From 4a52a71956a8d46fcb7294ac71734504bb09bcc2 Mon Sep 17 00:00:00 2001
From: S. Solomon Darnell
Date: Fri, 28 Mar 2025 21:52:21 -0500
Subject: two version of R2R are here

---
 R2R/r2r/parsers/media/ppt_parser.py | 30 ++++++++++++++++++++++++++++++
 1 file changed, 30 insertions(+)
 create mode 100755 R2R/r2r/parsers/media/ppt_parser.py

(limited to 'R2R/r2r/parsers/media/ppt_parser.py')

diff --git a/R2R/r2r/parsers/media/ppt_parser.py b/R2R/r2r/parsers/media/ppt_parser.py
new file mode 100755
index 00000000..8f192840
--- /dev/null
+++ b/R2R/r2r/parsers/media/ppt_parser.py
@@ -0,0 +1,30 @@
+from io import BytesIO
+from typing import AsyncGenerator
+
+from r2r.base.abstractions.document import DataType
+from r2r.base.parsers.base_parser import AsyncParser
+
+
+class PPTParser(AsyncParser[DataType]):
+    """A parser for PPT data."""
+
+    def __init__(self):
+        try:
+            from pptx import Presentation
+
+            self.Presentation = Presentation
+        except ImportError:
+            raise ValueError(
+                "Error, `python-pptx` is required to run `PPTParser`. Please install it using `pip install python-pptx`."
+            )
+
+    async def ingest(self, data: DataType) -> AsyncGenerator[str, None]:
+        """Ingest PPT data and yield text from each slide."""
+        if isinstance(data, str):
+            raise ValueError("PPT data must be in bytes format.")
+
+        prs = self.Presentation(BytesIO(data))
+        for slide in prs.slides:
+            for shape in slide.shapes:
+                if hasattr(shape, "text"):
+                    yield shape.text
-- 
cgit v1.2.3