deepset-ai/haystack · error · ValueError

Parent document with id {parent_id} does not have any childr

Error message

Parent document with id {parent_id} does not have any children.

What it means

A parent document found in the store must declare its '__children_ids' meta so the retriever can merge siblings upward. If the parent exists but has no '__children_ids' (empty or missing), this ValueError is raised.

Source

Thrown at haystack/components/retrievers/auto_merging_retriever.py:135

        Recursively groups documents by their parents and merges them if they meet the threshold,
        continuing up the hierarchy until no more merges are possible.

        :param documents: List of leaf documents that were matched by a retriever
        :returns:
            List of documents (could be a mix of different hierarchy levels)
        """

        AutoMergingRetriever._check_valid_documents(documents)

        def _get_parent_doc(parent_id: str) -> Document:
            parent_docs = self.document_store.filter_documents({"field": "id", "operator": "==", "value": parent_id})
            if len(parent_docs) != 1:
                raise ValueError(f"Expected 1 parent document with id {parent_id}, found {len(parent_docs)}")

            parent_doc = parent_docs[0]
            if not parent_doc.meta.get("__children_ids"):
                raise ValueError(f"Parent document with id {parent_id} does not have any children.")

            return parent_doc

        def _try_merge_level(docs_to_merge: list[Document], docs_to_return: list[Document]) -> list[Document]:
            parent_doc_id_to_child_docs: dict[str, list[Document]] = defaultdict(list)  # to group documents by parent

            for doc in docs_to_merge:
                if doc.meta.get("__parent_id"):  # only docs that have parents
                    parent_doc_id_to_child_docs[doc.meta["__parent_id"]].append(doc)
                else:
                    docs_to_return.append(doc)  # keep docs that have no parents

            # Process each parent group
            merged_docs = []
            for parent_doc_id, child_docs in parent_doc_id_to_child_docs.items():
                parent_doc = _get_parent_doc(parent_doc_id)

                # Calculate merge score

View on GitHub (pinned to e318778c9b)

Solutions

1. Write the full hierarchy (parent + children with '__children_ids') to the store
2. Add '__children_ids' to the parent's meta and update the store
3. Re-index with HierarchicalDocumentSplitter

Example fix

// before
parent = Document(content="...")
store.write_documents([parent])
// after
parent = Document(content="...", meta={"__children_ids": [c.id for c in children]})
store.write_documents([parent])

for pid in {d.meta.get("__parent_id") for d in leaves}:
    parents = store.filter_documents({"field": "id", "operator": "==", "value": pid})
    if parents and not parents[0].meta.get("__children_ids"):
        raise ValueError(f"Parent {pid} has no __children_ids")

try:
    result = retriever.run(documents=leaves)
except ValueError as e:
    logger.warning("Parent without children meta: %s — returning leaf documents", e)
    result = {"documents": leaves}

When it happens

Trigger: Merging where the fetched parent document lacks meta['__children_ids'] — e.g. parent stored outside the hierarchical pipeline or meta overwritten.

Common situations: Parents written by a custom writer that skips splitter meta; documents re-serialized and losing dunder meta keys.

Related errors

• Expected 1 parent document with id {parent_id}, found {len(p
• The matched leaf documents do not have the required meta fie
• The matched leaf documents do not have the required meta fie
• The matched leaf documents do not have the required meta fie
• document_store must be an instance of InMemoryDocumentStore

AI-assisted analysis of deepset-ai/haystack@e318778c9b (2026-08-30). Data as JSON: /api/errors/eeb4f82ea9eaa017. Report an issue: GitHub.