412 lines
8.1 KiB
Ruby
412 lines
8.1 KiB
Ruby
require "fileutils"
|
|
require "open3"
|
|
|
|
Results = Struct.new(:stdout, :stderr, :status)
|
|
|
|
describe Propane do
|
|
def write_grammar(grammar, options = {})
|
|
options[:name] ||= ""
|
|
File.write("spec/run/testparser#{options[:name]}.propane", grammar)
|
|
end
|
|
|
|
def build_parser(options = {})
|
|
options[:name] ||= ""
|
|
command = %W[./propane.sh spec/run/testparser#{options[:name]}.propane spec/run/testparser#{options[:name]}.d --log spec/run/testparser#{options[:name]}.log]
|
|
if (options[:capture])
|
|
stdout, stderr, status = Open3.capture3(*command)
|
|
Results.new(stdout, stderr, status)
|
|
else
|
|
result = system(*command)
|
|
expect(result).to be_truthy
|
|
end
|
|
end
|
|
|
|
def compile(test_files, options = {})
|
|
test_files = Array(test_files)
|
|
options[:parsers] ||= [""]
|
|
parsers = options[:parsers].map do |name|
|
|
"spec/run/testparser#{name}.d"
|
|
end
|
|
result = system(*%w[ldc2 --unittest -of spec/run/testparser -Ispec], *parsers, *test_files)
|
|
expect(result).to be_truthy
|
|
end
|
|
|
|
def run
|
|
stdout, stderr, status = Open3.capture3("spec/run/testparser")
|
|
File.binwrite("spec/run/.stderr", stderr)
|
|
File.binwrite("spec/run/.stdout", stdout)
|
|
stderr.sub!(/^.*modules passed unittests\n/, "")
|
|
Results.new(stdout, stderr, status)
|
|
end
|
|
|
|
def lines(str)
|
|
str.lines.map(&:chomp)
|
|
end
|
|
|
|
def verify_lines(lines, patterns)
|
|
if lines.is_a?(String)
|
|
lines = lines.lines.map(&:chomp)
|
|
end
|
|
patterns.each_with_index do |pattern, i|
|
|
found_index =
|
|
if pattern.is_a?(Regexp)
|
|
lines.find_index {|line| line =~ pattern}
|
|
else
|
|
lines.find_index do |line|
|
|
line.chomp == pattern.chomp
|
|
end
|
|
end
|
|
unless found_index
|
|
$stderr.puts "Lines:"
|
|
$stderr.puts lines
|
|
raise "A line matching #{pattern.inspect} (index #{i}) was not found."
|
|
end
|
|
end
|
|
end
|
|
|
|
before(:each) do
|
|
FileUtils.rm_rf("spec/run")
|
|
FileUtils.mkdir_p("spec/run")
|
|
end
|
|
|
|
it "generates a D lexer" do
|
|
write_grammar <<EOF
|
|
token int /\\d+/;
|
|
token plus /\\+/;
|
|
token times /\\*/;
|
|
drop /\\s+/;
|
|
Start -> Foo;
|
|
Foo -> int <<
|
|
>>
|
|
Foo -> plus <<
|
|
>>
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_d_lexer.d")
|
|
results = run
|
|
expect(results.stderr).to eq ""
|
|
expect(results.status).to eq 0
|
|
end
|
|
|
|
it "detects a lexer error when an unknown character is seen" do
|
|
write_grammar <<EOF
|
|
ptype int;
|
|
token int /\\d+/ <<
|
|
int v;
|
|
foreach (c; match)
|
|
{
|
|
v *= 10;
|
|
v += (c - '0');
|
|
}
|
|
$$ = v;
|
|
>>
|
|
Start -> int <<
|
|
$$ = $1;
|
|
>>
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_lexer_unknown_character.d")
|
|
results = run
|
|
expect(results.stderr).to eq ""
|
|
expect(results.status).to eq 0
|
|
end
|
|
|
|
it "generates a parser" do
|
|
write_grammar <<EOF
|
|
token plus /\\+/;
|
|
token times /\\*/;
|
|
token zero /0/;
|
|
token one /1/;
|
|
Start -> E;
|
|
E -> E times B;
|
|
E -> E plus B;
|
|
E -> B;
|
|
B -> zero;
|
|
B -> one;
|
|
EOF
|
|
build_parser
|
|
end
|
|
|
|
it "generates an SLR parser" do
|
|
write_grammar <<EOF
|
|
token one /1/;
|
|
Start -> E;
|
|
E -> one E;
|
|
E -> one;
|
|
EOF
|
|
build_parser
|
|
end
|
|
|
|
it "distinguishes between multiple identical rules with lookahead symbol" do
|
|
write_grammar <<EOF
|
|
token a;
|
|
token b;
|
|
Start -> R1 a;
|
|
Start -> R2 b;
|
|
R1 -> a b;
|
|
R2 -> a b;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_d_parser_identical_rules_lookahead.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
end
|
|
|
|
it "handles reducing a rule that could be arrived at from multiple states" do
|
|
write_grammar <<EOF
|
|
token a;
|
|
token b;
|
|
drop /\\s+/;
|
|
Start -> a R1;
|
|
Start -> b R1;
|
|
R1 -> b;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_d_parser_rule_from_multiple_states.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
end
|
|
|
|
it "executes user code when matching lexer token" do
|
|
write_grammar <<EOF
|
|
token abc <<
|
|
writeln("abc!");
|
|
>>
|
|
token def;
|
|
Start -> Abcs def;
|
|
Abcs -> ;
|
|
Abcs -> abc Abcs;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_user_code.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
verify_lines(results.stdout, [
|
|
"abc!",
|
|
"pass1",
|
|
"abc!",
|
|
"abc!",
|
|
"pass2",
|
|
])
|
|
end
|
|
|
|
it "supports a pattern statement" do
|
|
write_grammar <<EOF
|
|
token abc;
|
|
/def/ <<
|
|
writeln("def!");
|
|
>>
|
|
Start -> abc;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_pattern.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
verify_lines(results.stdout, [
|
|
"def!",
|
|
"pass1",
|
|
"def!",
|
|
"def!",
|
|
"pass2",
|
|
])
|
|
end
|
|
|
|
it "supports returning tokens from pattern code blocks" do
|
|
write_grammar <<EOF
|
|
token abc;
|
|
/def/ <<
|
|
writeln("def!");
|
|
>>
|
|
/ghi/ <<
|
|
writeln("ghi!");
|
|
return $token(abc);
|
|
>>
|
|
Start -> abc;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_return_token_from_pattern.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
verify_lines(results.stdout, [
|
|
"def!",
|
|
"ghi!",
|
|
"def!",
|
|
])
|
|
end
|
|
|
|
it "supports lexer modes" do
|
|
write_grammar <<EOF
|
|
token abc;
|
|
token def;
|
|
tokenid string;
|
|
drop /\\s+/;
|
|
/"/ <<
|
|
writeln("begin string mode");
|
|
$mode(string);
|
|
>>
|
|
string: /[^"]+/ <<
|
|
writeln("captured string");
|
|
>>
|
|
string: /"/ <<
|
|
$mode(default);
|
|
return $token(string);
|
|
>>
|
|
Start -> abc string def;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_lexer_modes.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
verify_lines(results.stdout, [
|
|
"begin string mode",
|
|
"captured string",
|
|
"pass1",
|
|
"begin string mode",
|
|
"captured string",
|
|
"pass2",
|
|
])
|
|
end
|
|
|
|
it "executes user code associated with a parser rule" do
|
|
write_grammar <<EOF
|
|
token a;
|
|
token b;
|
|
Start -> A B <<
|
|
writeln("Start!");
|
|
>>
|
|
A -> a <<
|
|
writeln("A!");
|
|
>>
|
|
B -> b <<
|
|
writeln("B!");
|
|
>>
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_parser_rule_user_code.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
verify_lines(results.stdout, [
|
|
"A!",
|
|
"B!",
|
|
"Start!",
|
|
])
|
|
end
|
|
|
|
it "parses lists" do
|
|
write_grammar <<EOF
|
|
ptype uint;
|
|
token a;
|
|
Start -> As <<
|
|
$$ = $1;
|
|
>>
|
|
As -> <<
|
|
$$ = 0u;
|
|
>>
|
|
As -> As a <<
|
|
$$ = $1 + 1u;
|
|
>>
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_parsing_lists.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
expect(results.stderr).to eq ""
|
|
end
|
|
|
|
it "fails to generate a parser for a LR(1) grammar that is not LALR" do
|
|
write_grammar <<EOF
|
|
token a;
|
|
token b;
|
|
token c;
|
|
token d;
|
|
token e;
|
|
Start -> a E c;
|
|
Start -> a F d;
|
|
Start -> b F c;
|
|
Start -> b E d;
|
|
E -> e;
|
|
F -> e;
|
|
EOF
|
|
results = build_parser(capture: true)
|
|
expect(results.status).to_not eq 0
|
|
expect(results.stderr).to match %r{reduce/reduce conflict.*\(E\).*\(F\)}
|
|
end
|
|
|
|
it "provides matched text to user code blocks" do
|
|
write_grammar <<EOF
|
|
token id /[a-zA-Z_][a-zA-Z0-9_]*/ <<
|
|
writeln("Matched token is ", match);
|
|
>>
|
|
Start -> id;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_lexer_match_text.d")
|
|
results = run
|
|
expect(results.status).to eq 0
|
|
verify_lines(results.stdout, [
|
|
"Matched token is identifier_123",
|
|
"pass1",
|
|
])
|
|
end
|
|
|
|
it "allows storing a result value for the lexer" do
|
|
write_grammar <<EOF
|
|
ptype ulong;
|
|
token word /[a-z]+/ <<
|
|
$$ = match.length;
|
|
>>
|
|
Start -> word <<
|
|
$$ = $1;
|
|
>>
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_lexer_result_value.d")
|
|
results = run
|
|
expect(results.stderr).to eq ""
|
|
expect(results.status).to eq 0
|
|
end
|
|
|
|
it "tracks position of parser errors" do
|
|
write_grammar <<EOF
|
|
token a;
|
|
token num /\\d+/;
|
|
drop /\\s+/;
|
|
Start -> a num Start;
|
|
Start -> a num;
|
|
EOF
|
|
build_parser
|
|
compile("spec/test_error_positions.d")
|
|
results = run
|
|
expect(results.stderr).to eq ""
|
|
expect(results.status).to eq 0
|
|
end
|
|
|
|
it "allows creating a JSON parser" do
|
|
write_grammar(File.read("spec/json_parser.propane"))
|
|
build_parser
|
|
compile(["spec/test_parsing_json.d", "spec/json_types.d"])
|
|
end
|
|
|
|
it "allows generating multiple parsers in the same program" do
|
|
write_grammar(<<EOF, name: "myp1")
|
|
prefix myp1_;
|
|
token a;
|
|
token num /\\d+/;
|
|
drop /\\s+/;
|
|
Start -> a num;
|
|
EOF
|
|
build_parser(name: "myp1")
|
|
write_grammar(<<EOF, name: "myp2")
|
|
prefix myp2_;
|
|
token b;
|
|
token c;
|
|
Start -> b c b;
|
|
EOF
|
|
build_parser(name: "myp2")
|
|
compile("spec/test_multiple_parsers.d", parsers: %w[myp1 myp2])
|
|
results = run
|
|
expect(results.stderr).to eq ""
|
|
expect(results.status).to eq 0
|
|
end
|
|
end
|