fix(cloud-init): disable PyYAML line wrapping in scripts

[soyacz]

Long bash commands in write_files were being split by PyYAML's default 80-character limit, causing script execution failures on Azure VMs. Applied width=10000 and default_flow_style=False.

fixes: https://scylladb.atlassian.net/browse/SCT-62

Testing

• - https://jenkins.scylladb.com/view/staging/job/scylla-staging/job/lukasz/job/2025/job/artifacts-azure-image-test/6/

PR pre-checks (self review)

• I added the relevant backport labels
• I didn't leave commented-out/debugging code

Reminders

• Add New configuration option and document them (in sdcm/sct_config.py)
• Add unit tests to cover my changes (under unit-test/ folder)
• Update the Readme/doc folder relevant to this change (if needed)

[soyacz]

@dimakr this one could also mess with Vector installation on Azure

[fruch]

seems like ruamel can convert to Literal Operator a bit better

from ruamel.yaml.scalarstring import walk_tree

data = { ... large nested dict ... }

# This utility finds all multi-line strings and wraps them for you
walk_tree(data)

yaml.dump(data, sys.stdout)

while in pyyaml is global and ugly:

import yaml

def force_literal_dump(data):
    # Define a representer that forces '|' for any string with a newline
    yaml.SafeDumper.org_represent_str = yaml.SafeDumper.represent_str

    def repr_str(dumper, data):
        if '\n' in data:
            return dumper.represent_scalar('tag:yaml.org,2002:str', data, style='|')
        return dumper.represent_scalar('tag:yaml.org,2002:str', data)

    yaml.add_representer(str, repr_str)
    output = yaml.dump(data)
    
    # It's good practice to reset the representer back to default afterwards
    yaml.add_representer(str, yaml.SafeDumper.org_represent_str)
    return output

but that sound like a better way then width=10000
I think going forward we should consider ruamel for anything yaml...