From 496381c2f07031043c2330844bbcd09a0fc10b9d Mon Sep 17 00:00:00 2001
From: Joanne <zehan_ding@brown.edu>
Date: Fri, 27 Jun 2025 11:03:40 -0400
Subject: updated pdf upload and linking fixes for chatbot

---
 src/server/ApiManagers/AssistantManager.ts |  2 +-
 src/server/chunker/requirements.txt        | 37 +++++++++++++++++++++++-------
 2 files changed, 30 insertions(+), 9 deletions(-)

(limited to 'src')

diff --git a/src/server/ApiManagers/AssistantManager.ts b/src/server/ApiManagers/AssistantManager.ts
index c7c347c71..07c970a4e 100644
--- a/src/server/ApiManagers/AssistantManager.ts
+++ b/src/server/ApiManagers/AssistantManager.ts
@@ -1135,7 +1135,7 @@ function spawnPythonProcess(jobId: string, file_path: string, doc_id: string) {
         console.log('Virtual environment not found. Creating and setting up...');
 
         // Create venv
-        const createVenvProcess = spawn('python', ['-m', 'venv', venvPath]);
+        const createVenvProcess = spawn('python3.10', ['-m', 'venv', venvPath]);
 
         createVenvProcess.on('close', code => {
             if (code !== 0) {
diff --git a/src/server/chunker/requirements.txt b/src/server/chunker/requirements.txt
index 20bd486e5..3df3cdd24 100644
--- a/src/server/chunker/requirements.txt
+++ b/src/server/chunker/requirements.txt
@@ -1,15 +1,36 @@
+# Prefer official CPU wheels from the PyTorch index
+--extra-index-url https://download.pytorch.org/whl/cpu
+
+###############################################################################
+#  Stable env for pdf_chunker.py                                              #
+###############################################################################
+
+# ─── LLM clients ─────────────────────────────────────────────────────────────
+openai==1.40.6
+httpx==0.27.2          # <0.28 → avoids "proxies=" crash
 anthropic==0.34.0
 cohere==5.8.0
-python-dotenv==1.0.1
+
+# ─── Torch stack (CPU) ───────────────────────────────────────────────────────
+torch==2.5.1
+torchvision==0.20.1       # matches torch 2.5.x
+torchaudio==2.5.1
+
+# ─── Vision / OCR / PDF processing ───────────────────────────────────────────
+ultralyticsplus==0.0.28
+easyocr==1.7.0
 pymupdf==1.22.2
-lxml==5.3.0
+PyPDF2==3.0.1
+pytesseract==0.3.10
+Pillow==10.4.0
 layoutparser==0.3.4
+lxml==5.3.0
+
+# ─── ML / maths ──────────────────────────────────────────────────────────────
 numpy==1.26.4
-openai==1.40.6
-Pillow==10.4.0
-pytesseract==0.3.10
-PyPDF2==3.0.1
 scikit-learn==1.5.1
+
+# ─── Utilities ──────────────────────────────────────────────────────────────
 tqdm==4.66.5
-ultralyticsplus==0.0.28
-easyocr==1.7.0
\ No newline at end of file
+python-dotenv==1.0.1
+packaging==24.0
\ No newline at end of file
-- 
cgit v1.2.3-70-g09d2