Preprocessor in Python?

[flodolo]

I'm trying to figure out how to write a preprocessor in Python, given that it's listed in the example, but I can't even get started.

1. A simple script that read stdin and write to stdout without any changes

import sys

sys.stdout.writelines(sys.stdin)
sys.stdout.flush()

Results in

2019-07-08 10:35:25 [ERROR] (mdbook::utils): Error: Unable to parse the preprocessed book
2019-07-08 10:35:25 [ERROR] (mdbook::utils): 	Caused By: invalid type: map, expected a sequence at line 1 column 1

Clearly I'm missing something obvious.

2. Path to the script. I can call a script living in the same path as book.toml, but relative paths don't seem to work?

[ehuss]

Preprocessors use a particular JSON api to exchange input and output. I believe the input is an array of [PreprocessorContext, Book] and the output is just a serialized Book. This isn't really documented, and I'm not sure if that is intentional.

[flodolo]

import json
import sys

# Read book data from stdin
book_data = json.load(sys.stdin)

This fails with mdbook build(json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)). But, if I write to file the content of stdin once, then pipe it into this script, the JSON is correctly interpreted. And I have no clue why it would happen.

[ehuss]

A minimal preprocessor in Python would look like this:

#!/usr/bin/env python

import json
import sys

if len(sys.argv) > 1:
    if sys.argv[1] == 'supports':
        # sys.argv[2] is the renderer name
        sys.exit(0)

context, book = json.load(sys.stdin)
json.dump(book, sys.stdout)

[henryiii]

This was exactly what I was looking for. The docs does not really make it clear what the output should be. I've implemented a preprocessor in Python that sets up a tabbed interface for code or other content here.

[winternet-studio]

I was also confused by this - first that I only had to return the second array entry (which I now do see that the documentation states!) - but more importantly this thing about calling the script twice, which isn't specifically mentioned (only if you sit and think long and hard about the example code).

Here is my renderer in PHP I finally got working:

<?php
if (@$argv[1] == 'supports') {
	// name of renderer is in $argv[2]
	exit(0);  //currently support all renderers
}

$content = '';
while (false !== ($line = fgets(STDIN))) {
	$content .= $line;
}

$data = json_decode($content);
$secondEntry = $data[1];

// Do you transformations to $secondEntry here...

echo json_encode($secondEntry);

[LunarLambda]

import json
import sys

# Read book data from stdin
book_data = json.load(sys.stdin)

This fails with mdbook build(json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)). But, if I write to file the content of stdin once, then pipe it into this script, the JSON is correctly interpreted. And I have no clue why it would happen.

I'm also getting this error with the code outlined in the documentation.

book.toml has

[preprocessor.xnos]
command = "python3 ./xnos.py --"

Removing the -- fixed it. Eh?