# Path function appropriate for **URL:** \ **Category:** Questions & Answers\ **Created:** [November 24, 2021, 4:22pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246 "2021-11-24T16:22:51Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![Don](https://avatars.discourse-cdn.com/v4/letter/d/ed655f/32.png) [@Don](https://racket.discourse.group/u/Don)\ **Post date:** [November 24, 2021, 4:22pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/1 "2021-11-24T16:22:51Z")
I can imagine defining file paths something like this: (define file1 (string-append (path-\>string (expand-user-path)) () "file1")) 1. What is recommended for ? 2. My current racket installation has: (string-append (path-\>string (expand-user-path)) ".plt-scheme/4.2.1/collects/DG/" "\<...\>" "file1")) To make use of should I: a) move my dirs of modules here: ? OR b) re-install Racket & place my dirs of modules here: ? ---
**Author:** ![dstorrs](https://avatars.discourse-cdn.com/v4/letter/d/898d66/32.png) [@dstorrs](https://racket.discourse.group/u/dstorrs)\ **Post date:** [November 24, 2021, 5:01pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/2 "2021-11-24T17:01:37Z")
What are you trying to accomplish? If we know the goal we can give better help. Here are a couple possibilities and my suggestions, which someone else might have better answers to: 1. How do I include a module into my code so that I can use its functionality? Check out 'require': [https://docs.racket-lang.org/guide/module-paths.html](https://docs.racket-lang.org/guide/module-paths.html) [https://docs.racket-lang.org/reference/require.html](https://docs.racket-lang.org/reference/require.html) Simple examples, see the docs for caveats, exceptions, and additional features: (require "main.rkt") ; load the file 'main.rkt' which is in the same directory as the current file (require "../utils/hash.rkt") ; load the file 'hash.rkt' which is up one level then down into the 'utils' directory (require struct-plus-plus) ; Load the 'struct-plus-plus' module. Racket will figure out where it is 1. How do I refer to a file from a known location relative to the current file when that file is not a module? ; Assume that current directory is /Users/dstorrs/code #lang racket (require racket/runtime-path) (define here ".") (define-runtime-path this-dir ".") (define without-runtime-path (build-path here 'up "projects" "main.rkt")) (define with-runtime-path (build-path this-dir 'up "projects" "main.rkt")) (displayln (~a "Using define, you get: " (~v without-runtime-path))) (displayln (~a "Using define-runtime-path, you get: " (~v with-runtime-path))) (path-\>string with-runtime-path) (build-path this-dir "../foo/bar/") ; you can build paths in one chunk too Output: Using define, you get: #\ Using define-runtime-path, you get: #\ "/Users/dstorrs/code/./../projects/main.rkt" #\ ---
**Author:** ![Don](https://avatars.discourse-cdn.com/v4/letter/d/ed655f/32.png) [@Don](https://racket.discourse.group/u/Don)\ **Post date:** [November 24, 2021, 6:41pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/3 "2021-11-24T18:41:12Z")
Maybe I just need to create a package with the path: ".plt-scheme/4.2.1/collects/DG/" Then the path definition would be like this: (define file1 (string-append (path-\>string (expand-user-path)) "file1")) ---
**Author:** ![Don](https://avatars.discourse-cdn.com/v4/letter/d/ed655f/32.png) [@Don](https://racket.discourse.group/u/Don)\ **Post date:** [November 24, 2021, 6:46pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/4 "2021-11-24T18:46:11Z")
No wonder there is difficulty in understanding what I have written since Discourse is removing chunks of it. Here is another try: Maybe I just need to create a package with the path: ".plt-scheme/4.2.1/collects/DG/" Then the path definition would be something like this: (define file1 (string-append (path-\>string (expand-user-path)) pkg "file1")) I assume that if I create a package then there is some function, maybe just the name of the package, that will return the path of the package. That is all that I am after. Thanks ---
**Author:** ![jbclements](https://yyz2.discourse-cdn.com/free1/user_avatar/racket.discourse.group/jbclements/32/11_2.png) [@jbclements](https://racket.discourse.group/u/jbclements)\ **Post date:** [November 24, 2021, 7:04pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/5 "2021-11-24T19:04:42Z")
It sounds like you want to refer to files that are in a known location with respect to the source code. So, for instance, I might have a file called "email-processor.rkt" with a file called "email-rules.csv" that's in the same directory. In this case, I would echo Ryan Culpepper's suggestion that you use the runtime-path functions. here's a snippet of code that illustrates what I mean: ``` #lang racket (require racket/runtime-path) (define-runtime-path here ".") (file->string (build-path here "email-rules.csv")) ``` ---
**Author:** ![dstorrs](https://avatars.discourse-cdn.com/v4/letter/d/898d66/32.png) [@dstorrs](https://racket.discourse.group/u/dstorrs)\ **Post date:** [November 24, 2021, 7:34pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/6 "2021-11-24T19:34:29Z")
> [@Don](#): > > No wonder there is difficulty in understanding what I have written since Discourse is removing chunks of it. > Here is another try: > Maybe I just need to create a package with the path: ".plt-scheme/4.2.1/collects/DG/" > Then the path definition would be something like this: > (define file1 (string-append (path-\>string (expand-user-path)) pkg "file1")) > > I assume that if I create a package then there is some function, maybe just the name of the package, that will return the path of the package. That is all that I am after. > Thanks Again, it would help if you would **_tell us what you are trying to achieve._** What is the purpose of this code? Are you loading configuration data, calling out to a shell script, or what? If you want to find the path to an installed module, then install the fc package: [raco fc: Finding Collections](https://docs.racket-lang.org/raco-fc/index.html) > $ raco pkg install raco-find-collection Once it's installed, you can do this to find the path of a module: > $ raco fc struct-plus-plus > /Users/dstorrs/projects/struct-plus-plus > > $ raco fc module-i-do-not-have > raco-find-collection: could not find the collection path "module-i-do-not-have" ---
**Author:** ![Don](https://avatars.discourse-cdn.com/v4/letter/d/ed655f/32.png) [@Don](https://racket.discourse.group/u/Don)\ **Post date:** [November 24, 2021, 9:25pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/7 "2021-11-24T21:25:24Z")
## Currently, my system returns these: (require setup/dirs) (find-user-collects-dir) ;=\> #path:/home/don/.racket/8.1/collects ; but I want it to return: #.plt-scheme/4.2.1/collects\> (find-collects-dir) ;=\> #path:/home/don/racket/collects/ ; but I want it to return: #.plt-scheme/4.2.1/collects\> (define-runtime-path here ".") (file-\>string (build-path here "email-rules.csv")) I believe I should use: (collection-file-path "alpha-bool.ss" "DG" "all") ;returns: #path:/home/don/.plt-scheme/4.2.1/collects/DG/all/alpha-bool.ss instead of using: (define-runtime-path here "/home/don/.plt-scheme/4.2.1/collects/DG/all/") (file-\>string (build-path here "email-rules.csv")) However, before I start using collection-file-path, I want to be confident in my ability to manage the collects paths so that I only have one available. Since my system currently has collect paths that I am not using, I'd like to remove them. Am I correct in thinking that the only way to remove them is to uninstall Racket installations that I am not using? I was previously looking for a function that returned only: #path:/home/don/.plt-scheme/4.2.1/collects (get-collects-search-dirs) returns multiple paths associated with multiple installations: '(#path:/home/don/.racket/8.1/collects #path:/home/don/racket/collects/ #path:/home/don/.plt-scheme/4.2.1/collects) I'd prefer that this or a similar function returned the one path: '(#path:/home/don/.plt-scheme/4.2.1/collects) Does this mean that I'd have to uninstall racket installations that are associated with the other paths? Then (get-collects-search-dirs) would not list the unwanted collects paths? (I have never made use of the collects paths that I want to eliminate.) I am using racket 8.1 so the question arises: How do I configure racket 8.1 to use a collects path that is not the default: #path:/home/don/.racket/8.1/collects ---
**Author:** ![dstorrs](https://avatars.discourse-cdn.com/v4/letter/d/898d66/32.png) [@dstorrs](https://racket.discourse.group/u/dstorrs)\ **Post date:** [November 24, 2021, 10:30pm UTC](https://racket.discourse.group/t/path-function-appropriate-for/246/8 "2021-11-24T22:30:54Z")
Wow. I have PLT Scheme v4.2.1 on my machine but it's from 2009, which I suppose is the first time I tried to pick a Lisp. You've been doing this a while, eh? It sounds like you want to do two things: 1. Specify where you want collections to load from 2. Eliminate all loading options aside from the one you specify For #1 you can set the environment variable "PLTCOLLECTS" [18.2 Libraries and Collections](https://docs.racket-lang.org/reference/collects.html) For #2 I don't know.