Bugfix - JSON connector doesnt retain unicode characters during load

[msmygit]

Fixes #508

[absurdfarce]

So rather than creating a new test class for this we should be able to add this to the JsonNodeToStringCodecTest. We already have a should_convert_from_valid_external method which verifies the ability of that codec to handle legit incoming JsonNodes... and IIUC this is just another case of that. Something like the following should do it:

      String unicodeEscapeString = "\\u001a\\u001aL\\\\";
      String mixedContentString = "Text with \\\\u001a unicode \\\\u001aL\\\\\\\\ escapes";
      assertThat(codec)
              .convertsFromExternal(nodeFactory.textNode(unicodeEscapeString))
              .toInternal(unicodeEscapeString)
              .convertsFromExternal(nodeFactory.textNode(mixedContentString))
              .toInternal(mixedContentString);

Thing is... if I add that test to 1.x and run it (with no additional modifications, specifically without the changes to JsonNodeToStringCodec also included in this PR) the test already passes. So are we sure there isn't something else going on here?

[absurdfarce]

Hey @msmygit, I built a distribution from the tag of 1.11.0 + your change and I'm still seeing bogus results:

CREATE KEYSPACE test508 WITH replication = {'class': 'SimpleStrategy', 'replication_factor': '1'}  AND durable_writes = true;

CREATE TABLE test508.codec_test (
    i text PRIMARY KEY,
    j text
);

$ more ../508.json 
[{"i":"json","j":"NO\u001a\u001aL\\"}]
$ more ../508.csv 
i,j
csv,"NO\u001a\u001aL\\"

bin/dsbulk load -k test508 -t codec_test -c json -url ../508.json  --dsbulk.connector.json.mode SINGLE_DOCUMENT
bin/dsbulk load -k test508 -t codec_test -c csv -url ../508.csv

[cqlsh 6.2.0 | Cassandra 5.0.4 | CQL spec 3.4.7 | Native protocol v5]
Use HELP for help.
cqlsh> select * from test508.codec_test ;

 i	| j
------+------------------
  csv | NO\u001a\u001aL\
 json | 	NO\x1a\x1aL\

(2 rows)

Did I muck something up in the above?

[absurdfarce]

It looks like I did miss the --dsbulk.codec.binary HEX param but even if I include that the results don't seem to change

[absurdfarce]

Closing per discussion on the corresponding issue