fabricatio 0.2.8.dev2__cp312-cp312-manylinux_2_34_x86_64.whl → 0.2.8.dev4__cp312-cp312-manylinux_2_34_x86_64.whl

fabricatio/_rust.pyi

@@ -171,4 +171,4 @@ class BibManager:
 
         Returns:
             Field value if found, None otherwise
-        """
+        """

fabricatio/actions/article.py

@@ -5,20 +5,24 @@ from pathlib import Path
 from typing import Any, Callable, List, Optional
 
 from fabricatio._rust import BibManager
-from fabricatio.capabilities.advanced_judge import AdvancedJudge
+from fabricatio.capabilities.censor import Censor
+from fabricatio.capabilities.correct import Correct
+from fabricatio.capabilities.propose import Propose
 from fabricatio.fs import safe_text_read
 from fabricatio.journal import logger
 from fabricatio.models.action import Action
-from fabricatio.models.extra.article_base import ArticleRefPatch
+from fabricatio.models.extra.article_base import ArticleRefSequencePatch
 from fabricatio.models.extra.article_essence import ArticleEssence
 from fabricatio.models.extra.article_main import Article
 from fabricatio.models.extra.article_outline import ArticleOutline
 from fabricatio.models.extra.article_proposal import ArticleProposal
+from fabricatio.models.extra.rule import RuleSet
 from fabricatio.models.task import Task
-from fabricatio.models.utils import ok
+from fabricatio.utils import ok
+from more_itertools import filter_map
 
 
-class ExtractArticleEssence(Action):
+class ExtractArticleEssence(Action, Propose):
     """Extract the essence of article(s) in text format from the paths specified in the task dependencies.
 
     Notes:
@@ -32,39 +36,65 @@ class ExtractArticleEssence(Action):
     async def _execute(
         self,
         task_input: Task,
-        reader: Callable[[str], str] = lambda p: Path(p).read_text(encoding="utf-8"),
+        reader: Callable[[str], Optional[str]] = lambda p: Path(p).read_text(encoding="utf-8"),
         **_,
-    ) -> Optional[List[ArticleEssence]]:
+    ) -> List[ArticleEssence]:
         if not task_input.dependencies:
             logger.info(err := "Task not approved, since no dependencies are provided.")
             raise RuntimeError(err)
-
+        logger.info(f"Extracting article essence from {len(task_input.dependencies)} files.")
         # trim the references
-        contents = ["References".join(c.split("References")[:-1]) for c in map(reader, task_input.dependencies)]
-        return await self.propose(ArticleEssence, contents, **self.prepend_sys_msg())
+        contents = list(filter_map(reader, task_input.dependencies))
+        logger.info(f"Read {len(task_input.dependencies)} to get {len(contents)} contents.")
+
+        out = []
+
+        for ess in await self.propose(
+            ArticleEssence,
+            [
+                f"{c}\n\n\nBased the provided academic article above, you need to extract the essence from it."
+                for c in contents
+            ],
+        ):
+            if ess is None:
+                logger.warning("Could not extract article essence")
+            else:
+                out.append(ess)
+        logger.info(f"Extracted {len(out)} article essence from {len(task_input.dependencies)} files.")
+        return out
 
 
 class FixArticleEssence(Action):
     """Fix the article essence based on the bibtex key."""
 
+    output_key: str = "fixed_article_essence"
+    """The key of the output data."""
+
     async def _execute(
         self,
         bib_mgr: BibManager,
         article_essence: List[ArticleEssence],
         **_,
-    ) -> None:
+    ) -> List[ArticleEssence]:
+        out = []
+        count = 0
         for a in article_essence:
             if key := (bib_mgr.get_cite_key(a.title) or bib_mgr.get_cite_key_fuzzy(a.title)):
                 a.title = bib_mgr.get_title_by_key(key) or a.title
                 a.authors = bib_mgr.get_author_by_key(key) or a.authors
                 a.publication_year = bib_mgr.get_year_by_key(key) or a.publication_year
                 a.bibtex_cite_key = key
-                logger.info(f'Updated {a.title} with {key}')
+                logger.info(f"Updated {a.title} with {key}")
+                out.append(a)
             else:
                 logger.warning(f"No key found for {a.title}")
+                count += 1
+        if count:
+            logger.warning(f"{count} articles have no key")
+        return out
 
 
-class GenerateArticleProposal(Action):
+class GenerateArticleProposal(Action, Propose):
     """Generate an outline for the article based on the extracted essence."""
 
     output_key: str = "article_proposal"
@@ -97,7 +127,6 @@ class GenerateArticleProposal(Action):
                         )
                     )
                 ),
-                **self.prepend_sys_msg(),
             ),
             "Could not generate the proposal.",
         ).update_ref(briefing)
@@ -107,7 +136,7 @@ class GenerateArticleProposal(Action):
         return proposal
 
 
-class GenerateInitialOutline(Action):
+class GenerateInitialOutline(Action, Propose):
     """Generate the initial article outline based on the article proposal."""
 
     output_key: str = "initial_article_outline"
@@ -122,44 +151,34 @@ class GenerateInitialOutline(Action):
             await self.propose(
                 ArticleOutline,
                 article_proposal.as_prompt(),
-                **self.prepend_sys_msg(),
             ),
             "Could not generate the initial outline.",
         ).update_ref(article_proposal)
 
 
-class FixIntrospectedErrors(Action):
+class FixIntrospectedErrors(Action, Censor):
     """Fix introspected errors in the article outline."""
 
     output_key: str = "introspected_errors_fixed_outline"
     """The key of the output data."""
 
+    ruleset: Optional[RuleSet] = None
+    """The ruleset to use to fix the introspected errors."""
+
     async def _execute(
         self,
         article_outline: ArticleOutline,
-        supervisor_check: bool = False,
+        ruleset: Optional[RuleSet] = None,
         **_,
     ) -> Optional[ArticleOutline]:
-        introspect_manual = ok(
-            await self.draft_rating_manual(
-                topic=(
-                    intro_topic
-                    := "Fix the error in the article outline, make sure there is no more error in the article outline."
-                ),
-            ),
-            "Could not generate the rating manual.",
-        )
-
         while pack := article_outline.find_introspected():
             component, err = ok(pack)
             logger.warning(f"Found introspected error: {err}")
             corrected = ok(
-                await self.correct_obj(
+                await self.censor_obj(
                     component,
-                    reference=f"# Original Article Outline\n{article_outline.display()}\n# Error Need to be fixed\n{err}",
-                    topic=intro_topic,
-                    rating_manual=introspect_manual,
-                    supervisor_check=supervisor_check,
+                    ruleset=ok(ruleset or self.ruleset, "No ruleset provided"),
+                    reference=f"# Original Article Outline\n{article_outline.display()}\n# Some Basic errors found from `{component.title}` that need to be fixed\n{err}",
                 ),
                 "Could not correct the component.",
             )
@@ -168,141 +187,117 @@ class FixIntrospectedErrors(Action):
         return article_outline
 
 
-class FixIllegalReferences(Action):
+class FixIllegalReferences(Action, Censor):
     """Fix illegal references in the article outline."""
 
     output_key: str = "illegal_references_fixed_outline"
     """The key of the output data."""
 
+    ruleset: Optional[RuleSet] = None
+    """Ruleset to use to fix the illegal references."""
+
     async def _execute(
         self,
         article_outline: ArticleOutline,
-        supervisor_check: bool = False,
+        ruleset: Optional[RuleSet] = None,
         **_,
     ) -> Optional[ArticleOutline]:
-        ref_manual = ok(
-            await self.draft_rating_manual(
-                topic=(
-                    ref_topic
-                    := "Fix the internal referring error, make sure there is no more `ArticleRef` pointing to a non-existing article component."
-                ),
-            ),
-            "Could not generate the rating manual.",
-        )
-
-        while pack := article_outline.find_illegal_ref():
-            ref, err = ok(pack)
+        while pack := article_outline.find_illegal_ref(gather_identical=True):
+            refs, err = ok(pack)
             logger.warning(f"Found illegal referring error: {err}")
-            ok(
-                await self.correct_obj_inplace(
-                    ref,
-                    reference=f"# Original Article Outline\n{article_outline.display()}\n# Error Need to be fixed\n{err}",
-                    topic=ref_topic,
-                    rating_manual=ref_manual,
-                    supervisor_check=supervisor_check,
+            corrected_ref = ok(
+                await self.censor_obj(
+                    refs[0],  # pyright: ignore [reportIndexIssue]
+                    ruleset=ok(ruleset or self.ruleset, "No ruleset provided"),
+                    reference=f"# Original Article Outline\n{article_outline.display()}\n# Some Basic errors found that need to be fixed\n{err}",
                 )
             )
+            for ref in refs:
+                ref.update_from(corrected_ref)  # pyright: ignore [reportAttributeAccessIssue]
+
         return article_outline.update_ref(article_outline)
 
 
-class TweakOutlineBackwardRef(Action, AdvancedJudge):
-    """Tweak the backward references in the article outline.
+class TweakOutlineForwardRef(Action, Censor):
+    """Tweak the forward references in the article outline.
 
-    Ensures that the prerequisites of the current chapter are correctly referenced in the `depend_on` field.
+    Ensures that the conclusions of the current chapter effectively support the analysis of subsequent chapters.
     """
 
-    output_key: str = "article_outline_bw_ref_checked"
+    output_key: str = "article_outline_fw_ref_checked"
+    ruleset: Optional[RuleSet] = None
+    """Ruleset to use to fix the illegal references."""
 
-    async def _execute(self, article_outline: ArticleOutline, supervisor_check: bool = False, **cxt) -> ArticleOutline:
-        tweak_depend_on_manual = ok(
-            await self.draft_rating_manual(
-                topic := "Ensure prerequisites are correctly referenced in the `depend_on` field."
-            ),
-            "Could not generate the rating manual.",
+    async def _execute(
+        self, article_outline: ArticleOutline, ruleset: Optional[RuleSet] = None, **cxt
+    ) -> ArticleOutline:
+        return await self._inner(
+            article_outline,
+            ruleset=ok(ruleset or self.ruleset, "No ruleset provided"),
+            field_name="support_to",
         )
 
+    async def _inner(self, article_outline: ArticleOutline, ruleset: RuleSet, field_name: str) -> ArticleOutline:
         for a in article_outline.iter_dfs():
-            if await self.evidently_judge(
+            if judge := await self.evidently_judge(
                 f"{article_outline.as_prompt()}\n\n{a.display()}\n"
-                f"Does the `{a.__class__.__name__}`'s `depend_on` field need to be extended or tweaked?"
+                f"Does the `{a.__class__.__name__}`'s `{field_name}` field need to be extended or tweaked?"
             ):
-                patch = ArticleRefPatch.default()
-                patch.tweaked = a.depend_on
+                patch = ArticleRefSequencePatch.default()
+                patch.tweaked = getattr(a, field_name)
 
-                await self.correct_obj_inplace(
+                await self.censor_obj_inplace(
                     patch,
-                    topic=topic,
-                    reference=f"{article_outline.as_prompt()}\nThe Article component whose `depend_on` field needs to be extended or tweaked",
-                    rating_manual=tweak_depend_on_manual,
-                    supervisor_check=supervisor_check,
+                    ruleset=ruleset,
+                    reference=f"{article_outline.as_prompt()}\n"
+                    f"The Article component titled `{a.title}` whose `{field_name}` field needs to be extended or tweaked.\n"
+                    f"# Judgement\n{judge.display()}",
                 )
-
         return article_outline
 
 
-class TweakOutlineForwardRef(Action, AdvancedJudge):
-    """Tweak the forward references in the article outline.
+class TweakOutlineBackwardRef(TweakOutlineForwardRef):
+    """Tweak the backward references in the article outline.
 
-    Ensures that the conclusions of the current chapter effectively support the analysis of subsequent chapters.
+    Ensures that the prerequisites of the current chapter are correctly referenced in the `depend_on` field.
     """
 
-    output_key: str = "article_outline_fw_ref_checked"
+    output_key: str = "article_outline_bw_ref_checked"
+    ruleset: Optional[RuleSet] = None
 
-    async def _execute(self, article_outline: ArticleOutline, supervisor_check: bool = False, **cxt) -> ArticleOutline:
-        tweak_support_to_manual = ok(
-            await self.draft_rating_manual(
-                topic := "Ensure conclusions support the analysis of subsequent chapters, sections or subsections."
-            ),
-            "Could not generate the rating manual.",
+    async def _execute(
+        self, article_outline: ArticleOutline, ruleset: Optional[RuleSet] = None, **cxt
+    ) -> ArticleOutline:
+        return await self._inner(
+            article_outline,
+            ruleset=ok(ruleset or self.ruleset, "No ruleset provided"),
+            field_name="depend_on",
         )
 
-        for a in article_outline.iter_dfs():
-            if await self.evidently_judge(
-                f"{article_outline.as_prompt()}\n\n{a.display()}\n"
-                f"Does the `{a.__class__.__name__}`'s `support_to` field need to be extended or tweaked?"
-            ):
-                patch = ArticleRefPatch.default()
-                patch.tweaked = a.support_to
-
-                await self.correct_obj_inplace(
-                    patch,
-                    topic=topic,
-                    reference=f"{article_outline.as_prompt()}\nThe Article component whose `support_to` field needs to be extended or tweaked",
-                    rating_manual=tweak_support_to_manual,
-                    supervisor_check=supervisor_check,
-                )
-
-        return article_outline
 
-
-class GenerateArticle(Action):
+class GenerateArticle(Action, Censor):
     """Generate the article based on the outline."""
 
     output_key: str = "article"
     """The key of the output data."""
+    ruleset: Optional[RuleSet] = None
 
     async def _execute(
         self,
         article_outline: ArticleOutline,
-        supervisor_check: bool = False,
+        ruleset: Optional[RuleSet] = None,
         **_,
     ) -> Optional[Article]:
         article: Article = Article.from_outline(ok(article_outline, "Article outline not specified.")).update_ref(
             article_outline
         )
 
-        write_para_manual = ok(
-            await self.draft_rating_manual(w_topic := "write the following paragraph in the subsection.")
-        )
-
         await gather(
             *[
-                self.correct_obj_inplace(
+                self.censor_obj_inplace(
                     subsec,
+                    ruleset=ok(ruleset or self.ruleset, "No ruleset provided"),
                     reference=f"# Original Article Outline\n{article_outline.display()}\n# Error Need to be fixed\n{err}",
-                    topic=w_topic,
-                    rating_manual=write_para_manual,
-                    supervisor_check=supervisor_check,
                 )
                 for _, __, subsec in article.iter_subsections()
                 if (err := subsec.introspect())
@@ -313,18 +308,16 @@ class GenerateArticle(Action):
         return article
 
 
-class CorrectProposal(Action):
+class CorrectProposal(Action, Censor):
     """Correct the proposal of the article."""
 
     output_key: str = "corrected_proposal"
 
     async def _execute(self, article_proposal: ArticleProposal, **_) -> Any:
-        return (await self.censor_obj(article_proposal, reference=article_proposal.referenced)).update_ref(
-            article_proposal
-        )
+        raise  NotImplementedError("Not implemented.")
 
 
-class CorrectOutline(Action):
+class CorrectOutline(Action, Correct):
     """Correct the outline of the article."""
 
     output_key: str = "corrected_outline"
@@ -335,12 +328,10 @@ class CorrectOutline(Action):
         article_outline: ArticleOutline,
         **_,
     ) -> ArticleOutline:
-        return (await self.censor_obj(article_outline, reference=article_outline.referenced.as_prompt())).update_ref(
-            article_outline
-        )
+        raise NotImplementedError("Not implemented.")
 
 
-class CorrectArticle(Action):
+class CorrectArticle(Action, Correct):
     """Correct the article based on the outline."""
 
     output_key: str = "corrected_article"
@@ -352,4 +343,4 @@ class CorrectArticle(Action):
         article_outline: ArticleOutline,
         **_,
     ) -> Article:
-        return await self.censor_obj(article, reference=article_outline.referenced.as_prompt())
+        raise NotImplementedError("Not implemented.")

fabricatio/actions/article_rag.py

@@ -1,74 +1,86 @@
 """A module for writing articles using RAG (Retrieval-Augmented Generation) capabilities."""
 
 from asyncio import gather
-from typing import Dict, Optional
+from typing import Optional
 
+from fabricatio.capabilities.censor import Censor
 from fabricatio.capabilities.rag import RAG
 from fabricatio.models.action import Action
-from fabricatio.models.extra.article_main import Article, ArticleParagraphPatch, ArticleSubsection
-from fabricatio.models.utils import ok
+from fabricatio.models.extra.article_main import Article, ArticleParagraphSequencePatch, ArticleSubsection
+from fabricatio.models.extra.rule import RuleSet
+from fabricatio.utils import ok
 
 
-class TweakArticleRAG(Action, RAG):
-    """Write an article based on the provided outline."""
+class TweakArticleRAG(Action, RAG, Censor):
+    """Write an article based on the provided outline.
+
+    This class inherits from `Action`, `RAG`, and `Censor` to provide capabilities for writing and refining articles
+    using Retrieval-Augmented Generation (RAG) techniques. It processes an article outline, enhances subsections by
+    searching for related references, and applies censoring rules to ensure compliance with the provided ruleset.
+
+    Attributes:
+        output_key (str): The key used to store the output of the action.
+        ruleset (Optional[RuleSet]): The ruleset to be used for censoring the article.
+    """
 
     output_key: str = "rag_tweaked_article"
+    """The key used to store the output of the action."""
+
+    ruleset: Optional[RuleSet] = None
+    """The ruleset to be used for censoring the article."""
 
     async def _execute(
         self,
         article: Article,
         collection_name: str = "article_essence",
-        citation_requirement: str = "# Citation Format\n"
-        "Use correct citation format based on author count. Cite using author surnames and year:"
-        "For 3+ authors: 'Author1, Author2 et al. (YYYY)'"
-        "For 2 authors: 'Author1 & Author2 (YYYY)'"
-        "Single author: 'Author1 (YYYY)'"
-        "Multiple citations: 'Author1 (YYYY), Author2 (YYYY)'"
-        "Prioritize formulas from reference highlights."
-        "Specify authors/years only - no numeric citations"
-        "Paragraphs must exceed 2-3 sentences",
-        supervisor_check: bool = False,
+        ruleset: Optional[RuleSet] = None,
         parallel: bool = False,
         **cxt,
     ) -> Optional[Article]:
-        """Write an article based on the provided outline."""
-        criteria = await self.draft_rating_criteria(
-            topic := "choose appropriate reference to insert into the article, "
-            "making conclusions or reasoning based on concrete evidence instead of unreliable guesses."
-            "Extensively use formulas highlighted in reference文献 should be translated to 'references'."
-            "Only specify authors and years without numeric citation numbers (like [1],[2])."
-            "Each paragraph should not end with just 2-3 sentences for better readability.",
-            criteria_count=13,
-        )
+        """Write an article based on the provided outline.
 
-        tweak_manual = ok(
-            await self.draft_rating_manual(topic, criteria=criteria, co_extractor={"model": "openai/qwen-max"})
-        )
+        This method processes the article outline, either in parallel or sequentially, by enhancing each subsection
+        with relevant references and applying censoring rules.
+
+        Args:
+            article (Article): The article to be processed.
+            collection_name (str): The name of the collection to view for processing.
+            ruleset (Optional[RuleSet]): The ruleset to apply for censoring. If not provided, the class's ruleset is used.
+            parallel (bool): If True, process subsections in parallel. Otherwise, process them sequentially.
+            **cxt: Additional context parameters.
+
+        Returns:
+            Optional[Article]: The processed article with enhanced subsections and applied censoring rules.
+        """
         self.view(collection_name)
 
         if parallel:
             await gather(
                 *[
-                    self._inner(article, subsec, supervisor_check, citation_requirement, topic, tweak_manual)
+                    self._inner(article, subsec, ok(ruleset or self.ruleset, "No ruleset provided!"))
                     for _, __, subsec in article.iter_subsections()
                 ],
                 return_exceptions=True,
             )
         else:
             for _, __, subsec in article.iter_subsections():
-                await self._inner(article, subsec, supervisor_check, citation_requirement, topic, tweak_manual)
-
+                await self._inner(article, subsec, ok(ruleset or self.ruleset, "No ruleset provided!"))
         return article
 
-    async def _inner(
-        self,
-        article: Article,
-        subsec: ArticleSubsection,
-        supervisor_check: bool,
-        citation_requirement: str,
-        topic: str,
-        tweak_manual: Dict[str, str],
-    ) -> None:
+    async def _inner(self, article: Article, subsec: ArticleSubsection, ruleset: RuleSet) -> None:
+        """Enhance a subsection of the article with references and apply censoring rules.
+
+        This method refines the query for the subsection, retrieves related references, and applies censoring rules
+        to the subsection's paragraphs.
+
+        Args:
+            article (Article): The article containing the subsection.
+            subsec (ArticleSubsection): The subsection to be enhanced.
+            ruleset (RuleSet): The ruleset to apply for censoring.
+
+        Returns:
+            None
+        """
         refind_q = ok(
             await self.arefined_query(
                 f"{article.referenced.as_prompt()}\n"
@@ -79,12 +91,10 @@ class TweakArticleRAG(Action, RAG):
                 f"prioritizing both original article language and English usage",
             )
         )
-        patch = ArticleParagraphPatch.default()
+        patch = ArticleParagraphSequencePatch.default()
         patch.tweaked = subsec.paragraphs
-        await self.correct_obj_inplace(
+        await self.censor_obj_inplace(
             patch,
-            reference=f"{await self.aretrieve_compact(refind_q, final_limit=50)}\n{citation_requirement}",
-            topic=topic,
-            rating_manual=tweak_manual,
-            supervisor_check=supervisor_check,
+            ruleset=ruleset,
+            reference=await self.aretrieve_compact(refind_q, final_limit=30),
         )

fabricatio/actions/output.py

@@ -7,7 +7,7 @@ from fabricatio.journal import logger
 from fabricatio.models.action import Action
 from fabricatio.models.generic import FinalizedDumpAble, PersistentAble
 from fabricatio.models.task import Task
-from fabricatio.models.utils import ok
+from fabricatio.utils import ok
 
 
 class DumpFinalizedOutput(Action):
@@ -67,17 +67,18 @@ class PersistentAll(Action):
 
         for k, v in cxt.items():
             final_dir = persist_dir.joinpath(k)
-            final_dir.mkdir(parents=True, exist_ok=True)
             if isinstance(v, PersistentAble):
+                final_dir.mkdir(parents=True, exist_ok=True)
                 v.persist(final_dir)
                 count += 1
             if isinstance(v, Iterable) and any(
                 persistent_ables := (pers for pers in v if isinstance(pers, PersistentAble))
             ):
+                final_dir.mkdir(parents=True, exist_ok=True)
                 for per in persistent_ables:
                     per.persist(final_dir)
                     count += 1
-
+        logger.info(f"Persisted {count} objects to {persist_dir}")
         return count
 
 

fabricatio/actions/rag.py

@@ -5,7 +5,7 @@ from typing import List, Optional
 from fabricatio.capabilities.rag import RAG
 from fabricatio.journal import logger
 from fabricatio.models.action import Action
-from fabricatio.models.generic import PrepareVectorization
+from fabricatio.models.generic import Vectorizable
 from fabricatio.models.task import Task
 from questionary import text
 
@@ -15,7 +15,7 @@ class InjectToDB(Action, RAG):
 
     output_key: str = "collection_name"
 
-    async def _execute[T: PrepareVectorization](
+    async def _execute[T: Vectorizable](
         self, to_inject: Optional[T] | List[Optional[T]], collection_name: str = "my_collection",override_inject:bool=False, **_
     ) -> Optional[str]:
         if not isinstance(to_inject, list):
@@ -27,7 +27,7 @@ class InjectToDB(Action, RAG):
             [
                 t.prepare_vectorization(self.embedding_max_sequence_length)
                 for t in to_inject
-                if isinstance(t, PrepareVectorization)
+                if isinstance(t, Vectorizable)
             ],
         )