CODEWITHSUNDEEP
rshiny
Gopala
Gopala

Reputation: 10483

shiny reactivePoll re-use for multiple database table names

I have a dashboard that relies on 25+ database tables. I want to refresh this data periodically. Shiny reactivePoll that checks row count on each table, and then reads the table when row count increases is the correct thing for my case.

But, I don't want to write 25+ reactivePoll methods, one for each table. Instead, I want to define a 'generic' one and re-use it by simply supplying the database table name.

I can't find a way to do it. Any ideas? Conceptually, here is what I would like:

pollData <- reactivePoll(60000, session,
  checkFunc = function() {
    dbconn <- dbConnect(MySQL(), group = 'mysql')
    query <- dbSendQuery(dbconn,
                         '
                         SELECT
                             MAX(timeCreated) as lastCreated
                         FROM
                             <table name passed as argument>;
                         ')
    lastFeedback <- dbFetch(query, -1)
    dbClearResult(query)
    dbDisconnect(dbconn)

    lastFeedback$lastCreated
  },

  # This function returns the content of the logfile
  valueFunc = function() {
    dbconn <- dbConnect(MySQL(), group = 'mysql')
    query <- dbSendQuery(dbconn,
                         '
                         SELECT
                             *
                         FROM
                         <table name passed as argument>
                         ')
    data <- dbFetch(query, -1)
    dbClearResult(query)
    dbDisconnect(dbconn)

    data
  }
)

server <- function(input, output, session) {
    table1Data <- pollData(<table1 name>)
    table2Data <- pollData(<table2 name>)
    ....
    ....
}

NOTE: I understand that I can move the checkFunc and valueFunc out and make them generic, however, it will still require 25+ reactivePoll() definitions that invoke these other two functions. That is what I am trying to avoid.

Upvotes: 3

Views: 1142

Answers (1)

Scott Henwood
Scott Henwood

Reputation: 229

I found a more general answer at this question: Create a reactive function outside the shiny app

It cites: https://shiny.rstudio.com/articles/modules.html I believe you are trying to build a shiny module. In your case it could look something like follows (a function which calls a reactive function):

my.reactivePoll <- function(session, table_name, timeout = 60000)
{
  my.dataframe <- reactivePoll(timeout, session,
                   checkFunc = function() {
                           lastFeedback <- "testValue"
                           # checking logic here
                           lastFeedback
                         },

                         # This function returns the content of the logfile
                         valueFunc = function() {
                           data <- paste0("New value", table_name)
                           # value logic here
                           data
                         })
  my.dataframe
}

Then make calls in the server function, supplying the table_name 

server <- function(input, output, session) {
  table1Data <- my.reactivePoll(session, "table1")
  table2Data <- my.reactivePoll(session, "table2")
}

Upvotes: 4

Related Questions