Catch exceptions from opening invalid files

This ensures very badly invalid files are still recorded as such,
while allowing further files to be processed.

Author: colinpalmer

CHANGELOG.txt

@@ -6,7 +6,8 @@ Version 1.4.1 (2022-07-12)
 
 * Add file names to mrcfile-header output
 * Add file names and a success message for valid files to the output of the
-  validation functions (mrcfile-validate command and the validate() method)
+  validation functions (mrcfile-validate command and the mrcfile.validate()
+  function)
 * Add support for numpy 1.23 and Python 3.10 and 3.11
 
 Version 1.4.0 (2022-07-02)

mrcfile/validator.py

@@ -18,6 +18,7 @@
 
 import argparse
 import sys
+import traceback
 
 from . import load_functions
 
@@ -155,11 +156,15 @@ def validate(name, print_file=None):
             is not the same size as expected from the header.
     """
     print("Checking if {} is a valid MRC2014 file...".format(name), file=print_file)
-    with load_functions.open(name, permissive=True) as mrc:
-        result = mrc.validate(print_file=print_file)
-        if result:
-            print("File appears to be valid.", file=print_file)
-        return result
+    try:
+        with load_functions.open(name, permissive=True) as mrc:
+            result = mrc.validate(print_file=print_file)
+    except Exception:
+        result = False
+        traceback.print_exc(file=print_file)
+    if result:
+        print("File appears to be valid.", file=print_file)
+    return result
 
 
 if __name__ == '__main__':

tests/test_validation.py

@@ -9,14 +9,22 @@
 from __future__ import (absolute_import, division, print_function,
                         unicode_literals)
 
-import io
 import os
 import shutil
 import sys
 import tempfile
 import unittest
 import warnings
 
+# Avoid str/unicode issues when traceback.print_exc tries to write to StringIO
+# (see https://stackoverflow.com/a/34872005)
+try:
+    # Python 2
+    from cStringIO import StringIO
+except ImportError:
+    # Python 3
+    from io import StringIO
+
 import numpy as np
 
 import mrcfile
@@ -43,15 +51,16 @@ def setUp(self):
         self.ext_header_mrc_name = os.path.join(self.test_data, 'EMD-3001.map')
         self.fei1_ext_header_mrc_name = os.path.join(self.test_data, 'fei-extended.mrc')
         self.fei2_ext_header_mrc_name = os.path.join(self.test_data, 'epu2.9_example.mrc')
+        self.not_an_mrc_name = os.path.join(self.test_data, 'README.txt')
 
         # Set up stream to catch print output from validate()
-        self.print_stream = io.StringIO()
+        self.print_stream = StringIO()
 
         # Replace stdout and stderr to capture output for checking
         self.orig_stdout = sys.stdout
         self.orig_stderr = sys.stderr
-        sys.stdout = io.StringIO()
-        sys.stderr = io.StringIO()
+        sys.stdout = StringIO()
+        sys.stderr = StringIO()
 
     def tearDown(self):
         # Restore stdout and stderr
@@ -562,6 +571,7 @@ def test_validate_good_files(self):
 
     def test_validate_bad_files(self):
         bad_files = [
+            self.not_an_mrc_name,
             self.example_mrc_name,
             self.ext_header_mrc_name,
             self.gzip_mrc_name
@@ -570,11 +580,17 @@ def test_validate_bad_files(self):
         assert result == False
         print_output = self.print_stream.getvalue()
         assert len(print_output) > 0
+        assert "Checking if " + bad_files[0] + " is a valid MRC2014 file..." in print_output
+        assert "Checking if " + bad_files[1] + " is a valid MRC2014 file..." in print_output
+        assert "Checking if " + bad_files[2] + " is a valid MRC2014 file..." in print_output
+        assert "Checking if " + bad_files[3] + " is a valid MRC2014 file..." in print_output
+        assert "ValueError:" in print_output
         assert len(sys.stdout.getvalue()) == 0
         assert len(sys.stderr.getvalue()) == 0
 
     def test_validate_good_and_bad_files(self):
         files = self.create_good_files() + [
+            self.not_an_mrc_name,
             self.example_mrc_name,
             self.ext_header_mrc_name,
             self.gzip_mrc_name