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
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169 | class OnDemandLoaderTool(AsyncBaseTool):
"""
On-demand data loader tool.
Loads data with by calling the provided loader function,
stores in index, and queries for relevant data with a
natural language query string.
"""
def __init__(
self,
loader: Callable[..., List[Document]],
index_cls: Type[BaseIndex],
index_kwargs: Dict,
metadata: ToolMetadata,
use_query_str_in_loader: bool = False,
query_str_kwargs_key: str = "query_str",
) -> None:
"""Init params."""
self._loader = loader
self._index_cls = index_cls
self._index_kwargs = index_kwargs
self._use_query_str_in_loader = use_query_str_in_loader
self._metadata = metadata
self._query_str_kwargs_key = query_str_kwargs_key
@property
def metadata(self) -> ToolMetadata:
return self._metadata
@classmethod
def from_defaults(
cls,
reader: BaseReader,
index_cls: Optional[Type[BaseIndex]] = None,
index_kwargs: Optional[Dict] = None,
use_query_str_in_loader: bool = False,
query_str_kwargs_key: str = "query_str",
name: Optional[str] = None,
description: Optional[str] = None,
fn_schema: Optional[Type[BaseModel]] = None,
) -> "OnDemandLoaderTool":
"""From defaults."""
# NOTE: fn_schema should be specified if you want to use as langchain Tool
index_cls = index_cls or VectorStoreIndex
index_kwargs = index_kwargs or {}
if description is None:
description = f"Tool to load data from {reader.__class__.__name__}"
if fn_schema is None:
fn_schema = create_schema_from_function(
name or "LoadData",
reader.load_data,
[(query_str_kwargs_key, str, None)],
)
metadata = ToolMetadata(name=name, description=description, fn_schema=fn_schema)
return cls(
loader=reader.load_data,
index_cls=index_cls,
index_kwargs=index_kwargs,
use_query_str_in_loader=use_query_str_in_loader,
query_str_kwargs_key=query_str_kwargs_key,
metadata=metadata,
)
@classmethod
def from_tool(
cls,
tool: FunctionTool,
index_cls: Optional[Type[BaseIndex]] = None,
index_kwargs: Optional[Dict] = None,
use_query_str_in_loader: bool = False,
query_str_kwargs_key: str = "query_str",
name: Optional[str] = None,
description: Optional[str] = None,
return_direct: bool = False,
fn_schema: Optional[Type[BaseModel]] = None,
) -> "OnDemandLoaderTool":
"""From defaults."""
# NOTE: fn_schema should be specified if you want to use as langchain Tool
index_cls = index_cls or VectorStoreIndex
index_kwargs = index_kwargs or {}
if description is None:
description = f"Tool to load data from {tool.__class__.__name__}"
if fn_schema is None:
fn_schema = create_schema_from_function(
name or "LoadData", tool._fn, [(query_str_kwargs_key, str, None)]
)
metadata = ToolMetadata(
name=name,
description=description,
fn_schema=fn_schema,
return_direct=return_direct,
)
return cls(
loader=tool._fn,
index_cls=index_cls,
index_kwargs=index_kwargs,
use_query_str_in_loader=use_query_str_in_loader,
query_str_kwargs_key=query_str_kwargs_key,
metadata=metadata,
)
def _parse_args(self, *args: Any, **kwargs: Any) -> Tuple[str, List[Document]]:
if self._query_str_kwargs_key not in kwargs:
raise ValueError(
"Missing query_str in kwargs with parameter name: "
f"{self._query_str_kwargs_key}"
)
if self._use_query_str_in_loader:
query_str = kwargs[self._query_str_kwargs_key]
else:
query_str = kwargs.pop(self._query_str_kwargs_key)
docs = self._loader(*args, **kwargs)
return query_str, docs
def call(self, *args: Any, **kwargs: Any) -> ToolOutput:
"""Call."""
query_str, docs = self._parse_args(*args, **kwargs)
index = self._index_cls.from_documents(docs, **self._index_kwargs)
# TODO: add query kwargs
query_engine = index.as_query_engine()
response = query_engine.query(query_str)
return ToolOutput(
content=str(response),
tool_name=self.metadata.name,
raw_input={"query": query_str},
raw_output=response,
)
async def acall(self, *args: Any, **kwargs: Any) -> ToolOutput:
"""Async Call."""
query_str, docs = self._parse_args(*args, **kwargs)
index = self._index_cls.from_documents(docs, **self._index_kwargs)
# TODO: add query kwargs
query_engine = index.as_query_engine()
response = await query_engine.aquery(query_str)
return ToolOutput(
content=str(response),
tool_name=self.metadata.name,
raw_input={"query": query_str},
raw_output=response,
)
|
