fix: o1 model (#1992)

lua/avante/config.lua

@@ -612,13 +612,22 @@ function M.has_provider(provider_name) return vim.list_contains(M.provider_names
 ---@param provider_name avante.ProviderName
 function M.get_provider_config(provider_name)
   if not M.has_provider(provider_name) then error("No provider found: " .. provider_name, 2) end
-  if M._options[provider_name] ~= nil then
-    return vim.deepcopy(M._options[provider_name], true)
-  elseif M.vendors and M.vendors[provider_name] ~= nil then
-    return vim.deepcopy(M.vendors[provider_name], true)
-  else
-    error("Failed to find provider: " .. provider_name, 2)
+  local found = false
+  local config = {}
+
+  if M.vendors and M.vendors[provider_name] ~= nil then
+    found = true
+    config = vim.tbl_deep_extend("force", config, vim.deepcopy(M.vendors[provider_name], true))
   end
+
+  if M._options[provider_name] ~= nil then
+    found = true
+    config = vim.tbl_deep_extend("force", config, vim.deepcopy(M._options[provider_name], true))
+  end
+
+  if not found then error("Failed to find provider: " .. provider_name, 2) end
+
+  return config
 end
 
 M.BASE_PROVIDER_KEYS = {

lua/avante/diff.lua

@@ -272,6 +272,7 @@ local function set_cursor(position, side)
   if not position then return end
   local target = side == SIDES.OURS and position.current or position.incoming
   api.nvim_win_set_cursor(0, { target.range_start + 1, 0 })
+  vim.cmd("normal! zz")
 end
 
 local show_keybinding_hint_extmark_id = nil

lua/avante/llm.lua

@@ -335,9 +335,10 @@ function M.generate_prompts(opts)
     remaining_tokens = remaining_tokens - Utils.tokens.calculate_tokens(message.content)
   end
 
-  local dropped_history_messages = {}
-  if opts.prompt_opts and opts.prompt_opts.dropped_history_messages then
-    dropped_history_messages = vim.list_extend(dropped_history_messages, opts.prompt_opts.dropped_history_messages)
+  local pending_compaction_history_messages = {}
+  if opts.prompt_opts and opts.prompt_opts.pending_compaction_history_messages then
+    pending_compaction_history_messages =
+      vim.list_extend(pending_compaction_history_messages, opts.prompt_opts.pending_compaction_history_messages)
   end
 
   local cleaned_history_messages = history_messages
@@ -350,6 +351,7 @@ function M.generate_prompts(opts)
       if Config.history.max_tokens > 0 then
         remaining_tokens = math.min(Config.history.max_tokens, remaining_tokens)
       end
+
       -- Traverse the history in reverse, keeping only the latest history until the remaining tokens are exhausted and the first message role is "user"
       local retained_history_messages = {}
       for i = #cleaned_history_messages, 1, -1 do
@@ -368,11 +370,11 @@ function M.generate_prompts(opts)
           vim.list_slice(cleaned_history_messages, #cleaned_history_messages - 1, #cleaned_history_messages)
       end
 
-      dropped_history_messages =
+      pending_compaction_history_messages =
         vim.list_slice(cleaned_history_messages, 1, #cleaned_history_messages - #retained_history_messages)
 
-      dropped_history_messages = vim
-        .iter(dropped_history_messages)
+      pending_compaction_history_messages = vim
+        .iter(pending_compaction_history_messages)
         :filter(function(msg) return msg.is_dummy ~= true end)
         :totable()
 
@@ -411,7 +413,7 @@ function M.generate_prompts(opts)
     messages = messages,
     image_paths = image_paths,
     tools = tools,
-    dropped_history_messages = dropped_history_messages,
+    pending_compaction_history_messages = pending_compaction_history_messages,
   }
 end
 
@@ -464,13 +466,18 @@ function M.curl(opts)
     end
     local data_match = line:match("^data:%s*(.+)$")
     if data_match then
+      response_body = ""
       provider:parse_response(resp_ctx, data_match, current_event_state, handler_opts)
     else
       response_body = response_body .. line
       local ok, jsn = pcall(vim.json.decode, response_body)
       if ok then
+        if jsn.error then
+          handler_opts.on_stop({ reason = "error", error = jsn.error })
+        else
+          provider:parse_response(resp_ctx, response_body, current_event_state, handler_opts)
+        end
         response_body = ""
-        if jsn.error then handler_opts.on_stop({ reason = "error", error = jsn.error }) end
       end
     end
   end
@@ -485,16 +492,13 @@ function M.curl(opts)
 
   local temp_file = fn.tempname()
   local curl_body_file = temp_file .. "-request-body.json"
-  local resp_body_file = temp_file .. "-response-body.json"
+  local resp_body_file = temp_file .. "-response-body.txt"
+  local headers_file = temp_file .. "-response-headers.txt"
   local json_content = vim.json.encode(spec.body)
   fn.writefile(vim.split(json_content, "\n"), curl_body_file)
 
   Utils.debug("curl request body file:", curl_body_file)
-
   Utils.debug("curl response body file:", resp_body_file)
-
-  local headers_file = temp_file .. "-headers.txt"
-
   Utils.debug("curl headers file:", headers_file)
 
   local function cleanup()
@@ -609,7 +613,6 @@ function M.curl(opts)
               end
             end
           end
-          Utils.debug("result", result)
           local retry_after = 10
           if headers_map["retry-after"] then retry_after = tonumber(headers_map["retry-after"]) or 10 end
           handler_opts.on_stop({ reason = "rate_limit", retry_after = retry_after })
@@ -698,11 +701,11 @@ function M._stream(opts)
   local prompt_opts = M.generate_prompts(opts)
 
   if
-    prompt_opts.dropped_history_messages
-    and #prompt_opts.dropped_history_messages > 0
+    prompt_opts.pending_compaction_history_messages
+    and #prompt_opts.pending_compaction_history_messages > 0
     and opts.on_memory_summarize
   then
-    opts.on_memory_summarize(prompt_opts.dropped_history_messages)
+    opts.on_memory_summarize(prompt_opts.pending_compaction_history_messages)
     return
   end
 

lua/avante/llm_tools/dispatch_agent.lua

@@ -166,12 +166,16 @@ When you're done, provide a clear and concise summary of what you found.]]):gsub
     end,
   }
 
-  local function on_memory_summarize(dropped_history_messages)
-    Llm.summarize_memory(memory_content, dropped_history_messages or {}, function(memory)
+  local function on_memory_summarize(pending_compaction_history_messages)
+    Llm.summarize_memory(memory_content, pending_compaction_history_messages or {}, function(memory)
       if memory then stream_options.memory = memory.content end
       local new_history_messages = {}
       for _, msg in ipairs(history_messages) do
-        if vim.iter(dropped_history_messages):find(function(dropped_msg) return dropped_msg.uuid == msg.uuid end) then
+        if
+          vim
+            .iter(pending_compaction_history_messages)
+            :find(function(pending_compaction_msg) return pending_compaction_msg.uuid == msg.uuid end)
+        then
           goto continue
         end
         table.insert(new_history_messages, msg)

lua/avante/providers/openai.lua

@@ -245,30 +245,37 @@ function M:parse_response(ctx, data_stream, _, opts)
     opts.on_stop({ reason = "complete" })
     return
   end
-  if not data_stream:match('"delta":') then return end
-  ---@type AvanteOpenAIChatResponse
+  if data_stream == "[DONE]" then return end
   local jsn = vim.json.decode(data_stream)
-  if not jsn.choices or not jsn.choices[1] then return end
+  ---@cast jsn AvanteOpenAIChatResponse
+  if not jsn.choices then return end
   local choice = jsn.choices[1]
-  if choice.delta.reasoning_content and choice.delta.reasoning_content ~= vim.NIL then
+  if not choice then return end
+  local delta = choice.delta
+  if not delta then
+    local provider_conf = Providers.parse_config(self)
+    if provider_conf.model:match("o1") then delta = choice.message end
+  end
+  if not delta then return end
+  if delta.reasoning_content and delta.reasoning_content ~= vim.NIL then
     if ctx.returned_think_start_tag == nil or not ctx.returned_think_start_tag then
       ctx.returned_think_start_tag = true
       if opts.on_chunk then opts.on_chunk("<think>\n") end
     end
-    ctx.last_think_content = choice.delta.reasoning_content
-    self:add_thinking_message(ctx, choice.delta.reasoning_content, "generating", opts)
-    if opts.on_chunk then opts.on_chunk(choice.delta.reasoning_content) end
-  elseif choice.delta.reasoning and choice.delta.reasoning ~= vim.NIL then
+    ctx.last_think_content = delta.reasoning_content
+    self:add_thinking_message(ctx, delta.reasoning_content, "generating", opts)
+    if opts.on_chunk then opts.on_chunk(delta.reasoning_content) end
+  elseif delta.reasoning and delta.reasoning ~= vim.NIL then
     if ctx.returned_think_start_tag == nil or not ctx.returned_think_start_tag then
       ctx.returned_think_start_tag = true
       if opts.on_chunk then opts.on_chunk("<think>\n") end
     end
-    ctx.last_think_content = choice.delta.reasoning
-    self:add_thinking_message(ctx, choice.delta.reasoning, "generating", opts)
-    if opts.on_chunk then opts.on_chunk(choice.delta.reasoning) end
-  elseif choice.delta.tool_calls and choice.delta.tool_calls ~= vim.NIL then
+    ctx.last_think_content = delta.reasoning
+    self:add_thinking_message(ctx, delta.reasoning, "generating", opts)
+    if opts.on_chunk then opts.on_chunk(delta.reasoning) end
+  elseif delta.tool_calls and delta.tool_calls ~= vim.NIL then
     local choice_index = choice.index or 0
-    for idx, tool_call in ipairs(choice.delta.tool_calls) do
+    for idx, tool_call in ipairs(delta.tool_calls) do
       --- In Gemini's so-called OpenAI Compatible API, tool_call.index is nil, which is quite absurd! Therefore, a compatibility fix is needed here.
       if tool_call.index == nil then tool_call.index = choice_index + idx - 1 end
       if not ctx.tool_use_list then ctx.tool_use_list = {} end
@@ -290,7 +297,7 @@ function M:parse_response(ctx, data_stream, _, opts)
         self:add_tool_use_message(tool_use, "generating", opts)
       end
     end
-  elseif choice.delta.content then
+  elseif delta.content then
     if
       ctx.returned_think_start_tag ~= nil and (ctx.returned_think_end_tag == nil or not ctx.returned_think_end_tag)
     then
@@ -304,9 +311,9 @@ function M:parse_response(ctx, data_stream, _, opts)
       end
       self:add_thinking_message(ctx, "", "generated", opts)
     end
-    if choice.delta.content ~= vim.NIL then
-      if opts.on_chunk then opts.on_chunk(choice.delta.content) end
-      self:add_text_message(ctx, choice.delta.content, "generating", opts)
+    if delta.content ~= vim.NIL then
+      if opts.on_chunk then opts.on_chunk(delta.content) end
+      self:add_text_message(ctx, delta.content, "generating", opts)
     end
   end
   if choice.finish_reason == "stop" or choice.finish_reason == "eos_token" then

lua/avante/sidebar.lua

@@ -694,6 +694,11 @@ local function insert_conflict_contents(bufnr, snippets)
   for _, snippet in ipairs(snippets) do
     local start_line, end_line = unpack(snippet.range)
 
+    local first_line_content = lines[start_line]
+    local old_first_line_indentation = ""
+
+    if first_line_content then old_first_line_indentation = Utils.get_indentation(first_line_content) end
+
     local result = {}
     table.insert(result, "<<<<<<< HEAD")
     for i = start_line, end_line do
@@ -703,6 +708,14 @@ local function insert_conflict_contents(bufnr, snippets)
 
     local snippet_lines = vim.split(snippet.content, "\n")
 
+    if #snippet_lines > 0 then
+      local new_first_line_indentation = Utils.get_indentation(snippet_lines[1])
+      if #old_first_line_indentation > #new_first_line_indentation then
+        local line_indentation = old_first_line_indentation:sub(#new_first_line_indentation + 1)
+        snippet_lines = vim.iter(snippet_lines):map(function(line) return line_indentation .. line end):totable()
+      end
+    end
+
     vim.list_extend(result, snippet_lines)
 
     table.insert(result, ">>>>>>> Snippet")
@@ -1892,10 +1905,19 @@ function Sidebar:add_history_messages(messages)
   end
   self.chat_history.messages = history_messages
   Path.history.save(self.code.bufnr, self.chat_history)
-  if self.chat_history.title == "untitled" and #messages > 0 then
+  if
+    self.chat_history.title == "untitled"
+    and #messages > 0
+    and messages[1].just_for_display ~= true
+    and messages[1].state == "generated"
+  then
+    self.chat_history.title = "generating..."
     Llm.summarize_chat_thread_title(messages[1].message.content, function(title)
-      self:reload_chat_history()
-      if title then self.chat_history.title = title end
+      if title then
+        self.chat_history.title = title
+      else
+        self.chat_history.title = "untitled"
+      end
       Path.history.save(self.code.bufnr, self.chat_history)
     end)
   end
@@ -2371,14 +2393,7 @@ function Sidebar:create_input_container()
         if Config.behaviour.auto_apply_diff_after_generation then self:apply(false) end
       end, 0)
 
-      if self.chat_history.title == "untitled" then
-        Llm.summarize_chat_thread_title(request, function(title)
-          if title then self.chat_history.title = title end
-          Path.history.save(self.code.bufnr, self.chat_history)
-        end)
-      else
-        Path.history.save(self.code.bufnr, self.chat_history)
-      end
+      Path.history.save(self.code.bufnr, self.chat_history)
     end
 
     if request and request ~= "" then
@@ -2407,18 +2422,22 @@ function Sidebar:create_input_container()
         session_ctx = {},
       })
 
-      ---@param dropped_history_messages avante.HistoryMessage[]
-      local function on_memory_summarize(dropped_history_messages)
+      ---@param pending_compaction_history_messages avante.HistoryMessage[]
+      local function on_memory_summarize(pending_compaction_history_messages)
         local history_memory = self.chat_history.memory
-        Llm.summarize_memory(history_memory and history_memory.content, dropped_history_messages, function(memory)
-          if memory then
-            self.chat_history.memory = memory
-            Path.history.save(self.code.bufnr, self.chat_history)
-            stream_options.memory = memory.content
+        Llm.summarize_memory(
+          history_memory and history_memory.content,
+          pending_compaction_history_messages,
+          function(memory)
+            if memory then
+              self.chat_history.memory = memory
+              Path.history.save(self.code.bufnr, self.chat_history)
+              stream_options.memory = memory.content
+            end
+            stream_options.history_messages = self:get_history_messages_for_api()
+            Llm.stream(stream_options)
           end
-          stream_options.history_messages = self:get_history_messages_for_api()
-          Llm.stream(stream_options)
-        end)
+        )
       end
 
       stream_options.on_memory_summarize = on_memory_summarize

lua/avante/types.lua

@@ -114,7 +114,7 @@ vim.g.avante_login = vim.g.avante_login
 ---@field messages AvanteLLMMessage[]
 ---@field image_paths? string[]
 ---@field tools? AvanteLLMTool[]
----@field dropped_history_messages? avante.HistoryMessage[]
+---@field pending_compaction_history_messages? AvanteLLMMessage[]
 ---
 ---@class AvanteGeminiMessage
 ---@field role "user"
@@ -356,7 +356,7 @@ vim.g.avante_login = vim.g.avante_login
 ---@field tool_result? AvanteLLMToolResult
 ---@field tool_use? AvanteLLMToolUse
 ---
----@alias AvanteLLMMemorySummarizeCallback fun(dropped_history_messages: avante.HistoryMessage[]): nil
+---@alias AvanteLLMMemorySummarizeCallback fun(pending_compaction_history_messages: avante.HistoryMessage[]): nil
 ---
 ---@alias AvanteLLMToolUseState "generating" | "generated" | "running" | "succeeded" | "failed"
 ---@alias avante.GenerateState "generating" | "tool calling" | "failed" | "succeeded" | "cancelled" | "searching" | "thinking"