适配多个 LLM 源 (anthropic/gemini/mistral/groq/github) + SQLite 配置收敛 + 测试插件

internal/lua/adapters/anthropic.lua

@@ -0,0 +1,82 @@
+local adapter = {}
+
+adapter.name = "anthropic"
+adapter.version = "2.0.0"
+adapter.endpoint = "/v1/messages"
+adapter.headers = {
+    ["anthropic-version"] = "2023-06-01"
+}
+
+-- Anthropic Messages API: { model, messages[], max_tokens, system, stream }
+function adapter.transform_request(raw_body)
+    local ok, req = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local msgs = {}
+    local system = ""
+    for _, m in ipairs(req.messages or {}) do
+        if m.role == "system" then
+            system = system .. m.content .. "\n"
+        else
+            table.insert(msgs, { role = m.role, content = m.content })
+        end
+    end
+
+    local anthropic_req = {
+        model = req.model or "claude-sonnet-4-20250514",
+        max_tokens = req.max_tokens or 4096,
+        messages = msgs,
+        stream = req.stream or false,
+    }
+    if system ~= "" then
+        anthropic_req.system = system
+    end
+
+    return json.encode(anthropic_req)
+end
+
+function adapter.transform_response(raw_body)
+    local ok, resp = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local unified = {
+        content = "",
+        finish_reason = "",
+        token_usage = { prompt = 0, completion = 0, total = 0 }
+    }
+
+    if resp.usage then
+        unified.token_usage.prompt = resp.usage.input_tokens or 0
+        unified.token_usage.completion = resp.usage.output_tokens or 0
+        unified.token_usage.total = (resp.usage.input_tokens or 0) + (resp.usage.output_tokens or 0)
+    end
+
+    if resp.content and #resp.content > 0 then
+        for _, block in ipairs(resp.content) do
+            if block.type == "text" then
+                unified.content = unified.content .. (block.text or "")
+            end
+        end
+    end
+    unified.finish_reason = resp.stop_reason or ""
+
+    return json.encode(unified)
+end
+
+function adapter.transform_stream_chunk(raw_chunk)
+    local ok, chunk = pcall(json.decode, raw_chunk)
+    if not ok then return "" end
+    if chunk.type == "message_start" then return "" end
+    if chunk.type == "message_delta" then
+        return json.encode({ content = "", done = (chunk.delta and chunk.delta.stop_reason ~= nil) })
+    end
+    if chunk.type == "content_block_delta" and chunk.delta then
+        return json.encode({ content = chunk.delta.text or "", done = false })
+    end
+    if chunk.type == "message_stop" then
+        return json.encode({ content = "", done = true })
+    end
+    return ""
+end
+
+return adapter

internal/lua/adapters/gemini.lua

@@ -0,0 +1,89 @@
+local adapter = {}
+
+adapter.name = "gemini"
+adapter.version = "2.0.0"
+adapter.endpoint = "/v1/models"
+adapter.headers = {}
+
+-- Gemini API: POST /v1/models/{model}:generateContent
+-- Auth: API key in query param ?key=XXX or Authorization: Bearer XXX
+function adapter.transform_request(raw_body)
+    local ok, req = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local contents = {}
+    for _, m in ipairs(req.messages or {}) do
+        table.insert(contents, {
+            role = (m.role == "assistant") and "model" or m.role,
+            parts = { { text = m.content } }
+        })
+    end
+
+    local gemini_req = {
+        contents = contents,
+        generationConfig = {
+            temperature = req.temperature or 0.7,
+            maxOutputTokens = req.max_tokens or 4096,
+        }
+    }
+
+    if req.stream then
+        gemini_req.stream = true
+    end
+
+    return json.encode(gemini_req)
+end
+
+-- Gemini 的 endpoint 动态拼接：/v1/models/{model}:generateContent
+function adapter.transform_response(raw_body)
+    local ok, resp = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local unified = {
+        content = "",
+        finish_reason = "",
+        token_usage = { prompt = 0, completion = 0, total = 0 }
+    }
+
+    if resp.usageMetadata then
+        unified.token_usage.prompt = resp.usageMetadata.promptTokenCount or 0
+        unified.token_usage.completion = resp.usageMetadata.candidatesTokenCount or 0
+        unified.token_usage.total = resp.usageMetadata.totalTokenCount or 0
+    end
+
+    if resp.candidates and #resp.candidates > 0 then
+        local cand = resp.candidates[1]
+        if cand.content and cand.content.parts then
+            for _, part in ipairs(cand.content.parts) do
+                if part.text then
+                    unified.content = unified.content .. part.text
+                end
+            end
+        end
+        if cand.finishReason then
+            unified.finish_reason = cand.finishReason
+        end
+    end
+
+    return json.encode(unified)
+end
+
+function adapter.transform_stream_chunk(raw_chunk)
+    local ok, chunk = pcall(json.decode, raw_chunk)
+    if not ok then return "" end
+
+    if not chunk.candidates or #chunk.candidates == 0 then return "" end
+    local cand = chunk.candidates[1]
+    local content = ""
+    if cand.content and cand.content.parts then
+        for _, part in ipairs(cand.content.parts) do
+            content = content .. (part.text or "")
+        end
+    end
+    return json.encode({
+        content = content,
+        done = (cand.finishReason ~= nil)
+    })
+end
+
+return adapter

internal/lua/adapters/github.lua

@@ -0,0 +1,73 @@
+local adapter = {}
+
+adapter.name = "github"
+adapter.version = "2.0.0"
+adapter.endpoint = "/chat/completions"
+adapter.headers = {}
+
+-- GitHub Models: Azure-like endpoint, auth via Bearer token (PAT)
+-- BaseURL example: https://models.inference.ai.azure.com
+function adapter.transform_request(raw_body)
+    local ok, req = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+    req.model = req.model or "gpt-4o"
+    req.temperature = req.temperature or 0.7
+    req.max_tokens = req.max_tokens or 4096
+    req.stream = req.stream or false
+    return json.encode(req)
+end
+
+function adapter.transform_response(raw_body)
+    local ok, resp = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local unified = {
+        content = "",
+        finish_reason = "",
+        token_usage = { prompt = 0, completion = 0, total = 0 }
+    }
+
+    if resp.usage then
+        unified.token_usage.prompt = resp.usage.prompt_tokens or 0
+        unified.token_usage.completion = resp.usage.completion_tokens or 0
+        unified.token_usage.total = resp.usage.total_tokens or 0
+    end
+
+    if resp.choices and #resp.choices > 0 then
+        local ch = resp.choices[1]
+        if ch.message then
+            unified.content = ch.message.content or ""
+            if ch.message.tool_calls then
+                local tcs = {}
+                for _, tc in ipairs(ch.message.tool_calls) do
+                    local args_ok, args = pcall(json.decode, tc["function"].arguments)
+                    if not args_ok then args = {} end
+                    table.insert(tcs, {
+                        id = tc.id,
+                        type = tc.type or "function",
+                        name = tc["function"].name,
+                        arguments = args
+                    })
+                end
+                unified.tool_calls = tcs
+            end
+        end
+        unified.finish_reason = ch.finish_reason or ""
+    end
+
+    return json.encode(unified)
+end
+
+function adapter.transform_stream_chunk(raw_chunk)
+    local ok, chunk = pcall(json.decode, raw_chunk)
+    if not ok then return "" end
+    if not chunk.choices or #chunk.choices == 0 then return "" end
+    local delta = chunk.choices[1].delta or {}
+    local fr = chunk.choices[1].finish_reason
+    return json.encode({
+        content = delta.content or "",
+        done = (fr ~= nil)
+    })
+end
+
+return adapter

internal/lua/adapters/groq.lua

@@ -0,0 +1,72 @@
+local adapter = {}
+
+adapter.name = "groq"
+adapter.version = "2.0.0"
+adapter.endpoint = "/openai/v1/chat/completions"
+adapter.headers = {}
+
+-- Groq API is OpenAI-compatible
+function adapter.transform_request(raw_body)
+    local ok, req = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+    req.model = req.model or "llama3-70b-8192"
+    req.temperature = req.temperature or 0.7
+    req.max_tokens = req.max_tokens or 4096
+    req.stream = req.stream or false
+    return json.encode(req)
+end
+
+function adapter.transform_response(raw_body)
+    local ok, resp = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local unified = {
+        content = "",
+        finish_reason = "",
+        token_usage = { prompt = 0, completion = 0, total = 0 }
+    }
+
+    if resp.usage then
+        unified.token_usage.prompt = resp.usage.prompt_tokens or 0
+        unified.token_usage.completion = resp.usage.completion_tokens or 0
+        unified.token_usage.total = resp.usage.total_tokens or 0
+    end
+
+    if resp.choices and #resp.choices > 0 then
+        local ch = resp.choices[1]
+        if ch.message then
+            unified.content = ch.message.content or ""
+            if ch.message.tool_calls then
+                local tcs = {}
+                for _, tc in ipairs(ch.message.tool_calls) do
+                    local args_ok, args = pcall(json.decode, tc["function"].arguments)
+                    if not args_ok then args = {} end
+                    table.insert(tcs, {
+                        id = tc.id,
+                        type = tc.type or "function",
+                        name = tc["function"].name,
+                        arguments = args
+                    })
+                end
+                unified.tool_calls = tcs
+            end
+        end
+        unified.finish_reason = ch.finish_reason or ""
+    end
+
+    return json.encode(unified)
+end
+
+function adapter.transform_stream_chunk(raw_chunk)
+    local ok, chunk = pcall(json.decode, raw_chunk)
+    if not ok then return "" end
+    if not chunk.choices or #chunk.choices == 0 then return "" end
+    local delta = chunk.choices[1].delta or {}
+    local fr = chunk.choices[1].finish_reason
+    return json.encode({
+        content = delta.content or "",
+        done = (fr ~= nil)
+    })
+end
+
+return adapter

internal/lua/adapters/mistral.lua

@@ -0,0 +1,72 @@
+local adapter = {}
+
+adapter.name = "mistral"
+adapter.version = "2.0.0"
+adapter.endpoint = "/v1/chat/completions"
+adapter.headers = {}
+
+-- Mistral API is OpenAI-compatible, just passes through
+function adapter.transform_request(raw_body)
+    local ok, req = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+    req.model = req.model or "mistral-large-latest"
+    req.temperature = req.temperature or 0.7
+    req.max_tokens = req.max_tokens or 4096
+    req.stream = req.stream or false
+    return json.encode(req)
+end
+
+function adapter.transform_response(raw_body)
+    local ok, resp = pcall(json.decode, raw_body)
+    if not ok then return raw_body end
+
+    local unified = {
+        content = "",
+        finish_reason = "",
+        token_usage = { prompt = 0, completion = 0, total = 0 }
+    }
+
+    if resp.usage then
+        unified.token_usage.prompt = resp.usage.prompt_tokens or 0
+        unified.token_usage.completion = resp.usage.completion_tokens or 0
+        unified.token_usage.total = resp.usage.total_tokens or 0
+    end
+
+    if resp.choices and #resp.choices > 0 then
+        local ch = resp.choices[1]
+        if ch.message then
+            unified.content = ch.message.content or ""
+            if ch.message.tool_calls then
+                local tcs = {}
+                for _, tc in ipairs(ch.message.tool_calls) do
+                    local args_ok, args = pcall(json.decode, tc["function"].arguments)
+                    if not args_ok then args = {} end
+                    table.insert(tcs, {
+                        id = tc.id,
+                        type = tc.type or "function",
+                        name = tc["function"].name,
+                        arguments = args
+                    })
+                end
+                unified.tool_calls = tcs
+            end
+        end
+        unified.finish_reason = ch.finish_reason or ""
+    end
+
+    return json.encode(unified)
+end
+
+function adapter.transform_stream_chunk(raw_chunk)
+    local ok, chunk = pcall(json.decode, raw_chunk)
+    if not ok then return "" end
+    if not chunk.choices or #chunk.choices == 0 then return "" end
+    local delta = chunk.choices[1].delta or {}
+    local fr = chunk.choices[1].finish_reason
+    return json.encode({
+        content = delta.content or "",
+        done = (fr ~= nil)
+    })
+end
+
+return adapter