run-tests: ignore ENOENT failures when removing old .err results
When the same test runs in multiple threads and the previous run was a
failure, the threads can race to delete the error output. This fixes
that.
[('string', 'string value'), ('bool1', 'true'), ('bool2', 'false'), ('boolinvalid', 'foo'), ('int1', '42'), ('int2', '-42'), ('intinvalid', 'foo')]
[('list1', 'foo'), ('list2', 'foo bar baz'), ('list3', 'alice, bob'), ('list4', 'foo bar baz alice, bob'), ('list5', 'abc d"ef"g "hij def"'), ('list6', '"hello world", "how are you?"'), ('list7', 'Do"Not"Separate'), ('list8', '"Do"Separate'), ('list9', '"Do\\"NotSeparate"'), ('list10', 'string "with extraneous" quotation mark"'), ('list11', 'x, y'), ('list12', '"x", "y"'), ('list13', '""" key = "x", "y" """'), ('list14', ',,,, '), ('list15', '" just with starting quotation'), ('list16', '"longer quotation" with "no ending quotation'), ('list17', 'this is \\" "not a quotation mark"'), ('list18', '\n \n\nding\ndong')]
---
'string value'
'true'
'false'
None
---
values.string is not a boolean ('string value')
True
False
False
False
True
---
42
-42
---
['foo']
['foo', 'bar', 'baz']
['alice', 'bob']
['foo', 'bar', 'baz', 'alice', 'bob']
['foo', 'bar', 'baz', 'alice', 'bob']
['abc', 'd"ef"g', 'hij def']
['hello world', 'how are you?']
['Do"Not"Separate']
['Do', 'Separate']
['Do"NotSeparate']
['string', 'with extraneous', 'quotation', 'mark"']
['x', 'y']
['x', 'y']
['', ' key = ', 'x"', 'y', '', '"']
[]
['"', 'just', 'with', 'starting', 'quotation']
['longer quotation', 'with', '"no', 'ending', 'quotation']
['this', 'is', '"', 'not a quotation mark']
['ding', 'dong']
[]
[]
['foo']
['foo']
['foo', 'bar']
['foo', 'bar']
['foo bar']
['foo', 'bar']
None
True
boolinvalid
intinvalid