- Problem /...../mce/python/lib/python3.6/test/zip_cp437_header.zip - java.lang.IllegalArgumentException: malformed input off : 14, length : 1

[HoWeBrz]

I get this error message during the scanning of linus tool folders.
I am not sure if the log4J-script stops working or finishes its task.

java -jar log4j-detector-2021.12.17.jar /.../ > hits.txt
-- Problem /.../tools.lnx86/mce/python/lib/python3.6/test/zip_cp437_header.zip - java.lang.IllegalArgumentException: malformed input off : 14, length : 1

Is this a problem/bug and is there a solution, workaround ?

[juliusmusseau]

Probably it continues fine, but can you attach the problematic zip just so I can be sure?

[HoWeBrz]

Hi, Please find attached the zip-file.

[sker65]

I also discovered these errors a lot. Mostly in "inner zips -> https://pasteimg.com/image/image.fe2QH

[rgmz]

I also discovered these errors a lot. Mostly in "inner zips -> https://pasteimg.com/image/image.fe2QH

The JAR(s) in question are from pkg:maven/org.bytedeco/cpython@3.9.2-1.5.5, e.g. https://repo1.maven.org/maven2/org/bytedeco/cpython/3.9.2-1.5.5/cpython-3.9.2-1.5.5-linux-x86_64.jar.

Stack trace:

$ java -jar log4j-detector-2021.12.20.jar --verbose cpython-3.9.2-1.5.5-linux-x86_64.jar 
-- github.com/mergebase/log4j-detector v2021.12.22 (by mergebase.com) analyzing paths (could take a while).
-- Note: specify the '--verbose' flag to have every file examined printed to STDERR.
...
-- Examining /tmp/cpython-3.9.2-1.5.5-linux-x86_64.jar!/org/bytedeco/cpython/linux-x86_64/lib/python3.9/test/test_importlib/zipdata01/ziptestdata.zip... 
-- Examining /tmp/cpython-3.9.2-1.5.5-linux-x86_64.jar!/org/bytedeco/cpython/linux-x86_64/lib/python3.9/test/zip_cp437_header.zip... 
-- Problem /tmp/cpython-3.9.2-1.5.5-linux-x86_64.jar!/org/bytedeco/cpython/linux-x86_64/lib/python3.9/test/zip_cp437_header.zip - java.lang.IllegalArgumentException: malformed input off : 14, length : 1
java.lang.IllegalArgumentException: malformed input off : 14, length : 1
	at java.base/java.lang.StringCoding.throwMalformed(StringCoding.java:698)
	at java.base/java.lang.StringCoding.decodeUTF8_0(StringCoding.java:885)
	at java.base/java.lang.StringCoding.newStringUTF8NoRepl(StringCoding.java:978)
	at java.base/java.lang.System$2.newStringUTF8NoRepl(System.java:2205)
	at java.base/java.util.zip.ZipCoder$UTF8.toString(ZipCoder.java:60)
	at java.base/java.util.zip.ZipCoder.toString(ZipCoder.java:87)
	at java.base/java.util.zip.ZipInputStream.readLOC(ZipInputStream.java:303)
	at java.base/java.util.zip.ZipInputStream.getNextEntry(ZipInputStream.java:125)
	at com.mergebase.log4j.Log4JDetector.findLog4jRecursive(Log4JDetector.java:291)
	at com.mergebase.log4j.Log4JDetector.findLog4jRecursive(Log4JDetector.java:372)
	at com.mergebase.log4j.Log4JDetector.scan(Log4JDetector.java:617)
	at com.mergebase.log4j.Log4JDetector.analyze(Log4JDetector.java:724)
	at com.mergebase.log4j.Log4JDetector.main(Log4JDetector.java:160)
Caused by: java.nio.charset.MalformedInputException: Input length = 1
	... 13 more
...
-- No vulnerable Log4J 2.x samples found in supplied paths: [cpython-3.9.2-1.5.5-linux-x86_64.jar]
-- Congratulations, the supplied paths are not vulnerable to CVE-2021-44228 or CVE-2021-45046 !  :-) 

A cursory google search seems to indicate that the error is related to file encoding, but that may not be the case.

Relevant code:

log4j-detector/src/main/java/com/mergebase/log4j/Log4JDetector.java

Lines 289 to 294 in d8225c6

	while (true) {
	try {
	ze = zin.getNextEntry();
	} catch (Exception oops) {
	System.err.println("-- Problem " + zipPath + " - " + oops);
	if (verbose) {

[rgmz]

A cursory google search seems to indicate that the error is related to file encoding, but that may not be the case.

The archive in question does have a file with a non-ascii character in its name.

$ uchardet *
filename_without.txt: ASCII
filename_with_СoЖ.txt: ASCII
zip_cp437_header.zip: unknown

It seems like the fix for that is to either start with UTF-8 and try different encodings on java.nio.charset.MalformedInputException, or use something like ISO-8859-1 from the get-go.
https://stackoverflow.com/a/26268235