README.md

In AndreaSanchezTapia/sambaR: Killing saudade by translating Brazilian music

sambaR

Download and translate songs from the Genius database and the google translate API. This package was created out of saudade to explain some samba lyrics to our friends in the US, but can be used to translate any lyrics available in the Genius database. Say you like Turkish music but don’t understand 99.7% of the lyrics (not so random example, I love you Kalben).

This is basically package genius meets package googleLanguageR and I thank everybody involved.

The API for google translate must be configured, first in the cloud, then downloading the secret key in JSON format and saving it somewhere, then configuring .Renviron (usually in ~/.Renviron) with

GL_AUTH=path_to_that_json_file.json

The package with autenthicate automatically when loading.

I’ll try to detail this step better in the future but everything is in the googleLanguageR package documentation. The google translate API is paid and charges per character, but there's a 90 day free trial and sambaR saves everything on disk to avoid repeated API calls.

Installation

You can install the development version from GitHub with:

# install.packages("remotes")
remotes::install_github("liibre/sambaR")

Example

There’s only one function in the package, translate_lyrics()

library(sambaR)

## basic example code
hoje <- sambaR::translate_lyrics(
  artist = "caetano veloso",
  song = "é hoje",
  target = "es"
)
#> 
#> ── Column specification ────────────────────────────────────────────────────────
#> cols(
#>   track_title = col_character(),
#>   line = col_double(),
#>   lyric = col_character()
#> )
#> 
#> ── Column specification ────────────────────────────────────────────────────────
#> cols(
#>   translatedText = col_character(),
#>   detectedSourceLanguage = col_character(),
#>   text = col_character()
#> )

The function will return a two-column tibble with the lyrics and the translated lyrics.

hoje
#> # A tibble: 24 x 2
#>    lyric                                translation                           
#>    <chr>                                <chr>                                 
#>  1 A minha alegria atravessou o mar     Mi alegría cruzó el mar               
#>  2 E ancorou na passarela               Y anclado en la pasarela              
#>  3 Fez um desembarque fascinante        Hizo un aterrizaje fascinante         
#>  4 No maior show da terra               En el espectáculo más grande del mundo
#>  5 Será que eu serei o dono dessa festa ¿Seré el dueño de esta fiesta?        
#>  6 Um rei                               Un rey                                
#>  7 No meio de uma gente tão modesta     En medio de gente tan modesta         
#>  8 Eu vim descendo a serra              Bajé de la montaña                    
#>  9 Cheio de euforia para desfilar       Lleno de euforia para desfilar        
#> 10 O mundo inteiro espera               El mundo entero espera                
#> # … with 14 more rows

• If you do not set a target language, sambaR will only return the lyrics.
• If the target language is the same as the source language, it will not call the translation, to avoid charges and an unnecessary API call. Offline language detection if performed via cld2 package
• The search will automatically create a subfolder with the artist’s name and the lyrics, inside destdir (default: home ~). This is not optional for now and intends to avoid repeated queries to both Genius and google translate.
• I should add some error checks in case Genius doesn’t find the lyrics. Bear in mind that Genius doesn’t care about capitalization but it does about spelling but it does not work consistently with some non-ascii characters, so try your searches with different options.

Shiny app

A Shiny app was deployed and is available by running function runExample()

Read more

Genius tutorial: https://josiahparry.com/post/2019-05-08-genius-learnr-tutorial/ googleLanguageR - Analysing language through the Google Cloud Machine Learning APIs: https://ropensci.org/blog/2017/10/03/googlelanguager/ Package cld2: https://github.com/ropensci/cld2

AndreaSanchezTapia/sambaR documentation built on Dec. 5, 2022, 11:41 p.m.