Invalid UTF-8 middle byte 0x22

[Chuan Chuan Law]

Hi,

I am using the JVM Groovy module to generate body with regex like below:

 withBody{

                ComplianceDate{

                    Month (integer())

                    Year (integer())

                }

                BuildDate{}

                ListingType (~/\w+/)

                Odometer (integer())

                StockNumber (~/\w+/)

And occasionally when we run pact verify we get the error:

> com.fasterxml.jackson.core.JsonParseException: Invalid UTF-8 middle byte 0x22

Is there a way to ensure that the generated json file will be a valid UTF-8 char?

Please advice.

Thanks.

[Ronald Holshausen]

Hi Chuan,

UTF-8 is the default encoding on the JVM. Did you get this when running the pact verification, or in the consumer tests?

Can you post the full stacktrace?

--
You received this message because you are subscribed to the Google Groups "Pact" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pact-support...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Ronald Holshausen

DiUS Computing Pty Ltd

Level 10, 99 Queens Street
Melbourne, VIC 3000

Phone: +61 3 9008 5400
Mobile: +61 413 162 439

http://www.diuscomputing.com.au

[Chuan Chuan Law]

Hi Ronald,

I get this error sometimes when I run pact verification.

For eg the build failed this morning:

Verifying a pact between oag_consumer_es_industry and ElasticSearchCarProvider
  [Using file D:\programs\JenkinsSlave\workspace\Trade API Tests\TradeApiTests\target\pacts\OneAutoGate-Elastic Industry Search.json]
:pactVerify_ElasticSearchCarProvider FAILED

FAILURE: Build failed with an exception.

* What went wrong:
Execution failed for task ':pactVerify_ElasticSearchCarProvider'.
> com.fasterxml.jackson.core.JsonParseException: Invalid UTF-8 middle byte 0x22
 at [Source: D:\programs\JenkinsSlave\workspace\Trade API Tests\TradeApiTests\target\pacts\OneAutoGate-Elastic Industry Search.json; line: 49, column: 44]

* Try:
Run with --stacktrace option to get the stack trace. Run with --info or --debug option to get more log output.

Looks like sometimes the consumer test is generating some non-UTF8 chars.

Thanks.

Chuan

[Ronald Holshausen]

If you open the pact file in a text editor, what do you see at line line: 49, column: 44?

[Chuan Chuan Law]

I re-ran the build and the file got over written.

When I get this again, I shall post more information here.

Thanks Ronald.

Cheers,

Chuan

[Chuan Chuan Law]

Hi Ronald,

I just ran the test locally and is able to re-produce the issue.

The generated pact file contains weird character like this:

        }, {

          "url" : "w?",

          "photoAngle" : "w�?"

        } ]

      },

Thanks.

Chuan

On Tuesday, 23 June 2015 17:08:45 UTC+10, Ronald Holshausen wrote:

[Ronald Holshausen]

Ok, the regular expression are not being escaped correctly. This is a bug in the consumer library. You can use strings to get your tests working.

E.g.,

        withBody{

                ComplianceDate{

                    Month (integer())

                    Year (integer())

                }

                BuildDate{}

                 ListingType regexp('\\w+')

                Odometer integer()

                StockNumber regexp('\\w+')

[Ronald Holshausen]

Actually, it is the generated example values that are not UTF-8. You can give a second parameter for the example and this should fix your problem:

        withBody{

                ComplianceDate{

                    Month (integer())

                    Year (integer())

                }

                BuildDate{}

                 ListingType regexp('\\w+', 'ABCDEF')

                Odometer integer()

                StockNumber regexp('\\w+', '1234567')

[Chuan Chuan Law]

Hi Ronald,

I tried both 

   withBody {

                "spotId" (integer())

                "redbookCode" ('\\w+')

and

   withBody {

                "spotId" (integer())

                "redbookCode" ('\\w+', 'ABCDEF')

produces:

 "body" : {

        "spotId" : 1978289031,

        "redbookCode" : "\\w+",

It's generating "\\w+" in the Json file instead of producing a regex.

Thanks.

Chuan

[Ronald Holshausen]

You need to use  redbookCode regexp('\\w+', 'ABCDEF')