pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Feat: pipeline supports PPTX (#10167)

Browse Source 
### What problem does this PR solve?

Pipeline supports parsing PPTX naively (text only).

### Type of change

- [x] New Feature (non-breaking change which adds functionality)
This commit is contained in:
Yongteng Lei
2025-09-19 12:14:35 +08:00
committed by GitHub
parent 4fae40f66a
commit 5dfdbcce3a
4 changed files with 53 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
rag/flow/tokenizer/tokenizer.py
View File

@ -114,6 +114,9 @@ class Tokenizer(ProcessBase):
if from_upstream.chunks:
chunks = from_upstream.chunks
for i, ck in enumerate(chunks):
if ck.get("docnm_kwd"): # from presentation method
ck["title_tks"] = rag_tokenizer.tokenize(re.sub(r"\.[a-zA-Z]+$", "", ck["docnm_kwd"]))
ck["title_sm_tks"] = rag_tokenizer.fine_grained_tokenize(ck["title_tks"])
if ck.get("questions"):
ck["question_tks"] = rag_tokenizer.tokenize("\n".join(ck["questions"]))
if ck.get("keywords"):
@ -135,12 +138,18 @@ class Tokenizer(ProcessBase):
ck = {"text": payload}
if "full_text" in self._param.search_method:
if ck.get("docnm_kwd"): # from presentation method
ck["title_tks"] = rag_tokenizer.tokenize(re.sub(r"\.[a-zA-Z]+$", "", ck["docnm_kwd"]))
ck["title_sm_tks"] = rag_tokenizer.fine_grained_tokenize(ck["title_tks"])
ck["content_ltks"] = rag_tokenizer.tokenize(kwargs.get(kwargs["output_format"], ""))
ck["content_sm_ltks"] = rag_tokenizer.fine_grained_tokenize(ck["content_ltks"])
chunks = [ck]
else:
chunks = from_upstream.json_result
for i, ck in enumerate(chunks):
if ck.get("docnm_kwd"): # from presentation method
ck["title_tks"] = rag_tokenizer.tokenize(re.sub(r"\.[a-zA-Z]+$", "", ck["docnm_kwd"]))
ck["title_sm_tks"] = rag_tokenizer.fine_grained_tokenize(ck["title_tks"])
ck["content_ltks"] = rag_tokenizer.tokenize(ck["text"])
ck["content_sm_ltks"] = rag_tokenizer.fine_grained_tokenize(ck["content_ltks"])
if i % 100 == 99: