Merge pull request #462 from uhh-lt/improve-search

Improve search

backend/src/api/endpoints/analysis.py

@@ -5,22 +5,14 @@
 
 from api.dependencies import get_current_user, get_db_session
 from app.core.analysis.analysis_service import AnalysisService
-from app.core.analysis.annotated_images import (
-    AnnotatedImagesColumns,
-    find_annotated_images,
-    find_annotated_images_info,
-)
-from app.core.analysis.annotated_segments import (
-    AnnotatedSegmentsColumns,
-    find_annotated_segments,
-    find_annotated_segments_info,
-)
-from app.core.analysis.word_frequency import (
-    WordFrequencyColumns,
+from app.core.analysis.word_frequency_analysis.word_frequency import (
     word_frequency,
     word_frequency_export,
     word_frequency_info,
 )
+from app.core.analysis.word_frequency_analysis.word_frequency_columns import (
+    WordFrequencyColumns,
+)
 from app.core.authorization.authz_user import AuthzUser
 from app.core.data.doc_type import DocType
 from app.core.data.dto.analysis import (
@@ -32,9 +24,21 @@
     SampledSdocsResults,
     WordFrequencyResult,
 )
-from app.core.filters.columns import ColumnInfo
-from app.core.filters.filtering import Filter
-from app.core.filters.sorting import Sort
+from app.core.search.bbox_search.bbox_search import (
+    find_annotated_images,
+    find_annotated_images_info,
+)
+from app.core.search.bbox_search.bbox_search_columns import BBoxColumns
+from app.core.search.column_info import ColumnInfo
+from app.core.search.filtering import Filter
+from app.core.search.sorting import Sort
+from app.core.search.span_search.span_search import (
+    find_annotated_segments,
+    find_annotated_segments_info,
+)
+from app.core.search.span_search.span_search_columns import (
+    SpanColumns,
+)
 
 router = APIRouter(
     prefix="/analysis", dependencies=[Depends(get_current_user)], tags=["analysis"]
@@ -101,14 +105,14 @@ def annotation_occurrences(
 
 @router.post(
     "/annotated_segments_info",
-    response_model=List[ColumnInfo[AnnotatedSegmentsColumns]],
+    response_model=List[ColumnInfo[SpanColumns]],
     summary="Returns AnnotationSegments Info.",
 )
 def annotated_segments_info(
     *,
     project_id: int,
     authz_user: AuthzUser = Depends(),
-) -> List[ColumnInfo[AnnotatedSegmentsColumns]]:
+) -> List[ColumnInfo[SpanColumns]]:
     authz_user.assert_in_project(project_id)
     return find_annotated_segments_info(
         project_id=project_id,
@@ -124,10 +128,10 @@ def annotated_segments(
     *,
     project_id: int,
     user_id: int,
-    filter: Filter[AnnotatedSegmentsColumns],
+    filter: Filter[SpanColumns],
     page: Optional[int] = None,
     page_size: Optional[int] = None,
-    sorts: List[Sort[AnnotatedSegmentsColumns]],
+    sorts: List[Sort[SpanColumns]],
     authz_user: AuthzUser = Depends(),
 ) -> AnnotatedSegmentResult:
     authz_user.assert_in_project(project_id)
@@ -144,14 +148,14 @@ def annotated_segments(
 
 @router.post(
     "/annotated_images_info",
-    response_model=List[ColumnInfo[AnnotatedImagesColumns]],
+    response_model=List[ColumnInfo[BBoxColumns]],
     summary="Returns AnnotationSegments Info.",
 )
 def annotated_images_info(
     *,
     project_id: int,
     authz_user: AuthzUser = Depends(),
-) -> List[ColumnInfo[AnnotatedImagesColumns]]:
+) -> List[ColumnInfo[BBoxColumns]]:
     authz_user.assert_in_project(project_id)
     return find_annotated_images_info(
         project_id=project_id,
@@ -167,10 +171,10 @@ def annotated_images(
     *,
     project_id: int,
     user_id: int,
-    filter: Filter[AnnotatedImagesColumns],
+    filter: Filter[BBoxColumns],
     page: Optional[int] = None,
     page_size: Optional[int] = None,
-    sorts: List[Sort[AnnotatedImagesColumns]],
+    sorts: List[Sort[BBoxColumns]],
     authz_user: AuthzUser = Depends(),
 ) -> AnnotatedImageResult:
     authz_user.assert_in_project(project_id)

backend/src/api/endpoints/memo.py

@@ -5,7 +5,6 @@
 
 from api.dependencies import get_current_user, get_db_session
 from api.validation import Validate
-from app.core.analysis.memo import MemoColumns, memo_info, memo_search
 from app.core.authorization.authz_user import AuthzUser
 from app.core.data.crud import Crud
 from app.core.data.crud.memo import crud_memo
@@ -19,9 +18,11 @@
 )
 from app.core.data.dto.search import PaginatedElasticSearchDocumentHits
 from app.core.data.orm.util import get_parent_project_id
-from app.core.filters.columns import ColumnInfo
-from app.core.filters.filtering import Filter
-from app.core.filters.sorting import Sort
+from app.core.search.column_info import ColumnInfo
+from app.core.search.filtering import Filter
+from app.core.search.memo_search.memo_search import memo_info, memo_search
+from app.core.search.memo_search.memo_search_columns import MemoColumns
+from app.core.search.sorting import Sort
 
 router = APIRouter(
     prefix="/memo", dependencies=[Depends(get_current_user)], tags=["memo"]

backend/src/api/endpoints/project.py

@@ -32,7 +32,7 @@
 from app.core.data.dto.project_metadata import ProjectMetadataRead
 from app.core.data.dto.user import UserRead
 from app.core.data.orm.source_document import SourceDocumentORM
-from app.core.search.elasticsearch_service import ElasticSearchService
+from app.core.db.elasticsearch_service import ElasticSearchService
 from app.preprocessing.preprocessing_service import PreprocessingService
 
 router = APIRouter(

backend/src/api/endpoints/search.py

@@ -1,43 +1,46 @@
-from typing import List, Optional
+from typing import List, Optional, Union
 
 from fastapi import APIRouter, Depends
 
+import app.core.search.sdoc_search.sdoc_search as sdoc_search
 from api.dependencies import get_current_user
+from app.core.analysis.statistics import (
+    compute_code_statistics,
+    compute_keyword_statistics,
+    compute_tag_statistics,
+)
 from app.core.authorization.authz_user import AuthzUser
 from app.core.data.crud import Crud
 from app.core.data.dto.search import (
     PaginatedElasticSearchDocumentHits,
-    SearchColumns,
     SimSearchImageHit,
-    SimSearchQuery,
     SimSearchSentenceHit,
 )
 from app.core.data.dto.search_stats import KeywordStat, SpanEntityStat, TagStat
-from app.core.filters.columns import ColumnInfo
-from app.core.filters.filtering import Filter
-from app.core.filters.sorting import Sort
-from app.core.search.elasticsearch_service import ElasticSearchService
-from app.core.search.search_service import SearchService
+from app.core.db.elasticsearch_service import ElasticSearchService
+from app.core.search.column_info import ColumnInfo
+from app.core.search.filtering import Filter
+from app.core.search.sdoc_search.sdoc_search_columns import SdocColumns
+from app.core.search.sorting import Sort
 
 router = APIRouter(
     prefix="/search", dependencies=[Depends(get_current_user)], tags=["search"]
 )
 
-ss = SearchService()
 es = ElasticSearchService()
 
 
 @router.post(
     "/sdoc_info",
-    response_model=List[ColumnInfo[SearchColumns]],
+    response_model=List[ColumnInfo[SdocColumns]],
     summary="Returns Search Info.",
 )
 def search_sdocs_info(
     *, project_id: int, authz_user: AuthzUser = Depends()
-) -> List[ColumnInfo[SearchColumns]]:
+) -> List[ColumnInfo[SdocColumns]]:
     authz_user.assert_in_project(project_id)
 
-    return SearchService().search_info(project_id=project_id)
+    return sdoc_search.search_info(project_id=project_id)
 
 
 @router.post(
@@ -50,15 +53,15 @@ def search_sdocs(
     project_id: int,
     search_query: str,
     expert_mode: bool,
-    filter: Filter[SearchColumns],
-    sorts: List[Sort[SearchColumns]],
+    filter: Filter[SdocColumns],
+    sorts: List[Sort[SdocColumns]],
     highlight: bool,
     page_number: Optional[int] = None,
     page_size: Optional[int] = None,
     authz_user: AuthzUser = Depends(),
 ) -> PaginatedElasticSearchDocumentHits:
     authz_user.assert_in_project(project_id)
-    return SearchService().search(
+    return sdoc_search.search(
         search_query=search_query,
         expert_mode=expert_mode,
         highlight=highlight,
@@ -85,12 +88,12 @@ def search_code_stats(
     project_id: int,
     search_query: str,
     expert_mode: bool,
-    filter: Filter[SearchColumns],
-    sorts: List[Sort[SearchColumns]],
+    filter: Filter[SdocColumns],
+    sorts: List[Sort[SdocColumns]],
 ) -> List[SpanEntityStat]:
     # search for relevant sdoc_ids
     authz_user.assert_in_project(project_id)
-    search_result = SearchService().search(
+    search_result = sdoc_search.search(
         project_id=project_id,
         search_query=search_query,
         expert_mode=expert_mode,
@@ -104,9 +107,7 @@ def search_code_stats(
 
     # compute code stats
     authz_user.assert_in_same_project_as(Crud.CODE, code_id)
-    code_stats = SearchService().compute_code_statistics(
-        code_id=code_id, sdoc_ids=set(sdoc_ids)
-    )
+    code_stats = compute_code_statistics(code_id=code_id, sdoc_ids=set(sdoc_ids))
     if sort_by_global:
         code_stats.sort(key=lambda x: x.global_count, reverse=True)
     return code_stats
@@ -131,9 +132,7 @@ def filter_code_stats(
 
     # compute code stats
     authz_user.assert_in_same_project_as(Crud.CODE, code_id)
-    code_stats = SearchService().compute_code_statistics(
-        code_id=code_id, sdoc_ids=set(sdoc_ids)
-    )
+    code_stats = compute_code_statistics(code_id=code_id, sdoc_ids=set(sdoc_ids))
     if sort_by_global:
         code_stats.sort(key=lambda x: x.global_count, reverse=True)
     return code_stats
@@ -154,12 +153,12 @@ def search_keyword_stats(
     # search params
     search_query: str,
     expert_mode: bool,
-    filter: Filter[SearchColumns],
-    sorts: List[Sort[SearchColumns]],
+    filter: Filter[SdocColumns],
+    sorts: List[Sort[SdocColumns]],
 ) -> List[KeywordStat]:
     # search for relevant sdoc_ids
     authz_user.assert_in_project(project_id)
-    search_result = SearchService().search(
+    search_result = sdoc_search.search(
         project_id=project_id,
         search_query=search_query,
         expert_mode=expert_mode,
@@ -172,7 +171,7 @@ def search_keyword_stats(
         return []
 
     # compute keyword stats
-    keyword_stats = SearchService().compute_keyword_statistics(
+    keyword_stats = compute_keyword_statistics(
         proj_id=project_id, sdoc_ids=set(sdoc_ids), top_k=top_k
     )
     if sort_by_global:
@@ -199,7 +198,7 @@ def filter_keyword_stats(
         return []
 
     # compute keyword stats
-    keyword_stats = SearchService().compute_keyword_statistics(
+    keyword_stats = compute_keyword_statistics(
         proj_id=project_id, sdoc_ids=set(sdoc_ids), top_k=top_k
     )
     if sort_by_global:
@@ -221,12 +220,12 @@ def search_tag_stats(
     project_id: int,
     search_query: str,
     expert_mode: bool,
-    filter: Filter[SearchColumns],
-    sorts: List[Sort[SearchColumns]],
+    filter: Filter[SdocColumns],
+    sorts: List[Sort[SdocColumns]],
 ) -> List[TagStat]:
     # search for relevant sdoc_ids
     authz_user.assert_in_project(project_id)
-    search_result = SearchService().search(
+    search_result = sdoc_search.search(
         project_id=project_id,
         search_query=search_query,
         expert_mode=expert_mode,
@@ -239,7 +238,7 @@ def search_tag_stats(
         return []
 
     # compute tag stats
-    tag_stats = SearchService().compute_tag_statistics(sdoc_ids=set(sdoc_ids))
+    tag_stats = compute_tag_statistics(sdoc_ids=set(sdoc_ids))
     if sort_by_global:
         tag_stats.sort(key=lambda x: x.global_count, reverse=True)
     return tag_stats
@@ -262,7 +261,7 @@ def filter_tag_stats(
         return []
 
     # compute tag stats
-    tag_stats = SearchService().compute_tag_statistics(sdoc_ids=set(sdoc_ids))
+    tag_stats = compute_tag_statistics(sdoc_ids=set(sdoc_ids))
     if sort_by_global:
         tag_stats.sort(key=lambda x: x.global_count, reverse=True)
     return tag_stats
@@ -274,11 +273,18 @@ def filter_tag_stats(
     summary="Returns similar sentences according to a textual or visual query.",
 )
 def find_similar_sentences(
-    query: SimSearchQuery, authz_user: AuthzUser = Depends()
+    proj_id: int,
+    query: Union[str, List[str], int],
+    top_k: int,
+    threshold: float,
+    filter: Filter[SdocColumns],
+    authz_user: AuthzUser = Depends(),
 ) -> List[SimSearchSentenceHit]:
-    authz_user.assert_in_project(query.proj_id)
+    authz_user.assert_in_project(proj_id)
 
-    return ss.find_similar_sentences(query=query)
+    return sdoc_search.find_similar_sentences(
+        proj_id=proj_id, query=query, top_k=top_k, threshold=threshold, filter=filter
+    )
 
 
 @router.post(
@@ -287,8 +293,15 @@ def find_similar_sentences(
     summary="Returns similar images according to a textual or visual query.",
 )
 def find_similar_images(
-    query: SimSearchQuery, authz_user: AuthzUser = Depends()
+    proj_id: int,
+    query: Union[str, List[str], int],
+    top_k: int,
+    threshold: float,
+    filter: Filter[SdocColumns],
+    authz_user: AuthzUser = Depends(),
 ) -> List[SimSearchImageHit]:
-    authz_user.assert_in_project(query.proj_id)
+    authz_user.assert_in_project(proj_id)
 
-    return ss.find_similar_images(query=query)
+    return sdoc_search.find_similar_images(
+        proj_id=proj_id, query=query, top_k=top_k, threshold=threshold, filter=filter
+    )

backend/src/api/endpoints/timeline_analysis.py

@@ -4,11 +4,13 @@
 from sqlalchemy.orm import Session
 
 from api.dependencies import get_current_user, get_db_session
-from app.core.analysis.timeline import (
-    TimelineAnalysisColumns,
+from app.core.analysis.timeline_analysis.timeline import (
     timeline_analysis,
     timeline_analysis_info,
 )
+from app.core.analysis.timeline_analysis.timeline_analysis_columns import (
+    TimelineAnalysisColumns,
+)
 from app.core.authorization.authz_user import AuthzUser
 from app.core.data.crud import Crud
 from app.core.data.crud.timeline_analysis import crud_timeline_analysis
@@ -19,8 +21,8 @@
     TimelineAnalysisRead,
     TimelineAnalysisUpdate,
 )
-from app.core.filters.columns import ColumnInfo
-from app.core.filters.filtering import Filter
+from app.core.search.column_info import ColumnInfo
+from app.core.search.filtering import Filter
 
 router = APIRouter(
     prefix="/timelineAnalysis",