[vertesy]

Summary

Adds a few targeted comments (the row-names reset trick in column.2.row.names()), and normalizes two stray bare-string "comments" into real #/# TODO: comments.

Bugs found while reading (not fixed, just flagged with # NOTE: comments)

• read.simple.ssv(): references asTibble, which isn't a parameter of this function -- errors by default (wRownames = TRUE).
• write.simple.xlsx(): its internal assignRownames helper uses df (undefined) instead of its own argument x -- errors whenever has_row_names = FALSE.
• qs.2.table(): broken throughout -- qs:qread (single colon, should be ::), stringi::stri_detect isn't a real function, and base_filename/out_path are used but never defined on most branches. This function doesn't currently work for any input.

Flagging these for your call on whether/how to fix; didn't want to bundle behavior changes into a comments-only PR.

[vertesy]

@claude

Understand this is an annotation request, but fix the following 3 bugs that have been identified:

1. read.simple.ssv(): references asTibble, which isn't a parameter of this function -- errors by default (wRownames = TRUE).
2. write.simple.xlsx(): its internal assignRownames helper uses df (undefined) instead of its own argument x -- errors whenever has_row_names = FALSE.
3. qs.2.table(): broken throughout -- qs:qread (single colon, should be ::), stringi::stri_detect isn't a real function, and base_filename/out_path are used but never defined on most branches. This function doesn't currently work for any input.

[claude]

Claude finished @vertesy's task in 2m 44s —— View job

---

Tasks

• Gather context (read PR diff, locate the 3 flagged functions)
• Fix read.simple.ssv(): remove/fix bad asTibble reference
• Fix write.simple.xlsx(): assignRownames should use x not df
• Fix qs.2.table(): qs::qread, real string-detect function, define base_filename/out_path
• Commit and push
• Final summary

Summary

Fixed all 3 bugs you flagged, on top of the annotation-only diff, in commit 1ef3f08:

1. read.simple.ssv() (R/ReadWriter.R): added the missing asTibble = FALSE parameter to the function signature (and its @param doc) — matching how read.simple.tsv()/read.simple.csv() already declare it. Previously !asTibble referenced an undefined object whenever wRownames = TRUE (the default), so the function always errored.

2. write.simple.xlsx(): the internal assignRownames <- function(x) column.2.row.names(df, ...) now correctly uses its own argument, column.2.row.names(x, ...), instead of the undefined df. Previously this errored whenever has_row_names = FALSE.

3. qs.2.table(): rewrote the broken parts so the function actually runs:

   • qs:qread → qs::qread
   • stringi::stri_detect(..., regex=) (not a real function) → stringi::stri_detect_regex(..., pattern=)
   • base_filename is now derived from path (strip the .qs suffix), and out_path is set consistently on every branch (tsv/csv/csv2/excel) instead of being left undefined on most of them
   • Also removed a redundant/dead duplicate excel branch that existed before the if/else if chain
   • Since stringi is now genuinely called, added it to DESCRIPTION (Imports) and NAMESPACE (importFrom(stringi, stri_detect_regex))

Not able to verify: R/Rscript isn't available in this sandbox, so I could not run R CMD check, devtools::load_all(), or the roxygen regeneration for NAMESPACE — I hand-edited NAMESPACE/DESCRIPTION to match what roxygen2::roxygenise() would produce for the new @importFrom stringi stri_detect_regex tag. Please run devtools::document() / R CMD check locally to confirm before merging.

The original comments-only annotation work from this PR is untouched; these are additive fixes layered on top per your explicit request in the trigger comment.