feat: chat ls logging improvements and infinite looping issue fix

[EzekialSA]

I had this ready to go, someone pushed a change and I had no idea that my PR was merged already. I went ahead and closed it and rebased my fork to match and am properly doing branches for features. I've been using this in my lab for a couple weeks with zero issues.

Replace query.GetDialogs() iterator with manual pagination using MessagesGetDialogs API
to fix multiple critical issues with dialog listing functionality.

Problems fixed:

• Index out of range panic when accessing empty Messages array in fetchTopics

  • Add empty array check before accessing messages for pagination offsets
  • Return empty Topics array instead of nil on fetch errors

• Incomplete dialog fetching

  • Implement proper pagination loop with totalCount stopping condition
  • Use MessagesDialogsSlice.Count to determine when all dialogs are fetched
  • Handle partial batches that can occur mid-stream

• Infinite loop causing flood waits due to incorrect offset calculation

  • Calculate offsets from dialog.TopMessage instead of messages array
  • Prevent duplicate messages from causing offset to get stuck

• Entity population failures preventing proper dialog processing

  • Accumulate users, chats, channels, and messages across all batches into global maps
  • Build single peer.Entities object from complete dataset after fetching

• Hard failures when encountering deleted/inaccessible channels

  • Use InputPeerEmpty for offsetPeer to avoid ExtractPeer failures during pagination
  • Gracefully skip dialogs with invalid/missing peers instead of failing
  • Add detailed logging for skipped dialogs with peer IDs and types

Additional improvements:

• Switch from fixed-width columns to tab-delimited output
• Remove runewidth dependency and unused trunc function
• Add comprehensive batch-level and summary logging
• Track and report non-Dialog types (e.g., DialogFolder)

[Zig-VS-TypeScript-VS]

@EzekialSA hello
I tried your two PR, and it caused a bug. It resulted in an incorrect order for reading the TDL ID.

join https://t.me/happylibrary >>> View discussion >>> join.

tdl-export.json

logs

tdl dl -f D:\88\tdl-export.json --template "{{ filenamify .FileName }}" -i epub --skip-same --takeout --pool 0 -d D:\88\ -t 6 -l 7

All files will be downloaded to 'D:\88\' dir
书库小说讨论群[NSFW](1285760920):961110 -> D:\88~ ... done! [1.11 MB in 7.752s; 146.33 KB/s]

It should start downloading from the latest to the earliest ID.

{"id":965682,"type":"message","file":"复兴之路 - wanglong - 266CHS.epub"},
{"id":93,"type":"message","file":"宰执天下-cuslaa.epub"}

D:\88>tdl dl -f D:\88\tdl-export.json --template "{{ filenamify .FileName }}" -i txt --skip-same --takeout --pool 0 -d D:\88\ -t 6 -l 7
All files will be downloaded to 'D:\88\' dir
书库小说讨论群[NSFW](1285760920):516933 -> D:\88~ ... done! [666.13 KB in 4.775s; 138.09 KB/s]

It should start downloading from the latest to the earliest ID.

{"id":966782,"type":"message","file":"sxsy_org发现清冷系校花偷喝我飞机杯里O液的这件事_作者：_FXY.txt"}
{"id":53,"type":"message","file":"大奉打更人番外（7篇）_soushu555_org_搜书吧网址.txt"}]

[Zig-VS-TypeScript-VS]

EzekialSA Sorry, this isn’t your PR’s fault; it’s a bug in TDL itself.

• [Bug] tdl-export.json was not downloaded in ID order #1128