No newline error at the end of the VCF

[elowy01]

Hi,

I am running the vcf-validator and it is considering a VCF as not valid and throwing the following error:

Error: There is no newline at the end of the file.

I had a look to the VCF and it seems to be correct, I think this could be caused by the Boost libraries that I think the validator uses. But I could be wrong.

Could you please help with this?

Thanks,

ernesto

[jmmut]

ok, what OS are you using? Linux, OSX or windows?

Depending on how you are checking that the newline is there, it can be confusing, as some editors are misleading. can you show the last bytes of the file? If you are in linux, something like tail your_file.vcf | xxd

If the newline character is there, you should see the 0a byte like this:

last line in a vcf:
DS562895.1  5128614 ss5009159425    T   C   .   .   AC=1

last line in xxd output:
0000060: 2e09 2e09 4143 3d31 0a                   ....AC=1.

[elowy01]

Hi, I am using Linux

When I do: zcat input.vcf.gz | tail -n 5 | xxd

I get (showing only the last lines): 00d4e0: 0930 7c30 0930 7c30 0930 7c30 0930 7c30 .0|0.0|0.0|0.0|0 000d4f0: 0930 7c30 0930 7c30 0930 7c30 0930 7c30 .0|0.0|0.0|0.0|0 000d500: 0930 7c30 0930 7c30 0930 7c30 0930 7c30 .0|0.0|0.0|0.0|0 000d510: 0930 7c30 0930 7c30 0930 7c30 0a .0|0.0|0.0|0.

So it seems that the new line is there,

Thanks,

e

[jmmut]

well that's strange. We will need to debug more thoroughly. Why do you say it might be because of the boost library?

Also how big is the file? we took care to make the validator work with huge files and lots of samples, but there could still be some bug. Can you tell approximate number of samples and lines? or number of samples and file size if counting the lines would take too long

[srbcheema1]

@elowy01 could you please tell the command you are using to validate the file. And also please tell which version of vcf-validator are you using. Once try it after uncompressing it and then passing to the validator. zcat input.vcf.gz | ./vcf_validator It would be nice if you could provide the results of using uncompressed file on v0.7 of the validator. That information will be helpful for us.

[elowy01]

Hi, I've had problems with other programs using Boost that did not read the last line of a compressed file. The VCF is not very big: 283M. Number of samples: 2698

[elowy01]

On the version used: vcf_validator version 0.8

The command I used is (running the validator directly on the compressed file): vcf_validator_linux -i input.vcf.gz

[elowy01]

Hi, I've tried using zcat input.vcf.gz |./vcf_validator and it works now,

So I guess that the issue I had is fixed by doing this,

Thanks for your help,

e.