mirror of
https://github.com/VectifyAI/PageIndex.git
synced 2026-06-12 19:55:17 +02:00
refactor(pifs): remove dead semantic retrieval code (#33)
This commit is contained in:
parent
d3034fa1b9
commit
0f71da3bc1
8 changed files with 24 additions and 691 deletions
|
|
@ -308,7 +308,7 @@ def test_browse_supports_fixed_size_one_based_pagination_and_metadata_filter(tmp
|
|||
assert filtered["data"][0]["summary"] == "summary for doc_10"
|
||||
|
||||
|
||||
def test_browse_scopes_semantic_search_before_candidate_limit(tmp_path):
|
||||
def test_browse_scopes_channel_candidates_before_candidate_limit(tmp_path):
|
||||
import json
|
||||
|
||||
from pageindex.filesystem import PIFSCommandExecutor, PageIndexFileSystem
|
||||
|
|
@ -738,20 +738,6 @@ def test_broad_recursive_grep_suggests_browse_not_removed_semantic_commands(tmp_
|
|||
assert "semantic-grep" not in rendered
|
||||
|
||||
|
||||
def test_semantic_search_scope_filters_explicit_source_type_facets():
|
||||
from pageindex.filesystem import PageIndexFileSystem
|
||||
|
||||
assert PageIndexFileSystem._semantic_filters_for_scope(
|
||||
{"folder_path": "/source_type=google-drive"}
|
||||
) == {"source_type": "google_drive"}
|
||||
assert PageIndexFileSystem._semantic_filters_for_scope(
|
||||
{"folder_path": "/semantic/source_type=google-drive"}
|
||||
) == {"source_type": "google_drive"}
|
||||
assert PageIndexFileSystem._semantic_filters_for_scope(
|
||||
{"folder_path": "/documents"}
|
||||
) == {}
|
||||
|
||||
|
||||
def test_grep_source_file_requires_terms_on_same_line(tmp_path):
|
||||
from pageindex.filesystem import PIFSCommandExecutor, PageIndexFileSystem
|
||||
|
||||
|
|
@ -904,7 +890,7 @@ def test_existing_summary_projection_index_dimension_mismatch_rejects_retrieval(
|
|||
filesystem.configure_existing_projection_retrieval()
|
||||
|
||||
|
||||
def test_default_semantic_search_uses_summary_projection_when_only_summary_available(tmp_path):
|
||||
def test_browse_semantic_files_uses_summary_projection_when_only_summary_available(tmp_path):
|
||||
from pageindex.filesystem import PageIndexFileSystem
|
||||
from pageindex.filesystem.hybrid_projection import HybridProjectionSearchBackend
|
||||
from pageindex.filesystem.metadata_generation import MetadataGenerationResult
|
||||
|
|
@ -961,9 +947,14 @@ def test_default_semantic_search_uses_summary_projection_when_only_summary_avail
|
|||
},
|
||||
)
|
||||
|
||||
assert filesystem.search("purchase order exposure", semantic=False) == []
|
||||
assert filesystem.search("purchase order exposure") == []
|
||||
|
||||
results = filesystem.search("purchase order exposure", semantic=True)
|
||||
results = filesystem.browse_semantic_files(
|
||||
"/documents",
|
||||
"purchase order exposure",
|
||||
recursive=True,
|
||||
page_size=5,
|
||||
)
|
||||
|
||||
assert [result.external_id for result in results] == ["doc_summary_only"]
|
||||
assert results[0].snippet == "summary_vector rank=1"
|
||||
assert [item["external_id"] for item in results["data"]] == ["doc_summary_only"]
|
||||
assert results["data"][0]["snippet"] == "summary_vector rank=1"
|
||||
|
|
|
|||
|
|
@ -158,13 +158,13 @@ def test_add_configures_semantic_retrieval_in_same_filesystem_instance(tmp_path)
|
|||
filesystem.add_file(source, "/documents")
|
||||
|
||||
assert filesystem.semantic_retrieval_channels() == ("summary",)
|
||||
results = filesystem.search_semantic_channel(
|
||||
"summary",
|
||||
results = filesystem.browse_semantic_files(
|
||||
"/documents",
|
||||
"semantic recall",
|
||||
scope={"folder_path": "/documents", "recursive": True},
|
||||
limit=5,
|
||||
recursive=True,
|
||||
page_size=5,
|
||||
)
|
||||
assert [result.source_path for result in results] == ["documents/semantic.txt"]
|
||||
assert [item["source_path"] for item in results["data"]] == ["documents/semantic.txt"]
|
||||
|
||||
|
||||
def test_add_markdown_builds_pageindex_tree_from_copied_artifact(tmp_path, monkeypatch):
|
||||
|
|
|
|||
|
|
@ -263,7 +263,7 @@ class PIFSAgentStreamTest(unittest.TestCase):
|
|||
):
|
||||
self.assertNotIn(old_command, prompt_surface)
|
||||
|
||||
def test_demo_prompt_uses_browse_strategy_and_not_legacy_semantic_search(self):
|
||||
def test_demo_prompt_uses_browse_strategy_and_not_old_vector_commands(self):
|
||||
demo_prompt = load_demo_agent_prompt()
|
||||
|
||||
self.assertIn("Start with ls or tree", demo_prompt)
|
||||
|
|
|
|||
|
|
@ -46,7 +46,6 @@ def test_descendant_folder_filter_treats_underscore_literally(tmp_path):
|
|||
folder_id = filesystem.folder_info("/proj_1")["folder_id"]
|
||||
scoped_results = filesystem.search(
|
||||
scope={"folder_id": folder_id, "recursive": True},
|
||||
semantic=False,
|
||||
limit=10,
|
||||
)
|
||||
ranked_folders = {
|
||||
|
|
@ -102,12 +101,10 @@ def test_metadata_contains_treats_percent_and_underscore_literally(tmp_path):
|
|||
|
||||
percent_results = filesystem.search(
|
||||
metadata_filter={"status": {"$contains": "100% done"}},
|
||||
semantic=False,
|
||||
limit=10,
|
||||
)
|
||||
underscore_results = filesystem.search(
|
||||
metadata_filter={"status": {"$contains": "build_alpha"}},
|
||||
semantic=False,
|
||||
limit=10,
|
||||
)
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue