Merge origin/dev into PR branch and resolve toolcall test conflicts

This commit is contained in:
CJACK.
2026-03-08 01:10:53 +08:00
parent 9b93badb57
commit 60e9d707d4
16 changed files with 586 additions and 25 deletions

View File

@@ -39,6 +39,9 @@ func ParseToolCallsDetailed(text string, availableToolNames []string) ToolCallPa
var parsed []ParsedToolCall
for _, candidate := range candidates {
tc := parseToolCallsPayload(candidate)
if len(tc) == 0 {
tc = parseXMLToolCalls(candidate)
}
if len(tc) == 0 {
tc = parseMarkupToolCalls(candidate)
}
@@ -49,7 +52,11 @@ func ParseToolCallsDetailed(text string, availableToolNames []string) ToolCallPa
}
}
if len(parsed) == 0 {
return result
parsed = parseXMLToolCalls(text)
if len(parsed) == 0 {
return result
}
result.SawToolCallSyntax = true
}
calls, rejectedNames := filterToolCallsDetailed(parsed, availableToolNames)
@@ -80,6 +87,9 @@ func ParseStandaloneToolCallsDetailed(text string, availableToolNames []string)
continue
}
parsed := parseToolCallsPayload(candidate)
if len(parsed) == 0 {
parsed = parseXMLToolCalls(candidate)
}
if len(parsed) == 0 {
parsed = parseMarkupToolCalls(candidate)
}

View File

@@ -0,0 +1,219 @@
package util
import (
"encoding/json"
"encoding/xml"
"regexp"
"strings"
)
var xmlToolCallPattern = regexp.MustCompile(`(?is)<tool_call>\s*(.*?)\s*</tool_call>`)
var functionCallPattern = regexp.MustCompile(`(?is)<function_call>\s*([^<]+?)\s*</function_call>`)
var functionParamPattern = regexp.MustCompile(`(?is)<function\s+parameter\s+name="([^"]+)"\s*>\s*(.*?)\s*</function\s+parameter>`)
var antmlFunctionCallPattern = regexp.MustCompile(`(?is)<(?:[a-z0-9_]+:)?function_call[^>]*(?:name|function)="([^"]+)"[^>]*>\s*(.*?)\s*</(?:[a-z0-9_]+:)?function_call>`)
var antmlArgumentPattern = regexp.MustCompile(`(?is)<(?:[a-z0-9_]+:)?argument\s+name="([^"]+)"\s*>\s*(.*?)\s*</(?:[a-z0-9_]+:)?argument>`)
var antmlParametersPattern = regexp.MustCompile(`(?is)<(?:[a-z0-9_]+:)?parameters\s*>\s*(\{.*?\})\s*</(?:[a-z0-9_]+:)?parameters>`)
var invokeCallPattern = regexp.MustCompile(`(?is)<invoke\s+name="([^"]+)"\s*>(.*?)</invoke>`)
var invokeParamPattern = regexp.MustCompile(`(?is)<parameter\s+name="([^"]+)"\s*>\s*(.*?)\s*</parameter>`)
func parseXMLToolCalls(text string) []ParsedToolCall {
matches := xmlToolCallPattern.FindAllString(text, -1)
out := make([]ParsedToolCall, 0, len(matches)+1)
for _, block := range matches {
call, ok := parseSingleXMLToolCall(block)
if !ok {
continue
}
out = append(out, call)
}
if len(out) > 0 {
return out
}
if call, ok := parseFunctionCallTagStyle(text); ok {
return []ParsedToolCall{call}
}
if calls := parseAntmlFunctionCallStyles(text); len(calls) > 0 {
return calls
}
if call, ok := parseInvokeFunctionCallStyle(text); ok {
return []ParsedToolCall{call}
}
return nil
}
func parseSingleXMLToolCall(block string) (ParsedToolCall, bool) {
inner := strings.TrimSpace(block)
inner = strings.TrimPrefix(inner, "<tool_call>")
inner = strings.TrimSuffix(inner, "</tool_call>")
inner = strings.TrimSpace(inner)
if strings.HasPrefix(inner, "{") {
var payload map[string]any
if err := json.Unmarshal([]byte(inner), &payload); err == nil {
name := strings.TrimSpace(asString(payload["tool"]))
if name == "" {
name = strings.TrimSpace(asString(payload["tool_name"]))
}
if name != "" {
input := map[string]any{}
if params, ok := payload["params"].(map[string]any); ok {
input = params
} else if params, ok := payload["parameters"].(map[string]any); ok {
input = params
}
return ParsedToolCall{Name: name, Input: input}, true
}
}
}
dec := xml.NewDecoder(strings.NewReader(block))
name := ""
params := map[string]any{}
inParams := false
inTool := false
for {
tok, err := dec.Token()
if err != nil {
break
}
switch t := tok.(type) {
case xml.StartElement:
tag := strings.ToLower(t.Name.Local)
switch tag {
case "tool":
inTool = true
for _, attr := range t.Attr {
if strings.EqualFold(strings.TrimSpace(attr.Name.Local), "name") && strings.TrimSpace(name) == "" {
name = strings.TrimSpace(attr.Value)
}
}
case "parameters":
inParams = true
case "tool_name", "name":
var v string
if err := dec.DecodeElement(&v, &t); err == nil && strings.TrimSpace(v) != "" {
name = strings.TrimSpace(v)
}
default:
if inParams || inTool {
var v string
if err := dec.DecodeElement(&v, &t); err == nil {
params[t.Name.Local] = strings.TrimSpace(v)
}
}
}
case xml.EndElement:
tag := strings.ToLower(t.Name.Local)
if tag == "parameters" {
inParams = false
}
if tag == "tool" {
inTool = false
}
}
}
if strings.TrimSpace(name) == "" {
return ParsedToolCall{}, false
}
return ParsedToolCall{Name: strings.TrimSpace(name), Input: params}, true
}
func parseFunctionCallTagStyle(text string) (ParsedToolCall, bool) {
m := functionCallPattern.FindStringSubmatch(text)
if len(m) < 2 {
return ParsedToolCall{}, false
}
name := strings.TrimSpace(m[1])
if name == "" {
return ParsedToolCall{}, false
}
input := map[string]any{}
for _, pm := range functionParamPattern.FindAllStringSubmatch(text, -1) {
if len(pm) < 3 {
continue
}
key := strings.TrimSpace(pm[1])
val := strings.TrimSpace(pm[2])
if key != "" {
input[key] = val
}
}
return ParsedToolCall{Name: name, Input: input}, true
}
func parseAntmlFunctionCallStyles(text string) []ParsedToolCall {
matches := antmlFunctionCallPattern.FindAllStringSubmatch(text, -1)
if len(matches) == 0 {
return nil
}
out := make([]ParsedToolCall, 0, len(matches))
for _, m := range matches {
if call, ok := parseSingleAntmlFunctionCallMatch(m); ok {
out = append(out, call)
}
}
if len(out) == 0 {
return nil
}
return out
}
func parseSingleAntmlFunctionCallMatch(m []string) (ParsedToolCall, bool) {
if len(m) < 3 {
return ParsedToolCall{}, false
}
name := strings.TrimSpace(m[1])
if name == "" {
return ParsedToolCall{}, false
}
body := strings.TrimSpace(m[2])
input := map[string]any{}
if strings.HasPrefix(body, "{") {
if err := json.Unmarshal([]byte(body), &input); err == nil {
return ParsedToolCall{Name: name, Input: input}, true
}
}
if pm := antmlParametersPattern.FindStringSubmatch(body); len(pm) >= 2 {
if err := json.Unmarshal([]byte(strings.TrimSpace(pm[1])), &input); err == nil {
return ParsedToolCall{Name: name, Input: input}, true
}
}
for _, am := range antmlArgumentPattern.FindAllStringSubmatch(body, -1) {
if len(am) < 3 {
continue
}
k := strings.TrimSpace(am[1])
v := strings.TrimSpace(am[2])
if k != "" {
input[k] = v
}
}
return ParsedToolCall{Name: name, Input: input}, true
}
func parseInvokeFunctionCallStyle(text string) (ParsedToolCall, bool) {
m := invokeCallPattern.FindStringSubmatch(text)
if len(m) < 3 {
return ParsedToolCall{}, false
}
name := strings.TrimSpace(m[1])
if name == "" {
return ParsedToolCall{}, false
}
input := map[string]any{}
for _, pm := range invokeParamPattern.FindAllStringSubmatch(m[2], -1) {
if len(pm) < 3 {
continue
}
k := strings.TrimSpace(pm[1])
v := strings.TrimSpace(pm[2])
if k != "" {
input[k] = v
}
}
return ParsedToolCall{Name: name, Input: input}, true
}
func asString(v any) string {
s, _ := v.(string)
return s
}

View File

@@ -138,6 +138,140 @@ func TestParseToolCallsAllowsPunctuationVariantToolName(t *testing.T) {
}
}
func TestParseToolCallsSupportsClaudeXMLToolCall(t *testing.T) {
text := `<tool_call><tool_name>Bash</tool_name><parameters><command>pwd</command><description>show cwd</description></parameters></tool_call>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsDetailedMarksXMLToolCallSyntax(t *testing.T) {
text := `<tool_call><tool_name>Bash</tool_name><parameters><command>pwd</command></parameters></tool_call>`
res := ParseToolCallsDetailed(text, []string{"bash"})
if !res.SawToolCallSyntax {
t.Fatalf("expected SawToolCallSyntax=true, got %#v", res)
}
if len(res.Calls) != 1 {
t.Fatalf("expected one parsed call, got %#v", res)
}
}
func TestParseToolCallsSupportsClaudeXMLJSONToolCall(t *testing.T) {
text := `<tool_call>{"tool":"Bash","params":{"command":"pwd","description":"show cwd"}}</tool_call>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsFunctionCallTagStyle(t *testing.T) {
text := `<function_call>Bash</function_call><function parameter name="command">ls -la</function parameter><function parameter name="description">list</function parameter>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "ls -la" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsAntmlFunctionCallStyle(t *testing.T) {
text := `<antml:function_calls><antml:function_call name="Bash">{"command":"pwd","description":"x"}</antml:function_call></antml:function_calls>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsAntmlArgumentStyle(t *testing.T) {
text := `<antml:function_calls><antml:function_call id="1" name="Bash"><antml:argument name="command">pwd</antml:argument><antml:argument name="description">x</antml:argument></antml:function_call></antml:function_calls>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsInvokeFunctionCallStyle(t *testing.T) {
text := `<function_calls><invoke name="Bash"><parameter name="command">pwd</parameter><parameter name="description">d</parameter></invoke></function_calls>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsNestedToolTagStyle(t *testing.T) {
text := `<tool_call><tool name="Bash"><command>pwd</command><description>show cwd</description></tool></tool_call>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsAntmlFunctionAttributeWithParametersTag(t *testing.T) {
text := `<antml:function_calls><antml:function_call id="x" function="Bash"><antml:parameters>{"command":"pwd"}</antml:parameters></antml:function_call></antml:function_calls>`
calls := ParseToolCalls(text, []string{"bash"})
if len(calls) != 1 {
t.Fatalf("expected 1 call, got %#v", calls)
}
if calls[0].Name != "bash" {
t.Fatalf("expected canonical tool name bash, got %q", calls[0].Name)
}
if calls[0].Input["command"] != "pwd" {
t.Fatalf("expected command argument, got %#v", calls[0].Input)
}
}
func TestParseToolCallsSupportsMultipleAntmlFunctionCalls(t *testing.T) {
text := `<antml:function_calls><antml:function_call id="1" function="Bash"><antml:parameters>{"command":"pwd"}</antml:parameters></antml:function_call><antml:function_call id="2" function="Read"><antml:parameters>{"file_path":"README.md"}</antml:parameters></antml:function_call></antml:function_calls>`
calls := ParseToolCalls(text, []string{"bash", "read"})
if len(calls) != 2 {
t.Fatalf("expected 2 calls, got %#v", calls)
}
if calls[0].Name != "bash" || calls[1].Name != "read" {
t.Fatalf("expected canonical names [bash read], got %#v", calls)
}
}
func TestParseToolCallsDoesNotAcceptMismatchedMarkupTags(t *testing.T) {
text := `<tool_call><name>read_file</function><arguments>{"path":"README.md"}</arguments></tool_call>`
calls := ParseToolCalls(text, []string{"read_file"})