From 36caf28e6baef9d9fe35b0013d18695221f4f4de Mon Sep 17 00:00:00 2001
From: wb200 <45331952+wb200@users.noreply.github.com>
Date: Sat, 30 Mar 2024 01:35:54 +0800
Subject: [PATCH] Update llama_parse_json_element.py to fix error exception
caused by certain PDF (#12402)
* Update llama_parse_json_element.py
Please refer to https://github.com/run-llama/llama_index/issues/12314 for a detailed explanation of the bug fix.
* cr
---------
Co-authored-by: Haotian Zhang <socool.king@gmail.com>
---
.../node_parser/relational/llama_parse_json_element.py | 7 ++++++-
1 file changed, 6 insertions(+), 1 deletion(-)
diff --git a/llama-index-core/llama_index/core/node_parser/relational/llama_parse_json_element.py b/llama-index-core/llama_index/core/node_parser/relational/llama_parse_json_element.py
index c92c7534eb..013db27743 100644
--- a/llama-index-core/llama_index/core/node_parser/relational/llama_parse_json_element.py
+++ b/llama-index-core/llama_index/core/node_parser/relational/llama_parse_json_element.py
@@ -260,7 +260,12 @@ class LlamaParseJsonNodeParser(BaseElementNodeParser):
and element.type == "text"
and merged_elements[-1].type == "text"
):
- merged_elements[-1].element += "\n" + element.element
+ if isinstance(element.element, list):
+ merged_elements[-1].element += "\n" + " ".join(
+ str(e) for e in element.element
+ )
+ else:
+ merged_elements[-1].element += "\n" + element.element
else:
merged_elements.append(element)
elements = merged_elements
--
GitLab