shk-readme-generator/createreadmes.py

from pathlib import Path
import json
import itertools
import yaml
import jsonschema
from typing import Any, TypedDict
import requests
try:
    from yachalk import chalk
    yachalk_imported = True
except ModuleNotFoundError:
    yachalk_imported = False

dataset_path = Path('dataset')
output_path = Path('pages')
dataset_info = dataset_path / Path('dataset.json')
token = "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"

def error(msg: str) -> Exception:
    print(chalk.red(msg) if yachalk_imported else "Error: {}".format(msg))
    return Exception(msg)

def warning(msg: str):
    print(chalk.yellow(msg) if yachalk_imported else "Warning: {}".format(msg))

def open_dataset() -> dict[str, Any]:
    with open(dataset_info, 'r') as f:
        return json.load(f)

def save_dataset(dataset: dict[str, Any]):
    with open(dataset_info, 'w') as f:
        json.dump(dataset, f, indent=4)

def get_json(uri: str):
    print(uri)
    resp = requests.get(url=uri, headers={"Authorization": f"Bearer {token}"})
    print(resp)
    if not resp.ok:
        try:
            resp_error = resp.json()['message']
        except Exception:
            resp_error = resp.text
        raise Exception(f"Invalid response: {resp_error}")
    return resp.json()

def get_repo(slug: str):
    return get_json(f"https://api.github.com/repos/{slug}")
    
def get_user(name: str):
    return get_json(f"https://api.github.com/users/{name}")

def get_file(slug: str, path: str):
    return get_json(f"https://api.github.com/repos/{slug}/contents/{path}")
    
def plural(amount: int, name: str, plural: str = 's'):
    return f"{amount} {name}{plural[:amount^1]}"

from typing import TypedDict

class Artifact(TypedDict):
    file: str
    lines: list[int]

class SecurityRule(TypedDict):
    status: str
    argument: str
    artifacts: None | list[Artifact]

rule_schema = yaml.safe_load("""type: object
additionalProperties: no
required:
  - status
  - argument
properties:
  status:
    type: string
    enum:
      - disregarded
      - observed
      - not applicable
      - unknown
  argument:
    type: string
  artifacts:
    type: array
    items:
      type: object
      properties:
        file:
          type: string
        lines:
          type: array
          items:
            type: integer""")

def check_security_rules(security_rules: dict[Any, Any] | None) -> dict[int, SecurityRule]:
    if security_rules is None:
        raise Exception("Security rules file is empty!")
    for n in range(1, 19):
        try:
            rule = security_rules.get(n, None)
            if rule is None: raise jsonschema.ValidationError(f"Rule {n} is not evaluated")
            jsonschema.validate(rule, rule_schema)
            rule: SecurityRule
            if rule["status"] == "unknown":
                warning(f"Rule {n} is still unknown!")
        except jsonschema.ValidationError as e:
            error("Security rule {n}: {msg} at $.{n}.{path}".format(n=n, msg=e.message, path=e.json_path))
            warning("Not checking further rules!")
            break
    return security_rules

update_dataset = False

def get_name(slug: str):
    return slug[slug.find('/')+1:]

def get_tag_slug(tag: str) -> str:
    return tag.lower().replace(' ', '_')

def write_model_readmes(dataset: dict[str, Any]):
    for model_id, info in dataset.items():
        dir = output_path / 'dataset'
        readme = dir / f'{model_id}.md'
        slug: str = info['slug']
        data = info.get('data')
        if not data:
            data = get_repo(slug)
            info['data'] = data
        owner_url = data.get('owner', {}).get('url')
        if not owner_url:
            raise Exception(f'No owner in repo {slug}!')
        owner = info.get('owner')
        if not owner:
            owner = get_json(owner_url)
            info['owner'] = owner
        owner_name = owner.get('name')
        if not owner_name:
            raise Exception(f'No owner name in repo {slug}!')
        stars = data['stargazers_count']
        forks = data['forks']
        owner_slug = owner['login']
        info['stars'] = stars
        info['forks'] = forks
        info['owner_name'] = owner_name
        info['owner_slug'] = owner_slug
        security_rules_file = dataset_path / model_id / 'security_rules.yaml'
        try:
            with open(security_rules_file, 'r') as f:
                security_rules = yaml.safe_load(f)
            security_rules = check_security_rules(security_rules)
        except FileNotFoundError:
            warning("Security rules file not found at {}".format(security_rules_file))
            security_rules = {}
        except Exception as e:
            warning("Security rules file at {} is invalid: {}".format(security_rules_file, e))
            security_rules = {}
        print(f"Writing readme file {readme}")
        dir.mkdir(exist_ok=True)
        with open(readme, 'w', encoding="utf-8") as f:
            f.write(f"""---
title: {slug}
keywords: model TODO
tags: [{', '.join(get_tag_slug(tech) for tech in info['tech'])}]
sidebar: datasetdoc_sidebar
permalink: {model_id}.html
---

## Repository Information

Repository: [GitHub](https://github.com/{slug})

Owner: [{owner_name}](https://github.com/{owner_slug})

The repository has {plural(stars, 'star')} and was forked {plural(forks, 'time')}. The codebase consists of {plural(info['l'], 'line')} of code and makes use of the following technologies:

{chr(10).join(f'- {tech}' for tech in info['tech'])}

## Data Flow Diagram

### Statistics

The Application consists of a total of {plural(info['t'], 'element')}:

Element | Count
-- | --
Services | {info['s']}
External Entities | {info['e']}
Information Flows | {info['i']}
Annotations | {info['a']}
Total Items | {info['t']}

### Diagram

The below diagram is generated from the corresponding [model file](../../dataset/{model_id}/{model_id}.py).

Formats:
- [PlantUML Model](../../dataset/{model_id}/{model_id}/{model_id}.txt)
- [SVG Vector Image](../../dataset/{model_id}/{model_id}/{model_id}.svg)
- [PNG Raster Image](../../dataset/{model_id}/{model_id}/{model_id}.png)

![Data Flow Diagram](../../dataset/{model_id}/{model_id}/{model_id}.svg)""")

def write_root_readme(dataset: dict[str, Any]):
    print(f"Writing main readme file")
    with open('index.md', 'w', encoding="utf-8") as f:
        f.write(f"""---
title: code2DFD Dataset
keywords: sample homepage
tags: [getting_started]
sidebar: datasetdoc_sidebar
permalink: index.html
summary: Dataset of dataflow diagrams of microservice applications.
---

# Dataset of Dataflow Diagrams

This repository contains of {len(dataset)} manually created dataflow diagrams (DFDs) of microservice applications found on GitHub. The dataset is published as an additional contribution to "Automatic Extraction of Security-Rich Dataflow Diagrams for Microservice Applications written in Java" [Simon Schneider, Riccardo Scandariato]. Each folder in the [`dataset`](dataset/) directory contains one DFD in a [CodeableModels](https://github.com/uzdun/CodeableModels)-format that can be executed to generate PNG, SVG and TXT files for the DFD. Each model refers to stereotypes and metaclasses from the [metamodel](microservice_dfds_metamodel.py) which needs to be imported. This repository already contains rendered versions for each model, thus setup and rendering is only necessary once changes to the models are made.

## Models

<div class="datatable-begin"></div>

Name | Source | LoC | Stars | Forks | DFD Items | Technologies
-- | -- | -- | -- | -- | -- | --
{chr(10).join(f"[{info['slug']}]({model_id}.html) | [GitHub](https://github.com/{info['slug']}) | {info['l']} | {info['stars']} | {info['forks']} | {info['t']} | {len(info['tech'])}" for model_id, info in dataset.items())}

<div class="datatable-end"></div>

## DFD Items

Do culpa deserunt est excepteur amet. Non pariatur ea elit ad eiusmod veniam exercitation nulla. Commodo do adipisicing amet et. Voluptate laboris commodo dolor eu mollit ipsum. Amet reprehenderit velit eu culpa amet exercitation. Elit esse ullamco duis mollit quis. Eiusmod qui reprehenderit sunt cupidatat Lorem anim occaecat enim sint eiusmod tempor.

## Use-Cases

Veniam culpa nostrud id laborum deserunt consectetur consectetur voluptate. Sint aute cupidatat velit irure elit laboris anim labore esse labore. Quis ullamco ut consequat amet. Enim sit laboris deserunt veniam duis aliqua irure proident.
""")

def write_tag_readme(dataset: dict[str, Any]):
    tag_dir = output_path / 'tags'
    known_tech = set(tech for model in dataset.values() for tech in model['tech'])
    print(f"Writing tag data file")
    with open('_data/tags.yml', 'r+') as f:
        tags = yaml.safe_load(f)
        tags['allowed-tags'] = list(sorted(set(itertools.chain(tags['allowed-tags'], (get_tag_slug(tech) for tech in known_tech)))))
        f.seek(0)
        yaml.dump(tags, f)
        f.truncate()

    for tech in known_tech:
        slug = get_tag_slug(tech)
        info_file = tag_dir / f'tag_{slug}.md'
        print(f"Writing tag file for {tech}")
        with open(info_file, 'w', encoding="utf-8") as f:
            f.write(f"""---
title: "{tech}"
tagName: {slug}
search: exclude
permalink: tag_{slug}.html
sidebar: datasetdoc_sidebar
folder: tags
---
{{% include taglogic.html %}}

{{% include links.html %}}
""")

def main():
    global known_tags
    dataset = open_dataset()
    write_tag_readme(dataset) 
    write_root_readme(dataset)
    write_model_readmes(dataset)
    if update_dataset:
        save_dataset(dataset)

yaml.dump
if __name__ == '__main__':
    main()