fabricatio 0.3.15.dev5__cp312-cp312-win_amd64.whl → 0.4.4__cp312-cp312-win_amd64.whl

fabricatio/models/action.py

@@ -1,263 +0,0 @@
-"""Module that contains the classes for defining and executing task workflows.
-
-This module provides the Action and WorkFlow classes for creating structured
-task execution pipelines. Actions represent atomic operations, while WorkFlows
-orchestrate sequences of actions with shared context and error handling.
-
-Classes:
-    Action: Base class for defining executable actions with context management.
-    WorkFlow: Manages action sequences, context propagation, and task lifecycle.
-"""
-
-import traceback
-from abc import abstractmethod
-from asyncio import Queue, create_task
-from typing import Any, ClassVar, Dict, Generator, Self, Sequence, Tuple, Type, Union, final
-
-from fabricatio.journal import logger
-from fabricatio.models.generic import WithBriefing
-from fabricatio.models.task import Task
-from fabricatio.utils import override_kwargs
-from pydantic import Field, PrivateAttr
-
-OUTPUT_KEY = "task_output"
-
-INPUT_KEY = "task_input"
-
-
-class Action(WithBriefing):
-    """Class that represents an action to be executed in a workflow.
-
-    Actions are the atomic units of work in a workflow. Each action performs
-    a specific operation and can modify the shared context data.
-    """
-
-    ctx_override: ClassVar[bool] = False
-    """Whether to override the instance attr by the context variable."""
-
-    name: str = Field(default="")
-    """The name of the action."""
-
-    description: str = Field(default="")
-    """The description of the action."""
-
-    personality: str = Field(default="")
-    """The personality traits or context for the action executor."""
-
-    output_key: str = Field(default="")
-    """The key used to store this action's output in the context dictionary."""
-
-    @final
-    def model_post_init(self, __context: Any) -> None:
-        """Initialize the action by setting default name and description if not provided.
-
-        Args:
-            __context: The context to be used for initialization.
-        """
-        self.name = self.name or self.__class__.__name__
-        self.description = self.description or self.__class__.__doc__ or ""
-
-    @abstractmethod
-    async def _execute(self, *_: Any, **cxt) -> Any:
-        """Implement the core logic of the action.
-
-        Args:
-            **cxt: Context dictionary containing input/output data.
-
-        Returns:
-            Result of the action execution to be stored in context.
-        """
-        pass
-
-    @final
-    async def act(self, cxt: Dict[str, Any]) -> Dict[str, Any]:
-        """Execute action and update context.
-
-        Args:
-            cxt (Dict[str, Any]): Shared context dictionary.
-
-        Returns:
-            Updated context dictionary with new/modified entries.
-        """
-        ret = await self._execute(**cxt)
-
-        if self.output_key:
-            logger.debug(f"Setting output: {self.output_key}")
-            cxt[self.output_key] = ret
-
-        return cxt
-
-    @property
-    def briefing(self) -> str:
-        """Generate formatted action description with personality context.
-
-        Returns:
-            Briefing text combining personality and action description.
-        """
-        if self.personality:
-            return f"## Your personality: \n{self.personality}\n# The action you are going to perform: \n{super().briefing}"
-        return f"# The action you are going to perform: \n{super().briefing}"
-
-    def to_task_output(self, to: Union[str, "WorkFlow"] = OUTPUT_KEY) -> Self:
-        """Set the output key to OUTPUT_KEY and return the action instance."""
-        self.output_key = to.task_output_key if isinstance(to, WorkFlow) else to
-        return self
-
-
-class WorkFlow(WithBriefing):
-    """Manages sequences of actions to fulfill tasks.
-
-    Handles context propagation between actions, error handling, and task lifecycle
-    events like cancellation and completion.
-    """
-
-    name: str = "WorkFlow"
-    """The name of the workflow, which is used to identify and describe the workflow."""
-    description: str = ""
-    """The description of the workflow, which describes the workflow's purpose and requirements."""
-
-    _context: Queue[Dict[str, Any]] = PrivateAttr(default_factory=lambda: Queue(maxsize=1))
-    """Queue for storing the workflow execution context."""
-
-    _instances: Tuple[Action, ...] = PrivateAttr(default_factory=tuple)
-    """Instantiated action objects to be executed in this workflow."""
-
-    steps: Sequence[Union[Type[Action], Action]] = Field(frozen=True)
-    """The sequence of actions to be executed, can be action classes or instances."""
-
-    task_input_key: ClassVar[str] = INPUT_KEY
-    """Key used to store the input task in the context dictionary."""
-
-    task_output_key: ClassVar[str] = OUTPUT_KEY
-    """Key used to extract the final result from the context dictionary."""
-
-    extra_init_context: Dict[str, Any] = Field(default_factory=dict, frozen=True)
-    """Additional initial context values to be included at workflow start."""
-
-    def model_post_init(self, __context: Any) -> None:
-        """Initialize the workflow by instantiating any action classes.
-
-        Args:
-            __context: The context to be used for initialization.
-
-        """
-        self.name = self.name or self.__class__.__name__
-        # Convert any action classes to instances
-        self._instances = tuple(step if isinstance(step, Action) else step() for step in self.steps)
-
-    def iter_actions(self) -> Generator[Action, None, None]:
-        """Iterate over action instances."""
-        yield from self._instances
-
-    def inject_personality(self, personality: str) -> Self:
-        """Set personality for actions without existing personality.
-
-        Args:
-            personality (str): Shared personality context
-
-        Returns:
-            Workflow instance with updated actions
-        """
-        for action in filter(lambda a: not a.personality, self._instances):
-            action.personality = personality
-        return self
-
-    def override_action_variable(self, action: Action, ctx: Dict[str, Any]) -> Self:
-        """Override action variable with context values."""
-        if action.ctx_override:
-            for k, v in ctx.items():
-                if hasattr(action, k):
-                    setattr(action, k, v)
-
-        return self
-
-    async def serve(self, task: Task) -> None:
-        """Execute workflow to complete given task.
-
-        Args:
-            task (Task): Task instance to be processed.
-
-        Steps:
-            1. Initialize context with task instance and extra data
-            2. Execute each action sequentially
-            3. Handle task cancellation and exceptions
-            4. Extract final result from context
-        """
-        logger.info(f"Start execute workflow: {self.name}")
-
-        await task.start()
-        await self._init_context(task)
-
-        current_action = None
-        try:
-            # Process each action in sequence
-            for i, step in enumerate(self._instances):
-                logger.info(f"Executing step [{i}] >> {(current_action := step.name)}")
-
-                # Get current context and execute action
-                context = await self._context.get()
-
-                self.override_action_variable(step, context)
-                act_task = create_task(step.act(context))
-                # Handle task cancellation
-                if task.is_cancelled():
-                    logger.warning(f"Workflow cancelled by task: {task.name}")
-                    act_task.cancel(f"Cancelled by task: {task.name}")
-                    break
-
-                # Update context with modified values
-                modified_ctx = await act_task
-                logger.success(f"Step [{i}] `{current_action}` execution finished.")
-                if step.output_key:
-                    logger.success(f"Setting action `{current_action}` output to `{step.output_key}`")
-                await self._context.put(modified_ctx)
-
-            logger.success(f"Workflow `{self.name}` execution finished.")
-
-            # Get final context and extract result
-            final_ctx = await self._context.get()
-            result = final_ctx.get(self.task_output_key)
-
-            if self.task_output_key not in final_ctx:
-                logger.warning(
-                    f"Task output key: `{self.task_output_key}` not found in the context, None will be returned. "
-                    f"You can check if `Action.output_key` is set the same as `WorkFlow.task_output_key`."
-                )
-
-            await task.finish(result)
-
-        except Exception as e:  # noqa: BLE001
-            logger.critical(f"Error during task: {current_action} execution: {e}")
-            logger.critical(traceback.format_exc())
-            await task.fail()
-
-    async def _init_context[T](self, task: Task[T]) -> None:
-        """Initialize workflow execution context.
-
-        Args:
-            task (Task[T]): Task being processed
-
-        Context includes:
-            - Task instance stored under task_input_key
-            - Any extra_init_context values
-        """
-        logger.debug(f"Initializing context for workflow: {self.name}")
-        ctx = override_kwargs(self.extra_init_context, **task.extra_init_context)
-        if self.task_input_key in ctx:
-            raise ValueError(
-                f"Task input key: `{self.task_input_key}`, which is reserved, is already set in the init context"
-            )
-
-        await self._context.put({self.task_input_key: task, **ctx})
-
-    def update_init_context(self, /, **kwargs) -> Self:
-        """Update the initial context with additional key-value pairs.
-
-        Args:
-            **kwargs: Key-value pairs to add to the initial context.
-
-        Returns:
-            Self: The workflow instance for method chaining.
-        """
-        self.extra_init_context.update(kwargs)
-        return self

fabricatio/models/adv_kwargs_types.py

@@ -1,63 +0,0 @@
-"""A module containing kwargs types for content correction and checking operations."""
-
-from importlib.util import find_spec
-from typing import NotRequired, Optional, TypedDict
-
-from fabricatio.models.extra.problem import Improvement
-from fabricatio.models.extra.rule import RuleSet
-from fabricatio.models.generic import SketchedAble
-from fabricatio.models.kwargs_types import ReferencedKwargs
-
-
-class CorrectKwargs[T: SketchedAble](ReferencedKwargs[T], total=False):
-    """Arguments for content correction operations.
-
-    Extends GenerateKwargs with parameters for correcting content based on
-    specific criteria and templates.
-    """
-
-    improvement: Improvement
-
-
-class CheckKwargs(ReferencedKwargs[Improvement], total=False):
-    """Arguments for content checking operations.
-
-    Extends GenerateKwargs with parameters for checking content against
-    specific criteria and templates.
-    """
-
-    ruleset: RuleSet
-
-
-if find_spec("pymilvus"):
-    from pymilvus import CollectionSchema
-    from pymilvus.milvus_client import IndexParams
-
-    class CollectionConfigKwargs(TypedDict, total=False):
-        """Configuration parameters for a vector collection.
-
-        These arguments are typically used when configuring connections to vector databases.
-        """
-
-        dimension: int | None
-        primary_field_name: str
-        id_type: str
-        vector_field_name: str
-        metric_type: str
-        timeout: float | None
-        schema: CollectionSchema | None
-        index_params: IndexParams | None
-
-    class FetchKwargs(TypedDict):
-        """Arguments for fetching data from vector collections.
-
-        Controls how data is retrieved from vector databases, including filtering
-        and result limiting parameters.
-        """
-
-        collection_name: NotRequired[str | None]
-        similarity_threshold: NotRequired[float]
-        result_per_query: NotRequired[int]
-        tei_endpoint: NotRequired[Optional[str]]
-        reranker_threshold: NotRequired[float]
-        filter_expr: NotRequired[str]

fabricatio/models/extra/__init__.py

@@ -1 +0,0 @@
-"""A module contains extra models for fabricatio."""

fabricatio/models/extra/advanced_judge.py

@@ -1,32 +0,0 @@
-"""Module containing the JudgeMent class for holding judgment results."""
-
-from typing import List
-
-from fabricatio.models.generic import SketchedAble
-
-
-class JudgeMent(SketchedAble):
-    """Represents a judgment result containing supporting/denying evidence and final verdict.
-
-    The class stores both affirmative and denies evidence, truth and reasons lists along with the final boolean judgment.
-    """
-
-    issue_to_judge: str
-    """The issue to be judged, including the original question and context"""
-
-    deny_evidence: List[str]
-    """List of clues supporting the denial."""
-
-    affirm_evidence: List[str]
-    """List of clues supporting the affirmation."""
-
-    final_judgement: bool
-    """The final judgment made according to all extracted clues. true for the `issue_to_judge` is correct and false for incorrect."""
-
-    def __bool__(self) -> bool:
-        """Return the final judgment value.
-
-        Returns:
-            bool: The stored final_judgement value indicating the judgment result.
-        """
-        return self.final_judgement

fabricatio/models/extra/aricle_rag.py

@@ -1,286 +0,0 @@
-"""A Module containing the article rag models."""
-
-import re
-from dataclasses import dataclass, field
-from itertools import groupby
-from pathlib import Path
-from typing import ClassVar, Dict, List, Optional, Self, Unpack
-
-from fabricatio.fs import safe_text_read
-from fabricatio.journal import logger
-from fabricatio.models.extra.rag import MilvusDataBase
-from fabricatio.models.generic import AsPrompt
-from fabricatio.models.kwargs_types import ChunkKwargs
-from fabricatio.rust import BibManager, blake3_hash, split_into_chunks
-from fabricatio.utils import ok, wrapp_in_block
-from more_itertools.more import first
-from more_itertools.recipes import flatten, unique
-from pydantic import Field
-
-
-class ArticleChunk(MilvusDataBase):
-    """The chunk of an article."""
-
-    etc_word: ClassVar[str] = "等"
-    and_word: ClassVar[str] = "与"
-    _cite_number: Optional[int] = None
-
-    head_split: ClassVar[List[str]] = [
-        "引 言",
-        "引言",
-        "绪 论",
-        "绪论",
-        "前言",
-        "INTRODUCTION",
-        "Introduction",
-    ]
-    tail_split: ClassVar[List[str]] = [
-        "参 考 文 献",
-        "参  考  文  献",
-        "参考文献",
-        "REFERENCES",
-        "References",
-        "Bibliography",
-        "Reference",
-    ]
-    chunk: str
-    """The segment of the article"""
-    year: int
-    """The year of the article"""
-    authors: List[str] = Field(default_factory=list)
-    """The authors of the article"""
-    article_title: str
-    """The title of the article"""
-    bibtex_cite_key: str
-    """The bibtex cite key of the article"""
-
-    @property
-    def reference_header(self) -> str:
-        """Get the reference header."""
-        return f"[[{ok(self._cite_number, 'You need to update cite number first.')}]] reference `{self.article_title}` from {self.as_auther_seq()}"
-
-    @property
-    def cite_number(self) -> int:
-        """Get the cite number."""
-        return ok(self._cite_number, "cite number not set")
-
-    def _prepare_vectorization_inner(self) -> str:
-        return self.chunk
-
-    @classmethod
-    def from_file[P: str | Path](
-            cls, path: P | List[P], bib_mgr: BibManager, **kwargs: Unpack[ChunkKwargs]
-    ) -> List[Self]:
-        """Load the article chunks from the file."""
-        if isinstance(path, list):
-            result = list(flatten(cls._from_file_inner(p, bib_mgr, **kwargs) for p in path))
-            logger.debug(f"Number of chunks created from list of files: {len(result)}")
-            return result
-
-        return cls._from_file_inner(path, bib_mgr, **kwargs)
-
-    @classmethod
-    def _from_file_inner(cls, path: str | Path, bib_mgr: BibManager, **kwargs: Unpack[ChunkKwargs]) -> List[Self]:
-        path = Path(path)
-
-        title_seg = path.stem.split(" - ").pop()
-
-        key = (
-                bib_mgr.get_cite_key_by_title(title_seg)
-                or bib_mgr.get_cite_key_by_title_fuzzy(title_seg)
-                or bib_mgr.get_cite_key_fuzzy(path.stem)
-        )
-        if key is None:
-            logger.warning(f"no cite key found for {path.as_posix()}, skip.")
-            return []
-        authors = ok(bib_mgr.get_author_by_key(key), f"no author found for {key}")
-        year = ok(bib_mgr.get_year_by_key(key), f"no year found for {key}")
-        article_title = ok(bib_mgr.get_title_by_key(key), f"no title found for {key}")
-
-        result = [
-            cls(chunk=c, year=year, authors=authors, article_title=article_title, bibtex_cite_key=key)
-            for c in split_into_chunks(cls.purge_numeric_citation(cls.strip(safe_text_read(path))), **kwargs)
-        ]
-
-        logger.debug(f"Number of chunks created from file {path.as_posix()}: {len(result)}")
-        return result
-
-    @classmethod
-    def strip(cls, string: str) -> str:
-        """Strip the head and tail of the string."""
-        logger.debug(f"String length before strip: {(original := len(string))}")
-        for split in (s for s in cls.head_split if s in string):
-            logger.debug(f"Strip head using {split}")
-            parts = string.split(split)
-            string = split.join(parts[1:]) if len(parts) > 1 else parts[0]
-            break
-        logger.debug(
-            f"String length after head strip: {(stripped_len := len(string))}, decreased by {(d := original - stripped_len)}"
-        )
-        if not d:
-            logger.warning("No decrease at head strip, which is might be abnormal.")
-        for split in (s for s in cls.tail_split if s in string):
-            logger.debug(f"Strip tail using {split}")
-            parts = string.split(split)
-            string = split.join(parts[:-1]) if len(parts) > 1 else parts[0]
-            break
-        logger.debug(f"String length after tail strip: {len(string)}, decreased by {(d := stripped_len - len(string))}")
-        if not d:
-            logger.warning("No decrease at tail strip, which is might be abnormal.")
-
-        return string
-
-    def as_typst_cite(self) -> str:
-        """As typst cite."""
-        return f"#cite(<{self.bibtex_cite_key}>)"
-
-    @staticmethod
-    def purge_numeric_citation(string: str) -> str:
-        """Purge numeric citation."""
-        import re
-
-        return re.sub(r"\[[\d\s,\\~–-]+]", "", string)
-
-    @property
-    def auther_lastnames(self) -> List[str]:
-        """Get the last name of the authors."""
-        return [n.split()[-1] for n in self.authors]
-
-    def as_auther_seq(self) -> str:
-        """Get the auther sequence."""
-        match len(self.authors):
-            case 0:
-                raise ValueError("No authors found")
-            case 1:
-                return f"（{self.auther_lastnames[0]}，{self.year}）{self.as_typst_cite()}"
-            case 2:
-                return f"（{self.auther_lastnames[0]}{self.and_word}{self.auther_lastnames[1]}，{self.year}）{self.as_typst_cite()}"
-            case 3:
-                return f"（{self.auther_lastnames[0]}，{self.auther_lastnames[1]}{self.and_word}{self.auther_lastnames[2]}，{self.year}）{self.as_typst_cite()}"
-            case _:
-                return f"（{self.auther_lastnames[0]}，{self.auther_lastnames[1]}{self.and_word}{self.auther_lastnames[2]}{self.etc_word}，{self.year}）{self.as_typst_cite()}"
-
-    def update_cite_number(self, cite_number: int) -> Self:
-        """Update the cite number."""
-        self._cite_number = cite_number
-        return self
-
-
-@dataclass
-class CitationManager(AsPrompt):
-    """Citation manager."""
-
-    article_chunks: List[ArticleChunk] = field(default_factory=list)
-    """Article chunks."""
-
-    pat: str = r"(\[\[([\d\s,-]*)]])"
-    """Regex pattern to match citations."""
-    sep: str = ","
-    """Separator for citation numbers."""
-    abbr_sep: str = "-"
-    """Separator for abbreviated citation numbers."""
-
-    def update_chunks(
-            self, article_chunks: List[ArticleChunk], set_cite_number: bool = True, dedup: bool = True
-    ) -> Self:
-        """Update article chunks."""
-        self.article_chunks.clear()
-        self.article_chunks.extend(article_chunks)
-        if dedup:
-            self.article_chunks = list(unique(self.article_chunks, lambda c: blake3_hash(c.chunk.encode())))
-        if set_cite_number:
-            self.set_cite_number_all()
-        return self
-
-    def empty(self) -> Self:
-        """Empty the article chunks."""
-        self.article_chunks.clear()
-        return self
-
-    def add_chunks(self, article_chunks: List[ArticleChunk], set_cite_number: bool = True, dedup: bool = True) -> Self:
-        """Add article chunks."""
-        self.article_chunks.extend(article_chunks)
-        if dedup:
-            self.article_chunks = list(unique(self.article_chunks, lambda c: blake3_hash(c.chunk.encode())))
-        if set_cite_number:
-            self.set_cite_number_all()
-        return self
-
-    def set_cite_number_all(self) -> Self:
-        """Set citation numbers for all article chunks."""
-        number_mapping = {a.bibtex_cite_key: 0 for a in self.article_chunks}
-
-        for i, k in enumerate(number_mapping.keys()):
-            number_mapping[k] = i
-
-        for a in self.article_chunks:
-            a.update_cite_number(number_mapping[a.bibtex_cite_key])
-        return self
-
-    def _as_prompt_inner(self) -> Dict[str, str]:
-        """Generate prompt inner representation."""
-        seg = []
-        for k, g_iter in groupby(self.article_chunks, key=lambda a: a.bibtex_cite_key):
-            g = list(g_iter)
-
-            logger.debug(f"Group [{k}]: {len(g)}")
-            seg.append(wrapp_in_block("\n\n".join(a.chunk for a in g), first(g).reference_header))
-        return {"References": "\n".join(seg)}
-
-    def apply(self, string: str) -> str:
-        """Apply citation replacements to the input string."""
-        for origin, m in re.findall(self.pat, string):
-            logger.info(f"Matching citation: {m}")
-            notations = self.convert_to_numeric_notations(m)
-            logger.info(f"Citing Notations: {notations}")
-            citation_number_seq = list(flatten(self.decode_expr(n) for n in notations))
-            logger.info(f"Citation Number Sequence: {citation_number_seq}")
-            dedup = self.deduplicate_citation(citation_number_seq)
-            logger.info(f"Deduplicated Citation Number Sequence: {dedup}")
-            string = string.replace(origin, self.unpack_cite_seq(dedup))
-        return string
-
-    def citation_count(self, string: str) -> int:
-        """Get the citation count in the string."""
-        count = 0
-        for _, m in re.findall(self.pat, string):
-            logger.info(f"Matching citation: {m}")
-            notations = self.convert_to_numeric_notations(m)
-            logger.info(f"Citing Notations: {notations}")
-            citation_number_seq = list(flatten(self.decode_expr(n) for n in notations))
-            logger.info(f"Citation Number Sequence: {citation_number_seq}")
-            count += len(dedup := self.deduplicate_citation(citation_number_seq))
-            logger.info(f"Deduplicated Citation Number Sequence: {dedup}")
-        return count
-
-    def citation_coverage(self, string: str) -> float:
-        """Get the citation coverage in the string."""
-        return self.citation_count(string) / len(self.article_chunks)
-
-    def decode_expr(self, string: str) -> List[int]:
-        """Decode citation expression into a list of integers."""
-        if self.abbr_sep in string:
-            start, end = string.split(self.abbr_sep)
-            return list(range(int(start), int(end) + 1))
-        return [int(string)]
-
-    def convert_to_numeric_notations(self, string: str) -> List[str]:
-        """Convert citation string into numeric notations."""
-        return [s.strip() for s in string.split(self.sep)]
-
-    def deduplicate_citation(self, citation_seq: List[int]) -> List[int]:
-        """Deduplicate citation sequence."""
-        chunk_seq = [a for a in self.article_chunks if a.cite_number in citation_seq]
-        deduped = unique(chunk_seq, lambda a: a.bibtex_cite_key)
-        return [a.cite_number for a in deduped]
-
-    def unpack_cite_seq(self, citation_seq: List[int]) -> str:
-        """Unpack citation sequence into a string."""
-        chunk_seq = {a.bibtex_cite_key: a for a in self.article_chunks if a.cite_number in citation_seq}
-        return "".join(a.as_typst_cite() for a in chunk_seq.values())
-
-    def as_milvus_filter_expr(self, blacklist: bool = True) -> str:
-        """Asynchronously fetches documents from a Milvus database based on input vectors."""
-        if blacklist:
-            return " and ".join(f'bibtex_cite_key != "{a.bibtex_cite_key}"' for a in self.article_chunks)
-        return " or ".join(f'bibtex_cite_key == "{a.bibtex_cite_key}"' for a in self.article_chunks)