From 4f5cacca468727b66ebdd31c0d2870a25377619a Mon Sep 17 00:00:00 2001
From: Hikari Haru <rene9917@gmail.com>
Date: Fri, 17 Mar 2023 13:10:07 +0100
Subject: [PATCH] Fix office file indexing (#230)

---
 models/index_model.py | 7 ++++++-
 pyproject.toml        | 3 ++-
 requirements.txt      | 1 +
 requirements_base.txt | 3 ++-
 4 files changed, 11 insertions(+), 3 deletions(-)

diff --git a/models/index_model.py b/models/index_model.py
index 855481b..e30e032 100644
--- a/models/index_model.py
+++ b/models/index_model.py
@@ -381,7 +381,12 @@ class Index_handler:
             "application/json": ".json",
             "image/png": ".png",
             "image/": ".jpg",
-            "vnd.": ".pptx",
+            "ms-powerpoint": ".ppt",
+            "presentationml.presentation": ".pptx",
+            "ms-excel": ".xls",
+            "spreadsheetml.sheet": ".xlsx",
+            "msword": ".doc",
+            "wordprocessingml.document": ".docx",
             "audio/": ".mp3",
             "video/": ".mp4",
             "epub": ".epub",
diff --git a/pyproject.toml b/pyproject.toml
index 515307b..b19344a 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -40,7 +40,8 @@ dependencies = [
 "python-pptx==0.6.21",
 "langchain==0.0.105",
 "unidecode==1.3.6",
-"tqdm==4.64.1"
+"tqdm==4.64.1",
+"docx2txt==0.8"
 ]
 
 dynamic = ["version"]
diff --git a/requirements.txt b/requirements.txt
index 08a6410..db526f0 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -23,3 +23,4 @@ langchain==0.0.105
 openai-whisper
 unidecode==1.3.6
 tqdm==4.64.1
+docx2txt==0.8
diff --git a/requirements_base.txt b/requirements_base.txt
index 1e6d9b6..d73968a 100644
--- a/requirements_base.txt
+++ b/requirements_base.txt
@@ -20,4 +20,5 @@ protobuf==3.20.2
 python-pptx==0.6.21
 langchain==0.0.105
 unidecode==1.3.6
-tqdm==4.64.1
\ No newline at end of file
+tqdm==4.64.1
+docx2txt==0.8
\ No newline at end of file