ingestlib/Makefile at main · LangModule/ingestlib · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
.PHONY: docs docs-build test test-all test-llm test-nova test-embedding test-rerank test-rerank-aws test-rerank-jina test-openai test-ollama test-ocr test-parse test-classify test-split test-extract test-s3 test-pinecone test-qdrant test-sqlite test-pgvector test-mongodb test-milvus test-opensearch test-weaviate test-services test-cli test-lifecycle test-mcp eval

# fast suite — every opt-in e2e group skips (RUN_* gates unset)
test:
	uv run pytest tests/

# the ingestlib CLI (init/doctor + corpus commands) — no gate
test-cli:
	uv run pytest tests/cli/

# lifecycle services (remove/sync/backfill) + replace-aware ingest — no gate
test-lifecycle:
	uv run pytest tests/services/lifecycle/ tests/services/ingest/test_replace.py

# MCP server — tools, read_only gating, http auth, CLI entrypoint — no gate
test-mcp:
	uv run pytest tests/mcp/

# entire suite including every opt-in group (needs VL server running + Bedrock access)
test-all:
	RUN_AWS_RERANK=1 RUN_OLLAMA_E2E=1 RUN_OCR_E2E=1 RUN_PARSE_E2E=1 RUN_CLASSIFY_E2E=1 RUN_SPLIT_E2E=1 RUN_EXTRACT_E2E=1 RUN_S3_E2E=1 RUN_PINECONE_E2E=1 RUN_QDRANT_E2E=1 RUN_PGVECTOR_E2E=1 RUN_MONGODB_E2E=1 RUN_MILVUS_E2E=1 RUN_OPENSEARCH_E2E=1 RUN_WEAVIATE_E2E=1 RUN_SERVICES_E2E=1 uv run pytest tests/

# --- llm layer (mirrors src/ingestlib/foundations/llm/) ---

test-llm:
	uv run pytest tests/foundations/llm/

test-nova:
	uv run pytest tests/foundations/llm/bedrock/nova/

test-embedding:
	uv run pytest tests/foundations/llm/bedrock/embedding/

# both providers — AWS-hitting tests skip without RUN_AWS_RERANK=1
test-rerank:
	uv run pytest tests/foundations/llm/bedrock/rerank/ tests/foundations/llm/jina/

# opt-in: hits amazon.rerank-v1:0 (2 RPM quota — expect ~1 min of throttle sleeps)
test-rerank-aws:
	RUN_AWS_RERANK=1 uv run pytest tests/foundations/llm/bedrock/rerank/

test-rerank-jina:
	uv run pytest tests/foundations/llm/jina/

# skips without OPENAI_API_KEY in .env
test-openai:
	uv run pytest tests/foundations/llm/openai/

# opt-in: needs a local Ollama with qwen3.5:9b + qwen3-embedding:0.6b pulled
test-ollama:
	RUN_OLLAMA_E2E=1 uv run pytest tests/foundations/llm/ollama/

# --- ocr layer (mirrors src/ingestlib/foundations/ocr/) — needs the VL inference server ---

test-ocr:
	RUN_OCR_E2E=1 uv run pytest tests/foundations/ocr/

# --- parse operation (mirrors src/ingestlib/operations/parse/) — needs VL server + the LLM provider ---

test-parse:
	RUN_PARSE_E2E=1 uv run pytest tests/operations/parse/

# --- classify operation (mirrors src/ingestlib/operations/classify/) — needs the LLM provider ---

test-classify:
	RUN_CLASSIFY_E2E=1 uv run pytest tests/operations/classify/

# --- split operation (mirrors src/ingestlib/operations/split/) — needs the LLM provider ---

test-split:
	RUN_SPLIT_E2E=1 uv run pytest tests/operations/split/

# --- extract operation (mirrors src/ingestlib/operations/extract/) — needs the LLM provider
#     (the scanned-document case also needs the VL server) ---

test-extract:
	RUN_EXTRACT_E2E=1 uv run pytest tests/operations/extract/

# --- storage (S3 artifacts) — needs AWS credentials ---

test-s3:
	RUN_S3_E2E=1 uv run pytest tests/storage/

# --- pinecone connector — needs PINECONE_API_KEY + the embedding provider ---

test-pinecone:
	RUN_PINECONE_E2E=1 uv run pytest tests/storage/pinecone/

# --- qdrant connector — needs a Qdrant server at QDRANT_URL + the embedding provider ---

test-qdrant:
	RUN_QDRANT_E2E=1 uv run pytest tests/storage/qdrant/

# --- sqlite connector — no gate: no server exists, in-process IS the real thing ---

test-sqlite:
	uv run pytest tests/storage/sqlite/

# --- pgvector connector — needs a Postgres at PGVECTOR_URL (synthetic vectors, no embedding provider) ---

test-pgvector:
	RUN_PGVECTOR_E2E=1 uv run pytest tests/storage/pgvector/

# --- mongodb connector — needs a MongoDB at MONGODB_URL with search (synthetic vectors, no embedding provider) ---

test-mongodb:
	RUN_MONGODB_E2E=1 uv run pytest tests/storage/mongodb/

# --- milvus connector — needs a Milvus at MILVUS_URL (synthetic vectors, no embedding provider) ---

test-milvus:
	RUN_MILVUS_E2E=1 uv run pytest tests/storage/milvus/

# --- opensearch connector — needs an OpenSearch domain/server at OPENSEARCH_URL ---

test-opensearch:
	RUN_OPENSEARCH_E2E=1 uv run pytest tests/storage/opensearch/

# --- weaviate connector — needs a Weaviate server at WEAVIATE_URL ---

test-weaviate:
	RUN_WEAVIATE_E2E=1 uv run pytest tests/storage/weaviate/

# --- services (ingest + retrieve) — needs the FULL stack ---

test-services:
	RUN_SERVICES_E2E=1 uv run pytest tests/services/

# --- retrieval quality eval — measurement, not a test (needs the full stack;
#     first run also needs the VL server to ingest the fixture corpus) ---

eval:
	uv run python evals/run_eval.py

# --- documentation — live preview and the strict build CI runs ---

docs:
	uv run --group docs mkdocs serve

docs-build:
	uv run --group docs mkdocs build --strict