import io

def writer(n, data):
    s = io.StringIO()
    for loop in range(n):
        for text in data:
            s.write(text)
    return s.getvalue()

def writer_reader(n, data):
    s = io.StringIO()
    for loop in range(n):
        for text in data:
            s.write(text)
    s.seek(0)
    for line in s:
        pass
    return s.getvalue()

def reader(data):
    s = io.StringIO(data)
    for line in s:
        pass

def run_benchmark(bench):
    bench_lines(bench, 10)
    bench_lines(bench, 10**3)
    bench_lines(bench, 10**5)

def bench_lines(bench, n):
    bench.start_group("%s lines" % n)
    for charset, text in (
        ('ascii', ('a', 'bb', 'cc\n')),
        ('latin1', ('\xe9', 'bb', 'cc\n')),
        ('bmp', ('\u20ac', 'bb', 'cc\n')),
        ('non-bmp', ('\U0010ffff', 'bb', 'cc\n')),
    ):
        bench.bench_func('reader short line %s' % charset, reader, ''.join(text) * n)
        bench.bench_func('writer short line %s' % charset, writer, n, text)
        bench.bench_func('writer-reader short line %s' % charset, writer_reader, n, text)

    k = 300
    for charset, text in (
        ('ascii', 'a' * k + '\n'),
        ('latin1', '\xe9' * k + '\n'),
        ('bmp', '\u20ac' * k + '\n'),
        ('non-bmp', '\U0010ffff' * k + '\n'),
    ):
        bench.bench_func('reader long lines %s' % charset, reader, text * k)
        bench.bench_func('writer long lines %s' % charset, writer, n // k, [text])
        bench.bench_func('writer-reader long lines %s' % charset, writer_reader, n // k, [text])

    k = 10
    for charset, text in (
        ('ascii', 'a' * (n // k) + '\n'),
        ('latin1', '\xe9' * (n // k) + '\n'),
        ('bmp', '\u20ac' * (n // k) + '\n'),
        ('non-bmp', '\U0010ffff' * (n // k) + '\n'),
    ):
        bench.bench_func('reader very long lines %s' % charset, reader, text * k)
        bench.bench_func('writer very long lines %s' % charset, writer, k, [text])
        bench.bench_func('writer-reader very long lines %s' % charset, writer_reader, k, [text])

    if 0:
        data = 'abc\n' * n
        bench.bench_func('reader ascii', reader, data)
        data = 'ab\xe9\n' * n
        bench.bench_func('reader latin1', reader, data)
        data = 'ab\xe9\u20ac\n' * n
        bench.bench_func('reader bmp', reader, data)
        data = 'ab\xe9\u20ac\U0010ffff\n' * n
        bench.bench_func('reader non-bmp', reader, data)