From 0b87a1b339e71d8d6dcecf1da0c80316c8d223f7 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=D0=A1=D0=B5=D1=80=D0=B3=D0=B5=D0=B9=20=D0=9C=D0=B0=D1=80?=
 =?UTF-8?q?=D0=B8=D0=BD=D0=BA=D0=B5=D0=B2=D0=B8=D1=87?=
 <sergey.marinkevich@eltex.loc>
Date: Tue, 4 Aug 2026 15:24:09 +0700
Subject: [PATCH] Skip idle channels in incremental export via last_post_at.

Avoids a posts?since= round-trip for channels with no new activity since the last dump.

Co-authored-by: Cursor <cursoragent@cursor.com>
---
 mattermost-dl.py | 43 ++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 42 insertions(+), 1 deletion(-)

diff --git a/mattermost-dl.py b/mattermost-dl.py
index 083ea9b..63d3348 100644
--- a/mattermost-dl.py
+++ b/mattermost-dl.py
@@ -386,6 +386,45 @@ def find_channels(channels: List[dict], channel_spec: str) -> List[dict]:
     return selected
 
 
+def _channel_export_filepath(output_base: str, channel: dict) -> pathlib.Path:
+    channel_name = channel["display_name"].replace("\\", "").replace("/", "")
+    filtered = "".join(ch for ch in channel_name if ch not in "?!/\\.;:*\"<>|")
+    return pathlib.Path(output_base) / channel_name / (filtered + ".json")
+
+
+def filter_channels_by_last_post(
+    channels: List[dict],
+    after: str,
+    output_base: str,
+    merge: bool = False,
+) -> List[dict]:
+    """Drop idle channels using Mattermost `last_post_at` before posts fetch.
+
+    Skips channels whose latest post is not newer than `after`, avoiding a
+    per-channel `posts?since=` round-trip. In merge mode, channels without a
+    local export are always kept (first-time full history). Note: `last_post_at`
+    does not advance on edits to older posts — those still need `posts?since=`.
+    """
+    after_ms = int(_parse_filter_timestamp(after) * 1000)
+    selected = []
+    skipped = 0
+    for channel in channels:
+        if merge and not _channel_export_filepath(output_base, channel).exists():
+            selected.append(channel)
+            continue
+        last_post_at = channel.get("last_post_at") or 0
+        if last_post_at > after_ms:
+            selected.append(channel)
+        else:
+            skipped += 1
+    if skipped:
+        print(
+            f"Skipping {skipped} channel(s) with last_post_at <= {after} "
+            f"({len(selected)} remaining)"
+        )
+    return selected
+
+
 def select_team(d: Driver, my_user_id: str) -> str:
     print("Downloading all team information... ", end="")
     teams = d.teams.get_user_teams(my_user_id)
@@ -460,7 +499,7 @@ def export_channel(d: Driver, channel: str, user_id_to_name: Dict[str, str], out
     after_ts = _parse_filter_timestamp(after) if after else None
     before_ts = _parse_filter_timestamp(before) if before else None
     filtered_channel_name = ''.join(filter(lambda ch: ch not in "?!/\\.;:*\"<>|", channel_name))
-    output_filepath = pathlib.Path(output_base) / channel_name / (filtered_channel_name + ".json")
+    output_filepath = _channel_export_filepath(output_base, channel)
     if merge and after_ts is not None and not output_filepath.exists():
         print("  no existing export for this channel, fetching full history")
         after_ts = None
@@ -898,6 +937,8 @@ if __name__ == '__main__':
     else:
         team = select_team(d, my_user_id)
         channels = select_channel(d, team, my_user_id, user_id_to_name)
+    if after:
+        channels = filter_channels_by_last_post(channels, after, output_base, merge=merge)
     for i_channel, channel in enumerate(channels):
         print(f"Start export of channel {i_channel + 1}/{len(channels)}")
         export_channel(d, channel, user_id_to_name, output_base, config["download_files"],