[lit] Fix UnicodeEncodeError when test commands contain non-ASCII chars

[mgorny]

Summary

Ensure that the bash script written by lit TestRunner is open with UTF-8
encoding when using Python 3. Otherwise, attempt to write non-ASCII
characters causes UnicodeEncodeError. This happened e.g. with
the following LLD test:

UNRESOLVED: lld :: ELF/format-binary-non-ascii.s (657 of 2119)
******************** TEST 'lld :: ELF/format-binary-non-ascii.s' FAILED ********************
Exception during script execution:
Traceback (most recent call last):
  File "/home/mgorny/llvm-project/llvm/utils/lit/lit/worker.py", line 63, in _execute_test
    result = test.config.test_format.execute(test, lit_config)
  File "/home/mgorny/llvm-project/llvm/utils/lit/lit/formats/shtest.py", line 25, in execute
    self.execute_external)
  File "/home/mgorny/llvm-project/llvm/utils/lit/lit/TestRunner.py", line 1644, in executeShTest
    res = _runShTest(test, litConfig, useExternalSh, script, tmpBase)
  File "/home/mgorny/llvm-project/llvm/utils/lit/lit/TestRunner.py", line 1590, in _runShTest
    res = executeScript(test, litConfig, tmpBase, script, execdir)
  File "/home/mgorny/llvm-project/llvm/utils/lit/lit/TestRunner.py", line 1157, in executeScript
    f.write('{ ' + '; } &&\n{ '.join(commands) + '; }')
UnicodeEncodeError: 'ascii' codec can't encode character '\xa3' in position 274: ordinal not in range(128)

[serge-sans-paille]

LGTM. Can you add a test case in lit/tests for that?

[mgorny]

Here's a proposed test case. I've verified that it failed (became unresolved) without my patch.

[rnk]

lgtm

llvm/utils/lit/lit/TestRunner.py

1139

I was thinking perhaps this should be if not elif, but I've run the ELF LLD test suite successfully on Windows with Python 3 and it works fine as is.

[mgorny]

llvm/utils/lit/lit/TestRunner.py

1139

This is elif because encoding only applies when mode is w and not wb.

[ruiu]

LGTM

llvm/utils/lit/lit/TestRunner.py

1137–1142

I'd perhaps write this this way,

f = None
if litConfig.isWindows and not isWin32CMDEXE:
  f = open(script, 'wb')
elif sys.version_info > (3,0):
  f = open(script, 'w', encoding='utf-8')
else
  f = open(script, 'w')

so that it is clear what values are passed to open, but that's probably my personal preference.

[mgorny]

llvm/utils/lit/lit/TestRunner.py

1137–1142

Sure but the code below relies on mode being defined :-(. I personally think the whole function here is a horrible hack but I currently don't have a good idea how to write it better. I mean, how does it make sense to have two copies of everything with different string types in order to control endlines as a side effect? Even conditionally putting \r\n would be cleaner than that. In fact, maybe I'll rewrite it to do just that…

[mgorny]

Anyway, committed as-is to resolve the issue at hand and I'll think how to rewrite it best.

@rnk, do you happen to know whether we are using UTF-8 on Windows these days or some system-native encoding?

llvm/utils/lit/lit/TestRunner.py

def executeScript(test, litConfig, tmpBase, commands, cwd):
     bashPath = litConfig.getBashPath()
     isWin32CMDEXE = (litConfig.isWindows and not bashPath)
     script = tmpBase + '.script'
     if isWin32CMDEXE:
         script += '.bat'
 
     # Write script file
     mode = 'w'
+    open_kwargs = {}
     if litConfig.isWindows and not isWin32CMDEXE:
         mode += 'b'  # Avoid CRLFs when writing bash scripts.
-    f = open(script, mode)
+    elif sys.version_info > (3,0):
+        open_kwargs['encoding'] = 'utf-8'
+    f = open(script, mode, **open_kwargs)
     if isWin32CMDEXE:
         for i, ln in enumerate(commands):
             commands[i] = re.sub(kPdbgRegex, "echo '\\1' > nul && ", ln)
         if litConfig.echo_all_commands:
             f.write('@echo on\n')
         else:
             f.write('@echo off\n')
         f.write('\n@if %ERRORLEVEL% NEQ 0 EXIT\n'.join(commands))
     else: