Trimming behavior of the Pollen Preprocessor #180

Hello,

I'm trying to use the Pollen preprocessor to augment a programming language I created. I have realized that the Pollen preprocessor trims leading whitespace of its input file. Is there a way to make the preprocessor skip the trimming?

The problem is that the trimming mixes up the line-positions in the source file which makes interpreting error messages a bit of a hassle.

Hello, I'm trying to use the Pollen preprocessor to augment a programming language I created. I have realized that the Pollen preprocessor trims leading whitespace of its input file. Is there a way to make the preprocessor skip the trimming? The problem is that the trimming mixes up the line-positions in the source file which makes interpreting error messages a bit of a hassle.

Possibly, but are you sure you want #lang pollen/pre and not #lang scribble/text?

Possibly, but are you sure you want `#lang pollen/pre` and not `#lang scribble/text`?

Thanks for the hint, I tried out #lang scribble/text. It also trims the leading new lines XD. While I get results, I had to get rid of all occurrences of th @ symbol in my source file to make scribble/text swallow the input.

So if in doubt, I would stay with Pollen because the lozenge syntax is pretty much exactly what I need since it does not interfere with anything else.

Anyway, I just wanted to know whether there is already something in Pollen to retain the line numbers in the output. I guess if you do trim the leading new-lines, you do it for a reason. I don't intend to break anyone's workflow.

Thanks for the hint, I tried out `#lang scribble/text`. It also trims the leading new lines XD. While I get results, I had to get rid of all occurrences of th `@` symbol in my source file to make `scribble/text` swallow the input. So if in doubt, I would stay with Pollen because the lozenge syntax is pretty much exactly what I need since it does not interfere with anything else. Anyway, I just wanted to know whether there is already something in Pollen to retain the line numbers in the output. I guess if you do trim the leading new-lines, you do it for a reason. I don't intend to break anyone's workflow.

OK. Would it suffice to add a setup value that controls whitespace trimming? Meaning, you’d need to supply a "pollen.rkt" with a setup submodule that overrides the value.

OK. Would it suffice to add a `setup` value that controls whitespace trimming? Meaning, you’d need to supply a `"pollen.rkt"` with a `setup` submodule that overrides the value.

I'm not sure if I truly understand the feature. Would the pollen.rkt file need to be located in a fixed position in my file system or would I supply it to raco along with the source file name?

In the setup submodule, I guess I would have to redefine a parameter. Something like (trim-leading-ws #f)?

I'm not sure if I truly understand the feature. Would the `pollen.rkt` file need to be located in a fixed position in my file system or would I supply it to raco along with the source file name? In the setup submodule, I guess I would have to redefine a parameter. Something like `(trim-leading-ws #f)`?

Right. When pollen runs, it looks in the setup submodule of "pollen.rkt" for configuration values. You would put the "pollen.rkt" in the same folder as the Pollen source files, with something like this:

#lang racket/base

(module setup racket/base
  (provide trim-whitespace?)
  (define trim-whitespace? #f))

Right. When `pollen` runs, it looks in the `setup` submodule of `"pollen.rkt"` for configuration values. You would put the `"pollen.rkt"` in the same folder as the Pollen source files, with something like this: ```racket #lang racket/base (module setup racket/base (provide trim-whitespace?) (define trim-whitespace? #f)) ```

Labels Milestones

Trimming behavior of the Pollen Preprocessor #180