gardel/lightrag
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'context_format_csv_to_json'

Browse Source
This commit is contained in:
yangdx
2025-04-22 12:25:50 +08:00
parent 6a727103d6 8064a2339f
commit 21c0bb7abf
2 changed files with 48 additions and 75 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
lightrag/operate.py
View File

@@ -14,7 +14,6 @@ from .utils import (
compute_mdhash_id,
Tokenizer,
is_float_regex,
list_of_list_to_csv,
normalize_extracted_info,
pack_user_ass_to_openai_messages,
split_string_by_multi_markers,
@@ -26,6 +25,7 @@ from .utils import (
CacheData,
get_conversation_turns,
use_llm_func_with_cache,
list_of_list_to_json,
)
from .base import (
BaseGraphStorage,
@@ -1333,21 +1333,26 @@ async def _build_query_context(
[hl_text_units_context, ll_text_units_context],
)
# not necessary to use LLM to generate a response
if not entities_context.strip() and not relations_context.strip():
if not entities_context and not relations_context:
return None
# 转换为 JSON 字符串
entities_str = json.dumps(entities_context, ensure_ascii=False)
relations_str = json.dumps(relations_context, ensure_ascii=False)
text_units_str = json.dumps(text_units_context, ensure_ascii=False)
result = f"""
-----Entities-----
```csv
{entities_context}
```json
{entities_str}
```
-----Relationships-----
```csv
{relations_context}
```json
{relations_str}
```
-----Sources-----
```csv
{text_units_context}
```json
{text_units_str}
```
""".strip()
return result
@@ -1453,7 +1458,7 @@ async def _get_node_data(
file_path,
]
)
entities_context = list_of_list_to_csv(entites_section_list)
entities_context = list_of_list_to_json(entites_section_list)
relations_section_list = [
[
@@ -1490,14 +1495,14 @@ async def _get_node_data(
file_path,
]
)
relations_context = list_of_list_to_csv(relations_section_list)
relations_context = list_of_list_to_json(relations_section_list)
text_units_section_list = [["id", "content", "file_path"]]
for i, t in enumerate(use_text_units):
text_units_section_list.append(
[i, t["content"], t.get("file_path", "unknown_source")]
)
text_units_context = list_of_list_to_csv(text_units_section_list)
text_units_context = list_of_list_to_json(text_units_section_list)
return entities_context, relations_context, text_units_context
@@ -1775,7 +1780,7 @@ async def _get_edge_data(
file_path,
]
)
relations_context = list_of_list_to_csv(relations_section_list)
relations_context = list_of_list_to_json(relations_section_list)
entites_section_list = [
["id", "entity", "type", "description", "rank", "created_at", "file_path"]
@@ -1800,12 +1805,12 @@ async def _get_edge_data(
file_path,
]
)
entities_context = list_of_list_to_csv(entites_section_list)
entities_context = list_of_list_to_json(entites_section_list)
text_units_section_list = [["id", "content", "file_path"]]
for i, t in enumerate(use_text_units):
text_units_section_list.append([i, t["content"], t.get("file_path", "unknown")])
text_units_context = list_of_list_to_csv(text_units_section_list)
text_units_context = list_of_list_to_json(text_units_section_list)
return entities_context, relations_context, text_units_context