How to SHA2 hash a string in USQL

问题

I am trying to run a one-way hash for a string column in USQL. Is there a way to do this inline? Most of the C# samples found online require multiple lines of code - which is tricky in USQL without a code-behind or compiled C# assembly.

回答1:

Option 1 (Inline formula):

The code below can be used to compile a SHA256 or MD5 on any string, and runs without any special dependencies and without needing a code-behind file.

CREATE TABLE master.dbo.Test_MyEmail_Hashes AS
SELECT
      cust.CustEmailAddr          AS Email
    , String.Concat(System.Security.Cryptography.SHA256.Create()
                    .ComputeHash(Encoding.UTF8.GetBytes(
                        cust.CustEmailAddr))
                    .Select(item => item.ToString("x2")))
                                  AS Email_SHA2
    , String.Concat(System.Security.Cryptography.MD5.Create()
                    .ComputeHash(Encoding.UTF8.GetBytes(
                        cust.CustEmailAddr))
                    .Select(item => item.ToString("x2")))
                                  AS Email_MD5
FROM master.dbo.Customers AS cust
;

Option 2 (using Lambda functions): (UPDATED)

Thanks to @MichaelRys for the pointer that USQL now supports Lambda functions and can be cleaned up as in the below:

// Generic get_hash() function
DECLARE @get_hash Func<string,System.Security.Cryptography.HashAlgorithm,string> =
     (raw_value, hasher) => String.Concat(hasher.ComputeHash(Encoding.UTF8.GetBytes(raw_value)));

// Short-hand functions for MD5 and SHA256:
DECLARE @md5    = System.Security.Cryptography.MD5.Create();
DECLARE @get_md5 Func<string,string> =
    (raw_value) => @get_hash(raw_value, @md5);
DECLARE @sha256 = System.Security.Cryptography.SHA256.Create();
DECLARE @get_sha256 Func<string,string> =
    (raw_value) => @get_hash(raw_value, @sha256);

// Core query:
CREATE TABLE master.dbo.Test_MyEmail_Hashes AS
SELECT
      cust.CustEmailAddr                AS Email
    , @get_sha256(cust.CustEmailAddr)   AS Email_SHA2
    , @get_md5(cust.CustEmailAddr)      AS Email_MD5
FROM master.dbo.Customers AS cust

回答2:

Actually I suggest you use the recently added "named lambdas" (Func<> typed variables) to use the multi-line C# samples. An example is here: https://github.com/Azure/AzureDataLake/blob/master/docs/Release_Notes/2018/2018_Spring/USQL_Release_Notes_2018_Spring.md#u-sql-adds-c-func-typed-variables-in-declare-statements-named-lambdas

来源：https://stackoverflow.com/questions/50590920/how-to-sha2-hash-a-string-in-usql

标签

hash

azure-data-lake

u-sql