Skip to content

Commit

Permalink
vignette with tmpdirs
Browse files Browse the repository at this point in the history
  • Loading branch information
Andreas Blätte authored and Andreas Blätte committed Jul 10, 2023
1 parent d6b0a3b commit 1d76ddc
Showing 1 changed file with 20 additions and 6 deletions.
26 changes: 20 additions & 6 deletions vignettes/vignette.Rmd
Expand Up @@ -79,25 +79,39 @@ annodata <- duplicates_as_annotation_data(


```{r encode, eval = FALSE}
regdata <- registry_file_parse(corpus = "REUTERS2", registry = registry())
require("RcppCWB")
registry_tmp <- fs::path(tempdir(), "registry_tmp")
data_dir_tmp <- fs::path(tempdir(), "reuters2_tmp")
dir.create(registry_tmp)
dir.create(data_dir_tmp)
corpus_copy(
corpus = "REUTERS2",
registry_dir = system.file(package = "duplicates", "extdata", "cwb", "registry"),
data_dir = system.file(package = "duplicates", "extdata", "cwb", "indexed_corpora", "reuters2"),
registry_dir_new = registry_tmp,
data_dir_new = data_dir_tmp
)
regdata <- registry_file_parse(corpus = "REUTERS2", registry = registry_tmp)
for (s_attr in c("is_duplicate", "duplicates")){
s_attribute_encode(
values = as.character(annodata[[s_attr]]),
data_dir = regdata$home,
data_dir = data_dir_tmp,
s_attribute = s_attr,
corpus = "REUTERS2",
region_matrix = as.matrix(annodata[, c("cpos_left", "cpos_right")]),
method = "R",
registry_dir = registry(),
registry_dir = registry_tmp,
encoding = regdata$properties["charset"],
delete = FALSE,
verbose = TRUE
)
}
# RcppCWB::cl_load_corpus("REUTERS2", registry = polmineR::registry())
RcppCWB::cl_load_corpus("REUTERS2", registry = registry_tmp)
# corpus("REUTERS2", registry_dir = registry()) %>%
# s_attributes()
corpus("REUTERS2", registry_dir = registry_tmp) %>%
s_attributes()
```

0 comments on commit 1d76ddc

Please sign in to comment.