pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2026-02-01 08:05:07 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refa: remove ppt image. (#12909)

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

remove `aspose`

### Type of change

- [x] Refactoring
This commit is contained in:
Kevin Hu
2026-01-30 13:35:42 +08:00
committed by GitHub
parent 73645e2f78
commit f1c2fac03e
5 changed files with 13 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
rag/app/presentation.py
View File

@ -23,36 +23,13 @@ from PIL import Image
from PyPDF2 import PdfReader as pdf2_read
from deepdoc.parser import PdfParser, PptParser, PlainParser
from deepdoc.parser.ppt_parser import RAGFlowPptParser
from rag.app.naive import by_plaintext, PARSERS
from common.parser_config_utils import normalize_layout_recognizer
from rag.nlp import rag_tokenizer
from rag.nlp import tokenize, is_english
class Ppt(PptParser):
def __call__(self, fnm, from_page, to_page, callback=None):
txts = super().__call__(fnm, from_page, to_page)
callback(0.5, "Text extraction finished.")
import aspose.slides as slides
import aspose.pydrawing as drawing
imgs = []
with slides.Presentation(BytesIO(fnm)) as presentation:
for i, slide in enumerate(presentation.slides[from_page:to_page]):
try:
with BytesIO() as buffered:
slide.get_thumbnail(0.1, 0.1).save(buffered, drawing.imaging.ImageFormat.jpeg)
buffered.seek(0)
imgs.append(Image.open(buffered).copy())
except RuntimeError as e:
raise RuntimeError(f"ppt parse error at page {i + 1}, original error: {str(e)}") from e
assert len(imgs) == len(txts), "Slides text and image do not match: {} vs. {}".format(len(imgs), len(txts))
callback(0.9, "Image extraction finished")
self.is_english = is_english(txts)
return [(txts[i], imgs[i]) for i in range(len(txts))]
class Pdf(PdfParser):
def __init__(self):
super().__init__()
@ -159,15 +136,14 @@ def chunk(filename, binary=None, from_page=0, to_page=100000, lang="Chinese", ca
doc["title_sm_tks"] = rag_tokenizer.fine_grained_tokenize(doc["title_tks"])
res = []
if re.search(r"\.pptx?$", filename, re.IGNORECASE):
ppt_parser = Ppt()
for pn, (txt, img) in enumerate(ppt_parser(filename if not binary else binary, from_page, 1000000, callback)):
ppt_parser = RAGFlowPptParser()
for pn, txt in enumerate(ppt_parser(filename if not binary else binary, from_page, 1000000, callback)):
d = copy.deepcopy(doc)
pn += from_page
d["image"] = img
d["doc_type_kwd"] = "image"
d["page_num_int"] = [pn + 1]
d["top_int"] = [0]
d["position_int"] = [(pn + 1, 0, img.size[0], 0, img.size[1])]
d["position_int"] = [(pn + 1, 0, 0, 0, 0)]
tokenize(d, txt, eng)
res.append(d)
return res