|
| 1 | +# -*- coding: utf-8 -*- |
| 2 | + |
| 3 | +# Copyright 2024 Google LLC |
| 4 | +# |
| 5 | +# Licensed under the Apache License, Version 2.0 (the "License"); |
| 6 | +# you may not use this file except in compliance with the License. |
| 7 | +# You may obtain a copy of the License at |
| 8 | +# |
| 9 | +# http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | +# |
| 11 | +# Unless required by applicable law or agreed to in writing, software |
| 12 | +# distributed under the License is distributed on an "AS IS" BASIS, |
| 13 | +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 14 | +# See the License for the specific language governing permissions and |
| 15 | +# limitations under the License. |
| 16 | +# |
| 17 | + |
| 18 | +from vertexai.preview.rag.utils.resources import ( |
| 19 | + RagCorpus, |
| 20 | + RagFile, |
| 21 | +) |
| 22 | +from google.cloud import aiplatform |
| 23 | +from google.cloud.aiplatform_v1beta1 import ( |
| 24 | + GoogleDriveSource, |
| 25 | + RagFileChunkingConfig, |
| 26 | + ImportRagFilesConfig, |
| 27 | + ImportRagFilesRequest, |
| 28 | + ImportRagFilesResponse, |
| 29 | + RagCorpus as GapicRagCorpus, |
| 30 | + RagFile as GapicRagFile, |
| 31 | + RagContexts, |
| 32 | + RetrieveContextsResponse, |
| 33 | +) |
| 34 | + |
| 35 | + |
| 36 | +TEST_PROJECT = "test-project" |
| 37 | +TEST_PROJECT_NUMBER = "12345678" |
| 38 | +TEST_REGION = "us-central1" |
| 39 | +TEST_CORPUS_DISPLAY_NAME = "my-corpus-1" |
| 40 | +TEST_CORPUS_DISCRIPTION = "My first corpus." |
| 41 | +TEST_RAG_CORPUS_ID = "generate-123" |
| 42 | +TEST_API_ENDPOINT = "us-central1-" + aiplatform.constants.base.API_BASE_PATH |
| 43 | +TEST_RAG_CORPUS_RESOURCE_NAME = f"projects/{TEST_PROJECT_NUMBER}/locations/{TEST_REGION}/ragCorpora/{TEST_RAG_CORPUS_ID}" |
| 44 | + |
| 45 | +# RagCorpus |
| 46 | +TEST_GAPIC_RAG_CORPUS = GapicRagCorpus( |
| 47 | + name=TEST_RAG_CORPUS_RESOURCE_NAME, |
| 48 | + display_name=TEST_CORPUS_DISPLAY_NAME, |
| 49 | + description=TEST_CORPUS_DISCRIPTION, |
| 50 | +) |
| 51 | +TEST_RAG_CORPUS = RagCorpus( |
| 52 | + name=TEST_RAG_CORPUS_RESOURCE_NAME, |
| 53 | + display_name=TEST_CORPUS_DISPLAY_NAME, |
| 54 | + description=TEST_CORPUS_DISCRIPTION, |
| 55 | +) |
| 56 | +TEST_PAGE_TOKEN = "test-page-token" |
| 57 | + |
| 58 | +# RagFiles |
| 59 | +TEST_PATH = "usr/home/my_file.txt" |
| 60 | +TEST_GCS_PATH = "gs://usr/home/data_dir/" |
| 61 | +TEST_FILE_DISPLAY_NAME = "my-file.txt" |
| 62 | +TEST_FILE_DESCRIPTION = "my file." |
| 63 | +TEST_HEADERS = {"X-Goog-Upload-Protocol": "multipart"} |
| 64 | +TEST_UPLOAD_REQUEST_URI = "https://{}/upload/v1beta1/projects/{}/locations/{}/ragCorpora/{}/ragFiles:upload".format( |
| 65 | + TEST_API_ENDPOINT, TEST_PROJECT_NUMBER, TEST_REGION, TEST_RAG_CORPUS_ID |
| 66 | +) |
| 67 | +TEST_RAG_FILE_ID = "generate-456" |
| 68 | +TEST_RAG_FILE_RESOURCE_NAME = ( |
| 69 | + TEST_RAG_CORPUS_RESOURCE_NAME + f"/ragFiles/{TEST_RAG_FILE_ID}" |
| 70 | +) |
| 71 | +TEST_UPLOAD_RAG_FILE_RESPONSE_CONTENT = "" |
| 72 | +TEST_RAG_FILE_JSON = { |
| 73 | + "ragFile": { |
| 74 | + "name": TEST_RAG_FILE_RESOURCE_NAME, |
| 75 | + "displayName": TEST_FILE_DISPLAY_NAME, |
| 76 | + } |
| 77 | +} |
| 78 | +TEST_RAG_FILE_JSON_ERROR = {"error": {"code": 13}} |
| 79 | +TEST_CHUNK_SIZE = 512 |
| 80 | +TEST_CHUNK_OVERLAP = 100 |
| 81 | +# GCS |
| 82 | +TEST_IMPORT_FILES_CONFIG_GCS = ImportRagFilesConfig() |
| 83 | +TEST_IMPORT_FILES_CONFIG_GCS.gcs_source.uris = [TEST_GCS_PATH] |
| 84 | +TEST_IMPORT_REQUEST_GCS = ImportRagFilesRequest( |
| 85 | + parent=TEST_RAG_CORPUS_RESOURCE_NAME, |
| 86 | + import_rag_files_config=TEST_IMPORT_FILES_CONFIG_GCS, |
| 87 | +) |
| 88 | +# Google Drive folders |
| 89 | +TEST_DRIVE_FOLDER_ID = "123" |
| 90 | +TEST_DRIVE_FOLDER = ( |
| 91 | + f"https://drive.google.com/corp/drive/folders/{TEST_DRIVE_FOLDER_ID}" |
| 92 | +) |
| 93 | +TEST_IMPORT_FILES_CONFIG_DRIVE_FOLDER = ImportRagFilesConfig() |
| 94 | +TEST_IMPORT_FILES_CONFIG_DRIVE_FOLDER.google_drive_source.resource_ids = [ |
| 95 | + GoogleDriveSource.ResourceId( |
| 96 | + resource_id=TEST_DRIVE_FOLDER_ID, |
| 97 | + resource_type=GoogleDriveSource.ResourceId.ResourceType.RESOURCE_TYPE_FOLDER, |
| 98 | + ) |
| 99 | +] |
| 100 | +TEST_IMPORT_REQUEST_DRIVE_FOLDER = ImportRagFilesRequest( |
| 101 | + parent=TEST_RAG_CORPUS_RESOURCE_NAME, |
| 102 | + import_rag_files_config=TEST_IMPORT_FILES_CONFIG_DRIVE_FOLDER, |
| 103 | +) |
| 104 | +# Google Drive files |
| 105 | +TEST_DRIVE_FILE_ID = "456" |
| 106 | +TEST_DRIVE_FILE = f"https://drive.google.com/file/d/{TEST_DRIVE_FILE_ID}" |
| 107 | +TEST_IMPORT_FILES_CONFIG_DRIVE_FILE = ImportRagFilesConfig( |
| 108 | + rag_file_chunking_config=RagFileChunkingConfig( |
| 109 | + chunk_size=TEST_CHUNK_SIZE, |
| 110 | + chunk_overlap=TEST_CHUNK_OVERLAP, |
| 111 | + ) |
| 112 | +) |
| 113 | +TEST_IMPORT_FILES_CONFIG_DRIVE_FILE.google_drive_source.resource_ids = [ |
| 114 | + GoogleDriveSource.ResourceId( |
| 115 | + resource_id=TEST_DRIVE_FILE_ID, |
| 116 | + resource_type=GoogleDriveSource.ResourceId.ResourceType.RESOURCE_TYPE_FILE, |
| 117 | + ) |
| 118 | +] |
| 119 | +TEST_IMPORT_REQUEST_DRIVE_FILE = ImportRagFilesRequest( |
| 120 | + parent=TEST_RAG_CORPUS_RESOURCE_NAME, |
| 121 | + import_rag_files_config=TEST_IMPORT_FILES_CONFIG_DRIVE_FILE, |
| 122 | +) |
| 123 | + |
| 124 | +TEST_IMPORT_RESPONSE = ImportRagFilesResponse(imported_rag_files_count=2) |
| 125 | + |
| 126 | +TEST_GAPIC_RAG_FILE = GapicRagFile( |
| 127 | + name=TEST_RAG_FILE_RESOURCE_NAME, |
| 128 | + display_name=TEST_FILE_DISPLAY_NAME, |
| 129 | + description=TEST_FILE_DESCRIPTION, |
| 130 | +) |
| 131 | +TEST_RAG_FILE = RagFile( |
| 132 | + name=TEST_RAG_FILE_RESOURCE_NAME, |
| 133 | + display_name=TEST_FILE_DISPLAY_NAME, |
| 134 | + description=TEST_FILE_DESCRIPTION, |
| 135 | +) |
| 136 | + |
| 137 | +# Retrieval |
| 138 | +TEST_QUERY_TEXT = "What happen to the fox and the dog?" |
| 139 | +TEST_CONTEXTS = RagContexts( |
| 140 | + contexts=[ |
| 141 | + RagContexts.Context( |
| 142 | + source_uri="https://drive.google.com/file/d/123/view?usp=drivesdk", |
| 143 | + text="The quick brown fox jumps over the lazy dog.", |
| 144 | + ), |
| 145 | + RagContexts.Context(text="The slow red fox jumps over the lazy dog."), |
| 146 | + ] |
| 147 | +) |
| 148 | +TEST_RETRIEVAL_RESPONSE = RetrieveContextsResponse(contexts=TEST_CONTEXTS) |
0 commit comments