DSVRowString and DSVRowAny return undefined for non-existing rows, and DSVParsedArray return undefined for columns of empty input.

[azoson]

Please fill in this template.

• Use a meaningful title for the pull request. Include the name of the package modified.
• Test the change in your own code. (Compile and run.)
• Follow the advice from the readme.
• Avoid common mistakes.
• Run npm run lint package-name (or tsc if no tslint.json is present).

Select one of these and delete the others:

If changing an existing definition:

• Provide a URL to documentation or source code which provides context for the suggested changes: <https://github.com/d3/d3-dsv/blob/master/README.md#dsv_parse>
• Increase the version number in the header if appropriate.
• If you are making substantial changes, consider adding a tslint.json containing { "extends": "dtslint/dt.json" }.

[dt-bot]

types/d3-dsv/index.d.ts

to authors (@tomwanzek @gustavderdrache @borisyankov). Could you review this PR?
👍 or 👎?

[typescript-bot]

@azoson Please fix the failures indicated in the Travis CI log.

[tomwanzek]

@azoson Thanks for the contribution. As it stands the definitions for d3-request have not been officially vetted for use with strictNullChecks. (see #11365).

As you are putting your finger on an open TODO, I would suggest to use this PR as a starting point to address this is on complete sweep. I.e. we should vet the whole definition, make the appropriate changes and activate strictNullChecks in the tsconfig.json for the d3-request definitions.

This may also entail updating some of the tests.

Would you be willing to tackle this jointly as part of this PR? Would be appreciated... 🥇

[tomwanzek]

To correct my mistake: For d3-request in my above comment, read: d3-dsv.

[azoson]

At first, I excuse that my poor English skill may lead misunderstanding and rude sentences.
(Even this sentence may be rude, but I cannot judge myself 😣)

I have the willingness to fix this errors in this PR.
I am very interested in your proposal.
As you said, now my changes cause some errors in the test when strictNullCheck activated.

I will update d3-dsv-test.ts.
Then check if the codes passed test.
Is there other to do?

[tomwanzek]

No worries about your English. All good 😄 Let me have a look at d3-dsv this weekend. I haven't looked at it in a while, so I can't tell you off the top of my head, what changes may be needed to make it viable for strictNullChecks.

I'll provide you input based on what I see. So thanks for staying tuned.

[tomwanzek]

@azoson My apologies for the brief delay. Upon closer look at the definitions for d3-dsv, I thought it would be easier to do a more complete sweep. So I am going to create a new PR as the basis for addressing both the strictNullChecks and complete JSDoc comments.

It seemed easier in this case, as the explanation of "requested/suggested" changes might have been longer and more cumbersome to follow.

I will cross-reference this PR to it, as I submit the request for review/merge. I will @-mention you alongside @gustavderdrache on the new PR for review/discussion.

Feel free to leave this PR open for now, until we align on the new PR.

[azoson]

Do not worry about the delay 😄
I am glad to hear that there is a better way to enable strictNullChecks.
I pushed my changes to fix the test script.
It may help your new PR.
I'm looking forward to your new PR and discussion.

[tomwanzek]

See #21162.

[tomwanzek]

@azoson Can this be closed now in favour of #21162? If you have any further comments/suggestions, you could add them to the new PR before it is merged. (I.p. see my point there about how to handle the columns property.) Cheers, T.

[tomwanzek]

@azoson Let's close this PR with a big thank you as it is superseded by #21162.

[RyanCavanaugh]

@azoson Please address comments from the code reviewers.

[rbuckton]

Closed as it was superseded by #21162