port-ocean 0.28.12__py3-none-any.whl → 0.28.14__py3-none-any.whl

integrations/_infra/Dockerfile.Deb

@@ -48,6 +48,7 @@ RUN apt-get update \
         curl \
         acl \
         sudo \
+        jq \
     && apt-get clean
 
 LABEL INTEGRATION_VERSION=${INTEGRATION_VERSION}

integrations/_infra/Dockerfile.local

@@ -26,6 +26,7 @@ RUN apt-get update \
         python3-pip \
         python3-poetry \
         build-essential\
+        jq \
         git \
         python3-venv \
         acl \

port_ocean/core/handlers/entity_processor/jq_entity_processor.py

@@ -1,12 +1,11 @@
 import asyncio
 from asyncio import Task
 from dataclasses import dataclass, field
-
 from functools import lru_cache
+import json
 from typing import Any, Optional
 import jq  # type: ignore
 from loguru import logger
-
 from port_ocean.context.ocean import ocean
 from port_ocean.core.handlers.entity_processor.base import BaseEntityProcessor
 from port_ocean.core.handlers.port_app_config.models import ResourceConfig
@@ -22,6 +21,11 @@ from port_ocean.core.utils.utils import (
 )
 from port_ocean.exceptions.core import EntityProcessorException
 from port_ocean.utils.queue_utils import process_in_queue
+from port_ocean.core.handlers.entity_processor.jq_input_evaluator import (
+    InputClassifyingResult,
+    classify_input,
+    can_expression_run_with_no_input,
+)
 
 
 class ExampleStates:
@@ -76,7 +80,7 @@ class MappedEntity:
 
     entity: dict[str, Any] = field(default_factory=dict)
     did_entity_pass_selector: bool = False
-    raw_data: Optional[dict[str, Any]] = None
+    raw_data: Optional[dict[str, Any] | tuple[dict[str, Any], str]] = None
     misconfigurations: dict[str, str] = field(default_factory=dict)
 
 
@@ -133,17 +137,35 @@ class JQEntityProcessor(BaseEntityProcessor):
             return await loop.run_in_executor(
                 None, self._stop_iterator_handler(func.first)
             )
+        except Exception as exc:
+            logger.error(
+                f"Search failed for pattern '{pattern}' in data: {data}, Error: {exc}"
+            )
+            return None
+
+    @lru_cache
+    async def _search_stringified(self, data: str, pattern: str) -> Any:
+        try:
+            loop = asyncio.get_event_loop()
+            compiled_pattern = self._compile(pattern)
+            func = compiled_pattern.input_text(data)
+            return await loop.run_in_executor(
+                None, self._stop_iterator_handler(func.first)
+            )
         except Exception as exc:
             logger.debug(
                 f"Search failed for pattern '{pattern}' in data: {data}, Error: {exc}"
             )
             return None
 
-    async def _search_as_bool(self, data: dict[str, Any], pattern: str) -> bool:
+    async def _search_as_bool(self, data: dict[str, Any] | str, pattern: str) -> bool:
         loop = asyncio.get_event_loop()
 
         compiled_pattern = self._compile(pattern)
-        func = compiled_pattern.input_value(data)
+        if isinstance(data, str):
+            func = compiled_pattern.input_text(data)
+        else:
+            func = compiled_pattern.input_value(data)
 
         value = await loop.run_in_executor(
             None, self._stop_iterator_handler(func.first)
@@ -156,7 +178,7 @@ class JQEntityProcessor(BaseEntityProcessor):
 
     async def _search_as_object(
         self,
-        data: dict[str, Any],
+        data: dict[str, Any] | str,
         obj: dict[str, Any],
         misconfigurations: dict[str, str] | None = None,
     ) -> dict[str, Any | None]:
@@ -188,7 +210,12 @@ class JQEntityProcessor(BaseEntityProcessor):
                     self._search_as_object(data, value, misconfigurations)
                 )
             else:
-                search_tasks[key] = asyncio.create_task(self._search(data, value))
+                if isinstance(data, str):
+                    search_tasks[key] = asyncio.create_task(
+                        self._search_stringified(data, value)
+                    )
+                else:
+                    search_tasks[key] = asyncio.create_task(self._search(data, value))
 
         result: dict[str, Any | None] = {}
         for key, task in search_tasks.items():
@@ -212,16 +239,18 @@ class JQEntityProcessor(BaseEntityProcessor):
 
     async def _get_mapped_entity(
         self,
-        data: dict[str, Any],
+        data: dict[str, Any] | tuple[dict[str, Any], str],
         raw_entity_mappings: dict[str, Any],
+        items_to_parse_key: str | None,
         selector_query: str,
         parse_all: bool = False,
     ) -> MappedEntity:
-        should_run = await self._search_as_bool(data, selector_query)
+        should_run = await self._should_map_entity(
+            data, selector_query, items_to_parse_key
+        )
         if parse_all or should_run:
-            misconfigurations: dict[str, str] = {}
-            mapped_entity = await self._search_as_object(
-                data, raw_entity_mappings, misconfigurations
+            misconfigurations, mapped_entity = await self._map_entity(
+                data, raw_entity_mappings, items_to_parse_key
             )
             return MappedEntity(
                 mapped_entity,
@@ -237,6 +266,77 @@ class JQEntityProcessor(BaseEntityProcessor):
             misconfigurations={},
         )
 
+    async def _map_entity(
+        self,
+        data: dict[str, Any] | tuple[dict[str, Any], str],
+        raw_entity_mappings: dict[str, Any],
+        items_to_parse_key: str | None,
+    ) -> tuple[dict[str, str], dict[str, Any]]:
+        if not items_to_parse_key:
+            # No items to parse, map the entity and return the misconfigurations and the mapped entity
+            misconfigurations: dict[str, str] = {}
+            data_to_search = data if isinstance(data, dict) else data[0]
+            mapped_entity = await self._search_as_object(
+                data_to_search, raw_entity_mappings, misconfigurations
+            )
+            return misconfigurations, mapped_entity
+
+        modified_data: tuple[dict[str, Any], str | dict[str, Any]] = (
+            data
+            if isinstance(data, tuple)
+            else (
+                {items_to_parse_key: data[items_to_parse_key]},
+                data,
+            )
+        )
+
+        misconfigurations_item: dict[str, str] = {}
+        misconfigurations_all: dict[str, str] = {}
+        # Map the entity with jq expressions that classified as single item expressions with the single item as input
+        mapped_entity_item = await self._search_as_object(
+            modified_data[0], raw_entity_mappings["item"], misconfigurations_item
+        )
+        # To Prevent misclassification data loss, we merge the expressions that classified as single item expressions and resulted as misconfigured
+        # into the expressions that classified as all expressions
+        if misconfigurations_item:
+            # The misconfigurations dict not contains the mapping expressions themselves, so we need to filter the original mapping by the misconfigured keys
+            filtered_item_mappings = self._filter_mappings_by_keys(
+                raw_entity_mappings["item"], list(misconfigurations_item.keys())
+            )
+            raw_entity_mappings["all"] = self._deep_merge(
+                raw_entity_mappings["all"], filtered_item_mappings
+            )
+        # Map the entity with jq expressions that classified as all expressions with the whole data as input
+        mapped_entity_all = await self._search_as_object(
+            modified_data[1], raw_entity_mappings["all"], misconfigurations_all
+        )
+        # Map the entity with jq expressions that classified as no input required expressions with empty object as input
+        mapped_entity_empty = await self._search_as_object(
+            {}, raw_entity_mappings["empty"], misconfigurations_all
+        )
+        # Merge the mapped entities
+        mapped_entity = self._deep_merge(mapped_entity_item, mapped_entity_all)
+        mapped_entity = self._deep_merge(mapped_entity, mapped_entity_empty)
+        return misconfigurations_all, mapped_entity
+
+    async def _should_map_entity(
+        self,
+        data: dict[str, Any] | tuple[dict[str, Any], str],
+        selector_query: str,
+        items_to_parse_key: str | None,
+    ) -> bool:
+        if can_expression_run_with_no_input(selector_query):
+            return await self._search_as_bool({}, selector_query)
+        if isinstance(data, tuple):
+            return await self._search_as_bool(
+                data[0], selector_query
+            ) or await self._search_as_bool(data[1], selector_query)
+        if items_to_parse_key:
+            return await self._search_as_bool(
+                data[items_to_parse_key], selector_query
+            ) or await self._search_as_bool(data, selector_query)
+        return await self._search_as_bool(data, selector_query)
+
     async def _calculate_entity(
         self,
         data: dict[str, Any],
@@ -246,9 +346,17 @@ class JQEntityProcessor(BaseEntityProcessor):
         selector_query: str,
         parse_all: bool = False,
     ) -> tuple[list[MappedEntity], list[Exception]]:
-        raw_data = [data.copy()]
-        if not ocean.config.yield_items_to_parse:
-            if items_to_parse:
+        raw_data: list[dict[str, Any]] | list[tuple[dict[str, Any], str]] = [
+            data.copy()
+        ]
+        items_to_parse_key = None
+        if items_to_parse:
+            items_to_parse_key = items_to_parse_name
+            if not ocean.config.yield_items_to_parse:
+                if isinstance(data, dict) and data.get("__type") == "path":
+                    file_path = data.get("file", {}).get("content", {}).get("path")
+                    with open(file_path, "r") as f:
+                        data["file"]["content"] = json.loads(f.read())
                 items = await self._search(data, items_to_parse)
                 if not isinstance(items, list):
                     logger.warning(
@@ -256,13 +364,28 @@ class JQEntityProcessor(BaseEntityProcessor):
                         f" Skipping..."
                     )
                     return [], []
-                raw_data = [{items_to_parse_name: item, **data} for item in items]
+                raw_all_payload_stringified = json.dumps(data)
+                raw_data = [
+                    ({items_to_parse_name: item}, raw_all_payload_stringified)
+                    for item in items
+                ]
+            single_item_mappings, all_items_mappings, empty_items_mappings = (
+                self._build_raw_entity_mappings(
+                    raw_entity_mappings, items_to_parse_name
+                )
+            )
+            raw_entity_mappings = {
+                "item": single_item_mappings,
+                "all": all_items_mappings,
+                "empty": empty_items_mappings,
+            }
 
         entities, errors = await gather_and_split_errors_from_results(
             [
                 self._get_mapped_entity(
                     raw,
                     raw_entity_mappings,
+                    items_to_parse_key,
                     selector_query,
                     parse_all,
                 )
@@ -275,6 +398,325 @@ class JQEntityProcessor(BaseEntityProcessor):
             )
         return entities, errors
 
+    def _build_raw_entity_mappings(
+        self, raw_entity_mappings: dict[str, Any], items_to_parse_name: str
+    ) -> tuple[dict[str, Any], dict[str, Any], dict[str, Any]]:
+        """
+        Build the raw entity mappings for the items to parse.
+        The mappings are grouped by the input classifying result.
+        There are 3 input classifying results:
+            - NONE: The expression can be executed with no input
+            - SINGLE: The expression can be executed on a single item
+            - ALL: The expression can be executed on all the data
+        """
+        mappings: dict[InputClassifyingResult, dict[str, Any]] = {
+            InputClassifyingResult.NONE: {},
+            InputClassifyingResult.SINGLE: {},
+            InputClassifyingResult.ALL: {},
+        }
+        for key, value in raw_entity_mappings.items():
+            if isinstance(value, str):
+                # Direct string values (identifier, title, icon, blueprint, team)
+                self.group_string_mapping_value(
+                    items_to_parse_name,
+                    mappings,
+                    key,
+                    value,
+                )
+            elif isinstance(value, dict):
+                # Complex objects (IngestSearchQuery for identifier/team, properties, relations)
+                self.group_complex_mapping_value(
+                    items_to_parse_name,
+                    mappings,
+                    key,
+                    value,
+                )
+        return (
+            mappings[InputClassifyingResult.SINGLE],
+            mappings[InputClassifyingResult.ALL],
+            mappings[InputClassifyingResult.NONE],
+        )
+
+    def group_complex_mapping_value(
+        self,
+        pattern: str,
+        mappings: dict[InputClassifyingResult, dict[str, Any]],
+        key: str,
+        value: dict[str, Any],
+    ) -> None:
+
+        if key in ["properties", "relations"]:
+            mapping_dicts: dict[InputClassifyingResult, dict[str, Any]] = {
+                InputClassifyingResult.SINGLE: {},
+                InputClassifyingResult.ALL: {},
+                InputClassifyingResult.NONE: {},
+            }
+            # For properties and relations, filter the dictionary values
+            for dict_key, dict_value in value.items():
+                if isinstance(dict_value, str):
+                    self.group_string_mapping_value(
+                        pattern,
+                        mapping_dicts,
+                        dict_key,
+                        dict_value,
+                    )
+                elif isinstance(dict_value, dict):
+                    # Handle IngestSearchQuery objects
+                    self.group_search_query_mapping_value(
+                        pattern,
+                        mapping_dicts[InputClassifyingResult.SINGLE],
+                        mapping_dicts[InputClassifyingResult.ALL],
+                        dict_key,
+                        dict_value,
+                    )
+            for input_classifying_result, mapping_dict in mapping_dicts.items():
+                if mapping_dict:
+                    mappings[input_classifying_result][key] = mapping_dict
+        else:
+            # For identifier/team IngestSearchQuery objects
+            self.group_search_query_mapping_value(
+                pattern,
+                mappings[InputClassifyingResult.SINGLE],
+                mappings[InputClassifyingResult.ALL],
+                key,
+                value,
+            )
+
+    def group_search_query_mapping_value(
+        self,
+        pattern: str,
+        single_item_dict: dict[str, Any],
+        all_item_dict: dict[str, Any],
+        dict_key: str,
+        dict_value: dict[str, Any],
+    ) -> None:
+        if self._classify_search_query(dict_value, pattern):
+            single_item_dict[dict_key] = dict_value
+        else:
+            all_item_dict[dict_key] = dict_value
+
+    def group_string_mapping_value(
+        self,
+        pattern: str,
+        mappings: dict[InputClassifyingResult, dict[str, Any]],
+        key: str,
+        value: str,
+    ) -> None:
+        input_evaluation_result = classify_input(value, pattern)
+        mappings[input_evaluation_result][key] = value
+
+    def _classify_search_query(self, query_dict: dict[str, Any], pattern: str) -> bool:
+        """
+        Classify the input required to run jq expressions of an IngestSearchQuery
+        If at least one rule contains a value that includes the pattern, return True
+        Otherwise, return False
+        Example:
+        The pattern is item
+        relations:
+            someRelation:
+              combinator: "and"
+              rules:
+                - operator: "="
+                  property: "someProperty"
+                  value: .field
+                - combinator: "or"
+                  rules:
+                    - operator: "="
+                      property: "anotherProperty"
+                      value: .item.something
+                    - operator: "="
+                      property: "yetAnotherProperty"
+                      value: .yetAnotherValue
+        One value is .item.something, which is a single item based expression, so it will be classified as SINGLE
+        """
+        if "rules" not in query_dict:
+            return False
+
+        rules = query_dict["rules"]
+        if not isinstance(rules, list):
+            return False
+
+        # Check if any rule contains a value that includes the pattern
+        for rule in rules:
+            if isinstance(rule, dict) and self._is_rule_or_query_contains_pattern(
+                rule, pattern
+            ):
+                return True
+        return False
+
+    def _is_rule_or_query_contains_pattern(
+        self, rule: dict[str, Any], pattern: str
+    ) -> bool:
+        """
+        Check if a rule or query contains a value that includes the pattern
+        If the value is a single item based expression, return True
+        Otherwise, return False
+        Example:
+        The pattern is item
+        The rule is:
+                - combinator: "or"
+                  rules:
+                    - operator: "="
+                      property: "anotherProperty"
+                      value: .item.something
+                    - operator: "="
+                      property: "yetAnotherProperty"
+                      value: .yetAnotherValue
+        This rule is not a single rule (not contains a value property) but a search query rule (contains a rules property)
+        so we need to recursively check the rules property to check if at least one rule contains a value that includes the pattern
+        In this case, one value is .item.something, which is a single item based expression, so it will be classified as SINGLE ==> true
+        """
+        if "value" in rule and isinstance(rule["value"], str):
+            # Use evaluate_input to check if the pattern is relevant for this value
+            input_evaluation_result = classify_input(rule["value"], pattern)
+            if input_evaluation_result == InputClassifyingResult.SINGLE:
+                return True
+        # Recursively check nested IngestSearchQuery objects
+        elif "rules" in rule:
+            if self._classify_search_query(rule, pattern):
+                return True
+        return False
+
+    def _filter_mappings_by_keys(
+        self, mappings: dict[str, Any], target_keys: list[str]
+    ) -> dict[str, Any]:
+        """
+        Filter mappings to preserve structure with only the specified keys present.
+        Recursively handles nested dictionaries and lists, searching for keys at any level.
+
+        Args:
+            mappings: The dictionary containing mapping configurations to filter
+            target_keys: List of keys to preserve in the filtered result
+
+        Returns:
+            A filtered dictionary containing only the specified keys and their nested structures
+
+        Examples:
+            # Basic filtering with direct keys
+            mappings = {
+                "name": "John",
+                "age": 30,
+                "city": "New York",
+                "country": "USA"
+            }
+            target_keys = ["name", "age"]
+            result = self._filter_mappings_by_keys(mappings, target_keys)
+            # Returns: {"name": "John", "age": 30}
+
+            # Nested dictionary filtering
+            mappings = {
+                "user": {
+                    "profile": {"name": "John", "email": "john@example.com"},
+                    "settings": {"theme": "dark", "notifications": True}
+                },
+                "metadata": {"created_at": "2023-01-01", "version": "1.0"}
+            }
+            target_keys = ["name", "email", "created_at"]
+            result = self._filter_mappings_by_keys(mappings, target_keys)
+            # Returns: {
+            #     "user": {
+            #         "profile": {"name": "John", "email": "john@example.com"}
+            #     },
+            #     "metadata": {"created_at": "2023-01-01"}
+            # }
+
+            # Empty target keys returns empty dict
+            result = self._filter_mappings_by_keys(mappings, [])
+            # Returns: {}
+        """
+        if not target_keys:
+            return {}
+
+        filtered_mappings: dict[str, Any] = {}
+
+        for key, value in mappings.items():
+            filtered_value = self._process_mapping_value(key, value, target_keys)
+
+            # Include if it's a direct match or contains nested target keys
+            if key in target_keys or filtered_value:
+                filtered_mappings[key] = filtered_value
+
+        return filtered_mappings
+
+    def _process_mapping_value(
+        self, key: str, value: Any, target_keys: list[str]
+    ) -> Any:
+        """
+        Process a single mapping value, handling different types recursively.
+
+        This helper method is used by _filter_mappings_by_keys to process individual
+        key-value pairs. It handles both simple values and nested dictionaries,
+        applying the filtering logic recursively to maintain the hierarchical structure.
+
+        Args:
+            key: The key of the current mapping item being processed
+            value: The value associated with the key (can be any type)
+            target_keys: List of keys to preserve in the filtered result
+
+        Returns:
+            The processed value if it should be included, None otherwise.
+            For dictionaries, returns the filtered dictionary or None if empty.
+            For simple values, returns the value if the key is in target_keys, None otherwise.
+
+        Examples:
+            # Simple value processing - key in target_keys
+            result = self._process_mapping_value("name", "John", ["name", "age"])
+            # Returns: "John"
+
+            # Simple value processing - key not in target_keys
+            result = self._process_mapping_value("city", "New York", ["name", "age"])
+            # Returns: None
+
+            # Dictionary processing with nested target keys
+            nested_dict = {
+                "profile": {"name": "John", "email": "john@example.com"},
+                "settings": {"theme": "dark"}
+            }
+            result = self._process_mapping_value("user", nested_dict, ["name", "email"])
+            # Returns: {"profile": {"name": "John", "email": "john@example.com"}}
+
+            # Dictionary processing with no matching target keys
+            result = self._process_mapping_value("user", nested_dict, ["version"])
+            # Returns: None
+        """
+        if isinstance(value, dict):
+            # Recursively filter nested dictionary
+            filtered_dict = self._filter_mappings_by_keys(value, target_keys)
+            return filtered_dict if filtered_dict else None
+        else:
+            # Return simple values as-is
+            return value if key in target_keys else None
+
+    def _deep_merge(
+        self, dict1: dict[str, Any], dict2: dict[str, Any]
+    ) -> dict[str, Any]:
+        """
+        Deep merge two dictionaries, preserving nested structures.
+        Values from dict2 override values from dict1 for the same keys.
+        """
+        result = dict1.copy()
+
+        for key, value in dict2.items():
+            if (
+                key in result
+                and isinstance(result[key], dict)
+                and isinstance(value, dict)
+            ):
+                # Recursively merge nested dictionaries
+                result[key] = self._deep_merge(result[key], value)
+            elif (
+                key in result
+                and isinstance(result[key], list)
+                and isinstance(value, list)
+            ):
+                # Merge lists by extending
+                result[key].extend(value)
+            else:
+                # Override with value from dict2
+                result[key] = value
+
+        return result
+
     @staticmethod
     async def _send_examples(data: list[dict[str, Any]], kind: str) -> None:
         try:
@@ -329,7 +771,10 @@ class JQEntityProcessor(BaseEntityProcessor):
                 and result.raw_data is not None
             ):
                 examples_to_send.add_example(
-                    result.did_entity_pass_selector, result.raw_data
+                    result.did_entity_pass_selector,
+                    self._get_raw_data_for_example(
+                        result.raw_data, mapping.port.items_to_parse_name
+                    ),
                 )
 
             if result.entity.get("identifier") and result.entity.get("blueprint"):
@@ -355,3 +800,13 @@ class JQEntityProcessor(BaseEntityProcessor):
             errors,
             misconfigured_entity_keys=entity_misconfigurations,
         )
+
+    def _get_raw_data_for_example(
+        self,
+        data: dict[str, Any] | tuple[dict[str, Any], str],
+        items_to_parse_name: str,
+    ) -> dict[str, Any]:
+        if isinstance(data, tuple):
+            raw_data = json.loads(data[1])
+            return {items_to_parse_name: data[0], **raw_data}
+        return data