Release gtrendsR 1.4.6

[PMassicotte]

Prepare for release:

• Check current CRAN check results
• devtools::check(remote = TRUE, manual = TRUE)
• devtools::check_win_devel()
• rhub::check_for_cran()
• revdepcheck::revdep_check(num_workers = 4)
• Update cran-comments.md
• Polish NEWS

Submit to CRAN:

• usethis::use_version('patch')
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()

[PMassicotte]

This will be a release to fix a CRAN issue due to failing tests.

https://cran.r-project.org/web/checks/check_results_gtrendsR.html

[eddelbuettel]

What drove the breakage?

• Change in tinytest? (Unlikely)
• Change in our code? (Possible)
• Change at Google ?(sadly, most likely)

[PMassicotte]

It was related to a use case where there was no rising data returned by Google. So it was just an unhandled use case in our code.

gtrendsR/R/related_topics.R

Lines 47 to 82 in cb93d9e

	start_top <- which(grepl("TOP", res))
	start_rising <- which(grepl("RISING", res))
	if (length(start_top) == 0 | length(start_rising) == 0) {
	return(NULL) ## No data returned
	}
	top <- read.csv(textConnection(res[start_top:(start_rising - 2)]), row.names = NULL)
	top$subject <- rownames(top)
	rownames(top) <- NULL
	top <- top[, c(2, 1)]
	names(top) <- c("subject", "top")
	top <- reshape(
	top,
	varying = "top",
	v.names = "value",
	direction = "long",
	timevar = "related_topics",
	times = "top"
	)
	rising <- read.csv(textConnection(res[start_rising:length(res)]), row.names = NULL)
	rising$subject <- rownames(rising)
	rownames(rising) <- NULL
	rising <- rising[, c(2, 1)]
	names(rising) <- c("subject", "rising")
	rising <- reshape(
	rising,
	varying = "rising",
	v.names = "value",
	direction = "long",
	timevar = "related_topics",
	times = "rising"
	)

[PMassicotte]

Hi @eddelbuettel

I am facing an issue that I do not understand with tinytest failing on Windows. In this test file, I only have 3 lines of code. If I run it manually, it works, but it fails when I call it from tinytest.

gtrendsR/inst/tinytest/test_encoding_does_not_change.R

Lines 2 to 3 in cb93d9e

	kw <- "赤"
	res <- gtrends(kw)

This works:

kw <- "赤"
res <- gtrends(kw)

This fails:

tinytest::run_test_file("inst/tinytest/test_encoding_does_not_change.R")

Any ideas?

PS. It works fine on any Linux distro.

[eddelbuettel]

Yes. It is spelled W. I. N. D. O. W. S.

It simply doesn't have the utf-8 support real computers have. Skip the test on that platform, I do these same in some repo. (See e.g. anytime for a bazillion tests and skips because working with timezones on different OSs is only for crazy people ....) Off for a run now, poke me later if this persists,

[eddelbuettel]

I think I even have a utf-8 test somewhere that I skip on Windows. Possibly in Rcpp.