feat(functions): parse broken JSON when we parse the raw results, use dynamic rules for grammar keys (#2912)

* feat(functions): enhance parsing with broken JSON when we parse the raw results

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* breaking: make function name by default

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* feat(grammar): dynamically generate grammars with mutating keys

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* refactor: simplify condition

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

* Update docs

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>

---------

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2024-07-18 17:52:22 +02:00 committed by GitHub
parent 35d55572ac
commit bf9dd1de7f
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
9 changed files with 279 additions and 202 deletions

View file

@ -16,7 +16,7 @@ var _ = Describe("LocalAI function parse tests", func() {
Context("when using grammars and single result expected", func() {
It("should parse the function name and arguments correctly", func() {
input := `{"function": "add", "arguments": {"x": 5, "y": 3}}`
input := `{"name": "add", "arguments": {"x": 5, "y": 3}}`
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(1))
@ -28,13 +28,22 @@ var _ = Describe("LocalAI function parse tests", func() {
Context("when not using grammars and regex is needed", func() {
It("should extract function name and arguments from the regex", func() {
input := `add({"x":5,"y":3})`
functionConfig.ResponseRegex = []string{`(?P<function>\w+)\s*\((?P<arguments>.*)\)`}
functionConfig.ResponseRegex = []string{`(?P<name>\w+)\s*\((?P<arguments>.*)\)`}
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(1))
Expect(results[0].Name).To(Equal("add"))
Expect(results[0].Arguments).To(Equal(`{"x":5,"y":3}`))
})
It("should extract function name and arguments from the regex", func() {
input := `add({"x":5,"y":3})`
functionConfig.ResponseRegex = []string{`(?P<function>\w+)\s*\((?P<arguments>.*)\)`}
functionConfig.FunctionNameKey = "function"
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(1))
Expect(results[0].Name).To(Equal("add"))
Expect(results[0].Arguments).To(Equal(`{"x":5,"y":3}`))
})
})
Context("when having invalid input", func() {
@ -53,7 +62,7 @@ var _ = Describe("LocalAI function parse tests", func() {
Context("when parallel calls are enabled", func() {
It("should handle multiple function calls", func() {
input := `[{"function": "add", "arguments": {"x": 5, "y": 3}}, {"function": "subtract", "arguments": {"x": 10, "y": 7}}]`
input := `[{"name": "add", "arguments": {"x": 5, "y": 3}}, {"name": "subtract", "arguments": {"x": 10, "y": 7}}]`
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(2))
@ -66,8 +75,8 @@ var _ = Describe("LocalAI function parse tests", func() {
Context("without grammars and without regex", func() {
It("should parse the function name and arguments correctly with the name key", func() {
input := `{"name": "add", "arguments": {"x": 5, "y": 3}}`
functionConfig.FunctionName = true
input := `{"function": "add", "arguments": {"x": 5, "y": 3}}`
functionConfig.FunctionNameKey = "function"
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(1))
@ -76,7 +85,7 @@ var _ = Describe("LocalAI function parse tests", func() {
})
It("should parse the function name and arguments correctly with the function key", func() {
input := `{"function": "add", "arguments": {"x": 5, "y": 3}}`
input := `{"name": "add", "arguments": {"x": 5, "y": 3}}`
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(1))
@ -87,7 +96,7 @@ var _ = Describe("LocalAI function parse tests", func() {
It("should parse the result by matching the JSONRegexMatch", func() {
input := `
<tool_call>
{"function": "add", "arguments": {"x": 5, "y": 3}}
{"name": "add", "arguments": {"x": 5, "y": 3}}
</tool_call>`
functionConfig.JSONRegexMatch = []string{`(?s)<tool_call>(.*?)</tool_call>`}
@ -100,7 +109,7 @@ var _ = Describe("LocalAI function parse tests", func() {
It("should parse the result by matching the JSONRegexMatch", func() {
input := `
{"function": "add", "arguments": {"x": 5, "y": 3}}
{"name": "add", "arguments": {"x": 5, "y": 3}}
</tool_call>`
functionConfig.JSONRegexMatch = []string{`(?s)(.*?)</tool_call>`}
@ -110,13 +119,21 @@ var _ = Describe("LocalAI function parse tests", func() {
Expect(results[0].Name).To(Equal("add"))
Expect(results[0].Arguments).To(Equal(`{"x":5,"y":3}`))
})
It("should parse the result even with invalid JSON", func() {
input := `{"name": "add", "arguments": {"x": 5, "y": 3}} invalid {"name": "add", "arguments": {"x": 5, "y": 3}}`
results := ParseFunctionCall(input, functionConfig)
Expect(results).To(HaveLen(2))
Expect(results[0].Name).To(Equal("add"))
Expect(results[0].Arguments).To(Equal(`{"x":5,"y":3}`))
})
})
Context("when using ReplaceResults to clean up input", func() {
It("should replace text before and after JSON blob", func() {
input := `
Some text before the JSON
{"function": "add", "arguments": {"x": 5, "y": 3}}
{"name": "add", "arguments": {"x": 5, "y": 3}}
Some text after the JSON
`
@ -134,7 +151,7 @@ Some text after the JSON
It("should replace text before and after array JSON blob", func() {
input := `
Some text before the JSON
[{"function": "add", "arguments": {"x": 5, "y": 3}}, {"function": "subtract", "arguments": {"x": 10, "y": 7}}]
[{"name": "add", "arguments": {"x": 5, "y": 3}}, {"name": "subtract", "arguments": {"x": 10, "y": 7}}]
Some text after the JSON
`
functionConfig.ReplaceFunctionResults = []ReplaceResult{
@ -153,7 +170,7 @@ Some text after the JSON
It("should convert single-quoted key-value pairs to double-quoted and escape double quotes within values", func() {
input := `
Some text before the JSON
{'function': '"add"', 'arguments': {'x': 5, 'z': '"v"', 'y': 'v"value"'}}
{'name': '"add"', 'arguments': {'x': 5, 'z': '"v"', 'y': 'v"value"'}}
Some text after the JSON
`
functionConfig.JSONRegexMatch = []string{`(?s)<tool_call>(.*?)</tool_call>`}
@ -186,7 +203,7 @@ Some text after the JSON
It("should convert single-quoted key-value pairs to double-quoted and escape double quotes within values", func() {
input := `
Some text before the JSON
<tool_call>{'function': '"add"', 'arguments': {'x': 5, 'z': '"v"', 'y': 'v"value"'}}</tool_call>
<tool_call>{'name': '"add"', 'arguments': {'x': 5, 'z': '"v"', 'y': 'v"value"'}}</tool_call>
Some text after the JSON
`
functionConfig.JSONRegexMatch = []string{`(?s)<tool_call>(.*?)</tool_call>`}
@ -219,8 +236,8 @@ Some text after the JSON
It("should detect multiple functions call where the JSONRegexMatch is repeated", func() {
input := `
Some text before the JSON
<tool_call>{"function": "add", "arguments": {"x": 5, "y": 3}}</tool_call>
<tool_call>{"function": "subtract", "arguments": {"x": 10, "y": 7}}</tool_call>
<tool_call>{"name": "add", "arguments": {"x": 5, "y": 3}}</tool_call>
<tool_call>{"name": "subtract", "arguments": {"x": 10, "y": 7}}</tool_call>
Some text after the JSON
`
functionConfig.JSONRegexMatch = []string{`(?s)<tool_call>(.*?)</tool_call>`}
@ -240,7 +257,7 @@ Some text after the JSON
<sketchpad>
roses are red
</sketchpad>
<tool_call>{"function": "subtract", "arguments": {"x": 10, "y": 7}}</tool_call>
<tool_call>{"name": "subtract", "arguments": {"x": 10, "y": 7}}</tool_call>
Some text after the JSON
`
functionConfig.CaptureLLMResult = []string{`(?s)<sketchpad>(.*?)</sketchpad>`}
@ -251,7 +268,7 @@ roses are red
It("Defaults to empty if doesn't catch any", func() {
input := `
Some text before the JSON
<tool_call>{"function": "subtract", "arguments": {"x": 10, "y": 7}}</tool_call>
<tool_call>{"name": "subtract", "arguments": {"x": 10, "y": 7}}</tool_call>
Some text after the JSON
`
functionConfig.CaptureLLMResult = []string{`(?s)<sketchpad>(.*?)</sketchpad>`}
@ -259,4 +276,74 @@ roses are red
Expect(results).To(Equal(""))
})
})
Context("ParseJSON - when given valid JSON strings", func() {
It("should parse multiple JSON objects", func() {
input := `{"key1": "value1"} {"key2": "value2"}`
expected := []map[string]any{
{"key1": "value1"},
{"key2": "value2"},
}
result, err := ParseJSON(input)
Expect(err).NotTo(HaveOccurred())
Expect(result).To(Equal(expected))
})
It("should parse a single JSON object with various types", func() {
input := `{"key1": "value1", "key2": 2}`
expected := []map[string]any{
{"key1": "value1", "key2": float64(2)},
}
result, err := ParseJSON(input)
Expect(err).NotTo(HaveOccurred())
Expect(result).To(Equal(expected))
})
It("should handle JSON without syntax errors gracefully", func() {
input := `{"key1": "value1"}`
expected := []map[string]any{
{"key1": "value1"},
}
result, err := ParseJSON(input)
Expect(err).NotTo(HaveOccurred())
Expect(result).To(Equal(expected))
})
It("should handle JSON without syntax errors gracefully", func() {
input := `[{"key1": "value1"}]`
expected := []map[string]any{
{"key1": "value1"},
}
result, err := ParseJSON(input)
Expect(err).NotTo(HaveOccurred())
Expect(result).To(Equal(expected))
})
})
Context("ParseJSON - when given invalid JSON strings", func() {
It("should return an error for completely invalid JSON", func() {
input := `invalid json`
result, err := ParseJSON(input)
Expect(err).To(HaveOccurred())
Expect(result).To(BeNil())
})
It("should skip invalid JSON parts and parse valid parts", func() {
input := `{"key1": "value1"} invalid {"key2": "value2"}`
expected := []map[string]any{
{"key1": "value1"},
{"key2": "value2"},
}
result, err := ParseJSON(input)
Expect(err).NotTo(HaveOccurred())
Expect(result).To(Equal(expected))
})
PIt("should handle JSON with syntax errors gracefully", func() {
input := `{"key1": "value1", "key2": }`
expected := []map[string]any{
{"key1": "value1"},
}
result, err := ParseJSON(input)
Expect(err).NotTo(HaveOccurred())
Expect(result).To(Equal(expected))
})
})
})