alita-sdk 0.3.293__py3-none-any.whl → 0.3.294__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- alita_sdk/runtime/langchain/document_loaders/constants.py +20 -15
- alita_sdk/tools/utils/content_parser.py +6 -5
- {alita_sdk-0.3.293.dist-info → alita_sdk-0.3.294.dist-info}/METADATA +1 -1
- {alita_sdk-0.3.293.dist-info → alita_sdk-0.3.294.dist-info}/RECORD +7 -7
- {alita_sdk-0.3.293.dist-info → alita_sdk-0.3.294.dist-info}/WHEEL +0 -0
- {alita_sdk-0.3.293.dist-info → alita_sdk-0.3.294.dist-info}/licenses/LICENSE +0 -0
- {alita_sdk-0.3.293.dist-info → alita_sdk-0.3.294.dist-info}/top_level.txt +0 -0
@@ -13,7 +13,6 @@
|
|
13
13
|
# limitations under the License.
|
14
14
|
|
15
15
|
from langchain_community.document_loaders import (
|
16
|
-
UnstructuredMarkdownLoader,
|
17
16
|
AirbyteJSONLoader, UnstructuredHTMLLoader,
|
18
17
|
PythonLoader)
|
19
18
|
|
@@ -26,43 +25,49 @@ from .AlitaPDFLoader import AlitaPDFLoader
|
|
26
25
|
from .AlitaPowerPointLoader import AlitaPowerPointLoader
|
27
26
|
from .AlitaTextLoader import AlitaTextLoader
|
28
27
|
from .AlitaMarkdownLoader import AlitaMarkdownLoader
|
28
|
+
from enum import Enum
|
29
|
+
|
30
|
+
class LoaderProperties(Enum):
|
31
|
+
LLM = 'llm'
|
32
|
+
PROMPT = 'prompt'
|
33
|
+
PROMPT_DEFAULT = 'prompt_default'
|
29
34
|
|
30
35
|
loaders_map = {
|
31
36
|
'.png': {
|
32
37
|
'class': AlitaImageLoader,
|
33
38
|
'is_multimodal_processing': True,
|
34
39
|
'kwargs': {},
|
35
|
-
'allowed_to_override': ['max_tokens']
|
40
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value],
|
36
41
|
},
|
37
42
|
'.jpg': {
|
38
43
|
'class': AlitaImageLoader,
|
39
44
|
'is_multimodal_processing': True,
|
40
45
|
'kwargs': {},
|
41
|
-
'allowed_to_override': ['max_tokens']
|
46
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
42
47
|
},
|
43
48
|
'.jpeg': {
|
44
49
|
'class': AlitaImageLoader,
|
45
50
|
'is_multimodal_processing': True,
|
46
51
|
'kwargs': {},
|
47
|
-
'allowed_to_override': ['max_tokens']
|
52
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
48
53
|
},
|
49
54
|
'.gif': {
|
50
55
|
'class': AlitaImageLoader,
|
51
56
|
'is_multimodal_processing': True,
|
52
57
|
'kwargs': {},
|
53
|
-
'allowed_to_override': ['max_tokens']
|
58
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
54
59
|
},
|
55
60
|
'.bmp': {
|
56
61
|
'class': AlitaImageLoader,
|
57
62
|
'is_multimodal_processing': True,
|
58
63
|
'kwargs': {},
|
59
|
-
'allowed_to_override': ['max_tokens']
|
64
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
60
65
|
},
|
61
66
|
'.svg': {
|
62
67
|
'class': AlitaImageLoader,
|
63
68
|
'is_multimodal_processing': True,
|
64
69
|
'kwargs': {},
|
65
|
-
'allowed_to_override': ['max_tokens']
|
70
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
66
71
|
},
|
67
72
|
'.txt': {
|
68
73
|
'class': AlitaTextLoader,
|
@@ -119,7 +124,7 @@ loaders_map = {
|
|
119
124
|
'raw_content': False,
|
120
125
|
'cleanse': False
|
121
126
|
},
|
122
|
-
'allowed_to_override': ['max_tokens']
|
127
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
123
128
|
},
|
124
129
|
'.xls': {
|
125
130
|
'class': AlitaExcelLoader,
|
@@ -128,13 +133,13 @@ loaders_map = {
|
|
128
133
|
'raw_content': False,
|
129
134
|
'cleanse': False
|
130
135
|
},
|
131
|
-
'allowed_to_override': ['max_tokens']
|
136
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
132
137
|
},
|
133
138
|
'.pdf': {
|
134
139
|
'class': AlitaPDFLoader,
|
135
140
|
'is_multimodal_processing': False,
|
136
141
|
'kwargs': {},
|
137
|
-
'allowed_to_override': ['max_tokens']
|
142
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
138
143
|
},
|
139
144
|
'.docx': {
|
140
145
|
'class': AlitaDocxMammothLoader,
|
@@ -142,7 +147,7 @@ loaders_map = {
|
|
142
147
|
'kwargs': {
|
143
148
|
'extract_images': True
|
144
149
|
},
|
145
|
-
'allowed_to_override': ['max_tokens', 'mode']
|
150
|
+
'allowed_to_override': ['max_tokens', 'mode', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
146
151
|
},
|
147
152
|
'.json': {
|
148
153
|
'class': AlitaJSONLoader,
|
@@ -160,13 +165,13 @@ loaders_map = {
|
|
160
165
|
'class': UnstructuredHTMLLoader,
|
161
166
|
'is_multimodal_processing': False,
|
162
167
|
'kwargs': {},
|
163
|
-
'allowed_to_override': ['max_tokens']
|
168
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
164
169
|
},
|
165
170
|
'.html': {
|
166
171
|
'class': UnstructuredHTMLLoader,
|
167
172
|
'is_multimodal_processing': False,
|
168
173
|
'kwargs': {},
|
169
|
-
'allowed_to_override': ['max_tokens']
|
174
|
+
'allowed_to_override': ['max_tokens', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
170
175
|
},
|
171
176
|
'.ppt': {
|
172
177
|
'class': AlitaPowerPointLoader,
|
@@ -174,7 +179,7 @@ loaders_map = {
|
|
174
179
|
'kwargs': {
|
175
180
|
'mode': 'paged'
|
176
181
|
},
|
177
|
-
'allowed_to_override': ['max_tokens', 'mode']
|
182
|
+
'allowed_to_override': ['max_tokens', 'mode', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
178
183
|
},
|
179
184
|
'.pptx': {
|
180
185
|
'class': AlitaPowerPointLoader,
|
@@ -182,7 +187,7 @@ loaders_map = {
|
|
182
187
|
'kwargs': {
|
183
188
|
'mode': 'paged'
|
184
189
|
},
|
185
|
-
'allowed_to_override': ['max_tokens', 'mode']
|
190
|
+
'allowed_to_override': ['max_tokens', 'mode', LoaderProperties.LLM.value, LoaderProperties.PROMPT.value, LoaderProperties.PROMPT_DEFAULT.value]
|
186
191
|
},
|
187
192
|
'.py': {
|
188
193
|
'class': PythonLoader,
|
@@ -7,7 +7,7 @@ from typing import Generator, List
|
|
7
7
|
from langchain_core.documents import Document
|
8
8
|
from langchain_core.tools import ToolException
|
9
9
|
|
10
|
-
from alita_sdk.runtime.langchain.document_loaders.constants import loaders_map
|
10
|
+
from alita_sdk.runtime.langchain.document_loaders.constants import loaders_map, LoaderProperties
|
11
11
|
from ...runtime.utils.utils import IndexerKeywords
|
12
12
|
|
13
13
|
logger = getLogger(__name__)
|
@@ -52,8 +52,6 @@ Highlight any visible details that could help in understanding the image.
|
|
52
52
|
Be as precise and thorough as possible in your responses. If something is unclear or illegible, state that explicitly.
|
53
53
|
'''
|
54
54
|
|
55
|
-
IMAGE_EXTENSIONS = ['jpg', 'jpeg', 'png', 'gif', 'bmp', 'tiff', 'webp', 'svg']
|
56
|
-
|
57
55
|
|
58
56
|
def parse_file_content(file_name=None, file_content=None, is_capture_image: bool = False, page_number: int = None,
|
59
57
|
sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False) -> str | ToolException:
|
@@ -234,8 +232,11 @@ def process_content_by_type(document: Document, content, extension_source: str,
|
|
234
232
|
if chunking_config and (users_config_for_extension := chunking_config.get(extension, {})):
|
235
233
|
for key in set(users_config_for_extension.keys()) & set(allowed_to_override):
|
236
234
|
loader_kwargs[key] = users_config_for_extension[key]
|
237
|
-
|
238
|
-
|
235
|
+
if LoaderProperties.LLM.value in loader_kwargs:
|
236
|
+
loader_kwargs[LoaderProperties.LLM.value] = llm
|
237
|
+
if LoaderProperties.PROMPT_DEFAULT.value in loader_kwargs:
|
238
|
+
loader_kwargs.pop(LoaderProperties.PROMPT_DEFAULT.value)
|
239
|
+
loader_kwargs[LoaderProperties.PROMPT.value] = image_processing_prompt
|
239
240
|
loader = loader_cls(file_path=temp_file_path, **loader_kwargs)
|
240
241
|
counter = 1
|
241
242
|
try:
|
@@ -1,6 +1,6 @@
|
|
1
1
|
Metadata-Version: 2.4
|
2
2
|
Name: alita_sdk
|
3
|
-
Version: 0.3.
|
3
|
+
Version: 0.3.294
|
4
4
|
Summary: SDK for building langchain agents using resources from Alita
|
5
5
|
Author-email: Artem Rozumenko <artyom.rozumenko@gmail.com>, Mikalai Biazruchka <mikalai_biazruchka@epam.com>, Roman Mitusov <roman_mitusov@epam.com>, Ivan Krakhmaliuk <lifedjik@gmail.com>, Artem Dubrovskiy <ad13box@gmail.com>
|
6
6
|
License-Expression: Apache-2.0
|
@@ -69,7 +69,7 @@ alita_sdk/runtime/langchain/document_loaders/AlitaTableLoader.py,sha256=o0SRFPZ-
|
|
69
69
|
alita_sdk/runtime/langchain/document_loaders/AlitaTextLoader.py,sha256=uNcV0En49_0u0RYB1sP1XfNspT2Xc5CacuJr9Jqv79Q,2972
|
70
70
|
alita_sdk/runtime/langchain/document_loaders/ImageParser.py,sha256=gao5yCCKdDai_Gx7YdEx5U6oMyJYzn69eYmEvWLh-fc,656
|
71
71
|
alita_sdk/runtime/langchain/document_loaders/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
72
|
-
alita_sdk/runtime/langchain/document_loaders/constants.py,sha256=
|
72
|
+
alita_sdk/runtime/langchain/document_loaders/constants.py,sha256=bnZSQM1QTAcp4Z8MDViX9QcfhroTWFQKCPpzpVujCYU,7110
|
73
73
|
alita_sdk/runtime/langchain/document_loaders/utils.py,sha256=9xghESf3axBbwxATyVuS0Yu-TWe8zWZnXgCD1ZVyNW0,2414
|
74
74
|
alita_sdk/runtime/langchain/interfaces/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
75
75
|
alita_sdk/runtime/langchain/interfaces/kwextractor.py,sha256=kSJA9L8g8UArmHu7Bd9dIO0Rrq86JPUb8RYNlnN68FQ,3072
|
@@ -326,7 +326,7 @@ alita_sdk/tools/testrail/__init__.py,sha256=0kETjWKLU7R6mugBWsjwEUsh10pipbAeNSGJ
|
|
326
326
|
alita_sdk/tools/testrail/api_wrapper.py,sha256=xKQbjwL602J55KZiAdMcMtsuzK2jky0DUcrrdsazj0A,32981
|
327
327
|
alita_sdk/tools/utils/__init__.py,sha256=155xepXPr4OEzs2Mz5YnjXcBpxSv1X2eznRUVoPtyK0,3268
|
328
328
|
alita_sdk/tools/utils/available_tools_decorator.py,sha256=IbrdfeQkswxUFgvvN7-dyLMZMyXLiwvX7kgi3phciCk,273
|
329
|
-
alita_sdk/tools/utils/content_parser.py,sha256=
|
329
|
+
alita_sdk/tools/utils/content_parser.py,sha256=duLnLVdjAfswpIQL763J1JU1dgggCshwKAZBMTjCGzE,14044
|
330
330
|
alita_sdk/tools/vector_adapters/VectorStoreAdapter.py,sha256=ypBEAkFRGHv5edW0N9rdo1yKurNGQ4pRVEWtrN_7SeA,17656
|
331
331
|
alita_sdk/tools/vector_adapters/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
332
332
|
alita_sdk/tools/xray/__init__.py,sha256=AsHDvRgyD-6vvGyAyQDPWHbOD2WoMJ5Llt029bLuu6U,4277
|
@@ -348,8 +348,8 @@ alita_sdk/tools/zephyr_scale/api_wrapper.py,sha256=2M6yFoOmfH5BwxMwTLQO76yD0Ydfv
|
|
348
348
|
alita_sdk/tools/zephyr_squad/__init__.py,sha256=0AI_j27xVO5Gk5HQMFrqPTd4uvuVTpiZUicBrdfEpKg,2796
|
349
349
|
alita_sdk/tools/zephyr_squad/api_wrapper.py,sha256=kmw_xol8YIYFplBLWTqP_VKPRhL_1ItDD0_vXTe_UuI,14906
|
350
350
|
alita_sdk/tools/zephyr_squad/zephyr_squad_cloud_client.py,sha256=R371waHsms4sllHCbijKYs90C-9Yu0sSR3N4SUfQOgU,5066
|
351
|
-
alita_sdk-0.3.
|
352
|
-
alita_sdk-0.3.
|
353
|
-
alita_sdk-0.3.
|
354
|
-
alita_sdk-0.3.
|
355
|
-
alita_sdk-0.3.
|
351
|
+
alita_sdk-0.3.294.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
|
352
|
+
alita_sdk-0.3.294.dist-info/METADATA,sha256=mLD2KHv1QpAm1Yebo7O-JDDhkR5NVeIT8o8ta1fdgxs,18897
|
353
|
+
alita_sdk-0.3.294.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
|
354
|
+
alita_sdk-0.3.294.dist-info/top_level.txt,sha256=0vJYy5p_jK6AwVb1aqXr7Kgqgk3WDtQ6t5C-XI9zkmg,10
|
355
|
+
alita_sdk-0.3.294.dist-info/RECORD,,
|
File without changes
|
File without changes
|
File without changes
|