alita-sdk 0.3.328__tar.gz → 0.3.330__tar.gz

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alita_sdk
-Version: 0.3.328
+Version: 0.3.330
 Summary: SDK for building langchain agents using resources from Alita
 Author-email: Artem Rozumenko <artyom.rozumenko@gmail.com>, Mikalai Biazruchka <mikalai_biazruchka@epam.com>, Roman Mitusov <roman_mitusov@epam.com>, Ivan Krakhmaliuk <lifedj27@gmail.com>, Artem Dubrovskiy <ad13box@gmail.com>
 License-Expression: Apache-2.0

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/alita_sdk/runtime/langchain/document_loaders/AlitaTextLoader.py

@@ -1,4 +1,4 @@
-from typing import Iterator
+from typing import Iterator, Generator
 
 from langchain_core.documents import Document
 
@@ -6,6 +6,9 @@ from langchain_community.document_loaders.base import BaseLoader
 from langchain_community.document_loaders.helpers import detect_file_encodings
 from langchain_core.tools import ToolException
 
+from alita_sdk.tools.chunkers import markdown_chunker
+
+
 class AlitaTextLoader(BaseLoader):
 
     def __init__(self, **kwargs):
@@ -19,6 +22,8 @@ class AlitaTextLoader(BaseLoader):
             raise ToolException("'file_path' or 'file_content' parameter should be provided.")
         self.encoding = kwargs.get('encoding', 'utf-8')
         self.autodetect_encoding = kwargs.get('autodetect_encoding', False)
+        self.max_tokens=kwargs.get('max_tokens', 1024)
+        self.token_overlap = kwargs.get('token_overlap', 10)
 
     def get_content(self):
         text = ""
@@ -59,8 +64,16 @@ class AlitaTextLoader(BaseLoader):
 
         return text
 
+    def generate_document(self, text, metadata) -> Generator[Document, None, None]:
+        yield Document(page_content=text, metadata=metadata)
+
     def lazy_load(self) -> Iterator[Document]:
         """Load from file path."""
         text = self.get_content()
         metadata = {"source": str(self.file_path) if hasattr(self, 'file_path') else self.file_name}
-        yield Document(page_content=text, metadata=metadata)
+        chunks = markdown_chunker(file_content_generator=self.generate_document(text, metadata),
+                                  config={
+                                      "max_tokens": self.max_tokens,
+                                      "token_overlap": self.token_overlap
+                                  })
+        yield from chunks

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/alita_sdk/tools/ado/work_item/ado_wrapper.py

@@ -1,5 +1,6 @@
 import json
 import logging
+import re
 import urllib.parse
 from typing import Dict, List, Generator, Optional
 
@@ -7,6 +8,7 @@ from azure.devops.connection import Connection
 from azure.devops.v7_1.core import CoreClient
 from azure.devops.v7_1.wiki import WikiClient
 from azure.devops.v7_1.work_item_tracking import TeamContext, Wiql, WorkItemTrackingClient
+from bs4 import BeautifulSoup
 from langchain_core.documents import Document
 from langchain_core.tools import ToolException
 from msrest.authentication import BasicAuthentication
@@ -15,6 +17,7 @@ from pydantic import model_validator
 from pydantic.fields import Field
 
 from alita_sdk.tools.non_code_indexer_toolkit import NonCodeIndexerToolkit
+from ...utils.content_parser import parse_file_content
 from ....runtime.utils.utils import IndexerKeywords
 
 logger = logging.getLogger(__name__)
@@ -53,7 +56,11 @@ ADOGetWorkItem = create_model(
     id=(int, Field(description="The work item id")),
     fields=(Optional[list[str]], Field(description="Comma-separated list of requested fields", default=None)),
     as_of=(Optional[str], Field(description="AsOf UTC date time string", default=None)),
-    expand=(Optional[str], Field(description="The expand parameters for work item attributes. Possible options are { None, Relations, Fields, Links, All }.", default=None))
+    expand=(Optional[str], Field(description="The expand parameters for work item attributes. Possible options are { None, Relations, Fields, Links, All }.", default=None)),
+    parse_attachments=(Optional[bool], Field(description="Value that defines is attachment should be parsed.", default=False)),
+    image_description_prompt=(Optional[str],
+                     Field(description="Prompt which is used for image description", default=None)),
+
 )
 
 ADOLinkWorkItem = create_model(
@@ -284,8 +291,23 @@ class AzureDevOpsApiWrapper(NonCodeIndexerToolkit):
             logger.error(f"Error searching work items: {e}")
             return ToolException(f"Error searching work items: {e}")
 
-
-    def get_work_item(self, id: int, fields: Optional[list[str]] = None, as_of: Optional[str] = None, expand: Optional[str] = None):
+    def parse_attachment_by_url(self, attachment_url, file_name=None, image_description_prompt=None):
+        match = re.search(r'attachments/([\w-]+)(?:\?fileName=([^&]+))?', attachment_url)
+        if match:
+            attachment_id = match.group(1)
+            if not file_name:
+                file_name = match.group(2)
+            if not file_name:
+                raise ToolException("File name must be provided either in the URL or as a parameter.")
+            return self.parse_attachment_by_id(attachment_id, file_name, image_description_prompt)
+        raise ToolException(f"Attachment '{attachment_url}' was not found.")
+
+    def parse_attachment_by_id(self, attachment_id, file_name, image_description_prompt):
+        file_content = self.get_attachment_content(attachment_id)
+        return parse_file_content(file_content=file_content, file_name=file_name,
+                                            llm=self.llm, prompt=image_description_prompt)
+
+    def get_work_item(self, id: int, fields: Optional[list[str]] = None, as_of: Optional[str] = None, expand: Optional[str] = None, parse_attachments=False, image_description_prompt=None):
         """Get a single work item by ID."""
         try:
             # Validate that the Azure DevOps client is initialized
@@ -313,6 +335,24 @@ class AzureDevOpsApiWrapper(NonCodeIndexerToolkit):
                 for relation in relations_data:
                     parsed_item['relations'].append(relation.as_dict())
 
+            if parse_attachments:
+                # describe images in work item fields if present
+                for field_name, field_value in fields_data.items():
+                    if isinstance(field_value, str):
+                        soup = BeautifulSoup(field_value, 'html.parser')
+                        images = soup.find_all('img')
+                        for img in images:
+                            src = img.get('src')
+                            if src:
+                                description = self.parse_attachment_by_url(src, image_description_prompt)
+                                img['image-description'] = description
+                        parsed_item[field_name] = str(soup)
+                # parse attached documents if present
+                if parsed_item['relations']:
+                    for attachment in parsed_item['relations']:
+                        attachment['content'] = self.parse_attachment_by_url(attachment['url'], attachment['attributes']['name'], image_description_prompt)
+
+
             return parsed_item
         except Exception as e:
             logger.error(f"Error getting work item: {e}")
@@ -522,10 +562,13 @@ class AzureDevOpsApiWrapper(NonCodeIndexerToolkit):
                 'attachment_ids': {rel.url.split('/')[-1]:rel.attributes.get('name', '') for rel in wi.relations or [] if rel.rel == 'AttachedFile'}
             })
 
+    def get_attachment_content(self, attachment_id):
+        content_generator = self._client.get_attachment_content(id=attachment_id, download=True)
+        return b"".join(content_generator)
+
     def _process_document(self, document: Document) -> Generator[Document, None, None]:
         for attachment_id, file_name in document.metadata.get('attachment_ids', {}).items():
-            content_generator = self._client.get_attachment_content(id=attachment_id, download=True)
-            content = b"".join(x for x in content_generator)
+            content = self.get_attachment_content(attachment_id=attachment_id)
             yield Document(page_content="", metadata={'id': attachment_id, IndexerKeywords.CONTENT_FILE_NAME.value: file_name, IndexerKeywords.CONTENT_IN_BYTES.value: content})
 
     def _index_tool_params(self):

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/alita_sdk/tools/figma/api_wrapper.py

@@ -61,11 +61,11 @@ class ArgsSchema(Enum):
         ),
         geometry=(
             Optional[str],
-            Field(description="Sets to 'paths' to export vector data"),
+            Field(description="Sets to 'paths' to export vector data", default=None),
         ),
         version=(
             Optional[str],
-            Field(description="Sets version of file"),
+            Field(description="Sets version of file", default=None),
         ),
         extra_params=(
             Optional[Dict[str, Union[str, int, None]]],
@@ -120,7 +120,8 @@ class ArgsSchema(Enum):
         client_meta=(
             Optional[dict],
             Field(
-                description="Positioning information of the comment (Vector, FrameOffset, Region, FrameOffsetRegion)"
+                description="Positioning information of the comment (Vector, FrameOffset, Region, FrameOffsetRegion)",
+                default=None,
             ),
         ),
         extra_params=(
@@ -147,26 +148,28 @@ class ArgsSchema(Enum):
             ),
         ),
         ids=(
-            str,
+            Optional[str],
             Field(
                 description="Specifies id of file images separated by comma",
                 examples=["8:6,1:7"],
+                default="0:0",
             ),
         ),
         scale=(
             Optional[str],
-            Field(description="A number between 0.01 and 4, the image scaling factor"),
+            Field(description="A number between 0.01 and 4, the image scaling factor", default=None),
         ),
         format=(
             Optional[str],
             Field(
                 description="A string enum for the image output format",
                 examples=["jpg", "png", "svg", "pdf"],
+                default=None,
             ),
         ),
         version=(
             Optional[str],
-            Field(description="A specific version ID to use"),
+            Field(description="A specific version ID to use", default=None),
         ),
         extra_params=(
             Optional[Dict[str, Union[str, int, None]]],
@@ -636,7 +639,7 @@ class FigmaApiWrapper(NonCodeIndexerToolkit):
     def get_file_images(
         self,
         file_key: str,
-        ids: str = "0:0",
+        ids: Optional[str] = "0:0",
         scale: Optional[str] = None,
         format: Optional[str] = None,
         version: Optional[str] = None,

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/alita_sdk/tools/utils/content_parser.py

@@ -1,4 +1,5 @@
 import os
+import re
 import tempfile
 from logging import getLogger
 from pathlib import Path
@@ -8,6 +9,7 @@ from langchain_core.documents import Document
 from langchain_core.tools import ToolException
 
 from alita_sdk.runtime.langchain.document_loaders.constants import loaders_map, LoaderProperties
+from ...runtime.langchain.document_loaders.AlitaTextLoader import AlitaTextLoader
 from ...runtime.utils.utils import IndexerKeywords
 
 logger = getLogger(__name__)
@@ -54,7 +56,7 @@ Be as precise and thorough as possible in your responses. If something is unclea
 
 
 def parse_file_content(file_name=None, file_content=None, is_capture_image: bool = False, page_number: int = None,
-                       sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False) -> str | ToolException:
+                       sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False, prompt=None) -> str | ToolException:
     """Parse the content of a file based on its type and return the parsed content.
 
     Args:
@@ -71,6 +73,8 @@ def parse_file_content(file_name=None, file_content=None, is_capture_image: bool
     Raises:
         ToolException: If the file type is not supported or if there is an error reading the file.
         """
+    if not prompt:
+        prompt = image_processing_prompt
     loader = prepare_loader(
         file_name=file_name,
         file_content=file_content,
@@ -79,7 +83,8 @@ def parse_file_content(file_name=None, file_content=None, is_capture_image: bool
         sheet_name=sheet_name,
         llm=llm,
         file_path=file_path,
-        excel_by_sheets=excel_by_sheets
+        excel_by_sheets=excel_by_sheets,
+        prompt=prompt
     )
 
     if not loader:
@@ -120,7 +125,7 @@ def load_file_docs(file_name=None, file_content=None, is_capture_image: bool = F
     return loader.load()
 
 def get_loader_kwargs(loader_object, file_name=None, file_content=None, is_capture_image: bool = False, page_number: int = None,
-                    sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False):
+                    sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False, prompt=None):
     loader_kwargs = loader_object['kwargs']
     loader_kwargs.update({
         "file_path": file_path,
@@ -131,13 +136,15 @@ def get_loader_kwargs(loader_object, file_name=None, file_content=None, is_captu
         "page_number": page_number,
         "sheet_name": sheet_name,
         "excel_by_sheets": excel_by_sheets,
+        "prompt": prompt,
         "row_content": True,
         "json_documents": False
     })
     return loader_kwargs
 
 def prepare_loader(file_name=None, file_content=None, is_capture_image: bool = False, page_number: int = None,
-                       sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False):
+                       sheet_name: str = None, llm=None, file_path: str = None, excel_by_sheets: bool = False,
+                   prompt=None):
         if (file_path and (file_name or file_content)) or (not file_path and (not file_name or file_content is None)):
             raise ToolException("Either (file_name and file_content) or file_path must be provided, but not both.")
 
@@ -146,7 +153,7 @@ def prepare_loader(file_name=None, file_content=None, is_capture_image: bool = F
         loader_object = loaders_map.get(extension)
         if not loader_object:
             return None
-        loader_kwargs = get_loader_kwargs(loader_object, file_name, file_content, is_capture_image, page_number, sheet_name, llm, file_path, excel_by_sheets)
+        loader_kwargs = get_loader_kwargs(loader_object, file_name, file_content, is_capture_image, page_number, sheet_name, llm, file_path, excel_by_sheets, prompt)
         loader = loader_object['class'](**loader_kwargs)
         return loader
 
@@ -226,7 +233,8 @@ def process_content_by_type(content, filename: str, llm=None, chunking_config=No
     """Process the content of a file based on its type using a configured loader."""
     temp_file_path = None
     try:
-        extension = "." + filename.split('.')[-1].lower()
+        match = re.search(r'\.([^.]+)$', filename)
+        extension = f".{match.group(1).lower()}" if match else ".txt"
 
         with tempfile.NamedTemporaryFile(mode='w+b', suffix=extension, delete=False) as temp_file:
             temp_file_path = temp_file.name

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/alita_sdk.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: alita_sdk
-Version: 0.3.328
+Version: 0.3.330
 Summary: SDK for building langchain agents using resources from Alita
 Author-email: Artem Rozumenko <artyom.rozumenko@gmail.com>, Mikalai Biazruchka <mikalai_biazruchka@epam.com>, Roman Mitusov <roman_mitusov@epam.com>, Ivan Krakhmaliuk <lifedj27@gmail.com>, Artem Dubrovskiy <ad13box@gmail.com>
 License-Expression: Apache-2.0

{alita_sdk-0.3.328 → alita_sdk-0.3.330}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "alita_sdk"
-version = "0.3.328"
+version = "0.3.330"
 description = "SDK for building langchain agents using resources from Alita"
 readme = "README.md"
 requires-python = ">=3.10"