Thanks aurimas for the quick response.

Unfortunately I have no ability to edit the translators.

The reason why I would prefer not to manually delete the unwanted columns is this: Suppose I exported the collection at Time 1, and manually deleted those columns. When the collection expands, and the additional items in the collection are exported at Time 2, and if I want to copy and paste those additional rows into the previous excel file, then I would have to be very careful to delete from the Time 2 file precisely the same columns that were deleted at Time 1. And then Time 3, 4, etc. Too much trouble for me.

On second thought, I think it might be worth looking into a macro that can delete multiple columns at one go ...

Thank you so much aurimas for spelling out the steps for editing the translator. I think I followed all the steps, but I got an error message while trying to export using the customized translator.

I am suspecting that I might have messed up step (5). I am pasting below my edited part, and the original part. If you could take a quick look to see if I made any mistakes, I'd really appreciate it. Thanks a lot! (I also saw that you were one of the creators of the translator. That's so great!!)

=================
MINE:
// Exported columns in order of export
var exportedFields = [
// "Important" metadata
"itemType","publicationYear","creators/author","title",
"publicationTitle",abstractNote",
// Other common fields
"tags/own","tags/automatic",
];

-------
ORIGINAL:
// Exported columns in order of export
var exportedFields = [
// "Important" metadata
"key","itemType","publicationYear","creators/author","title",
"publicationTitle","ISBN","ISSN","DOI","url","abstractNote","date",
"dateAdded","dateModified",
// Other common fields
"accessDate","pages","numPages","issue","volume","numberOfVolumes",
"journalAbbreviation","shortTitle","series","seriesNumber","seriesText",
"seriesTitle","publisher","place","language","rights","type","archive",
"archiveLocation","libraryCatalog","callNumber","extra","notes",
"attachments/path","attachments/url","tags/own","tags/automatic",
// Creators
"creators/editor","creators/seriesEditor","creators/translator",
"creators/contributor","creators/attorneyAgent","creators/bookAuthor",
"creators/castMember","creators/commenter","creators/composer",
"creators/cosponsor","creators/counsel","creators/interviewer",
"creators/producer","creators/recipient","creators/reviewedAuthor",
"creators/scriptwriter","creators/wordsBy","creators/guest",
// Other fields
"number","edition","runningTime","scale","medium","artworkSize",
"filingDate","applicationNumber","assignee","issuingAuthority","country",
"meetingName","conferenceName","court","references","reporter",
"legalStatus","priorityNumbers","programmingLanguage","version","system",
"code","codeNumber","section","session","committee","history",
"legislativeBody"
];

Also missing quote before abstractNote

Hey @yinbin @aurimas @adamsmith, thanks for this informative thread on editing the zotero csv translator.

In addition to the discussion above, I want to remove the headers and make sure that, when pasted, the references start from the first line.

I managed to (1) edit the columns to include (as described above), (2) remove the headers by:

a. Removing writeColumnHeaders in function exportNotes (line 109)

Now I want to make sure it pastes to the first line, but I don't manage to make it happen. I suspect I should change something in the 'doExport' function (lines 105-120). Any suggestion how to proceed?

I created a Gist with an overview of the edits I made for clarification: https://gist.github.com/floriandierickx/b70c26d0300f73b8b9cf0f55d4260323/revisions

----

var exportNotes;
function doExport() {
exportNotes = Zotero.getOption("exportNotes");
// Until we fix UTF-8xBOM export, we'll write the BOM manually
Zotero.write("\uFEFF");
// column header writing disabled
// writeColumnHeaders();
var item, line;
while (item = Zotero.nextItem()) {
if (item.itemType == "note" || item.itemType == "attachment") continue;
line = '';
for (var i=0; i<exportedFields.length; i++) {
line += (i ? fieldDelimiter : recordDelimiter)
+ getValue(item, exportedFields[i]);
}
Zotero.write(line);
}
}

If you know some javascript, this would be hard. It'd be along the lines of

case 'design':
if (item.extra && item.extra.includes("Design:") {
value = item.extra.match(/Design:(.+)/)[1];
}
break;

And then include design in the exported fields.
Without some coding, there's no option other than splitting up Extra in Excel.

Exporting to CSV is occasionally useful but not a major use case. Building a whole GUI around column selection (a feature that would only provide minimal time savings compared to just deleting the columns in Excel) wouldn't be a good use of developer time or GUI space.

If you create a custom translator, you'd give it a separate title and that would be available in the export dialog after restarting Zotero. For a one-off, you can also just edit the existing CSV translator, though those changes would be overwritten after the next update of either Zotero or the CSV translator.

And FWIW, every time someone has mentioned chatGPT in a forum post here, they had been provided completely useless and false information by the LLM. This is true in your case as well. As you can tell by the posts in this thread going almost 10 years back, this has never been possible. Please don't use chatGPT for Zotero support.

"It sounded valid." That is the primary problem with chatGPT, the Bing-chatbot and the like. "They" speak with authority but gather the information from who-knows-where. I'm seeing college essays and reports that were clearly not written by a human with imaginary journals in the reference list. And who but a bot would state that the primary cause of child drownings in residential swimming pools is exposure to acetylene gas?