public/crypto-price-bot
6
0
Fork 0
Code Issues 1 Pull Requests 2 Actions Packages Projects Releases Wiki Activity

fix: independent token source lookups #19

Merged
thuanle merged 3 commits from feat/token-message-rich into main 2026-04-26 18:51:18 +07:00
Conversation 2 Commits 3 Files Changed 4
+201 -54
thuanle commented 2026-04-26 18:13:41 +07:00
Owner
Copy Link
Copy Source

Summary

  • decouple token data collection into independent spot/future/alpha checks
  • keep rendering behavior when any one source is available
  • add command-layer tests for spot-only reachability and future-failure-with-spot fallback

Test Plan

  • go test ./internal/services/tele/commands -run CollectRichTokenData -v -count=1
  • go test ./internal/services/tele/view -run RenderRichTokenMessage -v -count=1
  • go test ./... -count=1
## Summary - decouple token data collection into independent spot/future/alpha checks - keep rendering behavior when any one source is available - add command-layer tests for spot-only reachability and future-failure-with-spot fallback ## Test Plan - [x] go test ./internal/services/tele/commands -run CollectRichTokenData -v -count=1 - [x] go test ./internal/services/tele/view -run RenderRichTokenMessage -v -count=1 - [x] go test ./... -count=1
thuanle added 1 commit 2026-04-26 18:13:41 +07:00
fix: decouple token spot/future/alpha lookups 6e8a2e8f68 
Collect token market data independently for spot, futures, and alpha so any available source can be rendered even if others fail.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
thuanle added 1 commit 2026-04-26 18:22:57 +07:00
feat: fetch alpha price by symbol ticker 82b0f2f1a8 
Keep alpha token existence from list cache and fetch live alpha price via symbol ticker endpoint for richer token response accuracy.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
thuanle commented 2026-04-26 18:37:38 +07:00
Author
Owner
Copy Link
Copy Source

Codex review results for PR #19.

FAIL

  1. [High] collectRichTokenData regresses existing symbol resolution by hardcoding both lookups to token + "USDT" in internal/services/tele/commands/token.go:78 and :88.
    That bypasses the existing resolver logic in internal/helper/binancex/symbol.go and the future-to-spot mapping in Future2SpotSymbol(...).
    The repo already supports prefixed futures (1000, 1M), USDC shorthand (c), and mapped contracts like LUNA2USDT -> LUNAUSDT. Those cases can still enter the handler through binancex.IsToken(...), but the new collector now probes only <TOKEN>USDT, so supported aliases/mapped contracts can lose future data, spot data, or both.

Validation:

  • git fetch origin
  • git worktree add ../.worktree/pr-19 origin/feat/token-message-rich
  • env GOCACHE=/tmp/go-build-pr19 go test ./internal/services/tele/commands -run CollectRichTokenData -v -count=1
  • env GOCACHE=/tmp/go-build-pr19 go test ./internal/services/tele/view -run RenderRichTokenMessage -v -count=1
  • env GOCACHE=/tmp/go-build-pr19 go test ./... -count=1
  • env GOCACHE=/tmp/go-build-pr19 go vet ./...

Test gap:

  • the new tests cover straight ABCUSDT / ETHUSDT paths only; they do not cover the existing alias/prefix/mapping behavior that this change now bypasses.
Codex review results for PR #19. FAIL 1. [High] `collectRichTokenData` regresses existing symbol resolution by hardcoding both lookups to `token + "USDT"` in `internal/services/tele/commands/token.go:78` and `:88`. That bypasses the existing resolver logic in `internal/helper/binancex/symbol.go` and the future-to-spot mapping in `Future2SpotSymbol(...)`. The repo already supports prefixed futures (`1000`, `1M`), `USDC` shorthand (`c`), and mapped contracts like `LUNA2USDT -> LUNAUSDT`. Those cases can still enter the handler through `binancex.IsToken(...)`, but the new collector now probes only `<TOKEN>USDT`, so supported aliases/mapped contracts can lose future data, spot data, or both. Validation: - `git fetch origin` - `git worktree add ../.worktree/pr-19 origin/feat/token-message-rich` - `env GOCACHE=/tmp/go-build-pr19 go test ./internal/services/tele/commands -run CollectRichTokenData -v -count=1` - `env GOCACHE=/tmp/go-build-pr19 go test ./internal/services/tele/view -run RenderRichTokenMessage -v -count=1` - `env GOCACHE=/tmp/go-build-pr19 go test ./... -count=1` - `env GOCACHE=/tmp/go-build-pr19 go vet ./...` Test gap: - the new tests cover straight `ABCUSDT` / `ETHUSDT` paths only; they do not cover the existing alias/prefix/mapping behavior that this change now bypasses.
thuanle added 1 commit 2026-04-26 18:39:54 +07:00
Merge branch 'main' into feat/token-message-rich a4644f5982
thuanle commented 2026-04-26 18:47:49 +07:00
Author
Owner
Copy Link
Copy Source

[codex] scope clarification for PR #19.

The concern from my previous review comment is about spot/future symbol resolution only. It is not about the new Alpha lookup itself.

More specifically, the follow-up concern is that the new independent collector now probes spot/future as <TOKEN>USDT directly, which can bypass the existing spot/future resolver rules for prefixes, aliases, and future->spot mappings.

I split that concern into a separate issue so it can be handled independently from Alpha lookup changes:

  • #21 Token collector bypasses existing spot/future symbol resolution rules

So to be explicit:

  • Alpha lookup direction in this PR is not the concern I am calling out here.
  • The extracted follow-up concern is only about spot/future resolution behavior.
[codex] scope clarification for PR #19. The concern from my previous review comment is about spot/future symbol resolution only. It is not about the new Alpha lookup itself. More specifically, the follow-up concern is that the new independent collector now probes spot/future as `<TOKEN>USDT` directly, which can bypass the existing spot/future resolver rules for prefixes, aliases, and future->spot mappings. I split that concern into a separate issue so it can be handled independently from Alpha lookup changes: - #21 Token collector bypasses existing spot/future symbol resolution rules So to be explicit: - Alpha lookup direction in this PR is not the concern I am calling out here. - The extracted follow-up concern is only about spot/future resolution behavior.
thuanle merged commit 1486681ef9 into main 2026-04-26 18:51:18 +07:00
Sign in to join this conversation.
Reviewers
No Reviewers
Labels
No items
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
antigravity claudecode codex devops renovatebot thuanle
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: public/crypto-price-bot#19
Delete Branch "feat/token-message-rich"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?