parsanol 1.3.54 → 1.3.57

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ea29107bfe3e1b3ce6b66701239beffe98c16a2cb7207da64e81b43ea179544d
-  data.tar.gz: 6c31ecaaa82edb959a17b8a43ce5ac7d3fa3819c175e28637efac3a35e215760
+  metadata.gz: dac1ca6df26172acdc4f7a676c018d1a9cf6107f5a2b6643b1274cbd56ba3b8f
+  data.tar.gz: bbeebb38922a236a170578538858e2f795e9da94679dd6c7a0cd81576d4542a7
 SHA512:
-  metadata.gz: f4e3f11ff5bf640b96c2ed79bee2db2d2ef3d68efc409a8da5acb684c9b366d4af902ededc6f6e115163a491c27297027c8d5c9fd10975d43b458ccea2b3548a
-  data.tar.gz: f5ae7316b64a7979558eed0d592e13e3cb5edb79644d1d90c9f997bb34689e71075022f4e8b89dc0c86d5c8e678997eae532e086f36e1bd20c5e70f3408fb9aa
+  metadata.gz: 9170d4b42f82853d41741475abb92fe26f126fde281b4c1b2f31b6b6004dab21fa7a9a9a23b580b83c86f89978dd021767d5464c79080ef19d9ba68771cd7d23
+  data.tar.gz: 633a3b087e6b6d6656e8ddd0db3c0548f6985ea72e43da51537a95d07c480957397b5fe6eda56b8651a2d302cb6b886b5661fc2446dfc4aceb2943a9c31ba2a5

data/Cargo.lock

@@ -54,6 +54,15 @@ version = "2.13.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3ded4057c258ba199e2d26386d3af3780957ecaee6c4ef4041c6b4b8b97c0b06"
 
+[[package]]
+name = "block-buffer"
+version = "0.12.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d2f6c7dbe95a6ed67ad9f18e57daf93a2f034c524b99fd2b76d18fdfeb6660aa"
+dependencies = [
+ "hybrid-array",
+]
+
 [[package]]
 name = "bumpalo"
 version = "3.20.3"
@@ -86,6 +95,41 @@ dependencies = [
  "libloading",
 ]
 
+[[package]]
+name = "const-oid"
+version = "0.10.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a6ef517f0926dd24a1582492c791b6a4818a4d94e789a334894aa15b0d12f55c"
+
+[[package]]
+name = "cpufeatures"
+version = "0.3.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5ca28b0ae3115b884660db4118d803791fd6756b6e88f39c0f3f7859060d7566"
+dependencies = [
+ "libc",
+]
+
+[[package]]
+name = "crypto-common"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ce6e4c961d6cd6c9a86db418387425e8bdeaf05b3c8bc1411e6dca4c252f1453"
+dependencies = [
+ "hybrid-array",
+]
+
+[[package]]
+name = "digest"
+version = "0.11.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f1dd6dbb5841937940781866fa1281a1ff7bd3bf827091440879f9994983d5c2"
+dependencies = [
+ "block-buffer",
+ "const-oid",
+ "crypto-common",
+]
+
 [[package]]
 name = "either"
 version = "1.18.0"
@@ -148,6 +192,15 @@ dependencies = [
  "foldhash",
 ]
 
+[[package]]
+name = "hybrid-array"
+version = "0.4.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "27f864f10dfb56725ce5ce5472bc52252c8f93a4ab86327122cebf62c5f59a17"
+dependencies = [
+ "typenum",
+]
+
 [[package]]
 name = "itertools"
 version = "0.13.0"
@@ -165,9 +218,9 @@ checksum = "8f42a60cbdf9a97f5d2305f08a87dc4e09308d1276d28c869c684d7777685682"
 
 [[package]]
 name = "js-sys"
-version = "0.3.105"
+version = "0.3.106"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ce57d20d1ea864ce2ac172ab472d409214f4fd359f0b2a2775abdf522e2af99e"
+checksum = "7883d941dae510fb2d978fc3fe018c71c9e2892fd38854de3e8b92c2e5ad9cc5"
 dependencies = [
  "cfg-if",
  "wasm-bindgen",
@@ -203,8 +256,9 @@ checksum = "f9f8bd3e56ce4dfc153cf470fffbfa98c7620958b312ca5c3a4b8d5181fd13c6"
 
 [[package]]
 name = "magnus"
-version = "0.9.0"
-source = "git+https://github.com/matsadler/magnus?rev=9407bd9#9407bd98afef6692edca706426c15ea6bef60d0e"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4c01bef153b8d0f086991f0e166474a1d57b738221f9a9c11405d498ac95a44d"
 dependencies = [
  "magnus-macros",
  "rb-sys",
@@ -214,8 +268,9 @@ dependencies = [
 
 [[package]]
 name = "magnus-macros"
-version = "0.9.0"
-source = "git+https://github.com/matsadler/magnus?rev=9407bd9#9407bd98afef6692edca706426c15ea6bef60d0e"
+version = "0.9.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9dafde2eeb66f090e0f326112323eb8324b12cb32b5d1210e541d3df6b0ce77e"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -252,9 +307,9 @@ checksum = "9f7c3e4beb33f85d45ae3e3a1792185706c8e16d043238c593331cc7cd313b50"
 
 [[package]]
 name = "parsanol"
-version = "0.8.8"
+version = "0.10.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5709d5bc971fa4e28fccf401709ab1d275d180e447e3ab82fde7ebef2ecc787b"
+checksum = "3201f3b6340113ce57871a1405679ad7520a420cb8958777983371a2d4e6bf55"
 dependencies = [
  "ahash",
  "getrandom 0.3.4",
@@ -263,15 +318,17 @@ dependencies = [
  "magnus",
  "parsanol-derive",
  "regex",
+ "regex-syntax",
  "serde",
  "serde_json",
+ "sha2",
 ]
 
 [[package]]
 name = "parsanol-derive"
-version = "0.8.8"
+version = "0.10.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "93258404b7032b85dffd9a60e3fb1ed003a268282b4f3385f1ab1aba98db34e5"
+checksum = "07bea1e5fb6a78314596116ac62e08ab5c73c201c62cd83715e75309c44e0977"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -280,7 +337,7 @@ dependencies = [
 
 [[package]]
 name = "parsanol_cdylib"
-version = "1.3.50"
+version = "1.3.52"
 dependencies = [
  "parsanol",
 ]
@@ -443,6 +500,17 @@ dependencies = [
  "zmij",
 ]
 
+[[package]]
+name = "sha2"
+version = "0.11.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "446ba717509524cb3f22f17ecc096f10f4822d76ab5c0b9822c5f9c284e825f4"
+dependencies = [
+ "cfg-if",
+ "cpufeatures",
+ "digest",
+]
+
 [[package]]
 name = "shell-words"
 version = "1.1.1"
@@ -477,6 +545,12 @@ dependencies = [
  "unicode-ident",
 ]
 
+[[package]]
+name = "typenum"
+version = "1.20.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b6f5e870be6c3b371b77fe0ee0bafb859fa4964b4404c27de1d380043c4dda20"
+
 [[package]]
 name = "unicode-ident"
 version = "1.0.26"
@@ -500,9 +574,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen"
-version = "0.2.128"
+version = "0.2.129"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "aecb87a33d3b0c5e3b7aa46336eaf486cffafbd281b195e4c8b80d50df2351bf"
+checksum = "9bb54f33acc68fd454578d9820b0bde1a1a3d17aa17bb7b6595806d02886d409"
 dependencies = [
  "cfg-if",
  "once_cell",
@@ -513,9 +587,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-macro"
-version = "0.2.128"
+version = "0.2.129"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a690d511e3c1a8b3a55e33511e3c2c00c78415cd23650f32b808627f5696b9ed"
+checksum = "2e29d0c35b16e224a7eeb5cd2d25e3e1968fbd65604117b44d3b789d00ee8535"
 dependencies = [
  "quote",
  "wasm-bindgen-macro-support",
@@ -523,9 +597,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-macro-support"
-version = "0.2.128"
+version = "0.2.129"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "411e4887f0071ef2d2164a9d5fdf2d20efbef78fccd3a78b0c10a1dc5295e48a"
+checksum = "6f501a8bc3719dba86ef8ae4728879c08001bea749eb1333ac5b91e040e2a6b7"
 dependencies = [
  "bumpalo",
  "proc-macro2",
@@ -536,9 +610,9 @@ dependencies = [
 
 [[package]]
 name = "wasm-bindgen-shared"
-version = "0.2.128"
+version = "0.2.129"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "81941cd78d0c92026c33e5e01312845a4cb1e9af3407f9134b100dd03144103e"
+checksum = "23f0c9c52aa7cd7d77769a4cfe2a9adb1b331f489a41d912ce14513d5ab995c6"
 dependencies = [
  "unicode-ident",
 ]
@@ -557,18 +631,18 @@ checksum = "1ebf944e87a7c253233ad6766e082e3cd714b5d03812acc24c318f549614536e"
 
 [[package]]
 name = "zerocopy"
-version = "0.8.57"
+version = "0.8.59"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d35102a9f36d089ccae9e4c6802bc118be4487b80aaffc0ab4e0cf5ce92d2873"
+checksum = "6df92bf3d9227be3d53173901ddbffac2babc27ae50f397776ffd6dc33f800cb"
 dependencies = [
  "zerocopy-derive",
 ]
 
 [[package]]
 name = "zerocopy-derive"
-version = "0.8.57"
+version = "0.8.59"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "146c01f5ab44258da43cf276c74a2763db2ff3969c9c652c3f2de07041d0b2bc"
+checksum = "ac4f328cf2f05d084e496c3e9c3f33ed0a183656a16e1fcec4d464d8373aec82"
 dependencies = [
  "proc-macro2",
  "quote",

data/exe/parsanol

@@ -0,0 +1,7 @@
+#!/usr/bin/env ruby
+# frozen_string_literal: true
+
+require "parsanol"
+require "parsanol/parg/cli"
+argv = ARGV[0] == "parg" ? ARGV[1..] : ARGV
+exit Parsanol::PARG::CLI.run(argv)

data/ext/parsanol_cdylib/Cargo.toml

@@ -12,4 +12,4 @@ crate-type = ["cdylib"]
 doc = false
 
 [dependencies]
-parsanol = { version = "0.8.8" }
+parsanol = { version = "0.10" }

data/ext/parsanol_native/Cargo.toml

@@ -27,8 +27,8 @@ rb-sys = { version = "0.9.124", features = ["global-allocator"] }
 # magnus for Ruby bindings (from crates.io, patched locally for Ruby 4.0 support)
 magnus = { version = "0.9" }
 
-# parsanol parser library (from git for latest features)
-parsanol = { version = "0.8.9", features = ["ruby"] }
+# parsanol parser library (crates.io; 0.10 carries the class fast-path fix)
+parsanol = { version = "0.10", features = ["ruby"] }
 
 # Logging
 log = "0.4"

data/ext/parsanol_native/build.rs

@@ -0,0 +1,9 @@
+fn main() {
+    // MRI loadables on macOS resolve ruby symbols from the host process
+    // at load time; cargo's cdylib link defaults to failing on undefined
+    // symbols, so the flag must come from the top-level package.
+    if cfg!(target_os = "macos") {
+        println!("cargo:rustc-link-arg=-Wl,-undefined,dynamic_lookup");
+    }
+    println!("cargo:rerun-if-changed=build.rs");
+}

data/ext/parsanol_native/extconf.rb

@@ -30,6 +30,16 @@ create_rust_makefile("parsanol/parsanol_native") do |r|
   # Create debug builds in dev, release in production
   r.profile = ENV.fetch("RB_SYS_CARGO_PROFILE", :dev).to_sym
 
+  # cargo rustc builds dependency cdylibs verbatim, and parsanol ships
+  # a cdylib crate-type (the ffi tier). On macOS every dylib in the
+  # build resolves ruby symbols from the host process at load time, so
+  # every link must tolerate undefined symbols. (extconf runs as a
+  # subprocess, so this must be baked into the Makefile — env changes
+  # here would not survive it.)
+  if RbConfig::CONFIG["host_os"].include?("darwin")
+    r.extra_rustflags = %w[-C link-arg=-Wl,-undefined,dynamic_lookup]
+  end
+
   # Enable stable API compiled fallback for ruby-head and older Ruby versions
   r.use_stable_api_compiled_fallback = true
 

data/lib/parsanol/atoms/base.rb

@@ -208,7 +208,7 @@ module Parsanol
 
       # Converts raw input to Source if needed.
       def normalize_input(source)
-        source.respond_to?(:line_and_column) ? source : Parsanol::Source.new(source)
+        source.is_a?(Parsanol::Source) ? source : Parsanol::Source.new(source)
       end
 
       # Detects if we're in a Parser context.

data/lib/parsanol/atoms/cut.rb

@@ -42,7 +42,7 @@ module Parsanol
         # 1. Mark the current position as a cut point
         # 2. Empty the backtrack stack (we won't backtrack past here)
         # 3. Aggressively evict cache entries before this position
-        context.cut!(source.bytepos) if context.respond_to?(:cut!)
+        context.cut!(source.bytepos)
 
         [success, value]
       end

data/lib/parsanol/atoms/repetition.rb

@@ -66,7 +66,7 @@ module Parsanol
       # @return [Array(Boolean, Object)] result
       def try(source, context, consume_all)
         # Check for tree memoization support
-        if context.respond_to?(:use_tree_memoization?) && context.use_tree_memoization?
+        if context.use_tree_memoization?
           return with_tree_cache(source, context, consume_all)
         end
 
@@ -183,6 +183,7 @@ module Parsanol
         last_failure = nil
 
         loop do
+          iteration_start = source.bytepos
           success, value = @parslet.apply(source, context, false)
 
           unless success
@@ -193,6 +194,11 @@ module Parsanol
           occurrence += 1
           result[occurrence] = value
 
+          # A body match that consumes nothing cannot make progress;
+          # looping again would re-match empty forever. Count the empty
+          # match, then stop — the min check below applies as usual.
+          break if source.bytepos == iteration_start
+
           break if @max && occurrence >= @max
         end
 
@@ -254,6 +260,7 @@ module Parsanol
         last_failure = nil
 
         loop do
+          iteration_start = source.bytepos
           success, value = @parslet.apply(source, context, false)
 
           unless success
@@ -265,6 +272,10 @@ module Parsanol
           result[occurrence] = value
           positions << source.bytepos
 
+          # Zero-width body match: count it, stop iterating (the min
+          # check below applies as usual) — mirrors try_general.
+          break if source.bytepos == iteration_start
+
           break if @max && occurrence >= @max
         end
 
@@ -287,7 +298,7 @@ module Parsanol
           context.store_tree_memo(cache_key, start_pos,
                                   result[1, occurrence], end_pos)
         end
-        context.release_array(positions) if context.respond_to?(:release_array)
+        context.release_array(positions)
 
         # Check minimum
         if occurrence < @min

data/lib/parsanol/cause.rb

@@ -74,7 +74,7 @@ module Parsanol
       line_num, col_num = @source.line_and_column(@position)
 
       formatted_msg = @message.map do |msg|
-        msg.respond_to?(:to_slice) ? msg.content.inspect : msg.to_s
+        msg.is_a?(Parsanol::Slice) ? msg.content.inspect : msg.to_s
       end.join
 
       "#{formatted_msg} at line #{line_num} char #{col_num}#{@parsing_label}."

data/lib/parsanol/error_reporter/contextual.rb

@@ -72,7 +72,7 @@ module Parsanol
         cause = super
 
         # Apply label if the atom has one
-        if atom.respond_to?(:label) && (lbl = atom.label)
+        if atom.is_a?(Parsanol::Atoms::Base) && (lbl = atom.label)
           maybe_update_label(lbl, src.pos.bytepos)
           cause.set_label(@active_label_text)
         end

data/lib/parsanol/lazy_result.rb

@@ -103,6 +103,15 @@ module Parsanol
 
     alias kind_of? is_a?
 
+    # The delegated interface: Array's full class contract (including
+    # Enumerable), computed once against the type rather than probed on
+    # instances.
+    ARRAY_METHODS = (
+      Array.instance_methods(true) + Enumerable.instance_methods + [:to_ary]
+    ).freeze
+
+    private_constant :ARRAY_METHODS
+
     # Respond to array methods.
     #
     # @param method [Symbol] Method name
@@ -110,7 +119,7 @@ module Parsanol
     # @return [Boolean] true if responds
     #
     def respond_to?(method, include_private = false)
-      super || to_a.respond_to?(method, include_private)
+      super || ARRAY_METHODS.include?(method)
     end
 
     # Delegate unknown methods to materialized array.
@@ -121,7 +130,7 @@ module Parsanol
     # @return [Object] Result of method call
     #
     def method_missing(method, ...)
-      if to_a.respond_to?(method)
+      if ARRAY_METHODS.include?(method)
         to_a.public_send(method, ...)
       else
         super
@@ -135,7 +144,7 @@ module Parsanol
     # @return [Boolean] true if method is supported
     #
     def respond_to_missing?(method, include_private = false)
-      to_a.respond_to?(method, include_private) || super
+      ARRAY_METHODS.include?(method) || super
     end
 
     # Compare with another object.

data/lib/parsanol/native/serializer.rb

@@ -95,9 +95,12 @@ module Parsanol
 
     def serialize_re(atom)
       # Ruby's Regexp#to_s produces "(?-mix:pattern)" format
-      # We need to extract just the pattern for the Rust parser
+      # The pattern must reach the Rust regex engine verbatim, including
+      # inline flag groups: stripping (?i:...) silently turned every
+      # case-insensitive regex into a case-sensitive one on the native
+      # path (the Ruby atom still compiled it with the flag, so the two
+      # engines disagreed). Rust regex understands the same inline groups.
       pattern = atom.match
-      pattern = ::Regexp.last_match(1) if pattern =~ /^\(\?[-mix]*:(.+)\)$/
       {
         "Re" => {
           "pattern" => pattern,

data/lib/parsanol/native/transformer.rb

@@ -391,15 +391,19 @@ module Parsanol
           return left.is_a?(Hash) ? left.merge(right) : left + right
         end
 
-        if left.respond_to?(:to_str) && right.respond_to?(:to_str)
-          return right if right.respond_to?(:to_slice)
-          return left if left.respond_to?(:to_slice)
-
-          return left.to_str + right.to_str
+        # String-like = String or Slice (both implement to_str). When
+        # only ONE side is string-like, the OTHER side wins — the
+        # original duck-typed priority, preserved exactly.
+        left_stringlike = left.is_a?(String) || left.is_a?(Parsanol::Slice)
+        right_stringlike = right.is_a?(String) || right.is_a?(Parsanol::Slice)
+        if left_stringlike && right_stringlike
+          return right if right.is_a?(Parsanol::Slice)
+          return left if left.is_a?(Parsanol::Slice)
+
+          return left + right
         end
-
-        return left if right.respond_to?(:to_str)
-        return right if left.respond_to?(:to_str)
+        return left if right_stringlike
+        return right if left_stringlike
 
         return left + [right] if right.is_a?(Hash)
         return [left] + right if left.is_a?(Hash)

data/lib/parsanol/native.rb

@@ -216,7 +216,9 @@ module Parsanol
         # cause from that directly — the failure path never needs a
         # reporter reparse. The single interpreter pass below stays only
         # to recover inputs from grammars native cannot express.
-        if grammar.respond_to?(:parse) && (m = error.message.match(NATIVE_POS_MARKER))
+        parsanol_grammar =
+          grammar.is_a?(Parsanol::Atoms::Base) || grammar.is_a?(Parsanol::Parser)
+        if parsanol_grammar && (m = error.message.match(NATIVE_POS_MARKER))
           source = Parsanol::Source.new(input)
           success, value = grammar.run_with_context(source, nil, true)
           return grammar.finalize_result(value) if success
@@ -227,7 +229,7 @@ module Parsanol
           raise Parsanol::ParseFailed.new(cause.to_s, cause)
         end
 
-        if grammar.respond_to?(:parse)
+        if parsanol_grammar
           # No native diagnostics (e.g. incomplete-input errors): one
           # interpreter pass with the reporter attached — a success
           # recovers the tree, a failure raises the cause tree.