annif.rest.list_projects() - Code Metrics - NatLibFi/Annif - Measure and Improve Code Quality continuously with Scrutinizer

annif.rest.list_projects() A
last analyzed 2025-08-15 15:10 UTC

↳ Parent: annif.rest

Complexity

Conditions

Size

Total Lines	10
Code Lines	6

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	1
eloc	6
nop	0
dl	0
loc	10
rs	10
c	0
b	0
f	0

"""Definitions for REST API operations. These are wired via Connexion to
methods defined in the OpenAPI specification."""

from __future__ import annotations

import importlib
from typing import TYPE_CHECKING, Any

import connexion

import annif.registry
import annif.simplemma_util
from annif.corpus import Document, DocumentList, SubjectSet
from annif.exception import AnnifException, NotEnabledException
from annif.project import Access

if TYPE_CHECKING:
    from connexion.lifecycle import ConnexionResponse

    from annif.corpus.subject import SubjectIndex
    from annif.suggestion import SubjectSuggestion, SuggestionResults


def project_not_found_error(project_id: str) -> ConnexionResponse:
    """return a Connexion error object when a project is not found"""

    return connexion.problem(
        status=404,
        title="Project not found",
        detail="Project '{}' not found".format(project_id),
    )


def learning_not_enabled_error(project_id) -> ConnexionResponse:
    """return a Connexion error object when a project is not configured for learning"""

    return connexion.problem(
        status=403,
        title="Learning not allowed",
        detail=f"Project '{project_id}' is not configured to allow learning via API",
    )


def server_error(
    err: AnnifException,
) -> ConnexionResponse:
    """return a Connexion error object when there is a server error (project
    or backend problem)"""

    return connexion.problem(
        status=503, title="Service unavailable", detail=err.format_message()
    )


def show_info() -> tuple:
    """return version of annif and a title for the api according to OpenAPI spec"""

    result = {"title": "Annif REST API", "version": importlib.metadata.version("annif")}
    return result, 200, {"Content-Type": "application/json"}


def language_not_supported_error(lang: str) -> ConnexionResponse:
    """return a Connexion error object when attempting to use unsupported language"""

    return connexion.problem(
        status=400,
        title="Bad Request",
        detail=f'language "{lang}" not supported by vocabulary',
    )


def list_vocabs() -> tuple:
    """return a dict with vocabularies formatted according to OpenAPI spec"""

    result = {
        "vocabs": [
            vocab.dump()
            for vocab in annif.registry.get_vocabs(min_access=Access.public).values()
        ]
    }
    return result, 200, {"Content-Type": "application/json"}


def list_projects() -> tuple:
    """return a dict with projects formatted according to OpenAPI spec"""

    result = {
        "projects": [
            proj.dump()
            for proj in annif.registry.get_projects(min_access=Access.public).values()
        ]
    }
    return result, 200, {"Content-Type": "application/json"}


def show_project(
    project_id: str,
) -> dict | ConnexionResponse:
    """return a single project formatted according to OpenAPI spec"""

    try:
        project = annif.registry.get_project(project_id, min_access=Access.hidden)
    except ValueError:
        return project_not_found_error(project_id)
    return project.dump(), 200, {"Content-Type": "application/json"}


def detect_language(body: dict[str, Any]):
    """return scores for detected languages formatted according to Swagger spec"""

    text = body.get("text")
    languages = body.get("languages")

    try:
        proportions = annif.simplemma_util.detect_language(text, tuple(languages))
    except ValueError:
        return connexion.problem(
            status=400,
            title="Bad Request",
            detail="unsupported candidate languages",
        )

    result = {
        "results": [
            {"language": lang if lang != "unk" else None, "score": score}
            for lang, score in proportions.items()
        ]
    }
    return result, 200, {"Content-Type": "application/json"}


def _suggestion_to_dict(
    suggestion: SubjectSuggestion, subject_index: SubjectIndex, language: str
) -> dict[str, str | float | None]:
    subject = subject_index[suggestion.subject_id]
    return {
        "uri": subject.uri,
        "label": subject.labels[language],
        "notation": subject.notation,
        "score": suggestion.score,
    }


def _hit_sets_to_list(
    hit_sets: SuggestionResults, subjects: SubjectIndex, lang: str
) -> list[dict[str, list]]:
    return [
        {"results": [_suggestion_to_dict(hit, subjects, lang) for hit in hits]}
        for hits in hit_sets
    ]


def _is_error(result: list[dict[str, list]] | ConnexionResponse) -> bool:
    return (
        isinstance(result, connexion.lifecycle.ConnexionResponse)
        and result.status_code >= 400
    )


def suggest(
    project_id: str, body: dict[str, Any]
) -> dict[str, list] | ConnexionResponse:
    """suggest subjects for the given text and return a dict with results
    formatted according to OpenAPI spec"""

    parameters = dict(
        (key, body[key]) for key in ["language", "limit", "threshold"] if key in body
    )
    metadata = {
        key[len("metadata_") :]: value
        for key, value in body.items()
        if key.startswith("metadata_")
    }
    documents = [{"text": body["text"], "metadata": metadata}]
    result = _suggest(project_id, documents, parameters)

    if _is_error(result):
        return result
    return result[0], 200, {"Content-Type": "application/json"}


def suggest_batch(
    project_id: str,
    body: dict[str, list],
    **query_parameters,
) -> list[dict[str, Any]] | ConnexionResponse:
    """suggest subjects for the given documents and return a list of dicts with results
    formatted according to OpenAPI spec"""

    documents = body["documents"]
    result = _suggest(project_id, documents, query_parameters)

    if _is_error(result):
        return result
    for document_results, document in zip(result, documents):
        document_results["document_id"] = document.get("document_id")
    return result, 200, {"Content-Type": "application/json"}


def _suggest(
    project_id: str,
    documents: list[dict[str, str]],
    parameters: dict[str, Any],
) -> list[dict[str, list]] | ConnexionResponse:
    corpus = _documents_to_corpus(documents, subject_index=None)
    try:
        project = annif.registry.get_project(project_id, min_access=Access.hidden)
    except ValueError:
        return project_not_found_error(project_id)

    try:
        lang = parameters.get("language") or project.vocab_lang
    except AnnifException as err:
        return server_error(err)

    if lang not in project.vocab.languages:
        return language_not_supported_error(lang)

    limit = parameters.get("limit", 10)
    threshold = parameters.get("threshold", 0.0)

    try:
        hit_sets = project.suggest_corpus(corpus).filter(limit, threshold)
    except AnnifException as err:
        return server_error(err)

    return _hit_sets_to_list(hit_sets, project.subjects, lang)


def _documents_to_corpus(
    documents: list[dict[str, Any]],
    subject_index: SubjectIndex | None,
) -> annif.corpus.document.DocumentList:
    if subject_index is not None:
        corpus = [
            Document(
                text=d["text"],
                subject_set=SubjectSet(
                    [subject_index.by_uri(subj["uri"]) for subj in d["subjects"]]
                ),
                metadata=d.get("metadata", {}),
            )
            for d in documents
            if "text" in d and "subjects" in d
        ]
    else:
        corpus = [
            Document(text=d["text"], subject_set=None, metadata=d.get("metadata", {}))
            for d in documents
            if "text" in d
        ]
    return DocumentList(corpus)


def learn(
    project_id: str,
    body: list[dict[str, Any]],
) -> ConnexionResponse | tuple[None, int]:
    """learn from documents and return an empty 204 response if succesful"""

    try:
        project = annif.registry.get_project(project_id, min_access=Access.hidden)
    except ValueError:
        return project_not_found_error(project_id)

    try:
        corpus = _documents_to_corpus(body, project.subjects)
        project.learn(corpus)
    except NotEnabledException:
        return learning_not_enabled_error(project_id)
    except AnnifException as err:
        return server_error(err)

    return None, 204, {"Content-Type": "application/json"}


1			"""Definitions for REST API operations. These are wired via Connexion to
2			methods defined in the OpenAPI specification."""
3
4			from __future__ import annotations
5
6			import importlib
7			from typing import TYPE_CHECKING, Any
8
9			import connexion
10
11			import annif.registry
12			import annif.simplemma_util
13			from annif.corpus import Document, DocumentList, SubjectSet
14			from annif.exception import AnnifException, NotEnabledException
15			from annif.project import Access
16
17			if TYPE_CHECKING:
18			from connexion.lifecycle import ConnexionResponse
19
20			from annif.corpus.subject import SubjectIndex
21			from annif.suggestion import SubjectSuggestion, SuggestionResults
22
23
24			def project_not_found_error(project_id: str) -> ConnexionResponse:
25			"""return a Connexion error object when a project is not found"""
26
27			return connexion.problem(
28			status=404,
29			title="Project not found",
30			detail="Project '{}' not found".format(project_id),
31			)
32
33
34			def learning_not_enabled_error(project_id) -> ConnexionResponse:
35			"""return a Connexion error object when a project is not configured for learning"""
36
37			return connexion.problem(
38			status=403,
39			title="Learning not allowed",
40			detail=f"Project '{project_id}' is not configured to allow learning via API",
41			)
42
43
44			def server_error(
45			err: AnnifException,
46			) -> ConnexionResponse:
47			"""return a Connexion error object when there is a server error (project
48			or backend problem)"""
49
50			return connexion.problem(
51			status=503, title="Service unavailable", detail=err.format_message()
52			)
53
54
55			def show_info() -> tuple:
56			"""return version of annif and a title for the api according to OpenAPI spec"""
57
58			result = {"title": "Annif REST API", "version": importlib.metadata.version("annif")}
59			return result, 200, {"Content-Type": "application/json"}
60
61
62			def language_not_supported_error(lang: str) -> ConnexionResponse:
63			"""return a Connexion error object when attempting to use unsupported language"""
64
65			return connexion.problem(
66			status=400,
67			title="Bad Request",
68			detail=f'language "{lang}" not supported by vocabulary',
69			)
70
71
72			def list_vocabs() -> tuple:
73			"""return a dict with vocabularies formatted according to OpenAPI spec"""
74
75			result = {
76			"vocabs": [
77			vocab.dump()
78			for vocab in annif.registry.get_vocabs(min_access=Access.public).values()
79			]
80			}
81			return result, 200, {"Content-Type": "application/json"}
82
83
84			def list_projects() -> tuple:
85			"""return a dict with projects formatted according to OpenAPI spec"""
86
87			result = {
88			"projects": [
89			proj.dump()
90			for proj in annif.registry.get_projects(min_access=Access.public).values()
91			]
92			}
93			return result, 200, {"Content-Type": "application/json"}
94
95
96			def show_project(
97			project_id: str,
98			) -> dict \| ConnexionResponse:
99			"""return a single project formatted according to OpenAPI spec"""
100
101			try:
102			project = annif.registry.get_project(project_id, min_access=Access.hidden)
103			except ValueError:
104			return project_not_found_error(project_id)
105			return project.dump(), 200, {"Content-Type": "application/json"}
106
107
108			def detect_language(body: dict[str, Any]):
109			"""return scores for detected languages formatted according to Swagger spec"""
110
111			text = body.get("text")
112			languages = body.get("languages")
113
114			try:
115			proportions = annif.simplemma_util.detect_language(text, tuple(languages))
116			except ValueError:
117			return connexion.problem(
118			status=400,
119			title="Bad Request",
120			detail="unsupported candidate languages",
121			)
122
123			result = {
124			"results": [
125			{"language": lang if lang != "unk" else None, "score": score}
126			for lang, score in proportions.items()
127			]
128			}
129			return result, 200, {"Content-Type": "application/json"}
130
131
132			def _suggestion_to_dict(
133			suggestion: SubjectSuggestion, subject_index: SubjectIndex, language: str
134			) -> dict[str, str \| float \| None]:
135			subject = subject_index[suggestion.subject_id]
136			return {
137			"uri": subject.uri,
138			"label": subject.labels[language],
139			"notation": subject.notation,
140			"score": suggestion.score,
141			}
142
143
144			def _hit_sets_to_list(
145			hit_sets: SuggestionResults, subjects: SubjectIndex, lang: str
146			) -> list[dict[str, list]]:
147			return [
148			{"results": [_suggestion_to_dict(hit, subjects, lang) for hit in hits]}
149			for hits in hit_sets
150			]
151
152
153			def _is_error(result: list[dict[str, list]] \| ConnexionResponse) -> bool:
154			return (
155			isinstance(result, connexion.lifecycle.ConnexionResponse)
156			and result.status_code >= 400
157			)
158
159
160			def suggest(
161			project_id: str, body: dict[str, Any]
162			) -> dict[str, list] \| ConnexionResponse:
163			"""suggest subjects for the given text and return a dict with results
164			formatted according to OpenAPI spec"""
165
166			parameters = dict(
167			(key, body[key]) for key in ["language", "limit", "threshold"] if key in body
168			)
169			metadata = {
170			key[len("metadata_") :]: value
171			for key, value in body.items()
172			if key.startswith("metadata_")
173			}
174			documents = [{"text": body["text"], "metadata": metadata}]
175			result = _suggest(project_id, documents, parameters)
176
177			if _is_error(result):
178			return result
179			return result[0], 200, {"Content-Type": "application/json"}
180
181
182			def suggest_batch(
183			project_id: str,
184			body: dict[str, list],
185			**query_parameters,
186			) -> list[dict[str, Any]] \| ConnexionResponse:
187			"""suggest subjects for the given documents and return a list of dicts with results
188			formatted according to OpenAPI spec"""
189
190			documents = body["documents"]
191			result = _suggest(project_id, documents, query_parameters)
192
193			if _is_error(result):
194			return result
195			for document_results, document in zip(result, documents):
196			document_results["document_id"] = document.get("document_id")
197			return result, 200, {"Content-Type": "application/json"}
198
199
200			def _suggest(
201			project_id: str,
202			documents: list[dict[str, str]],
203			parameters: dict[str, Any],
204			) -> list[dict[str, list]] \| ConnexionResponse:
205			corpus = _documents_to_corpus(documents, subject_index=None)
206			try:
207			project = annif.registry.get_project(project_id, min_access=Access.hidden)
208			except ValueError:
209			return project_not_found_error(project_id)
210
211			try:
212			lang = parameters.get("language") or project.vocab_lang
213			except AnnifException as err:
214			return server_error(err)
215
216			if lang not in project.vocab.languages:
217			return language_not_supported_error(lang)
218
219			limit = parameters.get("limit", 10)
220			threshold = parameters.get("threshold", 0.0)
221
222			try:
223			hit_sets = project.suggest_corpus(corpus).filter(limit, threshold)
224			except AnnifException as err:
225			return server_error(err)
226
227			return _hit_sets_to_list(hit_sets, project.subjects, lang)
228
229
230			def _documents_to_corpus(
231			documents: list[dict[str, Any]],
232			subject_index: SubjectIndex \| None,
233			) -> annif.corpus.document.DocumentList:
234			if subject_index is not None:
235			corpus = [
236			Document(
237			text=d["text"],
238			subject_set=SubjectSet(
239			[subject_index.by_uri(subj["uri"]) for subj in d["subjects"]]
240			),
241			metadata=d.get("metadata", {}),
242			)
243			for d in documents
244			if "text" in d and "subjects" in d
245			]
246			else:
247			corpus = [
248			Document(text=d["text"], subject_set=None, metadata=d.get("metadata", {}))
249			for d in documents
250			if "text" in d
251			]
252			return DocumentList(corpus)
253
254
255			def learn(
256			project_id: str,
257			body: list[dict[str, Any]],
258			) -> ConnexionResponse \| tuple[None, int]:
259			"""learn from documents and return an empty 204 response if succesful"""
260
261			try:
262			project = annif.registry.get_project(project_id, min_access=Access.hidden)
263			except ValueError:
264			return project_not_found_error(project_id)
265
266			try:
267			corpus = _documents_to_corpus(body, project.subjects)
268			project.learn(corpus)
269			except NotEnabledException:
270			return learning_not_enabled_error(project_id)
271			except AnnifException as err:
272			return server_error(err)
273
274			return None, 204, {"Content-Type": "application/json"}
275

NatLibFi / Annif

annif.rest.list_projects() A last analyzed 2025-08-15 15:10 UTC

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like

annif.rest.list_projects() A
last analyzed 2025-08-15 15:10 UTC