mirror of
https://github.com/DayuanJiang/next-ai-draw-io.git
synced 2026-01-02 22:32:27 +08:00
Compare commits
5 Commits
d3be96de79
...
refactor/e
| Author | SHA1 | Date | |
|---|---|---|---|
|
|
4dbda5ba8e | ||
|
|
1fab261cd0 | ||
|
|
7a4a04c263 | ||
|
|
0d2e7a7ad6 | ||
|
|
3218ccc909 |
@@ -27,6 +27,7 @@ import {
|
||||
TooltipTrigger,
|
||||
} from "@/components/ui/tooltip"
|
||||
import { useDiagram } from "@/contexts/diagram-context"
|
||||
import { useDiagramToolHandlers } from "@/hooks/use-diagram-tool-handlers"
|
||||
import { useDictionary } from "@/hooks/use-dictionary"
|
||||
import { getSelectedAIConfig, useModelConfig } from "@/hooks/use-model-config"
|
||||
import { getApiEndpoint } from "@/lib/base-path"
|
||||
@@ -34,8 +35,9 @@ import { findCachedResponse } from "@/lib/cached-responses"
|
||||
import { isPdfFile, isTextFile } from "@/lib/pdf-utils"
|
||||
import { type FileData, useFileProcessor } from "@/lib/use-file-processor"
|
||||
import { useQuotaManager } from "@/lib/use-quota-manager"
|
||||
import { formatXML, isMxCellXmlComplete, wrapWithMxFile } from "@/lib/utils"
|
||||
import { formatXML } from "@/lib/utils"
|
||||
import { ChatMessageDisplay } from "./chat-message-display"
|
||||
import { DevXmlSimulator } from "./dev-xml-simulator"
|
||||
|
||||
// localStorage keys for persistence
|
||||
const STORAGE_MESSAGES_KEY = "next-ai-draw-io-messages"
|
||||
@@ -76,6 +78,7 @@ interface ChatPanelProps {
|
||||
const TOOL_ERROR_STATE = "output-error" as const
|
||||
const DEBUG = process.env.NODE_ENV === "development"
|
||||
const MAX_AUTO_RETRY_COUNT = 1
|
||||
|
||||
const MAX_CONTINUATION_RETRY_COUNT = 2 // Limit for truncation continuation retries
|
||||
|
||||
/**
|
||||
@@ -212,9 +215,6 @@ export default function ChatPanel({
|
||||
chartXMLRef.current = chartXML
|
||||
}, [chartXML])
|
||||
|
||||
// Ref to hold stop function for use in onToolCall (avoids stale closure)
|
||||
const stopRef = useRef<(() => void) | null>(null)
|
||||
|
||||
// Ref to track consecutive auto-retry count (reset on user action)
|
||||
const autoRetryCountRef = useRef(0)
|
||||
// Ref to track continuation retry count (for truncation handling)
|
||||
@@ -237,6 +237,16 @@ export default function ChatPanel({
|
||||
> | null>(null)
|
||||
const LOCAL_STORAGE_DEBOUNCE_MS = 1000 // Save at most once per second
|
||||
|
||||
// Diagram tool handlers (display_diagram, edit_diagram, append_diagram)
|
||||
const { handleToolCall } = useDiagramToolHandlers({
|
||||
partialXmlRef,
|
||||
editDiagramOriginalXmlRef,
|
||||
chartXMLRef,
|
||||
onDisplayChart,
|
||||
onFetchChart,
|
||||
onExport,
|
||||
})
|
||||
|
||||
const {
|
||||
messages,
|
||||
sendMessage,
|
||||
@@ -249,311 +259,8 @@ export default function ChatPanel({
|
||||
transport: new DefaultChatTransport({
|
||||
api: getApiEndpoint("/api/chat"),
|
||||
}),
|
||||
async onToolCall({ toolCall }) {
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
`[onToolCall] Tool: ${toolCall.toolName}, CallId: ${toolCall.toolCallId}`,
|
||||
)
|
||||
}
|
||||
|
||||
if (toolCall.toolName === "display_diagram") {
|
||||
const { xml } = toolCall.input as { xml: string }
|
||||
|
||||
// DEBUG: Log raw input to diagnose false truncation detection
|
||||
console.log(
|
||||
"[display_diagram] XML ending (last 100 chars):",
|
||||
xml.slice(-100),
|
||||
)
|
||||
console.log("[display_diagram] XML length:", xml.length)
|
||||
|
||||
// Check if XML is truncated (incomplete mxCell indicates truncated output)
|
||||
const isTruncated = !isMxCellXmlComplete(xml)
|
||||
console.log("[display_diagram] isTruncated:", isTruncated)
|
||||
|
||||
if (isTruncated) {
|
||||
// Store the partial XML for continuation via append_diagram
|
||||
partialXmlRef.current = xml
|
||||
|
||||
// Tell LLM to use append_diagram to continue
|
||||
const partialEnding = partialXmlRef.current.slice(-500)
|
||||
addToolOutput({
|
||||
tool: "display_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Output was truncated due to length limits. Use the append_diagram tool to continue.
|
||||
|
||||
Your output ended with:
|
||||
\`\`\`
|
||||
${partialEnding}
|
||||
\`\`\`
|
||||
|
||||
NEXT STEP: Call append_diagram with the continuation XML.
|
||||
- Do NOT include wrapper tags or root cells (id="0", id="1")
|
||||
- Start from EXACTLY where you stopped
|
||||
- Complete all remaining mxCell elements`,
|
||||
})
|
||||
return
|
||||
}
|
||||
|
||||
// Complete XML received - use it directly
|
||||
// (continuation is now handled via append_diagram tool)
|
||||
const finalXml = xml
|
||||
partialXmlRef.current = "" // Reset any partial from previous truncation
|
||||
|
||||
// Wrap raw XML with full mxfile structure for draw.io
|
||||
const fullXml = wrapWithMxFile(finalXml)
|
||||
|
||||
// loadDiagram validates and returns error if invalid
|
||||
const validationError = onDisplayChart(fullXml)
|
||||
|
||||
if (validationError) {
|
||||
console.warn(
|
||||
"[display_diagram] Validation error:",
|
||||
validationError,
|
||||
)
|
||||
// Return error to model - sendAutomaticallyWhen will trigger retry
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] Adding tool output with state: output-error",
|
||||
)
|
||||
}
|
||||
addToolOutput({
|
||||
tool: "display_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `${validationError}
|
||||
|
||||
Please fix the XML issues and call display_diagram again with corrected XML.
|
||||
|
||||
Your failed XML:
|
||||
\`\`\`xml
|
||||
${finalXml}
|
||||
\`\`\``,
|
||||
})
|
||||
} else {
|
||||
// Success - diagram will be rendered by chat-message-display
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] Success! Adding tool output with state: output-available",
|
||||
)
|
||||
}
|
||||
addToolOutput({
|
||||
tool: "display_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
output: "Successfully displayed the diagram.",
|
||||
})
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] Tool output added. Diagram should be visible now.",
|
||||
)
|
||||
}
|
||||
}
|
||||
} else if (toolCall.toolName === "edit_diagram") {
|
||||
const { operations } = toolCall.input as {
|
||||
operations: Array<{
|
||||
type: "update" | "add" | "delete"
|
||||
cell_id: string
|
||||
new_xml?: string
|
||||
}>
|
||||
}
|
||||
|
||||
let currentXml = ""
|
||||
try {
|
||||
// Use the original XML captured during streaming (shared with chat-message-display)
|
||||
// This ensures we apply operations to the same base XML that streaming used
|
||||
const originalXml = editDiagramOriginalXmlRef.current.get(
|
||||
toolCall.toolCallId,
|
||||
)
|
||||
if (originalXml) {
|
||||
currentXml = originalXml
|
||||
} else {
|
||||
// Fallback: use chartXML from ref if streaming didn't capture original
|
||||
const cachedXML = chartXMLRef.current
|
||||
if (cachedXML) {
|
||||
currentXml = cachedXML
|
||||
} else {
|
||||
// Last resort: export from iframe
|
||||
currentXml = await onFetchChart(false)
|
||||
}
|
||||
}
|
||||
|
||||
const { applyDiagramOperations } = await import(
|
||||
"@/lib/utils"
|
||||
)
|
||||
const { result: editedXml, errors } =
|
||||
applyDiagramOperations(currentXml, operations)
|
||||
|
||||
// Check for operation errors
|
||||
if (errors.length > 0) {
|
||||
const errorMessages = errors
|
||||
.map(
|
||||
(e) =>
|
||||
`- ${e.type} on cell_id="${e.cellId}": ${e.message}`,
|
||||
)
|
||||
.join("\n")
|
||||
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Some operations failed:\n${errorMessages}
|
||||
|
||||
Current diagram XML:
|
||||
\`\`\`xml
|
||||
${currentXml}
|
||||
\`\`\`
|
||||
|
||||
Please check the cell IDs and retry.`,
|
||||
})
|
||||
// Clean up the shared original XML ref
|
||||
editDiagramOriginalXmlRef.current.delete(
|
||||
toolCall.toolCallId,
|
||||
)
|
||||
return
|
||||
}
|
||||
|
||||
// loadDiagram validates and returns error if invalid
|
||||
const validationError = onDisplayChart(editedXml)
|
||||
if (validationError) {
|
||||
console.warn(
|
||||
"[edit_diagram] Validation error:",
|
||||
validationError,
|
||||
)
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Edit produced invalid XML: ${validationError}
|
||||
|
||||
Current diagram XML:
|
||||
\`\`\`xml
|
||||
${currentXml}
|
||||
\`\`\`
|
||||
|
||||
Please fix the operations to avoid structural issues.`,
|
||||
})
|
||||
// Clean up the shared original XML ref
|
||||
editDiagramOriginalXmlRef.current.delete(
|
||||
toolCall.toolCallId,
|
||||
)
|
||||
return
|
||||
}
|
||||
onExport()
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
output: `Successfully applied ${operations.length} operation(s) to the diagram.`,
|
||||
})
|
||||
// Clean up the shared original XML ref
|
||||
editDiagramOriginalXmlRef.current.delete(
|
||||
toolCall.toolCallId,
|
||||
)
|
||||
} catch (error) {
|
||||
console.error("[edit_diagram] Failed:", error)
|
||||
|
||||
const errorMessage =
|
||||
error instanceof Error ? error.message : String(error)
|
||||
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Edit failed: ${errorMessage}
|
||||
|
||||
Current diagram XML:
|
||||
\`\`\`xml
|
||||
${currentXml || "No XML available"}
|
||||
\`\`\`
|
||||
|
||||
Please check cell IDs and retry, or use display_diagram to regenerate.`,
|
||||
})
|
||||
// Clean up the shared original XML ref even on error
|
||||
editDiagramOriginalXmlRef.current.delete(
|
||||
toolCall.toolCallId,
|
||||
)
|
||||
}
|
||||
} else if (toolCall.toolName === "append_diagram") {
|
||||
const { xml } = toolCall.input as { xml: string }
|
||||
|
||||
// Detect if LLM incorrectly started fresh instead of continuing
|
||||
// LLM should only output bare mxCells now, so wrapper tags indicate error
|
||||
const trimmed = xml.trim()
|
||||
const isFreshStart =
|
||||
trimmed.startsWith("<mxGraphModel") ||
|
||||
trimmed.startsWith("<root") ||
|
||||
trimmed.startsWith("<mxfile") ||
|
||||
trimmed.startsWith('<mxCell id="0"') ||
|
||||
trimmed.startsWith('<mxCell id="1"')
|
||||
|
||||
if (isFreshStart) {
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `ERROR: You started fresh with wrapper tags. Do NOT include wrapper tags or root cells (id="0", id="1").
|
||||
|
||||
Continue from EXACTLY where the partial ended:
|
||||
\`\`\`
|
||||
${partialXmlRef.current.slice(-500)}
|
||||
\`\`\`
|
||||
|
||||
Start your continuation with the NEXT character after where it stopped.`,
|
||||
})
|
||||
return
|
||||
}
|
||||
|
||||
// Append to accumulated XML
|
||||
partialXmlRef.current += xml
|
||||
|
||||
// Check if XML is now complete (last mxCell is complete)
|
||||
const isComplete = isMxCellXmlComplete(partialXmlRef.current)
|
||||
|
||||
if (isComplete) {
|
||||
// Wrap and display the complete diagram
|
||||
const finalXml = partialXmlRef.current
|
||||
partialXmlRef.current = "" // Reset
|
||||
|
||||
const fullXml = wrapWithMxFile(finalXml)
|
||||
const validationError = onDisplayChart(fullXml)
|
||||
|
||||
if (validationError) {
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Validation error after assembly: ${validationError}
|
||||
|
||||
Assembled XML:
|
||||
\`\`\`xml
|
||||
${finalXml.substring(0, 2000)}...
|
||||
\`\`\`
|
||||
|
||||
Please use display_diagram with corrected XML.`,
|
||||
})
|
||||
} else {
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
output: "Diagram assembly complete and displayed successfully.",
|
||||
})
|
||||
}
|
||||
} else {
|
||||
// Still incomplete - signal to continue
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `XML still incomplete (mxCell not closed). Call append_diagram again to continue.
|
||||
|
||||
Current ending:
|
||||
\`\`\`
|
||||
${partialXmlRef.current.slice(-500)}
|
||||
\`\`\`
|
||||
|
||||
Continue from EXACTLY where you stopped.`,
|
||||
})
|
||||
}
|
||||
}
|
||||
onToolCall: async ({ toolCall }) => {
|
||||
await handleToolCall({ toolCall }, addToolOutput)
|
||||
},
|
||||
onError: (error) => {
|
||||
// Handle server-side quota limit (429 response)
|
||||
@@ -717,9 +424,6 @@ Continue from EXACTLY where you stopped.`,
|
||||
},
|
||||
})
|
||||
|
||||
// Update stopRef so onToolCall can access it
|
||||
stopRef.current = stop
|
||||
|
||||
// Ref to track latest messages for unload persistence
|
||||
const messagesRef = useRef(messages)
|
||||
useEffect(() => {
|
||||
@@ -1337,6 +1041,14 @@ Continue from EXACTLY where you stopped.`,
|
||||
/>
|
||||
</main>
|
||||
|
||||
{/* Dev XML Streaming Simulator - only in development */}
|
||||
{DEBUG && (
|
||||
<DevXmlSimulator
|
||||
setMessages={setMessages}
|
||||
onDisplayChart={onDisplayChart}
|
||||
/>
|
||||
)}
|
||||
|
||||
{/* Input */}
|
||||
<footer
|
||||
className={`${isMobile ? "p-2" : "p-4"} border-t border-border/50 bg-card/50`}
|
||||
|
||||
350
components/dev-xml-simulator.tsx
Normal file
350
components/dev-xml-simulator.tsx
Normal file
@@ -0,0 +1,350 @@
|
||||
"use client"
|
||||
|
||||
import { useEffect, useRef, useState } from "react"
|
||||
import { wrapWithMxFile } from "@/lib/utils"
|
||||
|
||||
// Dev XML presets for streaming simulator
|
||||
const DEV_XML_PRESETS: Record<string, string> = {
|
||||
"Simple Box": `<mxCell id="2" value="Hello World" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#dae8fc;strokeColor=#6c8ebf;" vertex="1" parent="1">
|
||||
<mxGeometry x="120" y="100" width="120" height="60" as="geometry"/>
|
||||
</mxCell>`,
|
||||
"Two Boxes with Arrow": `<mxCell id="2" value="Start" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;" vertex="1" parent="1">
|
||||
<mxGeometry x="100" y="100" width="100" height="50" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="3" value="End" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#f8cecc;strokeColor=#b85450;" vertex="1" parent="1">
|
||||
<mxGeometry x="300" y="100" width="100" height="50" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="4" value="" style="endArrow=classic;html=1;" edge="1" parent="1" source="2" target="3">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>`,
|
||||
Flowchart: `<mxCell id="2" value="Start" style="ellipse;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;" vertex="1" parent="1">
|
||||
<mxGeometry x="160" y="40" width="80" height="40" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="3" value="Process A" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#dae8fc;strokeColor=#6c8ebf;" vertex="1" parent="1">
|
||||
<mxGeometry x="140" y="120" width="120" height="60" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="4" value="Decision" style="rhombus;whiteSpace=wrap;html=1;fillColor=#fff2cc;strokeColor=#d6b656;" vertex="1" parent="1">
|
||||
<mxGeometry x="150" y="220" width="100" height="80" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="5" value="Process B" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#dae8fc;strokeColor=#6c8ebf;" vertex="1" parent="1">
|
||||
<mxGeometry x="300" y="230" width="120" height="60" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="6" value="End" style="ellipse;whiteSpace=wrap;html=1;fillColor=#f8cecc;strokeColor=#b85450;" vertex="1" parent="1">
|
||||
<mxGeometry x="160" y="340" width="80" height="40" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="7" style="endArrow=classic;html=1;" edge="1" parent="1" source="2" target="3">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="8" style="endArrow=classic;html=1;" edge="1" parent="1" source="3" target="4">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="9" value="Yes" style="endArrow=classic;html=1;" edge="1" parent="1" source="4" target="6">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="10" value="No" style="endArrow=classic;html=1;exitX=1;exitY=0.5;exitDx=0;exitDy=0;" edge="1" parent="1" source="4" target="5">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>`,
|
||||
"Truncated (Error Test)": `<mxCell id="2" value="This cell is truncated" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#dae8fc;strokeColor=#6c8ebf;" vertex="1" parent="1">
|
||||
<mxGeometry x="120" y="100" width="120" height="60" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="3" value="Incomplete" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#f8cecc;strokeColor`,
|
||||
"HTML Escape + Cell Truncate": `<mxCell id="2" value="<b>Chain-of-Thought Prompting</b><br/><font size='12'>Eliciting Reasoning in Large Language Models</font>" style="rounded=0;whiteSpace=wrap;html=1;fillColor=#dae8fc;strokeColor=#6c8ebf;fontSize=16;fontStyle=1;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="40" width="720" height="60" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="3" value="<b>Problem: LLM Reasoning Limitations</b><br/>• Scaling parameters alone insufficient for logical tasks<br/>• Arithmetic, commonsense, symbolic reasoning challenges<br/>• Standard prompting fails on multi-step problems" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#ffe6cc;strokeColor=#d79b00;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="120" width="340" height="120" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="4" value="<b>Traditional Approaches</b><br/>1. <b>Finetuning:</b> Expensive, task-specific<br/>2. <b>Standard Few-Shot:</b> Input→Output pairs<br/> (No explanation of reasoning)" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#fff2cc;strokeColor=#d6b656;" vertex="1" parent="1">
|
||||
<mxGeometry x="420" y="120" width="340" height="120" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="5" value="<b>CoT Methodology</b><br/>• Add reasoning steps to few-shot examples<br/>• Natural language intermediate steps<br/>• No parameter updates needed<br/>• Model learns to generate own thought process" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="260" width="340" height="100" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="6" value="<b>Example Comparison</b><br/><b>Standard:</b><br/>Q: Roger has 5 balls. He buys 2 cans of 3 balls. How many?<br/>A: 11.<br/><br/><b>CoT:</b><br/>Q: Roger has 5 balls. He buys 2 cans of 3 balls. How many?<br/>A: Roger started with 5 balls. 2 cans of 3 tennis balls each is 6 tennis balls. 5 + 6 = 11. The answer is 11." style="rounded=1;whiteSpace=wrap;html=1;fillColor=#e1d5e7;strokeColor=#9673a6;" vertex="1" parent="1">
|
||||
<mxGeometry x="420" y="260" width="340" height="140" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="7" value="<b>Experimental Models</b><br/>• GPT-3 (175B)<br/>• LaMDA (137B)<br/>• PaLM (540B)<br/>• UL2 (20B)<br/>• Codex" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#f8cecc;strokeColor=#b85450;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="380" width="340" height="100" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="8" value="<b>Reasoning Domains Tested</b><br/>1. <b>Arithmetic:</b> GSM8K, SVAMP, ASDiv, AQuA, MAWPS<br/>2. <b>Commonsense:</b> CSQA, StrategyQA, Date Understanding, Sports Understanding<br/>3. <b>Symbolic:</b> Last Letter Concatenation, Coin Flip" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#f5f5f5;strokeColor=#666666;" vertex="1" parent="1">
|
||||
<mxGeometry x="420" y="420" width="340" height="100" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="9" value="<b>Key Results: Arithmetic</b><br/>• PaLM 540B + CoT: <b>56.9%</b> on GSM8K<br/> (vs 17.9% standard)<br/>• Surpassed finetuned GPT-3 (55%)<br/>• With calculator: <b>58.6%</b>" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="500" width="220" height="100" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="10" value="<b>Key Results: Commonsense</b><br/>• StrategyQA: <b>75.6%</b><br/> (vs 69.4% SOTA)<br/>• Sports Understanding: <b>95.4%</b><br/> (vs 84% human)" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;" vertex="1" parent="1">
|
||||
<mxGeometry x="280" y="500" width="220" height="100" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="11" value="<b>Key Results: Symbolic</b><br/>• OOD Generalization<br/>• Coin Flip: Trained on 2 flips<br/> Works on 3-4 flips with CoT<br/>• Standard prompting fails" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#d5e8d4;strokeColor=#82b366;" vertex="1" parent="1">
|
||||
<mxGeometry x="540" y="500" width="220" height="100" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="12" value="<b>Emergent Ability of Scale</b><br/>• Small models (<10B): No benefit, often harmful<br/>• Large models (100B+): Reasoning emerges<br/>• CoT gains increase dramatically with scale" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#ffe6cc;strokeColor=#d79b00;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="620" width="340" height="80" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="13" value="<b>Ablation Studies</b><br/>1. Equation only: Worse than CoT<br/>2. Variable compute (...): No improvement<br/>3. Answer first, then reasoning: Same as baseline<br/>→ Content matters, not just extra tokens" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#fff2cc;strokeColor=#d6b656;" vertex="1" parent="1">
|
||||
<mxGeometry x="420" y="620" width="340" height="80" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="14" value="<b>Error Analysis</b><br/>• Semantic understanding errors<br/>• One-step missing errors<br/>• Calculation errors<br/>• Larger models reduce semantic/missing-step errors" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#f8cecc;strokeColor=#b85450;" vertex="1" parent="1">
|
||||
<mxGeometry x="40" y="720" width="340" height="80" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="15" value="<b>Conclusion</b><br/>• CoT unlocks reasoning potential<br/>• Simple paradigm: "show your work"<br/>• Emergent capability of large models<br/>• No specialized architecture needed" style="rounded=1;whiteSpace=wrap;html=1;fillColor=#dae8fc;strokeColor=#6c8ebf;" vertex="1" parent="1">
|
||||
<mxGeometry x="420" y="720" width="340" height="80" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="16" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="3" target="5">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="17" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="4" target="6">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="18" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="5" target="7">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="19" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="6" target="8">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="20" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.25;entryY=0;" edge="1" parent="1" source="7" target="9">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="21" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="7" target="10">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="22" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.75;entryY=0;" edge="1" parent="1" source="7" target="11">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="23" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="9" target="12">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="24" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="10" target="13">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="25" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="11" target="14">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="26" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="12" target="15">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="27" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="13" target="15">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>
|
||||
<mxCell id="28" style="edgeStyle=orthogonalEdgeStyle;rounded=0;orthogonalLoop=1;jettySize=auto;html=1;exitX=0.5;exitY=1;entryX=0.5;entryY=0;" edge="1" parent="1" source="14" target="15">
|
||||
<mxGeometry relative="1" as="geometry"/>
|
||||
</mxCell>`,
|
||||
}
|
||||
|
||||
interface DevXmlSimulatorProps {
|
||||
setMessages: React.Dispatch<React.SetStateAction<any[]>>
|
||||
onDisplayChart: (xml: string) => void
|
||||
}
|
||||
|
||||
export function DevXmlSimulator({
|
||||
setMessages,
|
||||
onDisplayChart,
|
||||
}: DevXmlSimulatorProps) {
|
||||
const [devXml, setDevXml] = useState("")
|
||||
const [isSimulating, setIsSimulating] = useState(false)
|
||||
const [devIntervalMs, setDevIntervalMs] = useState(1)
|
||||
const [devChunkSize, setDevChunkSize] = useState(10)
|
||||
const devStopRef = useRef(false)
|
||||
const devXmlInitializedRef = useRef(false)
|
||||
|
||||
// Restore dev XML from localStorage on mount (after hydration)
|
||||
useEffect(() => {
|
||||
const saved = localStorage.getItem("dev-xml-simulator")
|
||||
if (saved) setDevXml(saved)
|
||||
devXmlInitializedRef.current = true
|
||||
}, [])
|
||||
|
||||
// Save dev XML to localStorage (only after initial load)
|
||||
useEffect(() => {
|
||||
if (devXmlInitializedRef.current) {
|
||||
localStorage.setItem("dev-xml-simulator", devXml)
|
||||
}
|
||||
}, [devXml])
|
||||
|
||||
const handleDevSimulate = async () => {
|
||||
if (!devXml.trim() || isSimulating) return
|
||||
|
||||
setIsSimulating(true)
|
||||
devStopRef.current = false
|
||||
const toolCallId = `dev-sim-${Date.now()}`
|
||||
const xml = devXml.trim()
|
||||
|
||||
// Add user message and initial assistant message with empty XML
|
||||
const userMsg = {
|
||||
id: `user-${Date.now()}`,
|
||||
role: "user" as const,
|
||||
parts: [
|
||||
{
|
||||
type: "text" as const,
|
||||
text: "[Dev] Simulating XML streaming",
|
||||
},
|
||||
],
|
||||
}
|
||||
const assistantMsg = {
|
||||
id: `assistant-${Date.now()}`,
|
||||
role: "assistant" as const,
|
||||
parts: [
|
||||
{
|
||||
type: "tool-display_diagram" as const,
|
||||
toolCallId,
|
||||
state: "input-streaming" as const,
|
||||
input: { xml: "" },
|
||||
},
|
||||
],
|
||||
}
|
||||
setMessages((prev) => [...prev, userMsg, assistantMsg] as any)
|
||||
|
||||
// Stream characters progressively
|
||||
for (let i = 0; i < xml.length; i += devChunkSize) {
|
||||
if (devStopRef.current) {
|
||||
setIsSimulating(false)
|
||||
return
|
||||
}
|
||||
|
||||
const chunk = xml.slice(0, i + devChunkSize)
|
||||
|
||||
setMessages((prev) => {
|
||||
const updated = [...prev]
|
||||
const lastMsg = updated[updated.length - 1] as any
|
||||
if (lastMsg?.role === "assistant" && lastMsg.parts?.[0]) {
|
||||
lastMsg.parts[0].input = { xml: chunk }
|
||||
}
|
||||
return updated
|
||||
})
|
||||
|
||||
await new Promise((r) => setTimeout(r, devIntervalMs))
|
||||
}
|
||||
|
||||
if (devStopRef.current) {
|
||||
setIsSimulating(false)
|
||||
return
|
||||
}
|
||||
|
||||
// Finalize: set state to output-available
|
||||
setMessages((prev) => {
|
||||
const updated = [...prev]
|
||||
const lastMsg = updated[updated.length - 1] as any
|
||||
if (lastMsg?.role === "assistant" && lastMsg.parts?.[0]) {
|
||||
lastMsg.parts[0].state = "output-available"
|
||||
lastMsg.parts[0].output = "Successfully displayed the diagram."
|
||||
lastMsg.parts[0].input = { xml }
|
||||
}
|
||||
return updated
|
||||
})
|
||||
|
||||
// Display the final diagram
|
||||
const fullXml = wrapWithMxFile(xml)
|
||||
onDisplayChart(fullXml)
|
||||
|
||||
setIsSimulating(false)
|
||||
}
|
||||
|
||||
return (
|
||||
<div className="border-t border-dashed border-orange-500/50 px-4 py-2 bg-orange-50/50 dark:bg-orange-950/30">
|
||||
<details>
|
||||
<summary className="text-xs text-orange-600 dark:text-orange-400 cursor-pointer font-medium">
|
||||
Dev: XML Streaming Simulator
|
||||
</summary>
|
||||
<div className="mt-2 space-y-2">
|
||||
<div className="flex items-center gap-2">
|
||||
<label className="text-xs text-muted-foreground whitespace-nowrap">
|
||||
Preset:
|
||||
</label>
|
||||
<select
|
||||
onChange={(e) => {
|
||||
if (e.target.value) {
|
||||
setDevXml(DEV_XML_PRESETS[e.target.value])
|
||||
}
|
||||
}}
|
||||
className="flex-1 text-xs p-1 border rounded bg-background"
|
||||
defaultValue=""
|
||||
>
|
||||
<option value="" disabled>
|
||||
Select a preset...
|
||||
</option>
|
||||
{Object.keys(DEV_XML_PRESETS).map((name) => (
|
||||
<option key={name} value={name}>
|
||||
{name}
|
||||
</option>
|
||||
))}
|
||||
</select>
|
||||
<button
|
||||
type="button"
|
||||
onClick={() => setDevXml("")}
|
||||
className="px-2 py-1 text-xs text-muted-foreground hover:text-foreground border rounded"
|
||||
>
|
||||
Clear
|
||||
</button>
|
||||
</div>
|
||||
<textarea
|
||||
value={devXml}
|
||||
onChange={(e) => setDevXml(e.target.value)}
|
||||
placeholder="Paste mxCell XML here or select a preset..."
|
||||
className="w-full h-24 text-xs font-mono p-2 border rounded bg-background"
|
||||
/>
|
||||
<div className="flex items-center gap-4">
|
||||
<div className="flex items-center gap-2 flex-1">
|
||||
<label className="text-xs text-muted-foreground whitespace-nowrap">
|
||||
Interval:
|
||||
</label>
|
||||
<input
|
||||
type="range"
|
||||
min="1"
|
||||
max="200"
|
||||
step="1"
|
||||
value={devIntervalMs}
|
||||
onChange={(e) =>
|
||||
setDevIntervalMs(Number(e.target.value))
|
||||
}
|
||||
className="flex-1 h-1 accent-orange-500"
|
||||
/>
|
||||
<span className="text-xs text-muted-foreground w-12">
|
||||
{devIntervalMs}ms
|
||||
</span>
|
||||
</div>
|
||||
<div className="flex items-center gap-2">
|
||||
<label className="text-xs text-muted-foreground whitespace-nowrap">
|
||||
Chars:
|
||||
</label>
|
||||
<input
|
||||
type="number"
|
||||
min="1"
|
||||
max="100"
|
||||
value={devChunkSize}
|
||||
onChange={(e) =>
|
||||
setDevChunkSize(
|
||||
Math.max(1, Number(e.target.value)),
|
||||
)
|
||||
}
|
||||
className="w-14 text-xs p-1 border rounded bg-background"
|
||||
/>
|
||||
</div>
|
||||
</div>
|
||||
<div className="flex gap-2">
|
||||
<button
|
||||
type="button"
|
||||
onClick={handleDevSimulate}
|
||||
disabled={isSimulating || !devXml.trim()}
|
||||
className="px-3 py-1 text-xs bg-orange-500 text-white rounded hover:bg-orange-600 disabled:opacity-50 disabled:cursor-not-allowed"
|
||||
>
|
||||
{isSimulating
|
||||
? "Streaming..."
|
||||
: `Simulate (${devChunkSize} chars/${devIntervalMs}ms)`}
|
||||
</button>
|
||||
{isSimulating && (
|
||||
<button
|
||||
type="button"
|
||||
onClick={() => {
|
||||
devStopRef.current = true
|
||||
}}
|
||||
className="px-3 py-1 text-xs bg-red-500 text-white rounded hover:bg-red-600"
|
||||
>
|
||||
Stop
|
||||
</button>
|
||||
)}
|
||||
</div>
|
||||
</div>
|
||||
</details>
|
||||
</div>
|
||||
)
|
||||
}
|
||||
383
hooks/use-diagram-tool-handlers.ts
Normal file
383
hooks/use-diagram-tool-handlers.ts
Normal file
@@ -0,0 +1,383 @@
|
||||
import type { MutableRefObject } from "react"
|
||||
import { isMxCellXmlComplete, wrapWithMxFile } from "@/lib/utils"
|
||||
|
||||
const DEBUG = process.env.NODE_ENV === "development"
|
||||
|
||||
interface ToolCall {
|
||||
toolCallId: string
|
||||
toolName: string
|
||||
input: unknown
|
||||
}
|
||||
|
||||
type AddToolOutputSuccess = {
|
||||
tool: string
|
||||
toolCallId: string
|
||||
state?: "output-available"
|
||||
output: string
|
||||
errorText?: undefined
|
||||
}
|
||||
|
||||
type AddToolOutputError = {
|
||||
tool: string
|
||||
toolCallId: string
|
||||
state: "output-error"
|
||||
output?: undefined
|
||||
errorText: string
|
||||
}
|
||||
|
||||
type AddToolOutputParams = AddToolOutputSuccess | AddToolOutputError
|
||||
|
||||
type AddToolOutputFn = (params: AddToolOutputParams) => void
|
||||
|
||||
interface DiagramOperation {
|
||||
type: "update" | "add" | "delete"
|
||||
cell_id: string
|
||||
new_xml?: string
|
||||
}
|
||||
|
||||
interface UseDiagramToolHandlersParams {
|
||||
partialXmlRef: MutableRefObject<string>
|
||||
editDiagramOriginalXmlRef: MutableRefObject<Map<string, string>>
|
||||
chartXMLRef: MutableRefObject<string>
|
||||
onDisplayChart: (xml: string, skipValidation?: boolean) => string | null
|
||||
onFetchChart: (saveToHistory?: boolean) => Promise<string>
|
||||
onExport: () => void
|
||||
}
|
||||
|
||||
/**
|
||||
* Hook that creates the onToolCall handler for diagram-related tools.
|
||||
* Handles display_diagram, edit_diagram, and append_diagram tools.
|
||||
*
|
||||
* Note: addToolOutput is passed at call time (not hook init) because
|
||||
* it comes from useChat which creates a circular dependency.
|
||||
*/
|
||||
export function useDiagramToolHandlers({
|
||||
partialXmlRef,
|
||||
editDiagramOriginalXmlRef,
|
||||
chartXMLRef,
|
||||
onDisplayChart,
|
||||
onFetchChart,
|
||||
onExport,
|
||||
}: UseDiagramToolHandlersParams) {
|
||||
const handleToolCall = async (
|
||||
{ toolCall }: { toolCall: ToolCall },
|
||||
addToolOutput: AddToolOutputFn,
|
||||
) => {
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
`[onToolCall] Tool: ${toolCall.toolName}, CallId: ${toolCall.toolCallId}`,
|
||||
)
|
||||
}
|
||||
|
||||
if (toolCall.toolName === "display_diagram") {
|
||||
await handleDisplayDiagram(toolCall, addToolOutput)
|
||||
} else if (toolCall.toolName === "edit_diagram") {
|
||||
await handleEditDiagram(toolCall, addToolOutput)
|
||||
} else if (toolCall.toolName === "append_diagram") {
|
||||
handleAppendDiagram(toolCall, addToolOutput)
|
||||
}
|
||||
}
|
||||
|
||||
const handleDisplayDiagram = async (
|
||||
toolCall: ToolCall,
|
||||
addToolOutput: AddToolOutputFn,
|
||||
) => {
|
||||
const { xml } = toolCall.input as { xml: string }
|
||||
|
||||
// DEBUG: Log raw input to diagnose false truncation detection
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] XML ending (last 100 chars):",
|
||||
xml.slice(-100),
|
||||
)
|
||||
console.log("[display_diagram] XML length:", xml.length)
|
||||
}
|
||||
|
||||
// Check if XML is truncated (incomplete mxCell indicates truncated output)
|
||||
const isTruncated = !isMxCellXmlComplete(xml)
|
||||
if (DEBUG) {
|
||||
console.log("[display_diagram] isTruncated:", isTruncated)
|
||||
}
|
||||
|
||||
if (isTruncated) {
|
||||
// Store the partial XML for continuation via append_diagram
|
||||
partialXmlRef.current = xml
|
||||
|
||||
// Tell LLM to use append_diagram to continue
|
||||
const partialEnding = partialXmlRef.current.slice(-500)
|
||||
addToolOutput({
|
||||
tool: "display_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Output was truncated due to length limits. Use the append_diagram tool to continue.
|
||||
|
||||
Your output ended with:
|
||||
\`\`\`
|
||||
${partialEnding}
|
||||
\`\`\`
|
||||
|
||||
NEXT STEP: Call append_diagram with the continuation XML.
|
||||
- Do NOT include wrapper tags or root cells (id="0", id="1")
|
||||
- Start from EXACTLY where you stopped
|
||||
- Complete all remaining mxCell elements`,
|
||||
})
|
||||
return
|
||||
}
|
||||
|
||||
// Complete XML received - use it directly
|
||||
// (continuation is now handled via append_diagram tool)
|
||||
const finalXml = xml
|
||||
partialXmlRef.current = "" // Reset any partial from previous truncation
|
||||
|
||||
// Wrap raw XML with full mxfile structure for draw.io
|
||||
const fullXml = wrapWithMxFile(finalXml)
|
||||
|
||||
// loadDiagram validates and returns error if invalid
|
||||
const validationError = onDisplayChart(fullXml)
|
||||
|
||||
if (validationError) {
|
||||
console.warn("[display_diagram] Validation error:", validationError)
|
||||
// Return error to model - sendAutomaticallyWhen will trigger retry
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] Adding tool output with state: output-error",
|
||||
)
|
||||
}
|
||||
addToolOutput({
|
||||
tool: "display_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `${validationError}
|
||||
|
||||
Please fix the XML issues and call display_diagram again with corrected XML.
|
||||
|
||||
Your failed XML:
|
||||
\`\`\`xml
|
||||
${finalXml}
|
||||
\`\`\``,
|
||||
})
|
||||
} else {
|
||||
// Success - diagram will be rendered by chat-message-display
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] Success! Adding tool output with state: output-available",
|
||||
)
|
||||
}
|
||||
addToolOutput({
|
||||
tool: "display_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
output: "Successfully displayed the diagram.",
|
||||
})
|
||||
if (DEBUG) {
|
||||
console.log(
|
||||
"[display_diagram] Tool output added. Diagram should be visible now.",
|
||||
)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
const handleEditDiagram = async (
|
||||
toolCall: ToolCall,
|
||||
addToolOutput: AddToolOutputFn,
|
||||
) => {
|
||||
const { operations } = toolCall.input as {
|
||||
operations: DiagramOperation[]
|
||||
}
|
||||
|
||||
let currentXml = ""
|
||||
try {
|
||||
// Use the original XML captured during streaming (shared with chat-message-display)
|
||||
// This ensures we apply operations to the same base XML that streaming used
|
||||
const originalXml = editDiagramOriginalXmlRef.current.get(
|
||||
toolCall.toolCallId,
|
||||
)
|
||||
if (originalXml) {
|
||||
currentXml = originalXml
|
||||
} else {
|
||||
// Fallback: use chartXML from ref if streaming didn't capture original
|
||||
const cachedXML = chartXMLRef.current
|
||||
if (cachedXML) {
|
||||
currentXml = cachedXML
|
||||
} else {
|
||||
// Last resort: export from iframe
|
||||
currentXml = await onFetchChart(false)
|
||||
}
|
||||
}
|
||||
|
||||
const { applyDiagramOperations } = await import("@/lib/utils")
|
||||
const { result: editedXml, errors } = applyDiagramOperations(
|
||||
currentXml,
|
||||
operations,
|
||||
)
|
||||
|
||||
// Check for operation errors
|
||||
if (errors.length > 0) {
|
||||
const errorMessages = errors
|
||||
.map(
|
||||
(e) =>
|
||||
`- ${e.type} on cell_id="${e.cellId}": ${e.message}`,
|
||||
)
|
||||
.join("\n")
|
||||
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Some operations failed:\n${errorMessages}
|
||||
|
||||
Current diagram XML:
|
||||
\`\`\`xml
|
||||
${currentXml}
|
||||
\`\`\`
|
||||
|
||||
Please check the cell IDs and retry.`,
|
||||
})
|
||||
// Clean up the shared original XML ref
|
||||
editDiagramOriginalXmlRef.current.delete(toolCall.toolCallId)
|
||||
return
|
||||
}
|
||||
|
||||
// loadDiagram validates and returns error if invalid
|
||||
const validationError = onDisplayChart(editedXml)
|
||||
if (validationError) {
|
||||
console.warn(
|
||||
"[edit_diagram] Validation error:",
|
||||
validationError,
|
||||
)
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Edit produced invalid XML: ${validationError}
|
||||
|
||||
Current diagram XML:
|
||||
\`\`\`xml
|
||||
${currentXml}
|
||||
\`\`\`
|
||||
|
||||
Please fix the operations to avoid structural issues.`,
|
||||
})
|
||||
// Clean up the shared original XML ref
|
||||
editDiagramOriginalXmlRef.current.delete(toolCall.toolCallId)
|
||||
return
|
||||
}
|
||||
onExport()
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
output: `Successfully applied ${operations.length} operation(s) to the diagram.`,
|
||||
})
|
||||
// Clean up the shared original XML ref
|
||||
editDiagramOriginalXmlRef.current.delete(toolCall.toolCallId)
|
||||
} catch (error) {
|
||||
console.error("[edit_diagram] Failed:", error)
|
||||
|
||||
const errorMessage =
|
||||
error instanceof Error ? error.message : String(error)
|
||||
|
||||
addToolOutput({
|
||||
tool: "edit_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Edit failed: ${errorMessage}
|
||||
|
||||
Current diagram XML:
|
||||
\`\`\`xml
|
||||
${currentXml || "No XML available"}
|
||||
\`\`\`
|
||||
|
||||
Please check cell IDs and retry, or use display_diagram to regenerate.`,
|
||||
})
|
||||
// Clean up the shared original XML ref even on error
|
||||
editDiagramOriginalXmlRef.current.delete(toolCall.toolCallId)
|
||||
}
|
||||
}
|
||||
|
||||
const handleAppendDiagram = (
|
||||
toolCall: ToolCall,
|
||||
addToolOutput: AddToolOutputFn,
|
||||
) => {
|
||||
const { xml } = toolCall.input as { xml: string }
|
||||
|
||||
// Detect if LLM incorrectly started fresh instead of continuing
|
||||
// LLM should only output bare mxCells now, so wrapper tags indicate error
|
||||
const trimmed = xml.trim()
|
||||
const isFreshStart =
|
||||
trimmed.startsWith("<mxGraphModel") ||
|
||||
trimmed.startsWith("<root") ||
|
||||
trimmed.startsWith("<mxfile") ||
|
||||
trimmed.startsWith('<mxCell id="0"') ||
|
||||
trimmed.startsWith('<mxCell id="1"')
|
||||
|
||||
if (isFreshStart) {
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `ERROR: You started fresh with wrapper tags. Do NOT include wrapper tags or root cells (id="0", id="1").
|
||||
|
||||
Continue from EXACTLY where the partial ended:
|
||||
\`\`\`
|
||||
${partialXmlRef.current.slice(-500)}
|
||||
\`\`\`
|
||||
|
||||
Start your continuation with the NEXT character after where it stopped.`,
|
||||
})
|
||||
return
|
||||
}
|
||||
|
||||
// Append to accumulated XML
|
||||
partialXmlRef.current += xml
|
||||
|
||||
// Check if XML is now complete (last mxCell is complete)
|
||||
const isComplete = isMxCellXmlComplete(partialXmlRef.current)
|
||||
|
||||
if (isComplete) {
|
||||
// Wrap and display the complete diagram
|
||||
const finalXml = partialXmlRef.current
|
||||
partialXmlRef.current = "" // Reset
|
||||
|
||||
const fullXml = wrapWithMxFile(finalXml)
|
||||
const validationError = onDisplayChart(fullXml)
|
||||
|
||||
if (validationError) {
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `Validation error after assembly: ${validationError}
|
||||
|
||||
Assembled XML:
|
||||
\`\`\`xml
|
||||
${finalXml.substring(0, 2000)}...
|
||||
\`\`\`
|
||||
|
||||
Please use display_diagram with corrected XML.`,
|
||||
})
|
||||
} else {
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
output: "Diagram assembly complete and displayed successfully.",
|
||||
})
|
||||
}
|
||||
} else {
|
||||
// Still incomplete - signal to continue
|
||||
addToolOutput({
|
||||
tool: "append_diagram",
|
||||
toolCallId: toolCall.toolCallId,
|
||||
state: "output-error",
|
||||
errorText: `XML still incomplete (mxCell not closed). Call append_diagram again to continue.
|
||||
|
||||
Current ending:
|
||||
\`\`\`
|
||||
${partialXmlRef.current.slice(-500)}
|
||||
\`\`\`
|
||||
|
||||
Continue from EXACTLY where you stopped.`,
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
return { handleToolCall }
|
||||
}
|
||||
159
lib/utils.ts
159
lib/utils.ts
@@ -36,29 +36,32 @@ const VALID_ENTITIES = new Set(["lt", "gt", "amp", "quot", "apos"])
|
||||
/**
|
||||
* Check if mxCell XML output is complete (not truncated).
|
||||
* Complete XML ends with a self-closing tag (/>) or closing mxCell tag.
|
||||
* Also handles function-calling wrapper tags that may be incorrectly included.
|
||||
* Uses a robust approach that handles any LLM provider's wrapper tags
|
||||
* by finding the last valid mxCell ending and checking if suffix is just closing tags.
|
||||
* @param xml - The XML string to check (can be undefined/null)
|
||||
* @returns true if XML appears complete, false if truncated or empty
|
||||
*/
|
||||
export function isMxCellXmlComplete(xml: string | undefined | null): boolean {
|
||||
let trimmed = xml?.trim() || ""
|
||||
const trimmed = xml?.trim() || ""
|
||||
if (!trimmed) return false
|
||||
|
||||
// Strip Anthropic function-calling wrapper tags if present
|
||||
// These can leak into tool input due to AI SDK parsing issues
|
||||
// Use loop because tags are nested: </mxCell></mxParameter></invoke>
|
||||
let prev = ""
|
||||
while (prev !== trimmed) {
|
||||
prev = trimmed
|
||||
trimmed = trimmed
|
||||
.replace(/<\/mxParameter>\s*$/i, "")
|
||||
.replace(/<\/invoke>\s*$/i, "")
|
||||
.replace(/<\/antml:parameter>\s*$/i, "")
|
||||
.replace(/<\/antml:invoke>\s*$/i, "")
|
||||
.trim()
|
||||
}
|
||||
// Find position of last complete mxCell ending (either /> or </mxCell>)
|
||||
const lastSelfClose = trimmed.lastIndexOf("/>")
|
||||
const lastMxCellClose = trimmed.lastIndexOf("</mxCell>")
|
||||
|
||||
return trimmed.endsWith("/>") || trimmed.endsWith("</mxCell>")
|
||||
const lastValidEnd = Math.max(lastSelfClose, lastMxCellClose)
|
||||
|
||||
// No valid ending found at all
|
||||
if (lastValidEnd === -1) return false
|
||||
|
||||
// Check what comes after the last valid ending
|
||||
// For />: add 2 chars, for </mxCell>: add 9 chars
|
||||
const endOffset = lastMxCellClose > lastSelfClose ? 9 : 2
|
||||
const suffix = trimmed.slice(lastValidEnd + endOffset)
|
||||
|
||||
// If suffix is empty or only contains closing tags (any provider's wrapper) or whitespace, it's complete
|
||||
// This regex matches any sequence of closing XML tags like </foo>, </bar>, </|DSML|xyz>
|
||||
return /^(\s*<\/[^>]+>)*\s*$/.test(suffix)
|
||||
}
|
||||
|
||||
/**
|
||||
@@ -262,6 +265,21 @@ export function convertToLegalXml(xmlString: string): string {
|
||||
"&",
|
||||
)
|
||||
|
||||
// Fix unescaped < and > in attribute values for XML parsing
|
||||
// HTML content in value attributes (e.g., <b>Title</b>) needs to be escaped
|
||||
// This is critical because DOMParser will fail on unescaped < > in attributes
|
||||
if (/=\s*"[^"]*<[^"]*"/.test(cellContent)) {
|
||||
cellContent = cellContent.replace(
|
||||
/=\s*"([^"]*)"/g,
|
||||
(_match, value) => {
|
||||
const escaped = value
|
||||
.replace(/</g, "<")
|
||||
.replace(/>/g, ">")
|
||||
return `="${escaped}"`
|
||||
},
|
||||
)
|
||||
}
|
||||
|
||||
// Indent each line of the matched block for readability.
|
||||
const formatted = cellContent
|
||||
.split("\n")
|
||||
@@ -306,6 +324,20 @@ export function wrapWithMxFile(xml: string): string {
|
||||
content = xml.replace(/<\/?root>/g, "").trim()
|
||||
}
|
||||
|
||||
// Strip trailing LLM wrapper tags (from any provider: Anthropic, DeepSeek, etc.)
|
||||
// Find the last valid mxCell ending and remove everything after it
|
||||
const lastSelfClose = content.lastIndexOf("/>")
|
||||
const lastMxCellClose = content.lastIndexOf("</mxCell>")
|
||||
const lastValidEnd = Math.max(lastSelfClose, lastMxCellClose)
|
||||
if (lastValidEnd !== -1) {
|
||||
const endOffset = lastMxCellClose > lastSelfClose ? 9 : 2
|
||||
const suffix = content.slice(lastValidEnd + endOffset)
|
||||
// If suffix is only closing tags (wrapper tags), strip it
|
||||
if (/^(\s*<\/[^>]+>)*\s*$/.test(suffix)) {
|
||||
content = content.slice(0, lastValidEnd + endOffset)
|
||||
}
|
||||
}
|
||||
|
||||
// Remove any existing root cells from content (LLM shouldn't include them, but handle it gracefully)
|
||||
// Use flexible patterns that match both self-closing (/>) and non-self-closing (></mxCell>) formats
|
||||
content = content
|
||||
@@ -910,6 +942,21 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
fixes.push("Removed CDATA wrapper")
|
||||
}
|
||||
|
||||
// 1b. Strip trailing LLM wrapper tags (DeepSeek, Anthropic, etc.)
|
||||
// These are closing tags after the last valid mxCell that break XML parsing
|
||||
const lastSelfClose = fixed.lastIndexOf("/>")
|
||||
const lastMxCellClose = fixed.lastIndexOf("</mxCell>")
|
||||
const lastValidEnd = Math.max(lastSelfClose, lastMxCellClose)
|
||||
if (lastValidEnd !== -1) {
|
||||
const endOffset = lastMxCellClose > lastSelfClose ? 9 : 2
|
||||
const suffix = fixed.slice(lastValidEnd + endOffset)
|
||||
// If suffix contains only closing tags (wrapper tags) or whitespace, strip it
|
||||
if (/^(\s*<\/[^>]+>)+\s*$/.test(suffix)) {
|
||||
fixed = fixed.slice(0, lastValidEnd + endOffset)
|
||||
fixes.push("Stripped trailing LLM wrapper tags")
|
||||
}
|
||||
}
|
||||
|
||||
// 2. Remove text before XML declaration or root element (only if it's garbage text, not valid XML)
|
||||
const xmlStart = fixed.search(/<(\?xml|mxGraphModel|mxfile)/i)
|
||||
if (xmlStart > 0 && !/^<[a-zA-Z]/.test(fixed.trim())) {
|
||||
@@ -1015,8 +1062,8 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
fixes.push("Removed quotes around color values in style")
|
||||
}
|
||||
|
||||
// 4. Fix unescaped < in attribute values
|
||||
// This is tricky - we need to find < inside quoted attribute values
|
||||
// 4. Fix unescaped < and > in attribute values
|
||||
// < is required to be escaped, > is not strictly required but we escape for consistency
|
||||
const attrPattern = /(=\s*")([^"]*?)(<)([^"]*?)(")/g
|
||||
let attrMatch
|
||||
let hasUnescapedLt = false
|
||||
@@ -1027,12 +1074,12 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
}
|
||||
}
|
||||
if (hasUnescapedLt) {
|
||||
// Replace < with < inside attribute values
|
||||
// Replace < and > with < and > inside attribute values
|
||||
fixed = fixed.replace(/=\s*"([^"]*)"/g, (_match, value) => {
|
||||
const escaped = value.replace(/</g, "<")
|
||||
const escaped = value.replace(/</g, "<").replace(/>/g, ">")
|
||||
return `="${escaped}"`
|
||||
})
|
||||
fixes.push("Escaped < characters in attribute values")
|
||||
fixes.push("Escaped <> characters in attribute values")
|
||||
}
|
||||
|
||||
// 5. Fix invalid character references (remove malformed ones)
|
||||
@@ -1120,7 +1167,8 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
}
|
||||
|
||||
// 8c. Remove non-draw.io tags (after typo fixes so lowercase variants are fixed first)
|
||||
// Valid draw.io tags: mxfile, diagram, mxGraphModel, root, mxCell, mxGeometry, mxPoint, Array, Object
|
||||
// IMPORTANT: Only remove tags at the element level, NOT inside quoted attribute values
|
||||
// Tags like <b>, <br> inside value="<b>text</b>" should be preserved (they're HTML content)
|
||||
const validDrawioTags = new Set([
|
||||
"mxfile",
|
||||
"diagram",
|
||||
@@ -1133,25 +1181,59 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
"Object",
|
||||
"mxRectangle",
|
||||
])
|
||||
|
||||
// Helper: Check if a position is inside a quoted attribute value
|
||||
// by counting unescaped quotes before that position
|
||||
const isInsideQuotes = (str: string, pos: number): boolean => {
|
||||
let inQuote = false
|
||||
let quoteChar = ""
|
||||
for (let i = 0; i < pos && i < str.length; i++) {
|
||||
const c = str[i]
|
||||
if (inQuote) {
|
||||
if (c === quoteChar) inQuote = false
|
||||
} else if (c === '"' || c === "'") {
|
||||
// Check if this quote is part of an attribute (preceded by =)
|
||||
// Look back for = sign
|
||||
let j = i - 1
|
||||
while (j >= 0 && /\s/.test(str[j])) j--
|
||||
if (j >= 0 && str[j] === "=") {
|
||||
inQuote = true
|
||||
quoteChar = c
|
||||
}
|
||||
}
|
||||
}
|
||||
return inQuote
|
||||
}
|
||||
|
||||
const foreignTagPattern = /<\/?([a-zA-Z][a-zA-Z0-9_]*)[^>]*>/g
|
||||
let foreignMatch
|
||||
const foreignTags = new Set<string>()
|
||||
const foreignTagPositions: Array<{
|
||||
tag: string
|
||||
start: number
|
||||
end: number
|
||||
}> = []
|
||||
|
||||
while ((foreignMatch = foreignTagPattern.exec(fixed)) !== null) {
|
||||
const tagName = foreignMatch[1]
|
||||
if (!validDrawioTags.has(tagName)) {
|
||||
foreignTags.add(tagName)
|
||||
}
|
||||
// Skip if this is a valid draw.io tag
|
||||
if (validDrawioTags.has(tagName)) continue
|
||||
// Skip if this tag is inside a quoted attribute value
|
||||
if (isInsideQuotes(fixed, foreignMatch.index)) continue
|
||||
|
||||
foreignTags.add(tagName)
|
||||
foreignTagPositions.push({
|
||||
tag: tagName,
|
||||
start: foreignMatch.index,
|
||||
end: foreignMatch.index + foreignMatch[0].length,
|
||||
})
|
||||
}
|
||||
if (foreignTags.size > 0) {
|
||||
console.log(
|
||||
"[autoFixXml] Step 8c: Found foreign tags:",
|
||||
Array.from(foreignTags),
|
||||
)
|
||||
for (const tag of foreignTags) {
|
||||
// Remove opening tags (with or without attributes)
|
||||
fixed = fixed.replace(new RegExp(`<${tag}[^>]*>`, "gi"), "")
|
||||
// Remove closing tags
|
||||
fixed = fixed.replace(new RegExp(`</${tag}>`, "gi"), "")
|
||||
|
||||
if (foreignTagPositions.length > 0) {
|
||||
// Remove tags from end to start to preserve indices
|
||||
foreignTagPositions.sort((a, b) => b.start - a.start)
|
||||
for (const { start, end } of foreignTagPositions) {
|
||||
fixed = fixed.slice(0, start) + fixed.slice(end)
|
||||
}
|
||||
fixes.push(
|
||||
`Removed foreign tags: ${Array.from(foreignTags).join(", ")}`,
|
||||
@@ -1202,6 +1284,7 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
|
||||
// 10b. Remove extra closing tags (more closes than opens)
|
||||
// Need to properly count self-closing tags (they don't need closing tags)
|
||||
// IMPORTANT: Only count tags at element level, NOT inside quoted attribute values
|
||||
const tagCounts = new Map<
|
||||
string,
|
||||
{ opens: number; closes: number; selfClosing: number }
|
||||
@@ -1210,12 +1293,18 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
const fullTagPattern = /<(\/?[a-zA-Z][a-zA-Z0-9]*)[^>]*>/g
|
||||
let tagCountMatch
|
||||
while ((tagCountMatch = fullTagPattern.exec(fixed)) !== null) {
|
||||
// Skip tags inside quoted attribute values (e.g., value="<b>Title</b>")
|
||||
if (isInsideQuotes(fixed, tagCountMatch.index)) continue
|
||||
|
||||
const fullMatch = tagCountMatch[0] // e.g., "<mxCell .../>" or "</mxCell>"
|
||||
const tagPart = tagCountMatch[1] // e.g., "mxCell" or "/mxCell"
|
||||
const isClosing = tagPart.startsWith("/")
|
||||
const isSelfClosing = fullMatch.endsWith("/>")
|
||||
const tagName = isClosing ? tagPart.slice(1) : tagPart
|
||||
|
||||
// Only count valid draw.io tags - skip partial/invalid tags like "mx" from streaming
|
||||
if (!validDrawioTags.has(tagName)) continue
|
||||
|
||||
let counts = tagCounts.get(tagName)
|
||||
if (!counts) {
|
||||
counts = { opens: 0, closes: 0, selfClosing: 0 }
|
||||
|
||||
@@ -459,7 +459,8 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
fixes.push("Removed quotes around color values in style")
|
||||
}
|
||||
|
||||
// 10. Fix unescaped < in attribute values
|
||||
// 10. Fix unescaped < and > in attribute values
|
||||
// < is required to be escaped, > is not strictly required but we escape for consistency
|
||||
const attrPattern = /(=\s*")([^"]*?)(<)([^"]*?)(")/g
|
||||
let attrMatch
|
||||
let hasUnescapedLt = false
|
||||
@@ -471,10 +472,10 @@ export function autoFixXml(xml: string): { fixed: string; fixes: string[] } {
|
||||
}
|
||||
if (hasUnescapedLt) {
|
||||
fixed = fixed.replace(/=\s*"([^"]*)"/g, (_match, value) => {
|
||||
const escaped = value.replace(/</g, "<")
|
||||
const escaped = value.replace(/</g, "<").replace(/>/g, ">")
|
||||
return `="${escaped}"`
|
||||
})
|
||||
fixes.push("Escaped < characters in attribute values")
|
||||
fixes.push("Escaped <> characters in attribute values")
|
||||
}
|
||||
|
||||
// 11. Fix invalid hex character references
|
||||
@@ -903,24 +904,30 @@ export function validateAndFixXml(xml: string): {
|
||||
|
||||
/**
|
||||
* Check if mxCell XML output is complete (not truncated).
|
||||
* Uses a robust approach that handles any LLM provider's wrapper tags
|
||||
* by finding the last valid mxCell ending and checking if suffix is just closing tags.
|
||||
* @param xml - The XML string to check (can be undefined/null)
|
||||
* @returns true if XML appears complete, false if truncated or empty
|
||||
*/
|
||||
export function isMxCellXmlComplete(xml: string | undefined | null): boolean {
|
||||
let trimmed = xml?.trim() || ""
|
||||
const trimmed = xml?.trim() || ""
|
||||
if (!trimmed) return false
|
||||
|
||||
// Strip wrapper tags if present
|
||||
let prev = ""
|
||||
while (prev !== trimmed) {
|
||||
prev = trimmed
|
||||
trimmed = trimmed
|
||||
.replace(/<\/mxParameter>\s*$/i, "")
|
||||
.replace(/<\/invoke>\s*$/i, "")
|
||||
.replace(/<\/antml:parameter>\s*$/i, "")
|
||||
.replace(/<\/antml:invoke>\s*$/i, "")
|
||||
.trim()
|
||||
}
|
||||
// Find position of last complete mxCell ending (either /> or </mxCell>)
|
||||
const lastSelfClose = trimmed.lastIndexOf("/>")
|
||||
const lastMxCellClose = trimmed.lastIndexOf("</mxCell>")
|
||||
|
||||
return trimmed.endsWith("/>") || trimmed.endsWith("</mxCell>")
|
||||
const lastValidEnd = Math.max(lastSelfClose, lastMxCellClose)
|
||||
|
||||
// No valid ending found at all
|
||||
if (lastValidEnd === -1) return false
|
||||
|
||||
// Check what comes after the last valid ending
|
||||
// For />: add 2 chars, for </mxCell>: add 9 chars
|
||||
const endOffset = lastMxCellClose > lastSelfClose ? 9 : 2
|
||||
const suffix = trimmed.slice(lastValidEnd + endOffset)
|
||||
|
||||
// If suffix is empty or only contains closing tags (any provider's wrapper) or whitespace, it's complete
|
||||
// This regex matches any sequence of closing XML tags like </foo>, </bar>, </|DSML|xyz>
|
||||
return /^(\s*<\/[^>]+>)*\s*$/.test(suffix)
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user