display timings

Author: heilgar

CHANGELOG.md

@@ -5,6 +5,19 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [1.3.0] 2025-04-03
+### Added
+* Request Profiling Capabilities
+  * Detailed timing metrics for HTTP requests
+  * DNS resolution, connection, TLS handshake, and transfer time tracking
+  * Configurable display in response window
+  * Toggle profiling with `:HttpProfiling` command
+  * New keybinding: `<leader>hp` to toggle profiling
+  * Configuration options:
+    * `profiling.enabled` - Enable/disable profiling (default: true)
+    * `profiling.show_in_response` - Show metrics in response (default: true)
+    * `profiling.detailed_metrics` - Show detailed breakdown (default: true)
+
 ## [1.2.2] 2025-26-01
 ### Fixed
 * Improved JSON response formatting

README.md

@@ -17,6 +17,7 @@ The core goal is to ensure compatibility with .http files from IntelliJ or VSCod
     - [Commands](#commands)
     - [Keybindings](#keybindings)
     - [Response handler](#response-handler)
+    - [Request Profiling](#request-profiling)
 - [Telescope Integration](#telescope-integration)
 - [Documentation](#documentation)
 - [Examples](#examples)
@@ -40,6 +41,7 @@ The core goal is to ensure compatibility with .http files from IntelliJ or VSCod
 - Syntax highlighting for .http files and response buffers
 - Verbose mode for debugging
 - Dry run capability for request inspection
+- Request profiling with detailed timing metrics
 - Telescope integration for environment selection
 - Compatible with [JetBrains HTTP Client](https://www.jetbrains.com/help/idea/http-client-in-product-code-editor.html) and [VSCode Restclient](https://github.com/Huachao/vscode-restclient)
 
@@ -76,13 +78,19 @@ local http_client = require("http_client")
 local config = {
   default_env_file = '.env.json',
   request_timeout = 30000, -- 30 seconds
+  profiling = {
+    enabled = true,
+    show_in_response = true,
+    detailed_metrics = true,
+  },
   keybindings = {
     select_env_file = "<leader>hf",
     set_env = "<leader>he",
     run_request = "<leader>hr",
     stop_request = "<leader>hs",
     dry_run = "<leader>hd",
-    toggle_verbose = "<leader>hv"
+    toggle_verbose = "<leader>hv",
+    toggle_profiling = "<leader>hp"
   },
 }
 
@@ -115,6 +123,7 @@ You can adjust these settings to your preferences.
 - `:HttpRunAll`: Run all HTTP requests in the current file.
 - `:HttpStop`: Stop the currently running HTTP request.
 - `:HttpVerbose`: Toggle verbose mode for debugging.
+- `:HttpProfiling`: Toggle request profiling.
 - `:HttpDryRun`: Perform a dry run of the request under the cursor.
 - `:HttpCopyCurl`: Copy the curl command for the HTTP request under the cursor.
 
@@ -136,6 +145,7 @@ The plugin comes with the following default keybindings:
 - `<leader>hr`: Run HTTP request under cursor
 - `<leader>hx`: Stop running HTTP request
 - `<leader>hv`: Toggle verbose mode
+- `<leader>hp`: Toggle request profiling
 - `<leader>hd`: Perform dry run
 - `<leader>hc`: Copy curl command for HTTP request under cursor
 
@@ -156,6 +166,7 @@ To customize these keybindings, you can add the following to your Neovim configu
         { "<leader>hr", "<cmd>HttpRun<cr>", desc = "Run HTTP request" },
         { "<leader>hx", "<cmd>HttpStop<cr>", desc = "Stop HTTP request" },
         { "<leader>hv", "<cmd>HttpVerbose<cr>", desc = "Toggle verbose mode" },
+        { "<leader>hp", "<cmd>HttpProfiling<cr>", desc = "Toggle request profiling" },
         { "<leader>hd", "<cmd>HttpDryRun<cr>", desc = "Perform dry run" },
     },
     cmd = {
@@ -164,6 +175,7 @@ To customize these keybindings, you can add the following to your Neovim configu
         "HttpRun",
         "HttpStop",
         "HttpVerbose",
+        "HttpProfiling",
         "HttpDryRun"
     },
 }
@@ -191,6 +203,38 @@ GET {{base_url}}/protected
 Authorization: Bearer {{auth_token}}
 ```
 
+### Request Profiling
+
+The plugin includes request profiling capabilities that show you detailed timing metrics for your HTTP requests. 
+
+#### Configuration
+
+You can configure profiling in your setup:
+
+```lua
+profiling = {
+  enabled = true,               -- Enable or disable profiling
+  show_in_response = true,      -- Show timing metrics in response output
+  detailed_metrics = true,      -- Show detailed breakdown of timings
+},
+```
+
+#### Viewing Timing Metrics
+
+When profiling is enabled, the response window will include a "Timing" section that shows:
+
+- Total request time
+- DNS resolution time
+- TCP connection time
+- TLS handshake time (for HTTPS requests)
+- Request sending time
+- Content transfer time
+
+#### Commands and Keybindings
+
+- `:HttpProfiling` - Toggle profiling on/off
+- `<leader>hp` - Default keybinding to toggle profiling
+
 ### Running Without Environment
 You can now run requests without selecting an environment file. If environment variables are needed but not set, the plugin will display a message suggesting to select an environment file or set properties via a response handler.
 

lua/http_client/config.lua

@@ -5,6 +5,11 @@ M.defaults = {
     request_timeout = 30000, -- 30 seconds
     split_direction = "right",
     create_keybindings = true,
+    profiling = {
+        enabled = true,
+        show_in_response = true,
+        detailed_metrics = true,
+    },
     keybindings = {
         select_env_file = "<leader>hf",
         set_env = "<leader>he",
@@ -14,6 +19,7 @@ M.defaults = {
         toggle_verbose = "<leader>hv",
         copy_curl = "<leader>hc",
         save_response = "<header>hs",
+        toggle_profiling = "<leader>hp",
     },
 }
 

lua/http_client/core/http_client.lua

@@ -2,6 +2,7 @@ local M = {}
 local curl = require('plenary.curl')
 local vvv = require('http_client.utils.verbose')
 local state = require('http_client.state')
+local profiling = require('http_client.utils.profiling')
 
 local current_request = nil
 
@@ -119,7 +120,9 @@ local function prepare_response(request, response)
             url = request.url,
             http_version = request.http_version or "N/A",
             test_name = request.test_name or "N/A",
-        }
+        },
+        request_id = request.request_id,
+        timing_metrics = response.timing_metrics or {}
     }
 
     state.store_response(pr)
@@ -129,6 +132,17 @@ end
 
 local function display_response(pr)
     local ui = require('http_client.ui.display')
+    local config = require('http_client.config')
+
+    -- Format timing metrics if available
+    local timing_str = ""
+    local profiling_config = config.get('profiling')
+    
+    if profiling_config and profiling_config.enabled and profiling_config.show_in_response then
+        if pr.timing_metrics and next(pr.timing_metrics) then
+            timing_str = "\n# Timing:\n" .. profiling.format_metrics(pr.timing_metrics)
+        end
+    end
 
     local content = string.format([[
 Response Information (%s):
@@ -137,7 +151,7 @@ Response Information (%s):
 # Status: %s
 
 # Headers:
-%s
+%s%s
 
 # Body (%s):
 %s
@@ -148,6 +162,7 @@ Response Information (%s):
         pr.request.http_version,
         pr.status,
         format_headers(pr.headers),
+        timing_str,
         pr.content_type,
         pr.formatted_body
     )
@@ -188,6 +203,17 @@ M.send_request = function(request)
         return
     end
 
+    local config = require('http_client.config')
+    local profiling_config = config.get('profiling')
+    local profiling_enabled = profiling_config and profiling_config.enabled
+
+    -- Generate request ID and add it to the request object
+    request.request_id = profiling.generate_request_id()
+    
+    if profiling_enabled then
+        profiling.start_metric(request.request_id, "total")
+    end
+
     vvv.debug_print("Headers:")
     for k, v in pairs(request.headers) do
         vvv.debug_print(string.format("  %s: %s", k, v))
@@ -210,6 +236,10 @@ M.send_request = function(request)
         body = request.body,
         headers = request.headers,
         callback = function(response)
+            if profiling_enabled then
+                profiling.end_metric(request.request_id, "total")
+            end
+            
             vvv.debug_print("Response received")
             vvv.debug_print(string.format("Status: %s", response.status))
             vvv.debug_print("Response headers:")
@@ -220,15 +250,54 @@ M.send_request = function(request)
             vvv.debug_print(response.body)
 
             current_request = nil
+            
+            if profiling_enabled then
+                response.timing_metrics = profiling.get_metrics(request.request_id)
+                response.timing_metrics.url = request.url
+                profiling.clear_metrics(request.request_id)
+            end
 
             vvv.debug_print("Calling ui.display_response")
             local pr = prepare_response(request, response)
             display_response(pr)
             handle_response(pr)
         end
-
     }
 
+    -- Add event handlers only if profiling is enabled
+    if profiling_enabled then
+        curl_options.on_start = function()
+            profiling.end_metric(request.request_id, "dns_resolution")
+            profiling.start_metric(request.request_id, "connection")
+        end
+        
+        curl_options.on_connect = function()
+            profiling.end_metric(request.request_id, "connection")
+            profiling.start_metric(request.request_id, "send_request")
+        end
+        
+        curl_options.on_first_byte = function()
+            profiling.end_metric(request.request_id, "send_request")
+            
+            -- Calculate server processing time
+            local metrics = profiling.get_metrics(request.request_id)
+            local start_time = 0
+            
+            if metrics.send_request and metrics.send_request.end_time then
+                start_time = metrics.send_request.end_time
+                
+                local server_metrics = {}
+                server_metrics.start_time = start_time
+                server_metrics.end_time = vim.loop.hrtime()
+                server_metrics.duration = (server_metrics.end_time - start_time) / 1000000
+                
+                metrics.server_processing = server_metrics
+            end
+            
+            profiling.start_metric(request.request_id, "content_transfer")
+        end
+    end
+
     -- Handle different HTTP versions
     if request.http_version then
         if request.http_version == "HTTP/2" then
@@ -246,11 +315,18 @@ M.send_request = function(request)
     if ssl_config.verifyHostCertificate == false then
         curl_options.insecure = true
     end
+    
+    if profiling_enabled then
+        profiling.start_metric(request.request_id, "dns_resolution")
+    end
 
     current_request = curl.request(curl_options)
 
     if not current_request then
         vvv.debug_print("Failed to initiate request")
+        if profiling_enabled then
+            profiling.clear_metrics(request.request_id)
+        end
         return
     end
 
@@ -276,6 +352,16 @@ M.get_current_request = function()
 end
 
 M.send_request_sync = function(request)
+    local config = require('http_client.config')
+    local profiling_config = config.get('profiling')
+    local profiling_enabled = profiling_config and profiling_config.enabled
+    
+    request.request_id = profiling.generate_request_id()
+    
+    if profiling_enabled then
+        profiling.start_metric(request.request_id, "total")
+    end
+    
     local response = {}
     local curl_options = {
         url = request.url,
@@ -300,8 +386,19 @@ M.send_request_sync = function(request)
     end
 
     response = curl.get(curl_options)
+    
+    if profiling_enabled then
+        profiling.end_metric(request.request_id, "total")
+        response.timing_metrics = profiling.get_metrics(request.request_id)
+        response.timing_metrics.url = request.url
+    end
+    
     local pr = prepare_response(request, response)
     handle_response(pr)
+    
+    if profiling_enabled then
+        profiling.clear_metrics(request.request_id)
+    end
 
     return response
 end

lua/http_client/health.lua

@@ -35,6 +35,16 @@ M.check = function()
     else
         health.error("curl is not available", "Install curl")
     end
+
+    -- Check if profiling is enabled
+    local profiling_config = M.config.get('profiling')
+    if profiling_config and profiling_config.enabled then
+        health.ok('Profiling: enabled')
+    else
+        health.info('Profiling: disabled (use :HttpProfiling to enable)')
+    end
+
+    -- Check other parts of the plugin
 end
 
 return M

lua/http_client/init.lua

@@ -35,6 +35,7 @@ local function set_keybindings()
                 vim.keymap.set("n", keybindings.toggle_verbose, ":HttpVerbose<CR>", opts)
                 vim.keymap.set("n", keybindings.copy_curl, ":HttpCopyCurl<CR>", opts)
                 vim.keymap.set("n", keybindings.save_response, ":HttpSaveResponse<CR>", opts)
+                vim.keymap.set("n", keybindings.toggle_profiling, ":HttpProfiling<CR>", opts)
             end,
         })
     end
@@ -92,6 +93,14 @@ M.setup = function(opts)
         desc = "Toggle verbose mode for HTTP request.",
     })
 
+    vim.api.nvim_create_user_command("HttpProfiling", function()
+        local profiling_enabled = M.config.get("profiling").enabled
+        M.config.options.profiling.enabled = not profiling_enabled
+        print(string.format("HTTP Client profiling %s", not profiling_enabled and "enabled" or "disabled"))
+    end, {
+        desc = "Toggle profiling for HTTP requests.",
+    })
+
     vim.api.nvim_create_user_command("HttpDryRun", function()
         M.dry_run.display_dry_run(M)
     end, {