Unexpected column_names processing when incorporated with line skip #178
Comments
|
The behaviour of Please open a new issue with a reprex if you have continuing difficulty after that. |
jennybc
added a commit
that referenced
this issue
Jan 31, 2017
#240) * Load cells at xlsx worksheet ingest; handle skipping and/or blank rows; fixes #224 * Actually these *should* be the same * Make skipping tests more challenging (blank row btwn col names and data, plus another embedded blank row) * Simplify return of 0x0 tibble for completely empty worksheet * Mark cells to start reading from at the time of worksheet construction * Remove vestigial, internal cell printing function * Be quiet about empty worksheet * More tests of nothingness * Inform worksheet about its own name * Tighten up the NEWs bullet for this PR * Make some accessor member functions const * Use skip() to issue note-to-future-self Expedient place to park a summary of what this entire PR does. * Improved parsing of sheet geometry for xlsx. (#240, @jennybc). - Better handling of leading and embedded blank rows and explicit row skipping. (#224, #194, #178, #156, #101) - Worksheets that are completely empty or that contain only column names no longer error, but return a tibble with zero rows. (#222, #144, #65) - Location is inferred for cells that do not declare their location (e.g. xlsx written by JMP). (#163, #102)
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
I would have expected when skipping lines before starting to read a sheet when column_names is set TRUE, the first line read would be the assumed column_names. If it is always expected that column_names appear in first row, then it would be good feature add have mechanism to specify the starting row for read rather than having it always assumed equal first row when case is that the header is not in first row.
The text was updated successfully, but these errors were encountered: