feats: better group setting

[nekohy]

Summary by CodeRabbit

Release Notes

• New Features
  • Added per-model group configuration support, allowing users to customize group mappings for individual models within channels.
  • Introduced a dedicated model group mapping editor in the channel configuration UI.
  • Extended multi-language support with new interface labels for group configuration management.

[coderabbitai]

Walkthrough

This change introduces per-model group configuration for channels. The backend adds a ModelGroupMapping field and getter method to resolve model-specific groups, updating ability creation/update logic accordingly. The frontend introduces a new ModelGroupMappingEditor component integrated into the channel edit modal, with English and Chinese i18n support.

Changes

Cohort / File(s)	Summary
Backend Model Grouping model/ability.go, model/channel.go	Added ModelGroupMapping field with JSON storage and GetModelGroupMapping() method. Updated ability resolution in both AddAbilities and UpdateAbilities to use per-model groups when available, falling back to channel defaults.
Frontend Editor Component web/src/components/table/channels/ModelGroupMappingEditor.jsx	New React component for managing per-model group configurations with model selector, multi-select group editor, and delete actions. Renders table with configured models and their assigned groups.
Modal Integration web/src/components/table/channels/modals/EditChannelModal.jsx	Integrated ModelGroupMappingEditor into channel edit flow with state management for modelGroupConfig, parsing of model_group_mapping on load, and serialization during submission.
Internationalization web/src/i18n/locales/en.json, web/src/i18n/locales/zh.json	Added six translation keys (Chinese and English) for model group configuration UI labels, descriptions, and empty states.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

• feat: 可视化模型重定向设置中，允许用户从拉取的模型列表里直接选择模型 #2610: Both PRs enhance the channel EditChannelModal UI with per-model/per-entry configuration controls, introducing new editor components for model-specific settings.

Suggested reviewers

• seefs001

Poem

🐰 Per-model groups now dance with glee,
Each model picks its own destiny,
The editor lets us map with care,
While translations float in the air—
A feature fluffy, precise, and fair! ✨

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 inconclusive)

Check name	Status	Explanation	Resolution
Title check	❓ Inconclusive	The title 'feats: better group setting' is vague and does not clearly convey the specific implementation of per-model group configurations. It lacks concrete details about the core change.	Consider using a more descriptive title like 'feat: add per-model group configuration support' to clearly communicate the primary change to reviewers scanning PR history.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

model/ability.go (1)

146-176: ⚠️ Potential issue | 🟡 Minor

Trim and skip empty group names when generating abilities.

Whitespace or empty entries in default/custom groups can create invalid group records and mismatched routing.

🧹 Suggested normalization in both AddAbilities and UpdateAbilities

-		for _, group := range groups {
+		for _, group := range groups {
+			group = strings.TrimSpace(group)
+			if group == "" {
+				continue
+			}
 			key := group + "|" + model
 			if _, exists := abilitySet[key]; exists {
 				continue
 			}

-		for _, group := range groups {
+		for _, group := range groups {
+			group = strings.TrimSpace(group)
+			if group == "" {
+				continue
+			}
 			key := group + "|" + model
 			if _, exists := abilitySet[key]; exists {
 				continue
 			}

Also applies to: 226-256

🤖 Fix all issues with AI agents

In `@model/channel.go`:
- Around line 106-122: GetModelGroupMapping currently swallows JSON unmarshal
errors; update it to log the invalid JSON and the error before returning nil so
misconfigurations are visible. In the GetModelGroupMapping method, when
common.Unmarshal returns an error, emit a log entry (e.g., using the standard
log package or your package logger) that includes the error and the raw
*channel.ModelGroupMapping string for context, then continue to return nil; keep
the rest of the function (cleaning empty keys and returning mapping) unchanged.

In `@web/src/components/table/channels/modals/EditChannelModal.jsx`:
- Around line 1392-1395: The model_group_mapping currently serializes all
modelGroupConfig.specificRules, including stale entries for removed models;
update the logic around localInputs.model_group_mapping (in
EditChannelModal.jsx) to first prune specificRules by keeping only keys present
in normalizedModels (i.e., filter modelGroupConfig.specificRules by
normalizedModels set), then JSON.stringify that pruned object and omit (delete
or set to undefined/empty) model_group_mapping when the pruned object is empty
so no empty/stale overrides are submitted.

🧹 Nitpick comments (1)

web/src/components/table/channels/modals/EditChannelModal.jsx (1)

566-582: Normalize parsed model_group_mapping before storing.

If persisted data contains arrays or non-array values, the editor state can become inconsistent. Consider filtering to a plain object of string arrays.

♻️ Suggested normalization

-          if (mgMap && typeof mgMap === 'object') {
-            // 直接使用 model_group_mapping 的内容作为 specificRules，不做过滤
-            mgConfig.specificRules = mgMap;
-          }
+          if (mgMap && typeof mgMap === 'object' && !Array.isArray(mgMap)) {
+            const cleanedRules = {};
+            Object.entries(mgMap).forEach(([model, groups]) => {
+              if (Array.isArray(groups)) {
+                cleanedRules[model] = groups
+                  .map((g) => String(g).trim())
+                  .filter(Boolean);
+              }
+            });
+            mgConfig.specificRules = cleanedRules;
+          }

[coderabbitai]

⚠️ Potential issue | 🟡 Minor

Log invalid model_group_mapping JSON to avoid silent fallback.

Right now a malformed JSON silently disables the feature. Emitting a log makes misconfigurations actionable.

🛠️ Proposed fix

-	if err := common.Unmarshal([]byte(*channel.ModelGroupMapping), &mapping); err != nil {
-		return nil
-	}
+	if err := common.Unmarshal([]byte(*channel.ModelGroupMapping), &mapping); err != nil {
+		common.SysLog(fmt.Sprintf("failed to unmarshal model_group_mapping: channel_id=%d, error=%v", channel.Id, err))
+		return nil
+	}

📝 Committable suggestion

‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.

Suggested change

	// GetModelGroupMapping 解析模型分组映射，返回 map[model][]groups
	func (channel *Channel) GetModelGroupMapping() map[string][]string {
	if channel.ModelGroupMapping == nil || *channel.ModelGroupMapping == "" {
	return nil
	}
	var mapping map[string][]string
	if err := common.Unmarshal([]byte(*channel.ModelGroupMapping), &mapping); err != nil {
	return nil
	}
	// 清理空值
	for model, groups := range mapping {
	if model == "" || len(groups) == 0 {
	delete(mapping, model)
	}
	}
	return mapping
	}
	// GetModelGroupMapping 解析模型分组映射，返回 map[model][]groups
	func (channel *Channel) GetModelGroupMapping() map[string][]string {
	if channel.ModelGroupMapping == nil || *channel.ModelGroupMapping == "" {
	return nil
	}
	var mapping map[string][]string
	if err := common.Unmarshal([]byte(*channel.ModelGroupMapping), &mapping); err != nil {
	common.SysLog(fmt.Sprintf("failed to unmarshal model_group_mapping: channel_id=%d, error=%v", channel.Id, err))
	return nil
	}
	// 清理空值
	for model, groups := range mapping {
	if model == "" || len(groups) == 0 {
	delete(mapping, model)
	}
	}
	return mapping
	}

🤖 Prompt for AI Agents

In `@model/channel.go` around lines 106 - 122, GetModelGroupMapping currently
swallows JSON unmarshal errors; update it to log the invalid JSON and the error
before returning nil so misconfigurations are visible. In the
GetModelGroupMapping method, when common.Unmarshal returns an error, emit a log
entry (e.g., using the standard log package or your package logger) that
includes the error and the raw *channel.ModelGroupMapping string for context,
then continue to return nil; keep the rest of the function (cleaning empty keys
and returning mapping) unchanged.

[coderabbitai]

⚠️ Potential issue | 🟡 Minor

Prune mappings for models that were removed.

Stale overrides remain hidden in the editor but still get serialized, so they can reappear if a model is re-added. Filter to normalizedModels before stringify and omit when empty.

🧹 Suggested pruning before submit

-    localInputs.model_group_mapping = JSON.stringify(modelGroupConfig.specificRules || {});
+    const modelGroupMapping = Object.fromEntries(
+      Object.entries(modelGroupConfig.specificRules || {}).filter(([model]) =>
+        normalizedModels.includes(model),
+      ),
+    );
+    localInputs.model_group_mapping =
+      Object.keys(modelGroupMapping).length > 0
+        ? JSON.stringify(modelGroupMapping)
+        : '';

🤖 Prompt for AI Agents

In `@web/src/components/table/channels/modals/EditChannelModal.jsx` around lines
1392 - 1395, The model_group_mapping currently serializes all
modelGroupConfig.specificRules, including stale entries for removed models;
update the logic around localInputs.model_group_mapping (in
EditChannelModal.jsx) to first prune specificRules by keeping only keys present
in normalizedModels (i.e., filter modelGroupConfig.specificRules by
normalizedModels set), then JSON.stringify that pruned object and omit (delete
or set to undefined/empty) model_group_mapping when the pruned object is empty
so no empty/stale overrides are submitted.