Revisions returned incorrectly

[karthik]

Revisions returned incorrectly

[scheidec]

Upon further investigation, it seems there are actually 15 columns returned from the drop_history function when the is_deleted column is true. The last column, is_deleted is either missing or true. When missing, the observation only has 14 variables. But when creating the datatable in the last line here:

drop_history <- function(path = NULL, dtoken = get_dropbox_token()) {
    rev_url <- "https://api.dropbox.com/1/revisions/auto/"
    assert_that(!is.null(path))
    rev_url <- paste0(rev_url, strip_slashes(path))
    res <- httr::GET(rev_url, , config(token = dtoken))
    httr::stop_for_status(res)
    results <- httr::content(res)
    res_df <- lapply(results, function(x) data.frame(t(unlist(x)), stringsAsFactors = FALSE))
    data.table::rbindlist(res_df, fill = TRUE)
}

it results in a dataframe with 15 columns if some of the observations have is_deleted as TRUE.

Is the expect_equal(ncol(x), 14) just checking to make sure that no files have been deleted?

Note: is_deleted is metadata that will be changing in v2:
https://www.dropbox.com/developers/reference/migration-guide#metadata-mapping