GitHub - ruby/lrama: Pure Ruby LALR parser generator (original) (raw)
Lrama
Lrama is LALR (1) parser generator written by Ruby. The first goal of this project is providing error tolerant parser for CRuby with minimal changes on CRuby parse.y file.
- Features
- Installation
- Usage
- Versions and Branches
- Supported Ruby version
- Development
- Release flow
- License
Features
- Bison style grammar file is supported with some assumptions
- b4_locations_if is always true
- b4_pure_if is always true
- b4_pull_if is always false
- b4_lac_if is always false
- Error Tolerance parser
- Subset of Repairing Syntax Errors in LR Parsers (Corchuelo et al.) algorithm is supported
- Parameterizing rules
- The definition of a non-terminal symbol can be parameterized with other (terminal or non-terminal) symbols.
- Providing a generic definition of parameterizing rules as a standard library.
- Inlining
- The %inline directive causes all references to symbols to be replaced with its definition.
- Resolve shift/reduce conflicts without artificially altering the grammar file.
Installation
From source codes,
$ cd "$(lrama root)" $ bundle install $ bundle exec rake install $ bundle exec lrama --version 0.5.0
Usage
"y.tab.c" and "y.tab.h" are generated
$ lrama -d sample/parse.y
"calc", "calc.c", and "calc.h" are generated
$ lrama -d sample/calc.y -o calc.c && gcc -Wall calc.c -o calc && ./calc Enter the formula: 1 => 1 1+2*3 => 7 (1+2)*3 => 9
Versions and Branches
v0_6 (master
branch)
This branch is for Ruby 3.4. lrama_0_6
branch is created from this branch, once Ruby 3.4 is released.
v0_5 (lrama_0_5
branch)
This branch is for Ruby 3.3.
v0_4 (lrama_0_4
branch)
This branch generates "parse.c" compatible with Bison 3.8.2 for ruby 3.0, 3.1, 3.2. The first version migrated to ruby is "0.4.0" therefore keep this branch for Bison compatible branch.
Supported Ruby version
Lrama is executed with BASERUBY when building ruby from source code. Therefore Lrama needs to support BASERUBY, currently 2.5, or later version.
This also requires Lrama to be able to run with only default gems because BASERUBY runs with --disable=gems
option.
Development
How to generate parser.rb
$ bundle exec rake build:parser
parser.rb
is generated from parser.y
by Racc. Run the rake command when you update parser.y
then commit changes of both files.
Test
Running tests:
$ bundle install $ bundle exec rspec
or
$ bundle exec rake spec
Running type check:
$ bundle install $ bundle exec rbs collection install $ bundle exec steep check
or
$ bundle exec rake steep
Running both of them:
$ bundle install $ bundle exec rake
Call-stack Profiling Lrama
1. Create parse.tmp.y in ruby/ruby
$ ruby tool/id2token.rb parse.y > parse.tmp.y $ cp parse.tmp.y dir/lrama/tmp
2. Enable Profiler
diff --git a/exe/lrama b/exe/lrama index ba5fb06..2497178 100755 --- a/exe/lrama +++ b/exe/lrama @@ -3,4 +3,6 @@ $LOAD_PATH << File.join(dir, "../lib") require "lrama"
-Lrama::Command.new.run(ARGV.dup) +Lrama::Report::Profile.report_profile do
- Lrama::Command.new.run(ARGV.dup) +end
3. Run Lrama
$ exe/lrama -o parse.tmp.c --header=parse.tmp.h tmp/parse.tmp.y
4. Generate Flamegraph
$ stackprof --d3-flamegraph tmp/stackprof-cpu-myapp.dump > tmp/flamegraph.html
Memory Profiling Lrama
1. Create parse.tmp.y in ruby/ruby
$ ruby tool/id2token.rb parse.y > parse.tmp.y $ cp parse.tmp.y dir/lrama/tmp
2. Enable Profiler
diff --git a/exe/lrama b/exe/lrama index 1aece5d141..f5f94cf7fa 100755 --- a/exe/lrama +++ b/exe/lrama @@ -3,5 +3,9 @@
$LOAD_PATH << File.join(dir, "../lib") require "lrama" +require 'memory_profiler'
-Lrama::Command.new.run(ARGV.dup) +report = MemoryProfiler.report do
- Lrama::Command.new.run(ARGV.dup) +end +report.pretty_print
3. Run Lrama
$ exe/lrama -o parse.tmp.c --header=parse.tmp.h tmp/parse.tmp.y > report.txt
Build Ruby
- Install Lrama
- Run
make main
Release flow
- Update
Lrama::VERSION
and NEWS.md - Release as a gem by
rake release
- Update Lrama in ruby/ruby by
cp -r LEGAL.md NEWS.md MIT exe lib template ruby/tool/lrama
- Create new release on GitHub
License
See LEGAL.md file.