> lexicon::hash_valence_shifters['fucking']
         x    y
1: fucking <NA>
> lexicon::hash_sentiment_jockers_rinker['fuckin']
        x  y
1: fuckin NA
> lexicon::hash_sentiment_jockers_rinker['fucking']
         x  y
1: fucking -1
> lexicon::hash_sentiment_jockers_rinker['fuckin']
        x  y
1: fuckin NA

#' Get Available \pkg{lexicon} Data
#'
#' See available \pkg{lexicon} data a data.frame.
#'
#' @param regex A regex to search for within the data columns.
#' @param \ldots Other arguments passed to \code{grep}.
#' @return Returns a data.frame
#' @export
#' @examples
#' available_data()
#' available_data('hash_')
#' available_data('hash_sentiment')
#' available_data('python')
#' available_data('prof')
#' available_data('English')
#' available_data('Stopword')
available_data <- function(regex = NULL, ...){


    results <- utils::data(package = 'lexicon')[["results"]]
    dat <- stats::setNames(data.frame(results[, 3:4, drop = FALSE],
        stringsAsFactors = FALSE), c("Data", "Description"))

    ns <- getNamespaceExports(loadNamespace('lexicon'))
    ns <- ns[!ns %in% c("available_data", 'grady_pos_feature')]
    dat <- rbind.data.frame(
        dat,
        data.frame(
            Data = ns,
            Description = c('Jockers Sentiment Polarity Table', 'Jockers Sentiment Data Set'),
            stringsAsFactors = FALSE
        )
    )
    dat <- dat[order(dat[['Data']]),]
    row.names(dat) <- NULL

    if (!is.null(regex)){
        locs <- sort(unique(unlist( lapply(dat, function(x){ grep(regex, x, ...) }) )))

        if (length(locs) > 0) {
            dat <- dat[locs,]
        } else {
            warning('`regex` not found, returning all available data')
        }
    }

    dat 

}

grep('[^ -~]', profanity_zac_anger, value = TRUE)
grep('[^ -~]', key_corporate_social_responsibility$regex, value = TRUE)
grep('[^ -~]', key_corporate_social_responsibility$token, value = TRUE)

if (!require("pacman")) install.packages("pacman")
pacman::p_load(readr, dplyr, textshape)


browseURL('https://github.com/zonination/perceptions/blob/master/probly.csv')

'https://raw.githubusercontent.com/zonination/perceptions/master/probly.csv' %>%
    read_csv() %>%
    tidy_list('phrase', 'probability') %>%
    mutate(probability = probability/100) %>%
    tbl_df()

## # A tibble: 782 x 2
##    phrase           probability
##    <chr>                  <dbl>
##  1 Almost Certainly       0.950
##  2 Almost Certainly       0.950
##  3 Almost Certainly       0.950
##  4 Almost Certainly       0.950
##  5 Almost Certainly       0.980
##  6 Almost Certainly       0.950
##  7 Almost Certainly       0.850
##  8 Almost Certainly       0.970
##  9 Almost Certainly       0.950
## 10 Almost Certainly       0.900
## # ... with 772 more rows

sentimentr:::update_polarity_table(hash_sentiment,
    data.frame(
        words = c('spot on', 'on time'),
        polarity = c(1, 1),
        stringsAsFactors = FALSE
    )
)

library(sentimentr)

"i was very unhappy" %>% sentiment()

dats <- c( 
    "crowdflower_deflategate", 
    "crowdflower_products", 
    "course_evaluations", 
    "crowdflower_self_driving_cars", 
    "crowdflower_weather", 
    "hotel_reviews", 
    "kaggle_movie_reviews", 
    "cannon_reviews", 
    "kotzias_reviews_amazon_cells"
) 


cdat <- combine_data(dats[1])


sdat <- get_sentences(cdat)
swears <- profanity(sdat, profanity_list = c( 'shit', 'shit'))

c("he", "her", "hers", "herself", "him", "himself", "his", "i", 
"it", "its", "me", "mine", "my", "myself", "our", "ours", "ourselves", 
"she", "thee", "their", "them", "themselves", "they", "thou", 
"thy", "thyself", "us", "we", "ye", "you", "your", "yours", "yourself",
"we"
)

> lexicon::hash_sentiment_inquirer
Error: 'hash_sentiment_inquirer' is not an exported object from 'namespace:lexicon'

combine_lexicon <- function(lexicons, ask.dupes = FALSE, ...){

    ## check that all lexicons are of the same type (grouped together)
    ## manual mapping?  or by naming convention?

    ## get internal data
    dats <- lapply(lexicons, function(x) eval(parse(text = paste0('lexicon::', x))))

    ## handling for atomic vs tabular
    if (is.atomic(dats)) {

    } else {

        ## for tabular lexicons


        if (ask.dupes){
            ## interactively as for which dupe to keep
        } else {
            ## keep first dupe
        }

        ## handling for unequal nuber of columns
        ## handling for data.table keyed lookups
    }

}