Fixes Jabref#7660 Unable to download some arXiv links if the "eprint" field is missing

[JavuesZhang]

Fixes #7660

Brief summary

1. Run EprintCleanup on a copy of the entry the ArXiv fetcher is fetching before getting arXiv id from the eprint field;
2. Add two test method. One finds full text with title containing colon and journal, while another finds full text with title containing colon and url.

Problem

When finding full text, this BibTeX reference works:

@Article{booth_bayes-trex_2020,
  author        = {Serena Booth and Yilun Zhou and Ankit Shah and Julie Shah},
  journal       = {arXiv:2002.10248v4 [cs]},
  title         = {Bayes-TrEx: a Bayesian Sampling Approach to Model Transparency by Example},
  year          = {2020},
  month         = dec,
  archiveprefix = {arXiv},
  eprint        = {2002.10248},
  url           = {http://arxiv.org/abs/2002.10248v4},
}

But when eprint field is missing, no full text will be found:

@Article{booth_bayes-trex_2020,
  author        = {Serena Booth and Yilun Zhou and Ankit Shah and Julie Shah},
  journal       = {arXiv:2002.10248v4 [cs]},
  title         = {Bayes-TrEx: a Bayesian Sampling Approach to Model Transparency by Example},
  year          = {2020},
  month         = dec,
  archiveprefix = {arXiv},
  url           = {http://arxiv.org/abs/2002.10248v4},
}

Solution

Since the title contains colon and arXiv uses colon to represent key and value, the title may be recognized mistakenly. So use other fields to get the eprint field to avoid this problem. Thanks to the advice from @tobiasdiez .

Screenshots

After fix:

• Change in CHANGELOG.md described in a way that is understandable for the average user (if applicable)
• Tests created for changes (if applicable)
• Manually tested changed features in running JabRef (always required)
• Screenshots added in PR description (for UI changes)
• Checked documentation: Is the information available and up to date? If not created an issue at https://github.com/JabRef/user-documentation/issues or, even better, submitted a pull request to the documentation repository.

[tobiasdiez]

Wow that was quick 🚀 . Code looks good to me! Thanks for your contribution.

[Siedlerchr]

Why do you clone the entry?

[JavuesZhang]

If we do clean up on the original entry, its fields will be changed but I think we needn't change fields here because it is just a search.

[JavuesZhang]

Wow that was quick 🚀 . Code looks good to me! Thanks for your contribution.

It is your advice that makes me efficient! At the beginning I try to ignore colons in title to solve this problem but it seems not so good.

[Siedlerchr]

Thanks for the quick fix!