hooks.py

import re
from collections import defaultdict

from inflection import camelize
from rest_framework.settings import api_settings

from drf_spectacular.drainage import warn
from drf_spectacular.plumbing import (
    ResolvedComponent, list_hash, load_enum_name_overrides, safe_ref,
)
from drf_spectacular.settings import spectacular_settings


def postprocess_schema_enums(result, generator, **kwargs):
    """
    simple replacement of Enum/Choices that globally share the same name and have
    the same choices. Aids client generation to not generate a separate enum for
    every occurrence. only takes effect when replacement is guaranteed to be correct.
    """

    def iter_prop_containers(schema, component_name=None):
        if not component_name:
            for component_name, schema in schema.items():
                if spectacular_settings.COMPONENT_SPLIT_PATCH:
                    component_name = re.sub('^Patched(.+)', r'\1', component_name)
                if spectacular_settings.COMPONENT_SPLIT_REQUEST:
                    component_name = re.sub('(.+)Request$', r'\1', component_name)
                yield from iter_prop_containers(schema, component_name)
        elif isinstance(schema, list):
            for item in schema:
                yield from iter_prop_containers(item, component_name)
        elif isinstance(schema, dict):
            if schema.get('properties'):
                yield component_name, schema['properties']
            yield from iter_prop_containers(schema.get('oneOf', []), component_name)
            yield from iter_prop_containers(schema.get('allOf', []), component_name)
            yield from iter_prop_containers(schema.get('anyOf', []), component_name)

    def create_enum_component(name, schema):
        component = ResolvedComponent(
            name=name,
            type=ResolvedComponent.SCHEMA,
            schema=schema,
            object=name,
        )
        generator.registry.register_on_missing(component)
        return component

    def extract_hash(schema):
        if 'x-spec-enum-id' in schema:
            # try to use the injected enum hash first as it generated from (name, value) tuples,
            # which prevents collisions on choice sets only differing in labels not values.
            return schema['x-spec-enum-id']
        else:
            # fall back to actual list hashing when we encounter enums not generated by us.
            # remove blank/null entry for hashing. will be reconstructed in the last step
            return list_hash([(i, i) for i in schema['enum'] if i not in ('', None)])

    schemas = result.get('components', {}).get('schemas', {})

    overrides = load_enum_name_overrides()

    prop_hash_mapping = defaultdict(set)
    hash_name_mapping = defaultdict(set)
    # collect all enums, their names and choice sets
    for component_name, props in iter_prop_containers(schemas):
        for prop_name, prop_schema in props.items():
            if prop_schema.get('type') == 'array':
                prop_schema = prop_schema.get('items', {})
            if 'enum' not in prop_schema:
                continue

            prop_enum_cleaned_hash = extract_hash(prop_schema)
            prop_hash_mapping[prop_name].add(prop_enum_cleaned_hash)
            hash_name_mapping[prop_enum_cleaned_hash].add((component_name, prop_name))

    # traverse all enum properties and generate a name for the choice set. naming collisions
    # are resolved and a warning is emitted. giving a choice set multiple names is technically
    # correct but potentially unwanted. also emit a warning there to make the user aware.
    enum_name_mapping = {}
    for prop_name, prop_hash_set in prop_hash_mapping.items():
        for prop_hash in prop_hash_set:
            if prop_hash in overrides:
                enum_name = overrides[prop_hash]
            elif len(prop_hash_set) == 1:
                # prop_name has been used exclusively for one choice set (best case)
                enum_name = f'{camelize(prop_name)}Enum'
            elif len(hash_name_mapping[prop_hash]) == 1:
                # prop_name has multiple choice sets, but each one limited to one component only
                component_name, _ = next(iter(hash_name_mapping[prop_hash]))
                enum_name = f'{camelize(component_name)}{camelize(prop_name)}Enum'
            else:
                enum_name = f'{camelize(prop_name)}{prop_hash[:3].capitalize()}Enum'
                warn(
                    f'enum naming encountered a non-optimally resolvable collision for fields '
                    f'named "{prop_name}". The same name has been used for multiple choice sets '
                    f'in multiple components. The collision was resolved with "{enum_name}". '
                    f'add an entry to ENUM_NAME_OVERRIDES to fix the naming.'
                )
            if enum_name_mapping.get(prop_hash, enum_name) != enum_name:
                warn(
                    f'encountered multiple names for the same choice set ({enum_name}). This '
                    f'may be unwanted even though the generated schema is technically correct. '
                    f'Add an entry to ENUM_NAME_OVERRIDES to fix the naming.'
                )
                del enum_name_mapping[prop_hash]
            else:
                enum_name_mapping[prop_hash] = enum_name
            enum_name_mapping[(prop_hash, prop_name)] = enum_name

    # replace all enum occurrences with a enum schema component. cut out the
    # enum, replace it with a reference and add a corresponding component.
    for _, props in iter_prop_containers(schemas):
        for prop_name, prop_schema in props.items():
            is_array = prop_schema.get('type') == 'array'
            if is_array:
                prop_schema = prop_schema.get('items', {})

            if 'enum' not in prop_schema:
                continue

            prop_enum_original_list = prop_schema['enum']
            prop_schema['enum'] = [i for i in prop_schema['enum'] if i not in ['', None]]
            prop_hash = extract_hash(prop_schema)
            # when choice sets are reused under multiple names, the generated name cannot be
            # resolved from the hash alone. fall back to prop_name and hash for resolution.
            enum_name = enum_name_mapping.get(prop_hash) or enum_name_mapping[prop_hash, prop_name]

            # split property into remaining property and enum component parts
            enum_schema = {k: v for k, v in prop_schema.items() if k in ['type', 'enum']}
            prop_schema = {k: v for k, v in prop_schema.items() if k not in ['type', 'enum', 'x-spec-enum-id']}

            # separate actual description from name-value tuples
            if spectacular_settings.ENUM_GENERATE_CHOICE_DESCRIPTION:
                if prop_schema.get('description', '').startswith('*'):
                    enum_schema['description'] = prop_schema.pop('description')
                elif '\n\n*' in prop_schema.get('description', ''):
                    _, _, post = prop_schema['description'].partition('\n\n*')
                    enum_schema['description'] = '*' + post

            components = [
                create_enum_component(enum_name, schema=enum_schema)
            ]
            if spectacular_settings.ENUM_ADD_EXPLICIT_BLANK_NULL_CHOICE:
                if '' in prop_enum_original_list:
                    components.append(create_enum_component('BlankEnum', schema={'enum': ['']}))
                if None in prop_enum_original_list:
                    if spectacular_settings.OAS_VERSION.startswith('3.1'):
                        components.append(create_enum_component('NullEnum', schema={'type': 'null'}))
                    else:
                        components.append(create_enum_component('NullEnum', schema={'enum': [None]}))

            # undo OAS 3.1 type list NULL construction as we cover this in a separate component already
            if spectacular_settings.OAS_VERSION.startswith('3.1') and isinstance(enum_schema['type'], list):
                enum_schema['type'] = [t for t in enum_schema['type'] if t != 'null'][0]

            if len(components) == 1:
                prop_schema.update(components[0].ref)
            else:
                prop_schema.update({'oneOf': [c.ref for c in components]})

            if is_array:
                props[prop_name]['items'] = safe_ref(prop_schema)
            else:
                props[prop_name] = safe_ref(prop_schema)

    # sort again with additional components
    result['components'] = generator.registry.build(spectacular_settings.APPEND_COMPONENTS)

    # remove remaining ids that were not part of this hook (operation parameters mainly)
    postprocess_schema_enum_id_removal(result, generator)

    return result


def postprocess_schema_enum_id_removal(result, generator, **kwargs):
    """
    Iterative modifying approach to scanning the whole schema and removing the
    temporary helper ids that allowed us to distinguish similar enums.
    """
    def clean(sub_result):
        if isinstance(sub_result, dict):
            for key in list(sub_result):
                if key == 'x-spec-enum-id':
                    del sub_result['x-spec-enum-id']
                else:
                    clean(sub_result[key])
        elif isinstance(sub_result, (list, tuple)):
            for item in sub_result:
                clean(item)

    clean(result)

    return result


def preprocess_exclude_path_format(endpoints, **kwargs):
    """
        preprocessing hook that filters out {format} suffixed paths, in case
        format_suffix_patterns is used and {format} path params are unwanted.
    """
    format_path = f'{{{api_settings.FORMAT_SUFFIX_KWARG}}}'
    return [
        (path, path_regex, method, callback)
        for path, path_regex, method, callback in endpoints
        if not (path.endswith(format_path) or path.endswith(format_path + '/'))
    ]