From e3abf4b37324f46a822816deb2124ba6c613335d Mon Sep 17 00:00:00 2001
From: Viktor Barzin <vbarzin@gmail.com>
Date: Wed, 14 May 2025 21:05:59 +0000
Subject: [PATCH] parameterize detect floorplan  step to work with custom data
 paths

---
 crawler/4_detect_floorplan.py | 8 +++++---
 crawler/main.py               | 9 ++++++---
 2 files changed, 11 insertions(+), 6 deletions(-)

diff --git a/crawler/4_detect_floorplan.py b/crawler/4_detect_floorplan.py
index 97dc8d9..525dc55 100644
--- a/crawler/4_detect_floorplan.py
+++ b/crawler/4_detect_floorplan.py
@@ -1,9 +1,10 @@
+import pathlib
 from data_access import Listing
 from tqdm import tqdm
 
 
-def detect_floorplan():
-    listings = Listing.get_all_listings()
+def detect_floorplan(listing_paths: list[str]):
+    listings = Listing.get_all_listings(listing_paths)
 
     for listing in tqdm(listings):
         tqdm.write(str(listing.identifier))
@@ -12,7 +13,8 @@ def detect_floorplan():
 
 
 def main():
-    detect_floorplan()
+    listing_paths = sorted(list(pathlib.Path("data/rs").glob("*/listing.json")))
+    detect_floorplan(listing_paths)
 
 
 if __name__ == "__main__":
diff --git a/crawler/main.py b/crawler/main.py
index 1d1ef62..6fe413c 100644
--- a/crawler/main.py
+++ b/crawler/main.py
@@ -75,9 +75,12 @@ def dump_images(ctx: click.core.Context):
 
 
 @cli.command()
-def detect_floorplan():
-    click.echo('Running detect_floorplan')
-    detect_floorplan_module.detect_floorplan()
+@click.pass_context
+def detect_floorplan(ctx: click.core.Context):
+    data_dir = ctx.obj['data_dir']
+    click.echo(f'Running detect_floorplan in {data_dir}')
+    listing_paths = sorted(list(pathlib.Path(data_dir).glob("*/listing.json")))
+    detect_floorplan_module.detect_floorplan(listing_paths)
 
 
 @cli.command()