gen.py

import re

def get_imports(source):
    imports = []
    i = 0
    while i < len(source)-1:
        print i
        # Find the next import
        i = source.find("#import", i, len(source))
        if i == -1:
            break
        clip_start = i
        i += + len("#import")
        
        # Start clip
        import_value = ""
        
        # Skip any whitespave
        while source[i] == ' ':
            i += 1

        # Match "
        print i
        assert source[i] == '"'
        
        # Extract value
        i += 1
        import_value = ""
        while source[i] != '"':
            import_value += source[i]
            i += 1
        i += 1

        # End clip
        clip_end = i
        print source[clip_start:clip_end]
        imports.append((clip_start, clip_end, import_value))
    
    return imports

source = ""
with open('src/doc.md') as source_doc:
    source = source_doc.read()

imports = get_imports(source)
offset = 0
for (start, end, value) in imports:
    value = 'src/' + value
    print "Importing: {}".format(value)
    with open(value, 'r') as import_file:
        insert = import_file.read()
        source = source[:start-offset] + insert + source[end-offset:]
        offset += end - start - len(insert)

with open('doc_gen.md', 'w') as doc:
    doc.write(source)

You might not need it, but I added a few extras to the hook to save me typing so much.
As well as adding the file type for highlighting, this adds the filename in bold above the code block, which saves writing it out.

It will also import single functions or subclasses, given their name and modifiers.
There's also #page which adds "<div style="page-break-after: always;">" to add a page break for PDFs.

https://gist.github.com/tpb1908/2e5b2eb2be50cdc95ed258d1dcad2b00