Website of the functional logic programming language Curry

Installation

1. Install stack
   • Stack is included in the haskell platform: https://www.haskell.org/platform/
   • Stack 1.9.3 should work
2. Install hakyll using stack: stack install hakyll
   • might require system packages gmp-devel or libgmp-dev to be installed for some dependencies to compile under linux

Build

1. Build the Haskell executable: stack build
   • This should install the required ghc version as well as all required dependencies, as configure in the curry-lang-org.cabal and stack.yaml
   • currently using stack resolver lts-14.15 which corresponds to ghc-8.6.5
2. Build the site using the compiled executable: stack exec site build
   • This generates the site into the _site subdirectory
3. (Optional) Check for broken links: stack exec site check

Development

To view the website simply build the project and start the built-in hakyll web server at the root directory of this repository with the following command.

stack build
stack exec site watch

Now open the browser of your choice and navigate to localhost:8000.

When any changes outside of the Haskell code are saved, all dependent pages will be rebuild automatically and be seen in the browser after a refresh.

After changes to the Haskell code more manual intervention is necessary. First stop the running stack exec site watch invocation. Then run the following commands.

stack build
stack exec site rebuild
stack exec site watch

It is also advisable to run stack exec site check to check for broken links.

Structure

Folder Structure

. project root
|
+ .gitignore
+ .gitlab-ci.yml                 YAML file containing the Gitlab CI/CD configuration
+ curry-lang-org.cabal           the cabal project file
+ index.html                     contains index.html for /
+ README.md                      this readme document
+ site.hs                        the haskell source defining the routes and resource processing
+ stack.yaml                     the stack project file 
|                                  
+-- assets                         
|   |                            
|   +-- js                       contains .js files
|   +-- css                      contains .css files
|   +-- img                      contains images files
|   +-- files                    contains other asset files
|                                  
+-- downloads                    contains the index.html for /downloads
+-- imprint                      contains the index.md for /imprint
+-- data                         contains files not directly used as routes or templates
|   |
|   +-- learn_more               contains the files that are used to generate the Features 
|   |   |                        and Ecosystem sections on the landingpage
|   |   |
|   |   +-- *_desc.html          one *_desc.html each defining the title 
|   |   |                        and highlight color for each section
|   |   |
|   |   +--  ecosystem/features  one folder containing the definitions for 
|   |                            the items of the corresponding section
|   |
|   +-- link_groups              contains the definitions for the generated header/footer links 
|   |   |
|   |   +-- footer               contains a definition for each footer group 
|   |   |                        and a folder each for the link definitions of that group
|   |   |
|   |   +-- header               contains a definition for each header link
|   |
|   +-- syntax_definition        contains the modified language definitions used by pandoc 
|   |  |
|   |  +-- curry.xml             modified version of the default kde curry language syntax definition 
|   |                            used for syntax highlighting
|   |
|   +-- versions                 contains one *_versions.html file 
|       |                        and folder for each table on the downloads page
|       |
|       +-- *_version.html       description for a table on the downloads page
|       +-- kics2/pakcs          version entries for the download page for each table respectively  
|
+-- privacy                      contains the index.md for /privacy
|                               
+-- templates                    contains various templates
|   |                           
|   +-- code                     contains emmbedded code as markdown templates
|                               
+-- test                         contains the source for experimental and test pages under /test
|                               
+-- tools                        contains the source for the pages under /tools 

Code Structure

The sites routes and used files are specified as part of the haskell program defined in site.hs.

The main function specifies the routes and route dependencies, it is separated from the remaining code by a line of dashes.

In the main function there is a match expression for each set of identically processed resources.

Other

stack exec site clean can be used to clean the _site and _cache directory.

Wiki -> Markdown Conversion

• Headlines

• replace = signes with markdown headline

• Main Headline underline with = or prefix with #

• Other headlines should be at least h2 headlines with ## prefix

• Code Examples

  • add backtick fence
  • decrease indentation
  • annotate language for correct highlighting
    • explicit default is different to no annotation
  • surround with lines containing only ___

• External Link Rewrite Regex

  • Find \[\[\s*(\S*)\s*\|\s*(\S*)\s*\]\]
  • Replace [$2]($1)

• Internal Links

  • manual
  • for links to wiki files (e.g. cass_longer.pdf) the file should be added to the assets and linked from there
  • for links that should still point to the wiki [title](link) should be used

• Images

• manual

  • for wiki image download image and add it to the assets then link from there
  • for external images just link to the external image

• Adding section

  • Start (first) Section

    <section>
    :::md:max-w-3xl

  • Change Section

    :::
    </section>  
    <section>
    :::md:max-w-3xl

  • End (last) Section

    :::
    </section>