Just a quick follow-up on the previous post about manually annotating random samples. Yesterday I was facing a situation where I had to go and find some broader contexts for concordances fetched from a corpus of Finnish newspaper texts. These corpora don't allow access for more than a sentence, maximum a paragraph at a time, and I needed to look at whole texts. Luckily, I figured out that the city library has access to the electronic archives of many of the newspapers included in the corpus.

So, I had a sample of 60 concordances I needed to get the full text for. The access to the full texts was provided only through the computers physically located in the library -- this meant that I, sadly, didn't have the possibility of working with just R all the time. The easiest solution I came up with was to quickly upload my data set to a Google sheet and then open that sheet in the browser of the PC at the library. Fortunately, there is the nice googlesheets library that makes this easy. All I had to do, was:

devtools::install_github("jennybc/googlesheets")
library(googlesheets)

Then I converted an existing data frame (well, a tibble, actually), to a Google sheet by just:

samp_gsheet <- gs_new("samp_press_fi",input=samp_press_fi)

...where samp_press_fi was the name of my tibble including the samples I wanted to get the contexts for. The command first takes you to Google authorization page, after which you're good to go. The nice thing about googlesheets is, that after I got the contexts I needed, the data was automatically updated in R by just, for instance:

> samp_gsheet %>% gs_read
Accessing worksheet titled 'Sheet1'.
Downloading: 8.2 kB     Parsed with column specification:
cols(
  sent = col_character(),
  sourcetext = col_character(),
  location3 = col_character(),
  group = col_character(),
  context = col_character()
)