How to hook a decode procedure as xexpr postprocessing #11

Héllo,

In the documentation, it's written that it's possible to hook a decode function to an xexpr. I did not find how. The documentation says it needs to be updated.

I want to use to do some typographic corrections, just like it's suggested in the decode documentation (replacing " with « in french for instance, or using dashes to create item lists à la markdown or just create p tags for paragraphs...

I have another problem, I use a template.html file with ◊->html{◊doc} but it embeds a root node, see example at http://hypermove.fr/2014/why-nerfed.html#doc the div#doc child is a root node.

My project is simple, the only "extra" feature I've written will browse a directory for blog posts and include them in the index page cf. index-year

Thanks.

Héllo, In the documentation, it's written that it's possible to hook a decode function to an xexpr. I did not find how. [The documentation says it needs to be updated](http://mbutterick.github.io/pollen/doc/Decode.html). I want to use to do some typographic corrections, just like it's suggested in the decode documentation (replacing `"` with `«` in french for instance, or using dashes to create item lists à la markdown or just create p tags for paragraphs... I have another problem, I use a template.html file with `◊->html{◊doc}` but it embeds a `root` node, see example at http://hypermove.fr/2014/why-nerfed.html#doc the `div#doc` child is a `root` node. My project is simple, the only "extra" feature I've written will browse a directory for blog posts and include them in the index page cf. [`index-year`](https://github.com/amirouche/racket-now/blob/master/amirouche/project-require.rkt#L41) Thanks.

I've updated the documentation with an example of how I use decode.

You should put decode (and other functions you write) in your "project-require.rkt" file (see Quick Tour for an explanation of this file)

Replacing French quotes: yes, it would work like smart-quotes, mais vous devez le traduire en français, bien sûr;

Detecting paragraphs: already part of the decode module, you can use that;

Using dashes to create lists: yes, but I wouldn't put this in the main decode function. I would create a special tag for the list (e.g., amirouche-list) so in the markup it appears as ◊amirouche-list{ ... }. Then, in "project-require.rkt", you add a function (define (amirouche-list . items) ... ). This function will automatically be applied to every ◊amirouche-list{...}. (You can use another decode within this function, if you like).

I've updated [the documentation](http://mbutterick.github.io/pollen/doc/Decode.html) with an example of how I use `decode`. You should put `decode` (and other functions you write) in your "project-require.rkt" file (see [Quick Tour](http://mbutterick.github.io/pollen/doc/quick-tour.html) for an explanation of this file) **Replacing French quotes**: yes, it would work like `smart-quotes`, mais vous devez le traduire en français, bien sûr; **Detecting paragraphs**: already part of the `decode` module, you can use that; **Using dashes to create lists**: yes, but I wouldn't put this in the main `decode` function. I would create a special tag for the list (e.g., `amirouche-list`) so in the markup it appears as `◊amirouche-list{ ... }`. Then, in "project-require.rkt", you add a function `(define (amirouche-list . items) ... )`. This function will automatically be applied to every `◊amirouche-list{...}`. (You can use another `decode` _within_ this function, if you like).

Yes, you are correct that ◊->html{◊doc} will embed a root node. That's just the default name for the main node of the data structure inside doc. You have two choices:

ignore it (it will just be treated as an inline tag by web browsers);
rename it in your main decode function. So rather than this:

(define (root . items)
  (decode (make-txexpr 'root '() items) 
  ... )

Try something like this:

(define (root . items)
  (decode (make-txexpr 'div '((id "doc")) items) 
  ... )

Yes, you are correct that `◊->html{◊doc}` will embed a `root` node. That's just the default name for the main node of the data structure inside `doc`. You have two choices: 1) ignore it (it will just be treated as an inline tag by web browsers); 2) rename it in your main `decode` function. So rather than this: ``` racket (define (root . items) (decode (make-txexpr 'root '() items) ... ) ``` Try something like this: ``` racket (define (root . items) (decode (make-txexpr 'div '((id "doc")) items) ... ) ```

The problem with my suggestion (2) is that it pushes template logic down into your code, which you may not want to do.

A better idea: I just pushed a fix for the ->html function that allows you to change the tag, or splice it into your page by passing keyword arguments. See the documentation for details.

So instead of writing ◊->html{◊doc}
You can rename the tag and id by doing this: ◊->html[#:tag 'div #:attrs '((id "doc"))]{◊doc}
Or strip the outer tag by doing this ◊->html[#:splice #t]{◊doc}

The problem with my suggestion (2) is that it pushes template logic down into your code, which you may not want to do. A better idea: I just pushed a fix for the `->html` function that allows you to change the tag, or splice it into your page by passing keyword arguments. See [the documentation](http://mbutterick.github.io/pollen/doc/Template.html) for details. So instead of writing `◊->html{◊doc}` You can rename the tag and id by doing this: `◊->html[#:tag 'div #:attrs '((id "doc"))]{◊doc}` Or strip the outer tag by doing this `◊->html[#:splice #t]{◊doc}`

I can't get paragraph displaying correctly. Everyline break is translated to a html br. I failed so far to rework pollen sources, can you help?

Can you show me your decode function?

Can you show me your `decode` function?

Here's a simple working example.

sample.html.pm

#lang pollen

Paragraph 1

Paragraph 2

project-require.rkt

#lang racket/base
(require pollen/decode)
(provide root)

(define (root . elements)
  (decode `(root ,@elements) #:txexpr-elements-proc detect-paragraphs))

Output from sample.html.pm in DrRacket:

'(root (p "Paragraph 1") (p "Paragraph 2"))

Here's a simple working example. **sample.html.pm** ``` racket #lang pollen Paragraph 1 Paragraph 2 ``` **project-require.rkt** ``` racket #lang racket/base (require pollen/decode) (provide root) (define (root . elements) (decode `(root ,@elements) #:txexpr-elements-proc detect-paragraphs)) ``` Output from sample.html.pm in DrRacket: ``` racket '(root (p "Paragraph 1") (p "Paragraph 2")) ```

Thanks for the help, maybe I put it to good use later. I created something "similar" to Pollen in python for the time being https://github.com/amirouche/azoufzouf

Labels Milestones

How to hook a decode procedure as xexpr postprocessing #11