[diffoscope] 03/03: Don't crash on invalid archives, printing a useful error instead. (Closes: #833697)

[Chris Lamb]

This is an automated email from the git hooks/post-receive script.

lamby pushed a commit to branch experimental
in repository diffoscope.

commit 82fbe27785f08d3623073eb98a3a4670b02fd2dc
Author: Chris Lamb <lamby at debian.org>
Date:   Fri Apr 14 23:00:10 2017 +0100

    Don't crash on invalid archives, printing a useful error instead. (Closes: #833697)
    
    Signed-off-by: Chris Lamb <lamby at debian.org>
---
 diffoscope/comparators/utils/file.py       |  9 ++++++++-
 diffoscope/comparators/utils/libarchive.py | 10 +++++++---
 diffoscope/exc.py                          |  5 +++++
 3 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/diffoscope/comparators/utils/file.py b/diffoscope/comparators/utils/file.py
index 611aa93..97d829b 100644
--- a/diffoscope/comparators/utils/file.py
+++ b/diffoscope/comparators/utils/file.py
@@ -24,7 +24,8 @@ import magic
 import logging
 import subprocess
 
-from diffoscope.exc import RequiredToolNotFound, OutputParsingError
+from diffoscope.exc import RequiredToolNotFound, OutputParsingError, \
+    ContainerExtractionError
 from diffoscope.tools import tool_required
 from diffoscope.profiling import profile
 from diffoscope.difference import Difference
@@ -274,6 +275,12 @@ class File(object, metaclass=abc.ABCMeta):
                     return None
                 difference.add_comment("Error parsing output of `%s` for %s" %
                         (e.command, e.object_class))
+            except ContainerExtractionError as e:
+                difference = self.compare_bytes(other, source=source)
+                if difference is None:
+                    return None
+                difference.add_comment("Error extracting '{}', falling back to "
+                    "binary comparison ('{}')".format(e.pathname, e.wrapped_exc))
             return difference
         return self.compare_bytes(other, source)
 
diff --git a/diffoscope/comparators/utils/libarchive.py b/diffoscope/comparators/utils/libarchive.py
index b8ba582..2271591 100644
--- a/diffoscope/comparators/utils/libarchive.py
+++ b/diffoscope/comparators/utils/libarchive.py
@@ -25,6 +25,7 @@ import logging
 import libarchive
 import collections
 
+from diffoscope.exc import ContainerExtractionError
 from diffoscope.excludes import any_excluded
 from diffoscope.tempfiles import get_temporary_directory
 
@@ -227,9 +228,12 @@ class LibarchiveContainer(Archive):
                 logger.debug("Extracting %s to %s", entry.pathname, dst)
 
                 os.makedirs(os.path.dirname(dst), exist_ok=True)
-                with open(dst, 'wb') as f:
-                    for block in entry.get_blocks():
-                        f.write(block)
+                try:
+                    with open(dst, 'wb') as f:
+                        for block in entry.get_blocks():
+                            f.write(block)
+                except Exception as exc:
+                    raise ContainerExtractionError(entry.pathname, exc)
 
         logger.debug(
             "Extracted %d entries from %s to %s",
diff --git a/diffoscope/exc.py b/diffoscope/exc.py
index daffbae..64c5810 100644
--- a/diffoscope/exc.py
+++ b/diffoscope/exc.py
@@ -38,3 +38,8 @@ class RequiredToolNotFound(Exception):
             return None
 
         return providers.get(get_current_os(), None)
+
+class ContainerExtractionError(Exception):
+    def __init__(self, pathname, wrapped_exc):
+        self.pathname = pathname
+        self.wrapped_exc = wrapped_exc

-- 
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/reproducible/diffoscope.git