Lazy Search in R Packages, Task Views, CRAN, the Web.

RWsearch

Installation

RWsearch stands for « Search in R packages, task views, CRAN and in the Web ». The package is available on CRAN (CRAN/package=RWsearch) and can be installed via:

install.packages("RWsearch")

If you find RWsearch useful and wish to use it everyday (as I do), then modify your R/etc/Rprofile.site file to launch RWsearch at every R start-up. Install also pacman as the two packages share the same syntax and complement each other. For instance:

local({r <- getOption("repos")
       r["CRAN"] <- "https://cloud.r-project.org"
       options(repos=r)
       old <- getOption("defaultPackages")
       options(defaultPackages = c(old, "pacman", "RWsearch"))
      })

Introduction

RWsearch is a management tool that serves several purposes:

1. Provide a simple non-standard evaluation instruction to read and evaluate non-standard content, mainly character vectors.
2. Download the files that list all available packages, archived packages, check_results and task views available on CRAN at a given date and rearrange them in convenient formats. The downloaded files are:
   • crandb_down() => crandb.rda (4.9 Mo),
   • archivedb_down() => CRAN-archive.html (3.3 Mo),
   • checkdb_down() => check_results.rds (7.6 Mo),
   • tvdb_down() => tvdb.rda (31 ko).
3. List the packages that has been added, updated and removed from CRAN between two downloads or two dates.
4. Search for packages that match one or several keywords in any column of the crandb data.frame and by default (with mode or, and, relax) in the Package name, Title, Description, Author and Maintainer columns (can be selected individually).
5. Display the results as a list or as a table in the console or in the browser and save them as txt, md, html, tex or pdf files.
6. In one instruction, download in one directory the whole documentation and the tar.gz files related to one or several packages. This is the perfect tool to read the documentation off-line and study the source code of a package.
7. Plot the graph of the package dependencies in an html page in your browser (one example is the image behind me in the video below).
8. List the parent and children dependencies of one or several packages, either the first level or the full list (termed recursive).
9. Use this information to reinstall or install packages and their dependencies in the right order, from the lowest number to the largest number of dependencies, and avoid mic-mac (very useful).
10. Check the CRAN check_results for one or several package and their (first level or recursive) dependencies.
11. Explore the packages that have been sent to the archives. Download the source code of the latest version or a specific one. Get their last CRAN check_result and the reason why they have been archived.
12. Provide tools for task view maintenance: Detect the packages recently added to or updated in CRAN, check if they match some keywords, check if they are already recorded in a given task views.
13. Use some bookmarks to various websites and search engines. Keywords can be typed directly in the R console and are sent to the search engine, then displayed in your browser. The list of the search engines is expected to grow (with your help).

Compare to other packages (packagefinder, websearchr) or web services (RDocumentation, rdrr) with similar objectives, the search options of RWsearch are more sophisticated and allow for a finer search. RWsearch also addresses a much larger number of web search engines. Non-standard evaluation (or evaluation of non-standard content) makes it user friendly.

Read the vignettes

5 vignettes are currently available (for the above topics):

1. Introduction (1, 2, 3, 4)
2. Print and download the documentation (5, 6)
3. Packages versions and dependencies + install packages (7, 8, 9)
4. Manage the task views (12, 5)
5. Bookmarks and web search engines (13)

RWsearch was presented at useR! 2019 in Toulouse, France.