From 7f517776dfb49338f0b4ad0c9ce0600204c5927a Mon Sep 17 00:00:00 2001
From: 0x1355 <0x1355@gmail.com>
Date: Fri, 20 Mar 2026 09:18:27 +0100
Subject: [PATCH] fix: auto-detect jobspy venv so contributors don't need
 PYTHON_PATH (#293)

---
 CONTRIBUTING.md              |  9 +++++++++
 extractors/jobspy/src/run.ts | 18 +++++++++++++++---
 2 files changed, 24 insertions(+), 3 deletions(-)

diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
index 560984a..68d138d 100644
--- a/CONTRIBUTING.md
+++ b/CONTRIBUTING.md
@@ -34,6 +34,15 @@ npm --workspace orchestrator run db:migrate
 npm --workspace orchestrator run dev
 ```
 
+If you are working with extractors that use Glassdoor, Indeed, or LinkedIn (powered by python-jobspy), set up the Python venv once:
+
+```bash
+python3 -m venv extractors/jobspy/.venv
+extractors/jobspy/.venv/bin/pip install -r extractors/jobspy/requirements.txt
+```
+
+The runner auto-detects the venv — no need to set `PYTHON_PATH`.
+
 If you are editing docs:
 
 ```bash
diff --git a/extractors/jobspy/src/run.ts b/extractors/jobspy/src/run.ts
index f2cdba4..3344bec 100644
--- a/extractors/jobspy/src/run.ts
+++ b/extractors/jobspy/src/run.ts
@@ -1,4 +1,5 @@
 import { spawn } from "node:child_process";
+import { existsSync } from "node:fs";
 import { mkdir, readFile, unlink } from "node:fs/promises";
 import { dirname, join } from "node:path";
 import { createInterface } from "node:readline";
@@ -180,11 +181,22 @@ export async function runJobSpy(
         const outputJson = join(OUTPUT_DIR, `jobspy_jobs_${suffix}.json`);
 
         await new Promise<void>((resolve, reject) => {
+          // Auto-detect venv if present, so contributors don't need to set
+          // PYTHON_PATH manually. The venv is created once with:
+          //   python3 -m venv .venv && .venv/bin/pip install -r requirements.txt
+          // In Docker, PYTHON_PATH is set explicitly to /usr/bin/python3.
+          const venvPython = join(
+            EXTRACTOR_DIR,
+            ".venv",
+            process.platform === "win32" ? "Scripts/python.exe" : "bin/python3",
+          );
           const pythonPath = process.env.PYTHON_PATH
             ? process.env.PYTHON_PATH
-            : process.platform === "win32"
-              ? "python"
-              : "python3";
+            : existsSync(venvPython)
+              ? venvPython
+              : process.platform === "win32"
+                ? "python"
+                : "python3";
 
           const child = spawn(pythonPath, [JOBSPY_SCRIPT], {
             cwd: EXTRACTOR_DIR,