Control: tags 1019652 + patch Control: tags 1019652 + pending Dear maintainer,
I've prepared an NMU for ruby-regexp-parser (versioned as 2.1.1-2.1) and uploaded it to DELAYED/15. Please feel free to tell me if I should cancel it. cu Adrian
diff -Nru ruby-regexp-parser-2.1.1/debian/changelog ruby-regexp-parser-2.1.1/debian/changelog --- ruby-regexp-parser-2.1.1/debian/changelog 2021-12-08 18:30:46.000000000 +0200 +++ ruby-regexp-parser-2.1.1/debian/changelog 2022-11-19 15:54:11.000000000 +0200 @@ -1,3 +1,10 @@ +ruby-regexp-parser (2.1.1-2.1) unstable; urgency=low + + * Non-maintainer upload. + * Add upstream fix for FTBFS with Ruby 3.1. (Closes: #1019652) + + -- Adrian Bunk <[email protected]> Sat, 19 Nov 2022 15:54:11 +0200 + ruby-regexp-parser (2.1.1-2) unstable; urgency=medium * Reupload to unstable diff -Nru ruby-regexp-parser-2.1.1/debian/patches/0001-Fix-build-on-Ruby-3.1.patch ruby-regexp-parser-2.1.1/debian/patches/0001-Fix-build-on-Ruby-3.1.patch --- ruby-regexp-parser-2.1.1/debian/patches/0001-Fix-build-on-Ruby-3.1.patch 1970-01-01 02:00:00.000000000 +0200 +++ ruby-regexp-parser-2.1.1/debian/patches/0001-Fix-build-on-Ruby-3.1.patch 2022-11-19 15:53:15.000000000 +0200 @@ -0,0 +1,226 @@ +From 09580ad00f1922401d4e793dd541bbeaf8d9b058 Mon Sep 17 00:00:00 2001 +From: =?UTF-8?q?Janosch=20Mu=CC=88ller?= <[email protected]> +Date: Sat, 4 Dec 2021 15:34:28 +0100 +Subject: Fix build on Ruby 3.1 ... + +As of Ruby 3.1, meta and control sequences are [pre-processed to hex escapes when used in Regexp literals]( https://github.com/ruby/ruby/commit/11ae581a4a7f5d5f5ec6378872eab8f25381b1b9 ), so they will only reach the scanner and will only be emitted if a String or a Regexp that has been built with the `::new` constructor is scanned. +--- + README.md | 9 +++-- + spec/parser/escapes_spec.rb | 74 +++++++++++++++++++++--------------- + spec/scanner/escapes_spec.rb | 47 ++++++++++++++--------- + 3 files changed, 77 insertions(+), 53 deletions(-) + +diff --git a/README.md b/README.md +index 5410d36..7b40476 100644 +--- a/README.md ++++ b/README.md +@@ -360,9 +360,9 @@ _Note that not all of these are available in all versions of Ruby_ + |   _**Reluctant** (Lazy)_ | `??`, `*?`, `+?`, `{m,M}?` | ✓ | + |   _**Possessive**_ | `?+`, `*+`, `++`, `{m,M}+` | ✓ | + | **String Escapes** | | ⋱ | +-|   _**Control**_ | `\C-C`, `\cD` | ✓ | ++|   _**Control** \[1\]_ | `\C-C`, `\cD` | ✓ | + |   _**Hex**_ | `\x20`, `\x{701230}` | ✓ | +-|   _**Meta**_ | `\M-c`, `\M-\C-C`, `\M-\cC`, `\C-\M-C`, `\c\M-C` | ✓ | ++|   _**Meta** \[1\]_ | `\M-c`, `\M-\C-C`, `\M-\cC`, `\C-\M-C`, `\c\M-C` | ✓ | + |   _**Octal**_ | `\0`, `\01`, `\012` | ✓ | + |   _**Unicode**_ | `\uHHHH`, `\u{H+ H+}` | ✓ | + | **Unicode Properties** | _<sub>([Unicode 11.0.0](http://www.unicode.org/versions/Unicode11.0.0/))</sub>_ | ⋱ | +@@ -374,6 +374,10 @@ _Note that not all of these are available in all versions of Ruby_ + |   _**Scripts**_ | `\p{Arabic}`, `\P{Hiragana}`, `\p{^Greek}` | ✓ | + |   _**Simple**_ | `\p{Dash}`, `\p{Extender}`, `\p{^Hyphen}` | ✓ | + ++**\[1\]**: As of Ruby 3.1, meta and control sequences are [pre-processed to hex escapes when used in Regexp literals]( ++ https://github.com/ruby/ruby/commit/11ae581a4a7f5d5f5ec6378872eab8f25381b1b9 ), so they will only reach the ++scanner and will only be emitted if a String or a Regexp that has been built with the `::new` constructor is scanned. ++ + ##### Inapplicable Features + + Some modifiers, like `o` and `s`, apply to the **Regexp** object itself and do not +@@ -387,7 +391,6 @@ expressions library (Onigmo). They are not supported by the scanner. + - **Quotes**: `\Q...\E` _[[See]](https://github.com/k-takata/Onigmo/blob/7911409/doc/RE#L499)_ + - **Capture History**: `(?@...)`, `(?@<name>...)` _[[See]](https://github.com/k-takata/Onigmo/blob/7911409/doc/RE#L550)_ + +- + See something missing? Please submit an [issue](https://github.com/ammar/regexp_parser/issues) + + _**Note**: Attempting to process expressions with unsupported syntax features can raise an error, +diff --git a/spec/parser/escapes_spec.rb b/spec/parser/escapes_spec.rb +index 25cc5eb..53fe6d9 100644 +--- a/spec/parser/escapes_spec.rb ++++ b/spec/parser/escapes_spec.rb +@@ -56,8 +56,20 @@ RSpec.describe('EscapeSequence parsing') do + expect { root[5].codepoint }.to raise_error(/#codepoints/) + end + ++ # Meta/control espaces ++ # ++ # After the following fix in Ruby 3.1, a Regexp#source containing meta/control ++ # escapes can only be set with the Regexp::new constructor. ++ # In Regexp literals, these escapes are now pre-processed to hex escapes. ++ # ++ # https://github.com/ruby/ruby/commit/11ae581a4a7f5d5f5ec6378872eab8f25381b1b9 ++ def parse_meta_control(regexp_body) ++ regexp = Regexp.new(regexp_body.force_encoding('ascii-8bit'), 'n') ++ RP.parse(regexp) ++ end ++ + specify('parse escape control sequence lower') do +- root = RP.parse(/a\\\c2b/) ++ root = parse_meta_control('a\\\\\c2b') + + expect(root[2]).to be_instance_of(EscapeSequence::Control) + expect(root[2].text).to eq '\\c2' +@@ -66,56 +78,56 @@ RSpec.describe('EscapeSequence parsing') do + end + + specify('parse escape control sequence upper') do +- root = RP.parse(/\d\\\C-C\w/) ++ root = parse_meta_control('\d\C-C\w') + +- expect(root[2]).to be_instance_of(EscapeSequence::Control) +- expect(root[2].text).to eq '\\C-C' +- expect(root[2].char).to eq "\x03" +- expect(root[2].codepoint).to eq 3 ++ expect(root[1]).to be_instance_of(EscapeSequence::Control) ++ expect(root[1].text).to eq '\\C-C' ++ expect(root[1].char).to eq "\x03" ++ expect(root[1].codepoint).to eq 3 + end + + specify('parse escape meta sequence') do +- root = RP.parse(/\Z\\\M-Z/n) ++ root = parse_meta_control('\Z\M-Z') + +- expect(root[2]).to be_instance_of(EscapeSequence::Meta) +- expect(root[2].text).to eq '\\M-Z' +- expect(root[2].char).to eq "\u00DA" +- expect(root[2].codepoint).to eq 218 ++ expect(root[1]).to be_instance_of(EscapeSequence::Meta) ++ expect(root[1].text).to eq '\\M-Z' ++ expect(root[1].char).to eq "\u00DA" ++ expect(root[1].codepoint).to eq 218 + end + + specify('parse escape meta control sequence') do +- root = RP.parse(/\A\\\M-\C-X/n) ++ root = parse_meta_control('\A\M-\C-X') + +- expect(root[2]).to be_instance_of(EscapeSequence::MetaControl) +- expect(root[2].text).to eq '\\M-\\C-X' +- expect(root[2].char).to eq "\u0098" +- expect(root[2].codepoint).to eq 152 ++ expect(root[1]).to be_instance_of(EscapeSequence::MetaControl) ++ expect(root[1].text).to eq '\\M-\\C-X' ++ expect(root[1].char).to eq "\u0098" ++ expect(root[1].codepoint).to eq 152 + end + + specify('parse lower c meta control sequence') do +- root = RP.parse(/\A\\\M-\cX/n) ++ root = parse_meta_control('\A\M-\cX') + +- expect(root[2]).to be_instance_of(EscapeSequence::MetaControl) +- expect(root[2].text).to eq '\\M-\\cX' +- expect(root[2].char).to eq "\u0098" +- expect(root[2].codepoint).to eq 152 ++ expect(root[1]).to be_instance_of(EscapeSequence::MetaControl) ++ expect(root[1].text).to eq '\\M-\\cX' ++ expect(root[1].char).to eq "\u0098" ++ expect(root[1].codepoint).to eq 152 + end + + specify('parse escape reverse meta control sequence') do +- root = RP.parse(/\A\\\C-\M-X/n) ++ root = parse_meta_control('\A\C-\M-X') + +- expect(root[2]).to be_instance_of(EscapeSequence::MetaControl) +- expect(root[2].text).to eq '\\C-\\M-X' +- expect(root[2].char).to eq "\u0098" +- expect(root[2].codepoint).to eq 152 ++ expect(root[1]).to be_instance_of(EscapeSequence::MetaControl) ++ expect(root[1].text).to eq '\\C-\\M-X' ++ expect(root[1].char).to eq "\u0098" ++ expect(root[1].codepoint).to eq 152 + end + + specify('parse escape reverse lower c meta control sequence') do +- root = RP.parse(/\A\\\c\M-X/n) ++ root = parse_meta_control('\A\c\M-X') + +- expect(root[2]).to be_instance_of(EscapeSequence::MetaControl) +- expect(root[2].text).to eq '\\c\\M-X' +- expect(root[2].char).to eq "\u0098" +- expect(root[2].codepoint).to eq 152 ++ expect(root[1]).to be_instance_of(EscapeSequence::MetaControl) ++ expect(root[1].text).to eq '\\c\\M-X' ++ expect(root[1].char).to eq "\u0098" ++ expect(root[1].codepoint).to eq 152 + end + end +diff --git a/spec/scanner/escapes_spec.rb b/spec/scanner/escapes_spec.rb +index 411d050..c3a9117 100644 +--- a/spec/scanner/escapes_spec.rb ++++ b/spec/scanner/escapes_spec.rb +@@ -35,25 +35,6 @@ RSpec.describe('Escape scanning') do + include_examples 'scan', 'a\u{640 0641}c', 1 => [:escape, :codepoint_list, '\u{640 0641}', 1, 13] + include_examples 'scan', 'a\u{10FFFF}c', 1 => [:escape, :codepoint_list, '\u{10FFFF}', 1, 11] + +- include_examples 'scan', /a\cBc/, 1 => [:escape, :control, '\cB', 1, 4] +- include_examples 'scan', /a\c^c/, 1 => [:escape, :control, '\c^', 1, 4] +- include_examples 'scan', /a\c\n/, 1 => [:escape, :control, '\c\n', 1, 5] +- include_examples 'scan', /a\c\\b/, 1 => [:escape, :control, '\c\\\\', 1, 5] +- include_examples 'scan', /a\C-bc/, 1 => [:escape, :control, '\C-b', 1, 5] +- include_examples 'scan', /a\C-^b/, 1 => [:escape, :control, '\C-^', 1, 5] +- include_examples 'scan', /a\C-\nb/, 1 => [:escape, :control, '\C-\n', 1, 6] +- include_examples 'scan', /a\C-\\b/, 1 => [:escape, :control, '\C-\\\\', 1, 6] +- include_examples 'scan', /a\c\M-Bc/n, 1 => [:escape, :control, '\c\M-B', 1, 7] +- include_examples 'scan', /a\C-\M-Bc/n, 1 => [:escape, :control, '\C-\M-B', 1, 8] +- +- include_examples 'scan', /a\M-Bc/n, 1 => [:escape, :meta_sequence, '\M-B', 1, 5] +- include_examples 'scan', /a\M-\cBc/n, 1 => [:escape, :meta_sequence, '\M-\cB', 1, 7] +- include_examples 'scan', /a\M-\c^/n, 1 => [:escape, :meta_sequence, '\M-\c^', 1, 7] +- include_examples 'scan', /a\M-\c\n/n, 1 => [:escape, :meta_sequence, '\M-\c\n', 1, 8] +- include_examples 'scan', /a\M-\c\\/n, 1 => [:escape, :meta_sequence, '\M-\c\\\\', 1, 8] +- include_examples 'scan', /a\M-\C-Bc/n, 1 => [:escape, :meta_sequence, '\M-\C-B', 1, 8] +- include_examples 'scan', /a\M-\C-\\/n, 1 => [:escape, :meta_sequence, '\M-\C-\\\\', 1, 9] +- + include_examples 'scan', 'ab\\\xcd', 1 => [:escape, :backslash, '\\\\', 2, 4] + include_examples 'scan', 'ab\\\0cd', 1 => [:escape, :backslash, '\\\\', 2, 4] + include_examples 'scan', 'ab\\\Kcd', 1 => [:escape, :backslash, '\\\\', 2, 4] +@@ -61,4 +42,32 @@ RSpec.describe('Escape scanning') do + include_examples 'scan', 'ab\^cd', 1 => [:escape, :bol, '\^', 2, 4] + include_examples 'scan', 'ab\$cd', 1 => [:escape, :eol, '\$', 2, 4] + include_examples 'scan', 'ab\[cd', 1 => [:escape, :set_open, '\[', 2, 4] ++ ++ # Meta/control espaces ++ # ++ # After the following fix in Ruby 3.1, a Regexp#source containing meta/control ++ # escapes can only be set with the Regexp::new constructor. ++ # In Regexp literals, these escapes are now pre-processed to hex escapes. ++ # ++ # https://github.com/ruby/ruby/commit/11ae581a4a7f5d5f5ec6378872eab8f25381b1b9 ++ n = ->(regexp_body){ Regexp.new(regexp_body.force_encoding('ascii-8bit'), 'n') } ++ ++ include_examples 'scan', 'a\cBc', 1 => [:escape, :control, '\cB', 1, 4] ++ include_examples 'scan', 'a\c^c', 1 => [:escape, :control, '\c^', 1, 4] ++ include_examples 'scan', 'a\c\n', 1 => [:escape, :control, '\c\n', 1, 5] ++ include_examples 'scan', 'a\c\\\\b', 1 => [:escape, :control, '\c\\\\', 1, 5] ++ include_examples 'scan', 'a\C-bc', 1 => [:escape, :control, '\C-b', 1, 5] ++ include_examples 'scan', 'a\C-^b', 1 => [:escape, :control, '\C-^', 1, 5] ++ include_examples 'scan', 'a\C-\nb', 1 => [:escape, :control, '\C-\n', 1, 6] ++ include_examples 'scan', 'a\C-\\\\b', 1 => [:escape, :control, '\C-\\\\', 1, 6] ++ include_examples 'scan', n.('a\c\M-Bc'), 1 => [:escape, :control, '\c\M-B', 1, 7] ++ include_examples 'scan', n.('a\C-\M-Bc'), 1 => [:escape, :control, '\C-\M-B', 1, 8] ++ ++ include_examples 'scan', n.('a\M-Bc'), 1 => [:escape, :meta_sequence, '\M-B', 1, 5] ++ include_examples 'scan', n.('a\M-\cBc'), 1 => [:escape, :meta_sequence, '\M-\cB', 1, 7] ++ include_examples 'scan', n.('a\M-\c^'), 1 => [:escape, :meta_sequence, '\M-\c^', 1, 7] ++ include_examples 'scan', n.('a\M-\c\n'), 1 => [:escape, :meta_sequence, '\M-\c\n', 1, 8] ++ include_examples 'scan', n.('a\M-\c\\\\'), 1 => [:escape, :meta_sequence, '\M-\c\\\\', 1, 8] ++ include_examples 'scan', n.('a\M-\C-Bc'), 1 => [:escape, :meta_sequence, '\M-\C-B', 1, 8] ++ include_examples 'scan', n.('a\M-\C-\\\\'), 1 => [:escape, :meta_sequence, '\M-\C-\\\\', 1, 9] + end +-- +2.30.2 + diff -Nru ruby-regexp-parser-2.1.1/debian/patches/series ruby-regexp-parser-2.1.1/debian/patches/series --- ruby-regexp-parser-2.1.1/debian/patches/series 1970-01-01 02:00:00.000000000 +0200 +++ ruby-regexp-parser-2.1.1/debian/patches/series 2022-11-19 15:54:10.000000000 +0200 @@ -0,0 +1 @@ +0001-Fix-build-on-Ruby-3.1.patch
