feat: improve runfiles handling

.github/workflows/ci.yml

@@ -83,9 +83,10 @@ jobs:
         uses: ruby/setup-ruby@v1
         with:
           ruby-version: '3.0.6'
-      - run: bazel test ...
+      - run: bazel build ...
       - run: bazel run lib/gem:add-numbers 2
       - run: bazel run lib/gem:print-version
+      - run: bazel test ...
       - if: failure() && runner.debug == '1'
         uses: mxschmitt/action-tmate@v3
 

examples/gem/lib/gem/BUILD

@@ -1,3 +1,4 @@
+load("@bazel_skylib//rules:run_binary.bzl", "run_binary")
 load("@rules_ruby//ruby:defs.bzl", "rb_binary", "rb_library")
 
 package(default_visibility = ["//:__subpackages__"])
@@ -32,3 +33,15 @@ rb_binary(
     args = ["lib/gem/version.rb"],
     deps = [":version"],
 )
+
+run_binary(
+    name = "perform-addition",
+    srcs = [":add"],
+    outs = ["addition-result"],
+    args = [
+        "1",
+        "2",
+        "$(location :addition-result)",
+    ],
+    tool = ":add-numbers",
+)

examples/gem/lib/gem/add.rb

@@ -17,7 +17,13 @@ def result
 end
 
 if __FILE__ == $PROGRAM_NAME
-  raise "Pass two numbers to sum: #{ARGV}" unless ARGV.size == 2
+  raise "Pass two numbers to sum: #{ARGV}" if ARGV.size < 2
 
-  puts GEM::Add.new(*ARGV.map(&:to_i)).result
+  one, two, output = *ARGV
+  result = GEM::Add.new(Integer(one), Integer(two)).result
+  if output
+    File.write(output, result)
+  else
+    puts result
+  end
 end

ruby/private/BUILD

@@ -1,12 +1,6 @@
 load("@bazel_skylib//:bzl_library.bzl", "bzl_library")
 
-exports_files([
-    "binary/binary.cmd.tpl",
-    "binary/binary.sh.tpl",
-    "gem_build/gem_builder.rb.tpl",
-])
-
-exports_files(glob(["**/*.bzl"]))
+exports_files(["gem_build/gem_builder.rb.tpl"])
 
 bzl_library(
     name = "binary",
@@ -15,6 +9,7 @@ bzl_library(
     deps = [
         ":library",
         ":providers",
+        "//ruby/private/binary:rlocation",
     ],
 )
 

ruby/private/binary.bzl

@@ -7,8 +7,10 @@ load(
     "get_bundle_env",
     "get_transitive_data",
     "get_transitive_deps",
+    "get_transitive_runfiles",
     "get_transitive_srcs",
 )
+load("//ruby/private/binary:rlocation.bzl", "BASH_RLOCATION_FUNCTION", "BATCH_RLOCATION_FUNCTION")
 
 ATTRS = {
     "main": attr.label(
@@ -30,19 +32,27 @@ Use a built-in `args` attribute to pass extra arguments to the script.
     ),
     "_binary_cmd_tpl": attr.label(
         allow_single_file = True,
-        default = "@rules_ruby//ruby/private:binary/binary.cmd.tpl",
+        default = "@rules_ruby//ruby/private/binary:binary.cmd.tpl",
     ),
     "_binary_sh_tpl": attr.label(
         allow_single_file = True,
-        default = "@rules_ruby//ruby/private:binary/binary.sh.tpl",
+        default = "@rules_ruby//ruby/private/binary:binary.sh.tpl",
+    ),
+    "_runfiles_library": attr.label(
+        allow_single_file = True,
+        default = "@bazel_tools//tools/bash/runfiles",
     ),
     "_windows_constraint": attr.label(
         default = "@platforms//os:windows",
     ),
 }
 
+_EXPORT_ENV_VAR_COMMAND = "{command} {name}={value}"
+_EXPORT_BATCH_COMMAND = "set"
+_EXPORT_BASH_COMMAND = "export"
+
 # buildifier: disable=function-docstring
-def generate_rb_binary_script(ctx, binary, bundler = False, args = []):
+def generate_rb_binary_script(ctx, binary, bundler = False, args = [], env = {}, java_bin = ""):
     windows_constraint = ctx.attr._windows_constraint[platform_common.ConstraintValueInfo]
     is_windows = ctx.target_platform_has_constraint(windows_constraint)
     toolchain = ctx.toolchains["@rules_ruby//ruby:toolchain_type"]
@@ -55,10 +65,14 @@ def generate_rb_binary_script(ctx, binary, bundler = False, args = []):
 
     if is_windows:
         binary_path = binary_path.replace("/", "\\")
-        toolchain_bindir = toolchain_bindir.replace("/", "\\")
+        export_command = _EXPORT_BATCH_COMMAND
+        rlocation_function = BATCH_RLOCATION_FUNCTION
         script = ctx.actions.declare_file("{}.rb.cmd".format(ctx.label.name))
+        toolchain_bindir = toolchain_bindir.replace("/", "\\")
         template = ctx.file._binary_cmd_tpl
     else:
+        export_command = _EXPORT_BASH_COMMAND
+        rlocation_function = BASH_RLOCATION_FUNCTION
         script = ctx.actions.declare_file("{}.rb.sh".format(ctx.label.name))
         template = ctx.file._binary_sh_tpl
 
@@ -70,6 +84,11 @@ def generate_rb_binary_script(ctx, binary, bundler = False, args = []):
     args = " ".join(args)
     args = ctx.expand_location(args)
 
+    environment = []
+    for (name, value) in env.items():
+        command = _EXPORT_ENV_VAR_COMMAND.format(command = export_command, name = name, value = value)
+        environment.append(command)
+
     ctx.actions.expand_template(
         template = template,
         output = script,
@@ -78,8 +97,11 @@ def generate_rb_binary_script(ctx, binary, bundler = False, args = []):
             "{args}": args,
             "{binary}": binary_path,
             "{toolchain_bindir}": toolchain_bindir,
+            "{env}": "\n".join(environment),
             "{bundler_command}": bundler_command,
             "{ruby_binary_name}": toolchain.ruby.basename,
+            "{java_bin}": java_bin,
+            "{rlocation_function}": rlocation_function,
         },
     )
 
@@ -89,36 +111,46 @@ def generate_rb_binary_script(ctx, binary, bundler = False, args = []):
 def rb_binary_impl(ctx):
     bundler = False
     env = {}
+    java_bin = ""
+
+    # TODO: avoid expanding the depset to a list, it may be expensive in a large graph
     transitive_data = get_transitive_data(ctx.files.data, ctx.attr.deps).to_list()
     transitive_deps = get_transitive_deps(ctx.attr.deps).to_list()
     transitive_srcs = get_transitive_srcs(ctx.files.srcs, ctx.attr.deps).to_list()
-    tools = []
-    java_toolchain = ctx.toolchains["@bazel_tools//tools/jdk:runtime_toolchain_type"]
-    ruby_toolchain = ctx.toolchains["@rules_ruby//ruby:toolchain_type"]
 
-    if not ctx.attr.main:
-        tools.append(ruby_toolchain.ruby)
+    ruby_toolchain = ctx.toolchains["@rules_ruby//ruby:toolchain_type"]
+    tools = [ruby_toolchain.ruby, ruby_toolchain.bundle, ruby_toolchain.gem]
 
     if ruby_toolchain.version.startswith("jruby"):
-        env["JAVA_HOME"] = java_toolchain.java_runtime.java_home_runfiles_path
+        java_toolchain = ctx.toolchains["@bazel_tools//tools/jdk:runtime_toolchain_type"]
+        tools.extend(ctx.files._runfiles_library)
         tools.extend(java_toolchain.java_runtime.files.to_list())
+        java_bin = java_toolchain.java_runtime.java_executable_runfiles_path[3:]
 
     for dep in transitive_deps:
         # TODO: Do not depend on workspace name to determine bundle
         if dep.label.workspace_name.endswith("bundle"):
             bundler = True
 
-    runfiles = ctx.runfiles(transitive_data + transitive_srcs + tools)
-
     bundle_env = get_bundle_env(ctx.attr.env, ctx.attr.deps)
     env.update(bundle_env)
     env.update(ctx.attr.env)
 
-    script = generate_rb_binary_script(ctx, ctx.executable.main, bundler)
+    script = generate_rb_binary_script(
+        ctx,
+        ctx.executable.main,
+        bundler = bundler,
+        env = env,
+        java_bin = java_bin,
+    )
+
+    runfiles = ctx.runfiles(transitive_srcs + transitive_data + tools)
+    runfiles = get_transitive_runfiles(runfiles, ctx.attr.srcs, ctx.attr.deps, ctx.attr.data)
 
     return [
         DefaultInfo(
             executable = script,
+            files = depset(transitive_srcs + transitive_data + tools),
             runfiles = runfiles,
         ),
         RubyFilesInfo(

ruby/private/binary/BUILD

@@ -0,0 +1,9 @@
+load("@bazel_skylib//:bzl_library.bzl", "bzl_library")
+
+exports_files(glob(["*.tpl"]))
+
+bzl_library(
+    name = "rlocation",
+    srcs = ["rlocation.bzl"],
+    visibility = ["//ruby:__subpackages__"],
+)

ruby/private/binary/binary.cmd.tpl

@@ -1,2 +1,16 @@
-@set PATH={toolchain_bindir};%PATH%
-@{bundler_command} {ruby_binary_name} {binary} {args} %*
+@echo off
+setlocal enableextensions enabledelayedexpansion
+
+:: Find location of JAVA_HOME in runfiles.
+if "{java_bin}" neq "" (
+  {rlocation_function}
+  set RUNFILES_MANIFEST_ONLY=1
+  call :rlocation {java_bin} java_bin
+  for %%a in ("%java_bin%\..\..") do set JAVA_HOME=%%~fa
+)
+
+:: Set environment variables.
+set PATH={toolchain_bindir};%PATH%
+{env}
+
+{bundler_command} {ruby_binary_name} {binary} {args} %*

ruby/private/binary/binary.sh.tpl

@@ -1,4 +1,13 @@
 #!/usr/bin/env bash
 
+# Find location of JAVA_HOME in runfiles.
+if [ -n "{java_bin}" ]; then
+  {rlocation_function}
+  export JAVA_HOME=$(dirname $(dirname $(rlocation "{java_bin}")))
+fi
+
+# Set environment variables.
 export PATH={toolchain_bindir}:$PATH
+{env}
+
 {bundler_command} {ruby_binary_name} {binary} {args} $@

ruby/private/binary/rlocation.bzl

@@ -0,0 +1,61 @@
+"Helpers to get location of files in runfiles. Vendored from aspect_bazel_lib"
+
+# https://github.com/aspect-build/bazel-lib/blob/ddac9c46c3bff4cf8d0118a164c75390dbec2da9/lib/paths.bzl
+BASH_RLOCATION_FUNCTION = r"""
+# --- begin runfiles.bash initialization v2 ---
+set -uo pipefail; f=bazel_tools/tools/bash/runfiles/runfiles.bash
+source "${RUNFILES_DIR:-/dev/null}/$f" 2>/dev/null || \
+source "$(grep -sm1 "^$f " "${RUNFILES_MANIFEST_FILE:-/dev/null}" | cut -f2- -d' ')" 2>/dev/null || \
+source "$0.runfiles/$f" 2>/dev/null || \
+source "$(grep -sm1 "^$f " "$0.runfiles_manifest" | cut -f2- -d' ')" 2>/dev/null || \
+source "$(grep -sm1 "^$f " "$0.exe.runfiles_manifest" | cut -f2- -d' ')" 2>/dev/null || \
+{ echo>&2 "ERROR: cannot find $f"; exit 1; }; f=; set -e
+# --- end runfiles.bash initialization v2 ---
+"""
+
+# https://github.com/aspect-build/bazel-lib/blob/ddac9c46c3bff4cf8d0118a164c75390dbec2da9/lib/windows_utils.bzl
+BATCH_RLOCATION_FUNCTION = r"""
+rem Usage of rlocation function:
+rem        call :rlocation <runfile_path> <abs_path>
+rem        The rlocation function maps the given <runfile_path> to its absolute
+rem        path and stores the result in a variable named <abs_path>.
+rem        This function fails if the <runfile_path> doesn't exist in mainifest
+rem        file.
+:: Start of rlocation
+goto :rlocation_end
+:rlocation
+if "%~2" equ "" (
+  echo>&2 ERROR: Expected two arguments for rlocation function.
+  exit 1
+)
+if "%RUNFILES_MANIFEST_ONLY%" neq "1" (
+  set %~2=%~1
+  exit /b 0
+)
+if exist "%RUNFILES_DIR%" (
+  set RUNFILES_MANIFEST_FILE=%RUNFILES_DIR%_manifest
+)
+if "%RUNFILES_MANIFEST_FILE%" equ "" (
+  set RUNFILES_MANIFEST_FILE=%~f0.runfiles\MANIFEST
+)
+if not exist "%RUNFILES_MANIFEST_FILE%" (
+  set RUNFILES_MANIFEST_FILE=%~f0.runfiles_manifest
+)
+set MF=%RUNFILES_MANIFEST_FILE:/=\%
+if not exist "%MF%" (
+  echo>&2 ERROR: Manifest file %MF% does not exist.
+  exit 1
+)
+set runfile_path=%~1
+for /F "tokens=2* usebackq" %%i in (`%SYSTEMROOT%\system32\findstr.exe /l /c:"!runfile_path! " "%MF%"`) do (
+  set abs_path=%%i
+)
+if "!abs_path!" equ "" (
+  echo>&2 ERROR: !runfile_path! not found in runfiles manifest
+  exit 1
+)
+set %~2=!abs_path!
+exit /b 0
+:rlocation_end
+:: End of rlocation
+"""