Open * ruby -v: ruby 3.5.0preview1 (2025-04-18 master d06ec25be4) +PRISM [x86_64-linux] * Backport: 3.2: UNKNOWN, 3.3: UNKNOWN, 3.4: UNKNOWN ---------------------------------------- # 再現手順 1. ruby-3.5.0-preview1.tar.xzをダウンロード 2. tar xfpvJ ruby-3.5.0-preview1.tar.xz 3. cd ruby-3.5.0-preview1 4. ./configure 5. fgrep ruby_version= config.log # 期待される結果 ``` % fgrep ruby_version= config.log ruby_version='3.5.0' % ``` # 実際に得られる結果 ``` % fgrep ruby_version= config.log ruby_version='3.5.0+0' % ``` 上記の手順4を`./configure --with-ruby-version=minor`にした場合にも同様の結果(ruby_versionの最後に"+0"が付く)となる # 環境以下の環境で問題が発生することを確認済 * Debian 12.10 amd64 * FreeBSD 14.2-RELEASE amd64 -- https://bugs.ruby-lang.org/

2 2

[ruby-dev:52153] [Ruby master Bug#21165] 引数なしで起動するとSegmentatino Fault
by taca (Takahiro Kambe) 02 Mar '25

02 Mar '25

Issue #21165 has been reported by taca (Takahiro Kambe). ---------------------------------------- Bug #21165: 引数なしで起動するとSegmentatino Fault https://bugs.ruby-lang.org/issues/21165 * Author: taca (Takahiro Kambe) * Status: Open * ruby -v: ruby 3.4.2 (2025-02-15 revision d2930f8e7a) +PRISM [x86_64-netbsd] * Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN, 3.4: UNKNOWN ---------------------------------------- 引数なしでrubyを起動すると、Segmentation Faultで落ちます。(そういう使い方をすることは稀だったため、気づいていませんでした。) https://mail-index.netbsd.org/pkgsrc-changes/2025/02/09/msg317142.html https://mail-index.netbsd.org/pkgsrc-users/2025/02/28/msg041167.html * Prismを使用していると発生します。 * pkgsrcの外でビルドした場合でも再現します。原因は prism.c の`pm_parse_stream_read()`の第3引数の関数ポインターの名前が`fgets`となっていて、`pm_parse_stream_read()`では、これを使って関数呼び出しをしています。ところが、何故か標準ライブラリのfgets(3)を呼び出してしまって落ちていたというのが真相でした。引数で渡された関数ポインターを使って呼び出すのが本来の動作とは思いますが、敢えて標準ライブラリと同じ名前を使用するのは、このような意図しない挙動を引き起こす可能性があるので、避けるべきではないかと思います。添付のパッチで問題は解消しました。 ``` % uname -rsmp NetBSD 10.1_STABLE amd64 x86_64 % sh configure --enable-pthread --without-git ... % make ... % gdb ruby ... (gdb) run Starting program: /var/tmp/ruby-3.4.2/ruby [New LWP 1009 of process 2755] `RubyGems' were not loaded. `error_highlight' was not loaded. `did_you_mean' was not loaded. `syntax_suggest' was not loaded. Thread 1 "" received signal SIGSEGV, Segmentation fault. rb_econv_close (ec=0x4) at transcode.c:1735 1735 if (ec->replacement_allocated) { (gdb) where #0 rb_econv_close (ec=0x4) at transcode.c:1735 #1 0x00000001dc969c5c in clear_readconv (fptr=0x7dfcda961040) at io.c:5636 #2 clear_codeconv (fptr=0x7dfcda961040) at io.c:5655 #3 io_encoding_set (fptr=fptr@entry=0x7dfcda961040, v1=v1@entry=4, v2=v2@entry=4, opt=opt@entry=4) at io.c:11759 #4 0x00000001dc96c0ec in rb_io_set_encoding (argc=<optimized out>, argc@entry=1, argv=argv@entry=0x7f7ffff88a30, io=138524952599320) at io.c:13519 #5 0x00000001dc97b0d6 in rb_stdio_set_default_encoding () at io.c:13540 #6 0x00000001dca70987 in process_options (argc=0, argc@entry=1, argv=<optimized out>, opt=opt@entry=0x7f7ffff89380) at ruby.c:2561 #7 0x00000001dca71f60 in ruby_process_options (argc=argc@entry=1, argv=argv@entry=0x7f7ffff895a8) at ruby.c:228 #8 0x00000001dc923a61 in ruby_options (argc=argc@entry=1, argv=argv@entry=0x7f7ffff895a8) at eval.c:117 #9 0x00000001dcbded22 in rb_main (argv=0x7f7ffff895a8, argc=1) at ./main.c:43 #10 main (argc=<optimized out>, argv=<optimized out>) at ./main.c:68 ``` ---Files-------------------------------- patch-prism_prism.c (2.7 KB) -- https://bugs.ruby-lang.org/

2 1

[ruby-dev:52150] [Ruby master Bug#21138] The modifier expression with "it" is parsed differently in parse.y and Prism, which is unexpected in both.
by kinoppyd (Yasuhiro Kinoshita) 14 Feb '25

14 Feb '25

Issue #21138 has been reported by kinoppyd (Yasuhiro Kinoshita). ---------------------------------------- Bug #21138: The modifier expression with "it" is parsed differently in parse.y and Prism, which is unexpected in both. https://bugs.ruby-lang.org/issues/21138 * Author: kinoppyd (Yasuhiro Kinoshita) * Status: Open * ruby -v: ruby 3.4.1 (2024-12-25 revision 48d4efcb85) [arm64-darwin21] * Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN, 3.4: UNKNOWN ---------------------------------------- ## Actual ### parse.y In parse.y, the block parameter "it" behaves as if it is not assigned. ```ruby # parse.y loop.rb:1:in 'block in <main>': undefined method '<' for nil (NoMethodError) 1.then { it *= 10 while it < 1000 } ^ from <internal:kernel>:126:in 'Kernel#then' from loop.rb:1:in '<main>' ``` ### prism In Prism, I may not understand what's happening due to my limited knowledge, but it results in a syntax error." ```ruby unexpected integer; expected an expression after the operator unexpected integer, expecting end-of-input > 1 1.then { it *= 10 while it < 1000 } loop.rb:1: syntax errors found (SyntaxError) > 1 | 1.then { it *= 10 while it < 1000 } | ^~ unexpected integer; expected an expression after the operator | ^~ unexpected integer, expecting end-of-input ``` ## Expected 'It' should be reassigned until it is greater than 1000. -- https://bugs.ruby-lang.org/

2 2

[ruby-dev:52146] [Ruby master Bug#21123] instance_exec with curried proc
by taichi730 (Taichi Ishitani) 09 Feb '25

09 Feb '25

Issue #21123 has been reported by taichi730 (Taichi Ishitani). ---------------------------------------- Bug #21123: instance_exec with curried proc https://bugs.ruby-lang.org/issues/21123 * Author: taichi730 (Taichi Ishitani) * Status: Open * ruby -v: 3.4.1 (2024-12-25 revision 48d4efcb85) +PRISM [x86_64-linux] * Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN, 3.4: UNKNOWN ---------------------------------------- I have a question about behavior of `#instance_exec` with a curried proc. When running `#instance_exex` with a curried proc, it appears that the given proc is executed on the context where it is created but not the receiver object. Example code: ```ruby a = Object.new def a.foo(n) p end b = proc { |n| foo(n) }.curry a.instance_exec(0, &b) ``` Execution result: ``` $ ruby test.rb test.rb:6:in 'block in <main>': undefined method 'foo' for main (NoMethodError) b = proc { |n| foo(n) }.curry ^^^ Did you mean? for from test.rb:7:in 'BasicObject#instance_exec' from test.rb:7:in '<main>' ``` Is this intentional behavior? I expect that the given curried proc is also executed on the context of the receiver object like a non-curried proc. -- https://bugs.ruby-lang.org/

3 3

[ruby-dev:52086] [Ruby master Bug#20489] Ractor behavior strange in ruby master
by nekoyama32767 (Jinsong Yu) 10 Jan '25

10 Jan '25

Issue #20489 has been reported by nekoyama32767 (Jinsong Yu). ---------------------------------------- Bug #20489: Ractor behavior strange in ruby master https://bugs.ruby-lang.org/issues/20489 * Author: nekoyama32767 (Jinsong Yu) * Status: Open * ruby -v: ruby 3.4.0dev (2024-05-14T01:58:31Z master 9d01f657b3) [x86_64-linux] * Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN ---------------------------------------- This is a tarai program Run`./ruby tarai_ractor.rb 2 8` is to use 2 thread to run 8 times tarai function total, that means 4 times tarai for each ractor(thread). ``` GC.disable def split_len(len, split) ret = [] mod = len % split head = 0 tail = 0 split.times do |i| if head >= len break end k = 0 if i < mod then k = 1 end tail = tail + (len/split) + k ret.append(head...tail) head = tail end return ret end def ary_split(ary, split) return split_len(ary.length,split) end def item_check(item) if item[0] != nil 1 + item_check(item[0]) + item_check(item[1]) else 1 end end def tarai(x, y, z) = x <= y ? y : tarai(tarai(x-1, y, z), tarai(y-1, z, x), tarai(z-1, x, y)) times = ARGV[0].to_i split = ARGV[1].to_i p split_len(times, split) split_len(times, split).each.map do |sp| Ractor.new (sp) { s = _1 s.each do tarai(13, 7, 0) end } end.each(&:take) ``` The problem is in ruby 3.1.2 and ruby 3.3 `./ruby tarai_ractor.rb 1 1` has simiular execute time with `./ruby tarai_ractor.rb 8 8` because each thread only run 1 time of tarai function, like follow: ruby 3.1.2p20 (2022-04-12 revision 4491bb740a) [x86_64-linux]: ``` time ruby exp_ractor_tarai.rb 1 1 [0...1] <internal:ractor>:267: warning: Ractor is experimental, and the behavior may change in future versions of Ruby! Also there are many implementation issues. real 0m1.442s user 0m1.429s sys 0m0.014s time ruby exp_ractor_tarai.rb 8 8 [0...1, 1...2, 2...3, 3...4, 4...5, 5...6, 6...7, 7...8] <internal:ractor>:267: warning: Ractor is experimental, and the behavior may change in future versions of Ruby! Also there are many implementation issues. real 0m1.857s user 0m13.817s sys 0m0.041s ``` But in ruby master(ruby 3.4.0dev) ruby 3.4.0dev (2024-05-14T01:58:31Z master 9d01f657b3) [x86_64-linux] 1 ractor 1 tarai: ``` time ../ruby exp_ractor_tarai.rb 1 1 `RubyGems' were not loaded. `error_highlight' was not loaded. `did_you_mean' was not loaded. `syntax_suggest' was not loaded. [0...1] exp_ractor_tarai.rb:47: warning: Ractor is experimental, and the behavior may change in future versions of Ruby! Also there are many implementation issues. real 0m1.671s user 0m1.666s sys 0m0.005s ``` 8 ractor 8 tarai: ``` time ../ruby exp_ractor_tarai.rb 8 8 `RubyGems' were not loaded. `error_highlight' was not loaded. `did_you_mean' was not loaded. `syntax_suggest' was not loaded. [0...1, 1...2, 2...3, 3...4, 4...5, 5...6, 6...7, 7...8] exp_ractor_tarai.rb:47: warning: Ractor is experimental, and the behavior may change in future versions of Ruby! Also there are many implementation issues. real 0m18.408s user 1m58.659s sys 0m0.021s ``` And in ruby 3.4.0dev when run `time ../ruby exp_ractor_tarai.rb 16 16` 16 thread should be used in system monitoring while only 8 threads are used. Ruby 3.3 and Ruby 3.1.2 do not have this problem. ---Files-------------------------------- thead16_16.png (168 KB) thread16_8.png (165 KB) -- https://bugs.ruby-lang.org/

5 13

[ruby-dev:52125] [Ruby master Bug#20888] TestGemRemoteFetcherLocalSSLServer tests or so began to fail
by mtasaka (Mamoru TASAKA) 06 Dec '24

06 Dec '24

Issue #20888 has been reported by mtasaka (Mamoru TASAKA). ---------------------------------------- Bug #20888: TestGemRemoteFetcherLocalSSLServer tests or so began to fail https://bugs.ruby-lang.org/issues/20888 * Author: mtasaka (Mamoru TASAKA) * Status: Open * ruby -v: ruby 3.4.0dev (2024-11-12 master a6fdd8883c) +PRISM [x86_64-linux] * Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN ---------------------------------------- With repeatedly trying to rebuild ruby master on Fedora copr build server, recently TestGemRemoteFetcherLocalSSLServer tests or so tests began to fail on x86_64 as: ``` 183) Failure: TestSocket_TCPSocket#test_initialize_with_connection_failure_after_hostname_resolution_failure [/builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:318]: [Errno::ECONNREFUSED] exception expected, not #<Socket::ResolutionError: getaddrinfo(3): Non-recoverable failure in name resolution>. 184) Failure: TestSocket_TCPSocket#test_initialize_resolv_timeout_with_connection_failure [/builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:284]: [Errno::ETIMEDOUT] exception expected, not #<Errno::ECONNREFUSED: Connection refused - connect(2) for "localhost" port 39177>. 185) Failure: TestSocket_TCPSocket#test_initialize_with_hostname_resolution_failure_after_connection_failure [/builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:303]: [Socket::ResolutionError] exception expected, not #<Errno::ECONNREFUSED: Connection refused - connect(2) for "localhost" port 40983>. 186) Error: TestGemRemoteFetcherLocalSSLServer#test_ssl_connection_allow_verify_none: Gem::RemoteFetcher::FetchError: Errno::ECONNREFUSED: Failed to open TCP connection to localhost:34229 (Connection refused - connect(2) for "localhost" port 34229) (https://localhost:34229/yaml) /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/lib/rubygems/remote_fetcher.rb:265:in 'Gem::RemoteFetcher#fetch_path' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/rubygems/test_gem_remote_fetcher_local_ssl_server.rb:103:in 'block in TestGemRemoteFetcherLocalSSLServer#test_ssl_connection_allow_verify_none' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/rubygems/test_gem_remote_fetcher_local_ssl_server.rb:145:in 'TestGemRemoteFetcherLocalSSLServer#with_configured_fetcher' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/rubygems/test_gem_remote_fetcher_local_ssl_server.rb:102:in 'TestGemRemoteFetcherLocalSSLServer#test_ssl_connection_allow_verify_none' 187) Error: TestSocket_TCPSocket#test_initialize_v6_hostname_resolution_failed_and_v4_hostname_resolution_is_success: Socket::ResolutionError: getaddrinfo(3): Non-recoverable failure in name resolution /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:266:in 'TCPSocket#initialize' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:266:in 'IO.new' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:266:in '<main>' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:259:in 'TestSocket_TCPSocket#test_initialize_v6_hostname_resolution_failed_and_v4_hostname_resolution_is_success' 188) Error: TestSocket_TCPSocket#test_initialize_v6_hostname_resolved_earlier_and_v6_server_is_not_listening: Errno::ECONNREFUSED: Connection refused - connect(2) for "localhost" port 47167 /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:222:in 'TCPSocket#initialize' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:222:in 'IO.new' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:222:in '<main>' /builddir/build/BUILD/ruby-3.4.0_20241112.1157gitfee706d9dd-build/ruby-3.4.0-fee706d9dd/test/socket/test_tcp.rb:213:in 'TestSocket_TCPSocket#test_initialize_v6_hostname_resolved_earlier_and_v6_server_is_not_listening' Finished tests in 692.000057s, 46.3801 tests/s, 9452.0484 assertions/s. 32095 tests, 6540818 assertions, 3 failures, 3 errors, 182 skips ``` ( From 1) to 182) are skipped tests ) Looks like * commit:821a5b966fbc2926dc3bf88b6ba09879fa35318e tests are successful * commit:a6fdd8883cab6b31357552db8a64cfb335678172 sees the above error * commit:fee706d9dd52d7f0444357259f33d335f21c980b sees the above error So looks like the culprit is between the above 2 commits. ( I suspect commit:4c270200dbc2a3a4511e8b793a033078afd6fb31 but not sure ) -- https://bugs.ruby-lang.org/

3 6

[ruby-dev:52133] [Ruby master Feature#11177] DATAでEOF文字以降が読めない
by YO4 (Yoshinao Muramatsu) 22 Nov '24

22 Nov '24

Issue #11177 has been updated by YO4 (Yoshinao Muramatsu). rubyの-xオプション付与でソースファイルをバイナリオープンさせることでも```\x1A```以降の読み込みが可能でした。こちらは ruby 3.3またはそれ以前でも可能のようです。 ``` >less -F data.rb #!ruby DATA.pos # posなしだと [] しか得られない p IO.open(DATA.fileno, "rt") {|f| f.readlines } __END__ textpart1 textpart1^Z binpart1 binpart2 >ruby -x data.rb ["textpart1\n", "textpart1\u001A\n", "binpart1\n", "binpart2\n"] ``` 同様の動作を仕込むことで EOF以降も DATAでアクセス可能にできそうですが、DATA.class が Fileでなく IOになっちゃいますね… ---------------------------------------- Feature #11177: DATAでEOF文字以降が読めない https://bugs.ruby-lang.org/issues/11177#change-110732 * Author: mame (Yusuke Endoh) * Status: Open ---------------------------------------- 遠藤です。 Windows で `__END__` 以降に EOF 文字 (`\x1A`) があったとき、それより先が読めないのは仕様でしょうか。 gen.rb: puts "p DATA.read" puts "__END__" puts "foo\x1Abar" 以下のように実行すると再現します。 > ruby gen.rb > t.rb > ruby t.rb "foo" `DATA.binmode.read` などとしてみても同じです。もちろん、Linux では先まで読めます。Windows でも、ソースコードをパイプで流し込んだ場合はなぜか読めます。 > ruby < t.rb "foo\x1Abar\n" さらに、EOF 以降に文字がいっぱいあった場合、EOF 以降の一部の文字が抜け落ちるような挙動になるようです。 gen2.rb: puts "p DATA.read" puts "__END__" puts "foo\x1A" + "X" * 8192 + "Z" > ruby gen2.rb > t.rb > ruby t.rb "fooXXXXXXXXXXXXXXXXXXXXXXXXXXXZ\r\n" バッファリングのバグっぽい挙動ですが、バグでしょうか。 -- Yusuke Endoh <mame(a)ruby-lang.org> -- https://bugs.ruby-lang.org/

1 0

[ruby-dev:52132] [Ruby master Feature#11177] DATAでEOF文字以降が読めない
by YO4 (Yoshinao Muramatsu) 21 Nov '24

21 Nov '24

Issue #11177 has been updated by YO4 (Yoshinao Muramatsu). 本件についてDATA.pos が__END__直後を示さないこと、Cのfseek(0, SEEK_CUR)っぽく DATA.seek(0, IO:SEEK_CUR)を行っても改善しないことから単純なワークアラウンドは存在しないと思っていましたが、 3.4-head で試したところ prismでの仕様の違いか別の理由からかは確認できていませんが DATA.pos が __END__直後を指していたので ``` IO.open(DATA.fileno, "rt") { _1.readlines } # 明示的にuniversal newlineを使用してopen ``` で```\x1A```込みで読み込み可能でした。 ``` >less -F data.rb p IO.open(DATA.fileno, "rt") { _1.readlines } __END__ textpart1 textpart1^Z binpart1 binpart2 >ruby data.rb ["textpart1\n", "textpart1\u001A\n", "binpart1\n", "binpart2\n"] ``` 上記にて、data.rb の改行は CR+LF です。 ---------------------------------------- Feature #11177: DATAでEOF文字以降が読めない https://bugs.ruby-lang.org/issues/11177#change-110720 * Author: mame (Yusuke Endoh) * Status: Open ---------------------------------------- 遠藤です。 Windows で `__END__` 以降に EOF 文字 (`\x1A`) があったとき、それより先が読めないのは仕様でしょうか。 gen.rb: puts "p DATA.read" puts "__END__" puts "foo\x1Abar" 以下のように実行すると再現します。 > ruby gen.rb > t.rb > ruby t.rb "foo" `DATA.binmode.read` などとしてみても同じです。もちろん、Linux では先まで読めます。Windows でも、ソースコードをパイプで流し込んだ場合はなぜか読めます。 > ruby < t.rb "foo\x1Abar\n" さらに、EOF 以降に文字がいっぱいあった場合、EOF 以降の一部の文字が抜け落ちるような挙動になるようです。 gen2.rb: puts "p DATA.read" puts "__END__" puts "foo\x1A" + "X" * 8192 + "Z" > ruby gen2.rb > t.rb > ruby t.rb "fooXXXXXXXXXXXXXXXXXXXXXXXXXXXZ\r\n" バッファリングのバグっぽい挙動ですが、バグでしょうか。 -- Yusuke Endoh <mame(a)ruby-lang.org> -- https://bugs.ruby-lang.org/

1 0