From 8e442f4f23fafe43521f3b7ef43bf65f56d03964 Mon Sep 17 00:00:00 2001
From: "codeflash-ai[bot]"
 <148906541+codeflash-ai[bot]@users.noreply.github.com>
Date: Thu, 30 Oct 2025 03:16:48 +0000
Subject: [PATCH] Optimize PPTParser.convert_ppt_to_pptx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The optimization introduces **caching for LibreOffice availability checking** using a class-level attribute `_libreoffice_found`. This eliminates the repeated `subprocess.run(["which", "libreoffice"])` call that was executed on every method invocation.

**Key changes:**
- **LibreOffice check caching**: The availability check now runs only once per class lifetime, storing the result in `self.__class__._libreoffice_found`
- **Direct return optimization**: Removed intermediate variable `pptx_content` and return file content directly

**Performance impact:**
The line profiler shows the LibreOffice check (`subprocess.run`) takes ~3.7ms and represents 78-93% of total execution time. By caching this check, subsequent calls skip this expensive operation entirely. The optimization is most effective for:

- **Batch processing scenarios**: When converting multiple PPT files in sequence, only the first call pays the LibreOffice check cost
- **Repeated conversions**: Applications that perform multiple conversions benefit immediately after the first successful check
- **High-frequency usage**: Services processing many PPT files see cumulative time savings

The 125% speedup (2.94ms → 1.30ms) demonstrates significant improvement, particularly valuable in production environments where PPT conversion happens repeatedly with the same parser instance.
---
 .../app/modules/parsers/pptx/ppt_parser.py    | 19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

diff --git a/backend/python/app/modules/parsers/pptx/ppt_parser.py b/backend/python/app/modules/parsers/pptx/ppt_parser.py
index 3ca2322590..f5d55d7d95 100644
--- a/backend/python/app/modules/parsers/pptx/ppt_parser.py
+++ b/backend/python/app/modules/parsers/pptx/ppt_parser.py
@@ -23,13 +23,24 @@ def convert_ppt_to_pptx(self, binary: bytes) -> bytes:
             FileNotFoundError: If the converted file is not found
             Exception: For other conversion errors
         """
-        with tempfile.TemporaryDirectory() as temp_dir:
+        if not hasattr(self.__class__, "_libreoffice_found"):
             try:
-                # Check if LibreOffice is installed
                 subprocess.run(
                     ["which", "libreoffice"], check=True, capture_output=True
                 )
+                self.__class__._libreoffice_found = True
+            except subprocess.CalledProcessError as e:
+                error_msg = "LibreOffice is not installed. Please install it using: sudo apt-get install libreoffice"
+                if e.stderr:
+                    error_msg += (
+                        f"\nError details: {e.stderr.decode('utf-8', errors='replace')}"
+                    )
+                raise subprocess.CalledProcessError(
+                    e.returncode, e.cmd, output=e.output, stderr=error_msg.encode()
+                )
 
+        with tempfile.TemporaryDirectory() as temp_dir:
+            try:
                 # Create input file path
                 temp_ppt = os.path.join(temp_dir, "input.ppt")
 
@@ -63,9 +74,7 @@ def convert_ppt_to_pptx(self, binary: bytes) -> bytes:
 
                 # Read the converted file into bytes
                 with open(pptx_file, "rb") as f:
-                    pptx_content = f.read()
-
-                return pptx_content
+                    return f.read()
 
             except subprocess.CalledProcessError as e:
                 error_msg = "LibreOffice is not installed. Please install it using: sudo apt-get install libreoffice"