383 lines
18 KiB
Python
383 lines
18 KiB
Python
# Copyright (c) 2025 Patrick Motsch
|
|
# All rights reserved.
|
|
|
|
import logging
|
|
from datetime import datetime, UTC
|
|
from modules.workflows.methods.methodBase import MethodBase
|
|
from modules.datamodels.datamodelWorkflowActions import WorkflowActionDefinition, WorkflowActionParameter
|
|
from modules.shared.frontendTypes import FrontendType
|
|
|
|
# Import helpers
|
|
from .helpers.csvProcessing import CsvProcessingHelper
|
|
|
|
# Import actions
|
|
from .actions.process import process
|
|
from .actions.webResearch import webResearch
|
|
from .actions.summarizeDocument import summarizeDocument
|
|
from .actions.translateDocument import translateDocument
|
|
from .actions.convert import convert
|
|
from .actions.convertDocument import convertDocument
|
|
from .actions.extractData import extractData
|
|
from .actions.generateDocument import generateDocument
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
class MethodAi(MethodBase):
|
|
"""AI processing methods."""
|
|
|
|
def __init__(self, services):
|
|
super().__init__(services)
|
|
self.name = "ai"
|
|
self.description = "AI processing methods"
|
|
|
|
# Initialize helper modules
|
|
self.csvProcessing = CsvProcessingHelper(self)
|
|
|
|
# RBAC-Integration: Action-Definitionen mit actionId
|
|
self._actions = {
|
|
"process": WorkflowActionDefinition(
|
|
actionId="ai.process",
|
|
description="Universal AI document processing action - accepts multiple input documents in any format and processes them together with a prompt",
|
|
parameters={
|
|
"aiPrompt": WorkflowActionParameter(
|
|
name="aiPrompt",
|
|
type="str",
|
|
frontendType=FrontendType.TEXTAREA,
|
|
required=True,
|
|
description="Instruction for the AI describing what processing to perform"
|
|
),
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=False,
|
|
description="Document reference(s) in any format to use as input/context"
|
|
),
|
|
"resultType": WorkflowActionParameter(
|
|
name="resultType",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["txt", "json", "md", "csv", "xml", "html", "pdf", "docx", "xlsx", "pptx", "png", "jpg"],
|
|
required=False,
|
|
default="txt",
|
|
description="Output file extension. All output documents will use this format"
|
|
)
|
|
},
|
|
execute=process.__get__(self, self.__class__)
|
|
),
|
|
"webResearch": WorkflowActionDefinition(
|
|
actionId="ai.webResearch",
|
|
description="Web research with two-step process: search for URLs, then crawl content",
|
|
parameters={
|
|
"prompt": WorkflowActionParameter(
|
|
name="prompt",
|
|
type="str",
|
|
frontendType=FrontendType.TEXTAREA,
|
|
required=True,
|
|
description="Natural language research instruction"
|
|
),
|
|
"urlList": WorkflowActionParameter(
|
|
name="urlList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.MULTISELECT,
|
|
required=False,
|
|
description="Specific URLs to crawl, if needed"
|
|
),
|
|
"country": WorkflowActionParameter(
|
|
name="country",
|
|
type="str",
|
|
frontendType=FrontendType.TEXT,
|
|
required=False,
|
|
description="Two-digit country code (lowercase, e.g., ch, us, de)"
|
|
),
|
|
"language": WorkflowActionParameter(
|
|
name="language",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["de", "en", "fr", "it", "es"],
|
|
required=False,
|
|
description="Language code (lowercase, e.g., de, en, fr)"
|
|
),
|
|
"researchDepth": WorkflowActionParameter(
|
|
name="researchDepth",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["fast", "general", "deep"],
|
|
required=False,
|
|
default="general",
|
|
description="Research depth"
|
|
)
|
|
},
|
|
execute=webResearch.__get__(self, self.__class__)
|
|
),
|
|
"summarizeDocument": WorkflowActionDefinition(
|
|
actionId="ai.summarizeDocument",
|
|
description="Summarize one or more documents, extracting key points and main ideas",
|
|
parameters={
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=True,
|
|
description="Document reference(s) to summarize"
|
|
),
|
|
"summaryLength": WorkflowActionParameter(
|
|
name="summaryLength",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["brief", "medium", "detailed"],
|
|
required=False,
|
|
default="medium",
|
|
description="Desired summary length"
|
|
),
|
|
"focus": WorkflowActionParameter(
|
|
name="focus",
|
|
type="str",
|
|
frontendType=FrontendType.TEXT,
|
|
required=False,
|
|
description="Specific aspect to focus on in the summary (e.g., financial data, key decisions)"
|
|
),
|
|
"resultType": WorkflowActionParameter(
|
|
name="resultType",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["txt", "md", "docx"],
|
|
required=False,
|
|
default="txt",
|
|
description="Output file extension"
|
|
)
|
|
},
|
|
execute=summarizeDocument.__get__(self, self.__class__)
|
|
),
|
|
"translateDocument": WorkflowActionDefinition(
|
|
actionId="ai.translateDocument",
|
|
description="Translate documents to a target language while preserving formatting and structure",
|
|
parameters={
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=True,
|
|
description="Document reference(s) to translate"
|
|
),
|
|
"targetLanguage": WorkflowActionParameter(
|
|
name="targetLanguage",
|
|
type="str",
|
|
frontendType=FrontendType.TEXT,
|
|
required=True,
|
|
description="Target language code or name (e.g., de, German, French, es)"
|
|
),
|
|
"sourceLanguage": WorkflowActionParameter(
|
|
name="sourceLanguage",
|
|
type="str",
|
|
frontendType=FrontendType.TEXT,
|
|
required=False,
|
|
description="Source language if known (e.g., en, English). If not provided, AI will detect"
|
|
),
|
|
"preserveFormatting": WorkflowActionParameter(
|
|
name="preserveFormatting",
|
|
type="bool",
|
|
frontendType=FrontendType.CHECKBOX,
|
|
required=False,
|
|
default=True,
|
|
description="Whether to preserve original formatting"
|
|
),
|
|
"resultType": WorkflowActionParameter(
|
|
name="resultType",
|
|
type="str",
|
|
frontendType=FrontendType.TEXT,
|
|
required=False,
|
|
description="Output file extension. If not specified, uses same format as input"
|
|
)
|
|
},
|
|
execute=translateDocument.__get__(self, self.__class__)
|
|
),
|
|
"convert": WorkflowActionDefinition(
|
|
actionId="ai.convert",
|
|
description="Convert documents/data between different formats with specific formatting options",
|
|
parameters={
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=True,
|
|
description="Document reference(s) to convert"
|
|
),
|
|
"inputFormat": WorkflowActionParameter(
|
|
name="inputFormat",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["json", "csv", "xlsx", "txt"],
|
|
required=True,
|
|
description="Source format"
|
|
),
|
|
"outputFormat": WorkflowActionParameter(
|
|
name="outputFormat",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["csv", "json", "xlsx", "txt"],
|
|
required=True,
|
|
description="Target format"
|
|
),
|
|
"columnsPerRow": WorkflowActionParameter(
|
|
name="columnsPerRow",
|
|
type="int",
|
|
frontendType=FrontendType.NUMBER,
|
|
required=False,
|
|
description="For CSV output, number of columns per row. Default: auto-detect",
|
|
validation={"min": 1, "max": 100}
|
|
),
|
|
"delimiter": WorkflowActionParameter(
|
|
name="delimiter",
|
|
type="str",
|
|
frontendType=FrontendType.TEXT,
|
|
required=False,
|
|
default=",",
|
|
description="For CSV output, delimiter character"
|
|
),
|
|
"includeHeader": WorkflowActionParameter(
|
|
name="includeHeader",
|
|
type="bool",
|
|
frontendType=FrontendType.CHECKBOX,
|
|
required=False,
|
|
default=True,
|
|
description="For CSV output, whether to include header row"
|
|
),
|
|
"language": WorkflowActionParameter(
|
|
name="language",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["de", "en", "fr"],
|
|
required=False,
|
|
default="en",
|
|
description="Language for output"
|
|
)
|
|
},
|
|
execute=convert.__get__(self, self.__class__)
|
|
),
|
|
"convertDocument": WorkflowActionDefinition(
|
|
actionId="ai.convertDocument",
|
|
description="Convert documents between different formats (PDF→Word, Excel→CSV, etc.)",
|
|
parameters={
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=True,
|
|
description="Document reference(s) to convert"
|
|
),
|
|
"targetFormat": WorkflowActionParameter(
|
|
name="targetFormat",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["docx", "pdf", "xlsx", "csv", "txt", "html", "json", "md"],
|
|
required=True,
|
|
description="Target format extension"
|
|
),
|
|
"preserveStructure": WorkflowActionParameter(
|
|
name="preserveStructure",
|
|
type="bool",
|
|
frontendType=FrontendType.CHECKBOX,
|
|
required=False,
|
|
default=True,
|
|
description="Whether to preserve document structure (headings, tables, etc.)"
|
|
)
|
|
},
|
|
execute=convertDocument.__get__(self, self.__class__)
|
|
),
|
|
"extractData": WorkflowActionDefinition(
|
|
actionId="ai.extractData",
|
|
description="Extract structured data from documents (key-value pairs, entities, facts, etc.)",
|
|
parameters={
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=True,
|
|
description="Document reference(s) to extract data from"
|
|
),
|
|
"dataStructure": WorkflowActionParameter(
|
|
name="dataStructure",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["flat", "nested", "list"],
|
|
required=False,
|
|
default="nested",
|
|
description="Desired data structure"
|
|
),
|
|
"fields": WorkflowActionParameter(
|
|
name="fields",
|
|
type="List[str]",
|
|
frontendType=FrontendType.MULTISELECT,
|
|
required=False,
|
|
description="Specific fields/properties to extract (e.g., [name, date, amount])"
|
|
),
|
|
"resultType": WorkflowActionParameter(
|
|
name="resultType",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["json", "csv", "xlsx"],
|
|
required=False,
|
|
default="json",
|
|
description="Output format"
|
|
)
|
|
},
|
|
execute=extractData.__get__(self, self.__class__)
|
|
),
|
|
"generateDocument": WorkflowActionDefinition(
|
|
actionId="ai.generateDocument",
|
|
description="Generate documents from scratch or based on templates/inputs",
|
|
parameters={
|
|
"prompt": WorkflowActionParameter(
|
|
name="prompt",
|
|
type="str",
|
|
frontendType=FrontendType.TEXTAREA,
|
|
required=True,
|
|
description="Description of the document to generate"
|
|
),
|
|
"documentList": WorkflowActionParameter(
|
|
name="documentList",
|
|
type="List[str]",
|
|
frontendType=FrontendType.DOCUMENT_REFERENCE,
|
|
required=False,
|
|
description="Template documents or reference documents to use as a guide"
|
|
),
|
|
"documentType": WorkflowActionParameter(
|
|
name="documentType",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["letter", "memo", "proposal", "contract", "report", "email"],
|
|
required=False,
|
|
description="Type of document"
|
|
),
|
|
"resultType": WorkflowActionParameter(
|
|
name="resultType",
|
|
type="str",
|
|
frontendType=FrontendType.SELECT,
|
|
frontendOptions=["docx", "pdf", "txt", "md"],
|
|
required=False,
|
|
default="docx",
|
|
description="Output format"
|
|
)
|
|
},
|
|
execute=generateDocument.__get__(self, self.__class__)
|
|
)
|
|
}
|
|
|
|
# Validate actions after definition
|
|
self._validateActions()
|
|
|
|
# Register actions as methods (optional, für direkten Zugriff)
|
|
self.process = process.__get__(self, self.__class__)
|
|
self.webResearch = webResearch.__get__(self, self.__class__)
|
|
self.summarizeDocument = summarizeDocument.__get__(self, self.__class__)
|
|
self.translateDocument = translateDocument.__get__(self, self.__class__)
|
|
self.convert = convert.__get__(self, self.__class__)
|
|
self.convertDocument = convertDocument.__get__(self, self.__class__)
|
|
self.extractData = extractData.__get__(self, self.__class__)
|
|
self.generateDocument = generateDocument.__get__(self, self.__class__)
|
|
|
|
def _format_timestamp_for_filename(self) -> str:
|
|
"""Format current timestamp as YYYYMMDD-hhmmss for filenames."""
|
|
return datetime.now(UTC).strftime("%Y%m%d-%H%M%S")
|
|
|