TL;DR this is an extensive answer explaining character and hex encoding, you can skip this and look at the code below
The sha256sum
and related commands are adding the dash: -
in the output. These commands have been made to show hash values of *files. A single dash simply means that the input was from the standard inpuIt stream (i.e. there is no file name). Unfortunately I don't see an option to suppress the output, so you have to remove it yourself to get to the actual hash value.
So the hash utilities do not only return the hash value. A SHA-256 hash value simply consists of 32 bytes. As humans cannot read binary the binary is displayed using hexadecimals, but the actual value should still be thought of as bytes. The hexadecimal characters are just a representation of those bytes.
The input of hash functions consist of bits or rather bytes as well. This means that any difference in encoding text will mean that the hash value will be different. This is especially tricky when it comes to white-space and end-of-line encoding. Instead of adding a trailing newline it is probably better to suppress it with the -n
command line option for the echo
command in the case of "hello" though.
Beware that hexadecimals themselves can also be displayed in different ways; you would make sure whitespace is not present and that the comparison is case-insensitive or that the representation of the bytes always uses the same case.
Shell code
Using sha256sum
:
echo -n "hello" | sha256sum | tr -d "[:space:]-"
Using OpenSSL command line:
echo -n hello | openssl sha256 -binary | od -An -tx1 | tr -d "[:space:]"
Here od -An -tx1
will show each byte separately, instead of grouping them which may lead to problems with endianness.
tr -d "[:space:]
will remove spaces from the hexadecimals as well as the trailing newline. For sha256sum
the dash file indicator is also removed (note the -
at the end). This way it is possible to perform a textual (case insensitive) compare.
Python code
In Python without the trailing end of line:
print(hashlib.sha256("hello").hexdigest(), end="")
Java code
In the case of Java you should also make sure that the text encoding matches the system default encoding or you may get into trouble. So you should change:
md.update(text.getBytes("UTF-8"));
to
md.update(text.getBytes());
to get to the platform character encoding. If you don't the compare will fail if the encoding of the platform is not compatible with UTF-8 for the string you want to compare.