qbr

CRAN status CRAN RStudio mirror downloads

The goal of qbr is to make it easy to interact with Quickbase’s JSON API.

Installation

You can install the development version of qbr like so:

 library(devtools)
 install_github("BHII-KSC/qbr")

Functions

API page Function Description
Users get_users Returns a tibble with details of each user in an account
User token clone_token Copy a usertoken
User token deactivate_token Deactivate a usertoken
User token delete_token Delete a usertoken
Apps get_app Get metadata for an app
Apps copy_app Copy an app
Apps delete_app Delete an app
Apps get_app_events Returns a tibble of triggerable events
Tables get_tables Get metadata for all tables in an app
Fields get_fields Get metadata for all fields in a table
Reports get_report Returns a named list of metadata for the specified report
Reports get_reports Returns a tibble of metadata for each report in a table
Reports run_report Returns a tibble containing all data in the specified report
Records delete_records Deletes records matching query conditions
N/A summarize_app Get metadata for an app and its users, tables, and fields

Usage

It is often cumbersome to manually download data from Quickbase to work on it in R. run_report makes it easy to extract report data via the Quickbase JSON API:

library(qbr)

# Get data from a Quickbase report as a tibble
run_report(subdomain = "bhi",
       auth = keyring::key_get("qb_example"),
       table_id = "bn9d8iesz",
       report_id = "7")
#> # A tibble: 6 × 5
#>   `Record ID#` `Date assessed` `Respondent type` Intuitive            Accessible
#>          <int> <chr>           <chr>             <chr>                <chr>     
#> 1            2 2018-12-19      Data analyst      4 - Somewhat agree   4 - Somew…
#> 2            1 2018-12-19      Data analyst      5 - Strongly agree   4 - Somew…
#> 3            3 2018-12-19      Evaluator         1 - Strongly disagr… 2 - Somew…
#> 4            4 2018-12-19      Evaluator         3 - Neutral          4 - Somew…
#> 5           20 2019-12-04      Data analyst      2 - Somewhat disagr… 3 - Neutr…
#> 6            5 2019-11-27      Data analyst      2 - Somewhat disagr… 4 - Somew…

Notice that this function returns a tibble even though the payload from Quickbase is non-tabular JSON. This function extracts data from the Quickbase report (recursively if needed to handle the API’s auto-pagination) and then makes the data tidy using ‘tidyverse’ principles.

If you don’t know the report ID of the report you want to retrieve data from, you can use get_reports to retrieve metadata about all reports in a table:

library(qbr)

get_reports(subdomain = "bhi",
            auth = keyring::key_get("qb_example"),
            table_id = "bn9d8iesz")
#> # A tibble: 6 × 13
#>   description        id    name  type  usedCount usedLast properties.displayOn…¹
#>   <chr>              <chr> <chr> <chr>     <int> <chr>    <lgl>                 
#> 1 ""                 6     Aspi… table        32 2023-08… FALSE                 
#> 2 ""                 5     Find… table        61 2023-08… FALSE                 
#> 3 ""                 1     List… table       124 2023-08… FALSE                 
#> 4 "Sorted by Date M… 2     List… table         0 <NA>     TRUE                  
#> 5 ""                 7     qbr … table        58 2023-08… FALSE                 
#> 6 ""                 8     qbr … table         4 2023-08… FALSE                 
#> # ℹ abbreviated name: ¹​properties.displayOnlyNewOrChangedRecords
#> # ℹ 6 more variables: query.fields <list>, query.filter <chr>,
#> #   query.formulaFields <list>, query.groupBy <list>, query.sortBy <list>,
#> #   query.tableId <chr>

It’s sometimes helpful to manage user tokens programmatically:

library(qbr)

# Clone a user token. The 'clone_name' must be unique. 
token <- clone_token(subdomain = "bhi", 
                     auth = keyring::key_get("qb_example"),
                     clone_name = "My new token",
                     clone_desc = "A token cloned by an R script")

# The token passed to 'auth' is deleted. Token supplied must be active.
delete_token(subdomain = "bhi", auth = token)
#> Token deleted

You can manage apps using the app functions:

library(qbr)

# Copy an app and print the new app's ID
app <- copy_app(subdomain = "bhi",
                auth = keyring::key_get("qb_example"),
                app_id = "bn9d8f78g",
                app_name = "R Testing copy",
                app_desc = "Used to test copy_app() from qbr package",
                keep_data = TRUE)

print(app$id)
#> [1] "btiuqitut"

# Delete the newly created app
delete_app(subdomain = "bhi",
           auth = keyring::key_get("qb_example"),
           app_id = app$id,
           app_name = app$name)
#> $deletedAppId
#> [1] "btiuqitut"

# Get the triggerable events of an app
get_app_events(subdomain = "bhi",
               auth = keyring::key_get("qb_example"),
               app_id = "bn9d8f78g")
#> # A tibble: 2 × 8
#>   type    isActive tableId  name  owner.email owner.id owner.name owner.userName
#>   <chr>   <lgl>    <chr>    <chr> <chr>       <chr>    <chr>      <chr>         
#> 1 webhook TRUE     bp5gg5b… Push… john.erdma… 5962446… John Erdm… jerdmann      
#> 2 webhook TRUE     bp84kms… GET … john.erdma… 5962446… John Erdm… jerdmann

Complex data types

This packages returns some field types slightly differently to a Quickbase report:

Field type Returned data type
Multi-select text Semicolon-separated text
User User’s email address
List-user Semicolon-separated email addresses

Limitations

Some extended ASCII characters (codes 128-255) cannot be properly parsed when converting from JSON at present, such as ’ (&#146). Reports containing text and rich-text fields are the most likely to suffer an error resulting from the use of these extended characters.

This package makes no attempt to convert data types. Number fields in Quickbase may well be interpreted in the resulting tibble as characters. The run_report function does provide a type_suffix argument to allow you to discern the intended datatype more easily.