rasa-pro 3.12.0.dev1__py3-none-any.whl

rasa/shared/core/training_data/story_writer/story_writer.py

@@ -0,0 +1,76 @@
+import typing
+from pathlib import Path
+from typing import List, Text, Union
+
+from ruamel import yaml
+
+if typing.TYPE_CHECKING:
+    from rasa.shared.core.events import Event
+    from rasa.shared.core.training_data.structures import StoryStep
+
+
+class StoryWriter:
+    """Writes story training data to file."""
+
+    def dumps(
+        self,
+        story_steps: List["StoryStep"],
+        is_appendable: bool = False,
+        is_test_story: bool = False,
+    ) -> Text:
+        """Turns Story steps into an string.
+
+        Args:
+            story_steps: Original story steps to be converted to the YAML.
+            is_appendable: Specify if result should not contain
+                           high level keys/definitions and can be appended to
+                           the existing story file.
+            is_test_story: Identifies if the stories should be exported in test stories
+                           format.
+
+        Returns:
+            String with story steps in the desired format.
+        """
+        raise NotImplementedError
+
+    def dump(
+        self,
+        target: Union[Text, Path, yaml.StringIO],
+        story_steps: List["StoryStep"],
+        is_appendable: bool = False,
+        is_test_story: bool = False,
+    ) -> None:
+        """Writes Story steps into a target file/stream.
+
+        Args:
+            target: name of the target file/stream to write the string to.
+            story_steps: Original story steps to be converted to the string.
+            is_appendable: Specify if result should not contain
+                           high level keys/definitions and can be appended to
+                           the existing story file.
+            is_test_story: Identifies if the stories should be exported in test stories
+                           format.
+        """
+        raise NotImplementedError
+
+    @staticmethod
+    def _filter_event(event: Union["Event", List["Event"]]) -> bool:
+        """Identifies if the event should be converted/written.
+
+        Args:
+            event: target event to check.
+
+        Returns:
+            `True` if the event should be converted/written, `False` otherwise.
+        """
+        from rasa.shared.core.training_data.structures import StoryStep
+
+        # This is an "OR" statement, so we accept it
+        if isinstance(event, list):
+            return True
+
+        return (
+            not StoryStep.is_action_listen(event)
+            and not StoryStep.is_action_unlikely_intent(event)
+            and not StoryStep.is_action_session_start(event)
+        )

rasa/shared/core/training_data/story_writer/yaml_story_writer.py

@@ -0,0 +1,444 @@
+from collections import OrderedDict
+from pathlib import Path
+from typing import Any, Dict, List, Text, Union, Optional
+
+from ruamel import yaml
+import json
+from ruamel.yaml.comments import CommentedMap
+from ruamel.yaml.scalarstring import DoubleQuotedScalarString, LiteralScalarString
+
+import rasa.shared.utils.io
+import rasa.shared.core.constants
+from rasa.shared.constants import LATEST_TRAINING_DATA_FORMAT_VERSION
+import rasa.shared.core.events
+from rasa.shared.core.constants import FLOW_HASHES_SLOT
+from rasa.shared.core.events import (
+    UserUttered,
+    ActionExecuted,
+    SlotSet,
+    ActiveLoop,
+    Event,
+    DialogueStackUpdated,
+)
+
+from rasa.shared.core.training_data.story_reader.yaml_story_reader import (
+    KEY_STORIES,
+    KEY_STORY_NAME,
+    KEY_USER_INTENT,
+    KEY_ENTITIES,
+    KEY_ACTION,
+    KEY_STEPS,
+    KEY_CHECKPOINT,
+    KEY_SLOT_NAME,
+    KEY_CHECKPOINT_SLOTS,
+    KEY_OR,
+    KEY_USER_MESSAGE,
+    KEY_ACTIVE_LOOP,
+    KEY_BOT_END_TO_END_MESSAGE,
+    KEY_RULES,
+    KEY_RULE_FOR_CONVERSATION_START,
+    KEY_WAIT_FOR_USER_INPUT_AFTER_RULE,
+    KEY_RULE_CONDITION,
+    KEY_RULE_NAME,
+    KEY_STACK_UPDATE,
+    KEY_COMMANDS,
+)
+
+from rasa.shared.core.training_data.story_writer.story_writer import StoryWriter
+from rasa.shared.core.training_data.structures import (
+    StoryStep,
+    Checkpoint,
+    STORY_START,
+    RuleStep,
+)
+from rasa.shared.utils.yaml import write_yaml
+
+
+class YAMLStoryWriter(StoryWriter):
+    """Writes Core training data into a file in a YAML format."""
+
+    def dumps(
+        self,
+        story_steps: List[StoryStep],
+        is_appendable: bool = False,
+        is_test_story: bool = False,
+    ) -> Text:
+        """Turns Story steps into an YAML string.
+
+        Args:
+            story_steps: Original story steps to be converted to the YAML.
+            is_appendable: Specify if result should not contain
+                           high level keys/definitions and can be appended to
+                           the existing story file.
+            is_test_story: Identifies if the stories should be exported in test stories
+                           format.
+
+        Returns:
+            String with story steps in the YAML format.
+        """
+        stream = yaml.StringIO()
+        self.dump(stream, story_steps, is_appendable, is_test_story)
+        return stream.getvalue()
+
+    def dump(
+        self,
+        target: Union[Text, Path, yaml.StringIO],
+        story_steps: List[StoryStep],
+        is_appendable: bool = False,
+        is_test_story: bool = False,
+    ) -> None:
+        """Writes Story steps into a target file/stream.
+
+        Args:
+            target: name of the target file/stream to write the YAML to.
+            story_steps: Original story steps to be converted to the YAML.
+            is_appendable: Specify if result should not contain
+                           high level keys/definitions and can be appended to
+                           the existing story file.
+            is_test_story: Identifies if the stories should be exported in test stories
+                           format.
+        """
+        result = self.stories_to_yaml(story_steps, is_test_story)
+        if is_appendable and KEY_STORIES in result:
+            result = result[KEY_STORIES]
+        write_yaml(result, target, True)
+
+    def stories_to_yaml(
+        self, story_steps: List[StoryStep], is_test_story: bool = False
+    ) -> Dict[Text, Any]:
+        """Converts a sequence of story steps into yaml format.
+
+        Args:
+            story_steps: Original story steps to be converted to the YAML.
+            is_test_story: `True` if the story is an end-to-end conversation test story.
+        """
+        from rasa.shared.utils.yaml import KEY_TRAINING_DATA_FORMAT_VERSION
+
+        self._is_test_story = is_test_story
+
+        stories = []
+        rules = []
+        for story_step in story_steps:
+            if isinstance(story_step, RuleStep):
+                rules.append(self.process_rule_step(story_step))
+            else:
+                stories.append(self.process_story_step(story_step))
+
+        result: OrderedDict[Text, Any] = OrderedDict()
+        result[KEY_TRAINING_DATA_FORMAT_VERSION] = DoubleQuotedScalarString(
+            LATEST_TRAINING_DATA_FORMAT_VERSION
+        )
+
+        if stories:
+            result[KEY_STORIES] = stories
+        if rules:
+            result[KEY_RULES] = rules
+
+        return result
+
+    def process_story_step(self, story_step: StoryStep) -> OrderedDict:
+        """Converts a single story step into an ordered dict.
+
+        Args:
+            story_step: A single story step to be converted to the dict.
+
+        Returns:
+            Dict with a story step.
+        """
+        result: OrderedDict[Text, Any] = OrderedDict()
+        result[KEY_STORY_NAME] = story_step.block_name
+        steps = self.process_checkpoints(story_step.start_checkpoints)
+        for event in story_step.events:
+            if not self._filter_event(event):
+                continue
+            processed = self.process_event(event)
+            if processed:
+                steps.append(processed)
+
+        steps.extend(self.process_checkpoints(story_step.end_checkpoints))
+
+        result[KEY_STEPS] = steps
+
+        return result
+
+    def process_event(self, event: Union[Event, List[Event]]) -> Optional[OrderedDict]:
+        """Process an event or list of events."""
+        if isinstance(event, list):
+            return self.process_or_utterances(event)
+        if isinstance(event, UserUttered):
+            return self.process_user_utterance(event, self._is_test_story)
+        if isinstance(event, ActionExecuted):
+            return self.process_action(event)
+        if isinstance(event, SlotSet):
+            return self.process_slot(event)
+        if isinstance(event, ActiveLoop):
+            return self.process_active_loop(event)
+        if isinstance(event, DialogueStackUpdated):
+            return self.process_stack(event)
+        return None
+
+    @staticmethod
+    def stories_contain_loops(stories: List[StoryStep]) -> bool:
+        """Checks if the stories contain at least one active loop.
+
+        Args:
+            stories: Stories steps.
+
+        Returns:
+            `True` if the `stories` contain at least one active loop.
+            `False` otherwise.
+        """
+        return any(
+            [
+                [event for event in story_step.events if isinstance(event, ActiveLoop)]
+                for story_step in stories
+            ]
+        )
+
+    @staticmethod
+    def process_user_utterance(
+        user_utterance: UserUttered, is_test_story: bool = False
+    ) -> OrderedDict:
+        """Converts a single user utterance into an ordered dict.
+
+        Args:
+            user_utterance: Original user utterance object.
+            is_test_story: Identifies if the user utterance should be added
+                           to the final YAML or not.
+
+        Returns:
+            Dict with a user utterance.
+        """
+        result = CommentedMap()
+        if user_utterance.intent_name and not user_utterance.use_text_for_featurization:
+            result[KEY_USER_INTENT] = (
+                user_utterance.full_retrieval_intent_name
+                if user_utterance.full_retrieval_intent_name
+                else user_utterance.intent_name
+            )
+
+        entities = []
+        if len(user_utterance.entities) and not is_test_story:
+            for entity in user_utterance.entities:
+                if "value" in entity:
+                    if hasattr(user_utterance, "inline_comment_for_entity"):
+                        # FIXME: to fix this type issue, WronglyClassifiedUserUtterance
+                        # needs to be imported but it's currently outside
+                        # of `rasa.shared`
+                        for predicted in user_utterance.predicted_entities:  # type: ignore[attr-defined]
+                            if predicted["start"] == entity["start"]:
+                                commented_entity = (
+                                    user_utterance.inline_comment_for_entity(
+                                        predicted, entity
+                                    )
+                                )
+                                if commented_entity:
+                                    entity_map = CommentedMap(
+                                        [(entity["entity"], entity["value"])]
+                                    )
+                                    entity_map.yaml_add_eol_comment(
+                                        commented_entity, entity["entity"]
+                                    )
+                                    entities.append(entity_map)
+                                else:
+                                    entities.append(
+                                        OrderedDict(
+                                            [(entity["entity"], entity["value"])]
+                                        )
+                                    )
+                    else:
+                        entities.append(
+                            OrderedDict([(entity["entity"], entity["value"])])
+                        )
+                else:
+                    entities.append(entity["entity"])
+            result[KEY_ENTITIES] = entities
+
+        if hasattr(user_utterance, "inline_comment"):
+            # FIXME: to fix this type issue, WronglyClassifiedUserUtterance needs to
+            # be imported but it's currently outside of `rasa.shared`
+            comment = user_utterance.inline_comment(
+                force_comment_generation=not entities
+            )
+            if comment:
+                result.yaml_add_eol_comment(comment, KEY_USER_INTENT)
+
+        if user_utterance.text and (
+            # We only print the utterance text if it was an end-to-end prediction
+            user_utterance.use_text_for_featurization
+            # or if we want to print a conversation test story.
+            or is_test_story
+        ):
+            result[KEY_USER_MESSAGE] = LiteralScalarString(
+                rasa.shared.core.events.format_message(
+                    user_utterance.text,
+                    user_utterance.intent_name,
+                    user_utterance.entities,
+                )
+            )
+
+        if user_utterance.commands:
+            result[KEY_COMMANDS] = user_utterance.commands
+        return result
+
+    @staticmethod
+    def process_action(action: ActionExecuted) -> Optional[OrderedDict]:
+        """Converts a single action into an ordered dict.
+
+        Args:
+            action: Original action object.
+
+        Returns:
+            Dict with an action.
+        """
+        if action.action_name == rasa.shared.core.constants.RULE_SNIPPET_ACTION_NAME:
+            return None
+
+        result = CommentedMap()
+        if action.action_name:
+            result[KEY_ACTION] = action.action_name
+        elif action.action_text:
+            result[KEY_BOT_END_TO_END_MESSAGE] = action.action_text
+
+        if hasattr(action, "inline_comment"):
+            # FIXME: to fix this type issue, WarningPredictedAction needs to
+            # be imported but it's currently outside of `rasa.shared`
+            comment = action.inline_comment()
+            if KEY_ACTION in result and comment:
+                result.yaml_add_eol_comment(comment, KEY_ACTION)
+            elif KEY_BOT_END_TO_END_MESSAGE in result and comment:
+                result.yaml_add_eol_comment(comment, KEY_BOT_END_TO_END_MESSAGE)
+
+        return result
+
+    @staticmethod
+    def process_slot(event: SlotSet) -> Optional[OrderedDict]:
+        """Converts a single `SlotSet` event into an ordered dict.
+
+        Args:
+            event: Original `SlotSet` event.
+
+        Returns:
+            OrderedDict with an `SlotSet` event.
+        """
+        if event.key == FLOW_HASHES_SLOT:
+            # this is a build in slot which should not be persisted
+            return None
+        return OrderedDict([(KEY_SLOT_NAME, [{event.key: event.value}])])
+
+    @staticmethod
+    def process_stack(event: DialogueStackUpdated) -> OrderedDict:
+        """Converts a stack event into an ordered dict.
+
+        Args:
+            event: Original stack event.
+
+        Returns:
+            OrderedDict with a stack event.
+        """
+        return OrderedDict([(KEY_STACK_UPDATE, json.loads(event.update))])
+
+    @staticmethod
+    def process_checkpoints(checkpoints: List[Checkpoint]) -> List[OrderedDict]:
+        """Converts checkpoints event into an ordered dict.
+
+        Args:
+            checkpoints: List of original checkpoint.
+
+        Returns:
+            List of converted checkpoints.
+        """
+        result = []
+        for checkpoint in checkpoints:
+            if checkpoint.name == STORY_START:
+                continue
+            next_checkpoint: OrderedDict[Text, Any] = OrderedDict(
+                [(KEY_CHECKPOINT, checkpoint.name)]
+            )
+            if checkpoint.conditions:
+                next_checkpoint[KEY_CHECKPOINT_SLOTS] = [
+                    {key: value} for key, value in checkpoint.conditions.items()
+                ]
+            result.append(next_checkpoint)
+        return result
+
+    def process_or_utterances(self, utterances: List[UserUttered]) -> OrderedDict:
+        """Converts user utterance containing the `OR` statement.
+
+        Args:
+            utterances: User utterances belonging to the same `OR` statement.
+
+        Returns:
+            Dict with converted user utterances.
+        """
+        return OrderedDict(
+            [
+                (
+                    KEY_OR,
+                    [
+                        self.process_user_utterance(utterance, self._is_test_story)
+                        for utterance in utterances
+                    ],
+                )
+            ]
+        )
+
+    @staticmethod
+    def process_active_loop(event: ActiveLoop) -> OrderedDict:
+        """Converts ActiveLoop event into an ordered dict.
+
+        Args:
+            event: ActiveLoop event.
+
+        Returns:
+            Converted event.
+        """
+        return OrderedDict([(KEY_ACTIVE_LOOP, event.name)])
+
+    def process_rule_step(self, rule_step: RuleStep) -> OrderedDict:
+        """Converts a RuleStep into an ordered dict.
+
+        Args:
+            rule_step: RuleStep object.
+
+        Returns:
+            Converted rule step.
+        """
+        result: OrderedDict[Text, Any] = OrderedDict()
+        result[KEY_RULE_NAME] = rule_step.block_name
+
+        condition_steps = []
+        condition_events = rule_step.get_rules_condition()
+        for event in condition_events:
+            processed = self.process_event(event)
+            if processed:
+                condition_steps.append(processed)
+        if condition_steps:
+            result[KEY_RULE_CONDITION] = condition_steps
+
+        normal_events = rule_step.get_rules_events()
+        if normal_events and not (
+            isinstance(normal_events[0], ActionExecuted)
+            and normal_events[0].action_name
+            == rasa.shared.core.constants.RULE_SNIPPET_ACTION_NAME
+        ):
+            result[KEY_RULE_FOR_CONVERSATION_START] = True
+
+        normal_steps = []
+        for event in normal_events:
+            processed = self.process_event(event)
+            if processed:
+                normal_steps.append(processed)
+        if normal_steps:
+            result[KEY_STEPS] = normal_steps
+
+        if len(normal_events) > 1:
+            last_event = normal_events[len(normal_events) - 1]
+            if (
+                isinstance(last_event, ActionExecuted)
+                and last_event.action_name
+                == rasa.shared.core.constants.RULE_SNIPPET_ACTION_NAME
+            ):
+                result[KEY_WAIT_FOR_USER_INPUT_AFTER_RULE] = False
+
+        return result