From eaf57df4181e48f280b6d22c8a6d3a456ab0beec Mon Sep 17 00:00:00 2001
From: localhost_frssoft <userquicker@disroot.org>
Date: Wed, 16 Nov 2022 20:59:20 +0300
Subject: [PATCH] Documentary movies will be dropped (Special for FMN, we not
 canceling this!)

---
 src/imdb_datasets_worker.py | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/src/imdb_datasets_worker.py b/src/imdb_datasets_worker.py
index cf67ab6..6f43109 100644
--- a/src/imdb_datasets_worker.py
+++ b/src/imdb_datasets_worker.py
@@ -27,6 +27,7 @@ def convert_tsv_to_db(title_basics_tsv):
                 original_name = line[3]
                 ru_name = None
                 year = line[5]
+                genres = line[-1].strip().split(',')
                 if year.startswith(r"\N"):
                     year = None
                 else:
@@ -35,7 +36,11 @@ def convert_tsv_to_db(title_basics_tsv):
                 if tt_type not in ("movie", "tvMovie", "video"):
                     original_name = None
                     year = None
-
+                if "Documentary" in genres:
+                    logger.debug(f'Документальный {original_name} отсеян')
+                    original_name = None
+                    year = None
+                    tt_type = "doc"
                 write_dataset.append((tt_id, tt_type, original_name, ru_name, year))
                 counter += 1
                 if counter >= chunk: