repo stringlengths 5 53 | pr_number int32 1 321k | task_type stringclasses 2
values | issue_text stringlengths 0 81.2k | pr_title stringlengths 1 319 | pr_body stringlengths 0 105k | base_sha stringlengths 40 40 | head_sha stringlengths 40 40 | gold_diff stringlengths 0 202M | changed_files listlengths 0 100 | review_threads listlengths 0 100 | test_patch stringlengths 0 23.4M | merged bool 1
class |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
ChromeDevTools/chrome-devtools-mcp | 2,086 | issue_to_patch | Impossible to start two Chrome instances from DevTools for agents MCP server when using Claude Code CLI
### Description of the bug
It looks like it's not possible to start multiple sessions using DevTools for agents MCP server with Claude Code CLI.
### Reproduction
- Open Claude Code CLI
- Insert prompt asking to ... | chore(main): release chrome-devtools-mcp 1.1.0 | :robot: I have created a release *beep* *boop*
---
## [1.1.0](https://github.com/ChromeDevTools/chrome-devtools-mcp/compare/chrome-devtools-mcp-v1.0.1...chrome-devtools-mcp-v1.1.0) (2026-05-26)
### 🎉 Features
* add extraHttpHeaders emulation to emulate tool ([#1176](https://github.com/ChromeDevTools/chrome-devtoo... | 329fbf1ac9b4cbec808c64312d3fc50aab1a80b8 | 4c7f393d97ef6c9f45818522d16ce3517b7dbd97 | diff --git a/.claude-plugin/marketplace.json b/.claude-plugin/marketplace.json
index 1cf2c12dd..ec77f2740 100644
--- a/.claude-plugin/marketplace.json
+++ b/.claude-plugin/marketplace.json
@@ -1,6 +1,6 @@
{
"name": "chrome-devtools-plugins",
- "version": "1.0.1",
+ "version": "1.1.0",
"description": "Bundled p... | [
".claude-plugin/marketplace.json",
".claude-plugin/plugin.json",
".cursor-plugin/plugin.json",
".github/plugin/plugin.json",
".release-please-manifest.json",
"CHANGELOG.md",
"gemini-extension.json",
"package-lock.json",
"package.json",
"server.json",
"src/version.ts"
] | [] | true | |
ChromeDevTools/chrome-devtools-mcp | 2,142 | issue_to_patch | fix(cli): have pageId as first argument | Refs #1777
This change makes pageId as the first argument for CLI if experimentalPageIdRouting flag is enabled | 0089ebdd37a6b16e52820a3fcb4af133214754a0 | 864ee64bfd475a4ebefba1ecc2cfdff6fb0bdd45 | diff --git a/src/ToolHandler.ts b/src/ToolHandler.ts
index c9bf56d9f..80950c97b 100644
--- a/src/ToolHandler.ts
+++ b/src/ToolHandler.ts
@@ -163,7 +163,7 @@ export class ToolHandler {
tool.pageScoped &&
serverArgs.experimentalPageIdRouting &&
!serverArgs.slim
- ? {...tool.schema, ...pageIdSc... | [
"src/ToolHandler.ts",
"src/tools/script.ts"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,135 | issue_to_patch | Releases should update the chrome-devtools-mcp version in plugins
Currently, all of our plugins use the latest MCP server by default https://github.com/ChromeDevTools/chrome-devtools-mcp/blob/main/.cursor-plugin/plugin.json#L14
we should update the release-please process to generate the version of the chrome-devtools... | fix: use pinned version for plugins | The regex used to match the versions is
```js
const regex = /(?<major>\d+)\.(?<minor>\d+)\.(?<patch>\d+)(-(?<preRelease>[\w.]+))?(\+(?<build>[-\w.]+))?/
```
And they replace what was matched.
Closes https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/2121 | 2eab50938093e848ca9786f2ce0c5bf371e63619 | c5c20b5a6bece6c8042259f6181995d63668a9b3 | diff --git a/.claude-plugin/plugin.json b/.claude-plugin/plugin.json
index 08f3e3dda..f6d3aa57f 100644
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@@ -6,7 +6,7 @@
"chrome-devtools": {
"command": "npx",
"args": [
- "chrome-devtools-mcp@latest"
+ "chrome-devtools-mcp... | [
".claude-plugin/plugin.json",
".cursor-plugin/plugin.json",
".github/plugin/plugin.json",
"gemini-extension.json",
"release-please-config.json"
] | [
{
"comment": "should this be `args[0]`?",
"path": "release-please-config.json",
"hunk": "@@ -40,15 +40,40 @@\n \"path\": \".claude-plugin/plugin.json\",\n \"jsonpath\": \"version\"\n },\n+ {\n+ \"type\": \"json\",\n+ \"path\": \".claude-plugin/plugin.... | true | |
ChromeDevTools/chrome-devtools-mcp | 2,145 | issue_to_patch | chore: remove flags from development commands | Local user must have Node =>v22.18 to use it without flags, consider we should be using v24 locally due to .nvm we can remove the flags and make it easier to read what the commands are doing. | 60be3e6bc157bd1121ea1d4b6ad59e37a73cac3e | 7d410e30758aadcc7ed3df15821a6c8ea22143fa | diff --git a/package.json b/package.json
index 3456d2e9c..a79a17208 100644
--- a/package.json
+++ b/package.json
@@ -9,28 +9,28 @@
},
"main": "./build/src/index.js",
"scripts": {
- "cli:generate": "node --experimental-strip-types scripts/generate-cli.ts",
+ "cli:generate": "node scripts/generate-cli.ts",... | [
"package.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,147 | issue_to_patch | Releases should update the chrome-devtools-mcp version in plugins
Currently, all of our plugins use the latest MCP server by default https://github.com/ChromeDevTools/chrome-devtools-mcp/blob/main/.cursor-plugin/plugin.json#L14
we should update the release-please process to generate the version of the chrome-devtools... | build: fix the updating of plugins | Closes #2121 | 4c886a1c8bf71cdf94beb278c1a0a77e529f5762 | a2e142d7f11353cab621d9ebd274610561faf779 | diff --git a/release-please-config.json b/release-please-config.json
index 5abfdd2af..578afe643 100644
--- a/release-please-config.json
+++ b/release-please-config.json
@@ -43,7 +43,7 @@
{
"type": "json",
"path": ".claude-plugin/plugin.json",
- "jsonpath": "mcpServers['chrome-dev... | [
"release-please-config.json"
] | [] | true | |
ChromeDevTools/chrome-devtools-mcp | 2,146 | issue_to_patch | test: drop node20 code in tests | 4c886a1c8bf71cdf94beb278c1a0a77e529f5762 | 84c18162039d1e2ac9798b61cc5911a03a28b95b | diff --git a/scripts/test.mjs b/scripts/test.mjs
index d545f392b..0d427847f 100644
--- a/scripts/test.mjs
+++ b/scripts/test.mjs
@@ -38,11 +38,7 @@ if (userArgs.length > 0) {
files.push(testPath);
}
} else {
- const isNode20 = process.version.startsWith('v20.');
if (flags.includes('--test-only')) {
- if... | [
"scripts/test.mjs",
"tests/McpContext.test.ts",
"tests/McpResponse.test.ts",
"tests/formatters/ConsoleFormatter.test.ts",
"tests/formatters/HeapSnapshotFormatter.test.ts",
"tests/formatters/IssueFormatter.test.ts",
"tests/formatters/NetworkFormatter.test.ts",
"tests/formatters/snapshotFormatter.test.t... | [] | diff --git a/tests/McpContext.test.ts b/tests/McpContext.test.ts
index 1d5d2cdcf..52e2293fe 100644
--- a/tests/McpContext.test.ts
+++ b/tests/McpContext.test.ts
@@ -148,7 +148,7 @@ describe('McpContext', () => {
response.setIncludeNetworkRequests(true);
const result = await response.handle('test', context... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,148 | issue_to_patch | chore(main): release chrome-devtools-mcp 1.1.1 | :robot: I have created a release *beep* *boop*
---
## [1.1.1](https://github.com/ChromeDevTools/chrome-devtools-mcp/compare/chrome-devtools-mcp-v1.1.0...chrome-devtools-mcp-v1.1.1) (2026-05-27)
### 🛠️ Fixes
* **cli:** have pageId as first argument ([#2142](https://github.com/ChromeDevTools/chrome-devtools-mcp/iss... | af17797f3b87b5d2123e47950d36cb01673abe38 | 7d9edb27b2553cb9080d6848a87c05abedfd8c4f | diff --git a/.claude-plugin/marketplace.json b/.claude-plugin/marketplace.json
index ec77f2740..8d2172818 100644
--- a/.claude-plugin/marketplace.json
+++ b/.claude-plugin/marketplace.json
@@ -1,6 +1,6 @@
{
"name": "chrome-devtools-plugins",
- "version": "1.1.0",
+ "version": "1.1.1",
"description": "Bundled p... | [
".claude-plugin/marketplace.json",
".claude-plugin/plugin.json",
".cursor-plugin/plugin.json",
".github/plugin/plugin.json",
".release-please-manifest.json",
"CHANGELOG.md",
"gemini-extension.json",
"package-lock.json",
"package.json",
"server.json",
"src/version.ts"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,131 | issue_to_patch | The Crux data returned is not accurate
### Description of the bug
There are several points to note regarding the CruX data using `performance_start_trace`:
- **It is not accurate**: there are noticeable differences compared to the CruX data returned by the [PageSpeed website.](https://pagespeed.web.dev/) Using the P... | chore: reflect emulated device in CrUX data | CrUX data returned with the performance tool now reflects the emulated device
Closes: #1813 | 7b5ec3a1b85bed929032348aeed78dd93614bbc1 | 64a94ea77478c758cdcc87d96620b159a02d0079 | diff --git a/src/McpResponse.ts b/src/McpResponse.ts
index 17d7a9f75..b00fba007 100644
--- a/src/McpResponse.ts
+++ b/src/McpResponse.ts
@@ -209,6 +209,10 @@ export class McpResponse implements Response {
#error?: Error;
#attachedWaitForResult?: WaitForEventsResult;
+ get #deviceScope(): DevTools.CrUXManager.D... | [
"src/McpResponse.ts",
"src/trace-processing/parse.ts",
"tests/tools/performance.test.ts"
] | [] | diff --git a/tests/tools/performance.test.ts b/tests/tools/performance.test.ts
index 229f1da23..b9d1c7897 100644
--- a/tests/tools/performance.test.ts
+++ b/tests/tools/performance.test.ts
@@ -387,16 +387,125 @@ describe('performance', () => {
{performanceCrux: false},
);
});
+
+ it('fetches CrU... | true |
ChromeDevTools/chrome-devtools-mcp | 2,152 | issue_to_patch | [Task]: Refactor path validation
### Task to do:
Currently we call validatePath in any tool with filePath inputs. Instead, we should add a new required annotation called `filePathFields: array<string>` requiring each tool to list field paths that require validation. The validation can then be done centrally when hand... | refactor: use validate files on the tool level | Fixes https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/2138
Closes #2150 | 50cecd810b9362f6b3fd3806bcceaca0e2b9d1fb | 6c455fa90f0c821bc074cff083757200ec4ebfb5 | diff --git a/src/ToolHandler.ts b/src/ToolHandler.ts
index 80950c97b..01e07d2a5 100644
--- a/src/ToolHandler.ts
+++ b/src/ToolHandler.ts
@@ -220,6 +220,12 @@ export class ToolHandler {
response.setRedactNetworkHeaders(this.serverArgs.redactNetworkHeaders);
try {
+ if (this.tool.verifyFilesSchema)... | [
"src/ToolHandler.ts",
"src/tools/ToolDefinition.ts",
"src/tools/console.ts",
"src/tools/emulation.ts",
"src/tools/extensions.ts",
"src/tools/input.ts",
"src/tools/lighthouse.ts",
"src/tools/memory.ts",
"src/tools/network.ts",
"src/tools/pages.ts",
"src/tools/performance.ts",
"src/tools/screenc... | [
{
"comment": "let's make it required so that it is less frequently overlooked",
"path": "src/tools/ToolDefinition.ts",
"hunk": "@@ -48,6 +48,7 @@ export interface BaseToolDefinition<\n };\n schema: Schema;\n blockedByDialog: boolean;\n+ verifyFilesSchema?: Array<keyof Schema>;",
"resolving_sh... | diff --git a/tests/ToolHandler.test.ts b/tests/ToolHandler.test.ts
index b86516d41..868d2ec7a 100644
--- a/tests/ToolHandler.test.ts
+++ b/tests/ToolHandler.test.ts
@@ -37,6 +37,7 @@ describe('ToolHandler', () => {
},
schema: {},
blockedByDialog: false,
+ verifyFilesSchema: [],
pageScop... | true |
ChromeDevTools/chrome-devtools-mcp | 2,152 | comment_to_fix | refactor: use validate files on the tool level | let's make it required so that it is less frequently overlooked | 50cecd810b9362f6b3fd3806bcceaca0e2b9d1fb | 6c455fa90f0c821bc074cff083757200ec4ebfb5 | diff --git a/src/tools/ToolDefinition.ts b/src/tools/ToolDefinition.ts
index dd8cba89b..7350b2b45 100644
--- a/src/tools/ToolDefinition.ts
+++ b/src/tools/ToolDefinition.ts
@@ -48,6 +48,7 @@ export interface BaseToolDefinition<
};
schema: Schema;
blockedByDialog: boolean;
+ verifyFilesSchema: Array<keyof Sche... | [
"src/tools/ToolDefinition.ts"
] | [
{
"comment": "let's make it required so that it is less frequently overlooked",
"path": "src/tools/ToolDefinition.ts",
"hunk": "@@ -48,6 +48,7 @@ export interface BaseToolDefinition<\n };\n schema: Schema;\n blockedByDialog: boolean;\n+ verifyFilesSchema?: Array<keyof Schema>;",
"resolving_sh... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,158 | issue_to_patch | chore(deps-dev): bump chrome-devtools-frontend from 1.0.1635648 to 1.0.1638082 in the bundled-devtools group | Bumps the bundled-devtools group with 1 update: [chrome-devtools-frontend](https://github.com/ChromeDevTools/devtools-frontend).
Updates `chrome-devtools-frontend` from 1.0.1635648 to 1.0.1638082
<details>
<summary>Commits</summary>
<ul>
<li><a href="https://github.com/ChromeDevTools/devtools-frontend/commit/0e1186138... | 2e039c09e1a273581d9b51081a0feb8a57791947 | 67b0f9d51aa81c1639b00e7980a71246643bc294 | diff --git a/package-lock.json b/package-lock.json
index a06a97544..505103376 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -28,7 +28,7 @@
"@types/yargs": "^17.0.33",
"@typescript-eslint/eslint-plugin": "^8.43.0",
"@typescript-eslint/parser": "^8.43.0",
- "chrome-devtools-f... | [
"package-lock.json",
"package.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,165 | issue_to_patch | refactor: change type of logger | This change aligns most of our logger allowing them to be omitted when needed meaning code inside the args will not be executed unconditionally - example `logger?.(JSON.stringy(X))`.
Currently only changes the type, while in a follow up I will remove the `debug` package and update the file logs. | 63304ef89624f1f89a6a2aad14bbc9a7ae6dde58 | 26c3ddc107a51c9345a1e762fecba951d94a8dec | diff --git a/src/McpContext.ts b/src/McpContext.ts
index 3855c02f4..1d15b3421 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -27,7 +27,6 @@ import {
type Browser,
type BrowserContext,
type ConsoleMessage,
- type Debugger,
type HTTPRequest,
type Page,
type ScreenRecorder,
@@ -41,6 +40,7 @@ ... | [
"src/McpContext.ts",
"src/McpPage.ts",
"src/PageCollector.ts",
"src/TextSnapshot.ts",
"src/ToolHandler.ts",
"src/WaitForHelper.ts",
"src/bin/chrome-devtools-mcp-main.ts",
"src/browser.ts",
"src/daemon/client.ts",
"src/daemon/daemon.ts",
"src/daemon/utils.ts",
"src/formatters/IssueFormatter.ts"... | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,166 | issue_to_patch | Include page title in `list_pages` output
### Is your feature request related to a problem? Please describe.
`list_pages` shows each page's URL but not its title, which is painful when pages share a host — e.g. several accounts of the same product, all under `app.example.com/u/0/…`, `/u/1/…`, `/u/2/…`. There's no way... | feat: include page title in list_pages output | ## Motivation
`list_pages` shows each page's URL but not its title, which is painful when multiple pages share a host — e.g. several tabs under `app.example.com/u/0/`, `/u/1/`, `/u/2/`. There's no way to tell which is which without visiting each one, even though every page has a usable `document.title`.
Fixes #21... | a97c642d43d19fc5198038a7544ff41528ddc316 | 403465713614e9ef9875e9667437357ad7229235 | diff --git a/src/McpResponse.ts b/src/McpResponse.ts
index b2abf7660..d0de129a8 100644
--- a/src/McpResponse.ts
+++ b/src/McpResponse.ts
@@ -759,7 +759,7 @@ export class McpResponse implements Response {
);
}
- format(
+ async format(
toolName: string,
context: McpContext,
data: {
@@ -777,7 +... | [
"src/McpResponse.ts",
"tests/McpResponse.test.js.snapshot",
"tests/tools/pages.test.js.snapshot"
] | [
{
"comment": "Please add a character limit as requested in https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/2156#issuecomment-4610053690\nIt would be good to extract this to a separate function since you'll need to do this in three places.",
"path": "src/McpResponse.ts",
"hunk": "@@ -849,10 ... | diff --git a/tests/McpResponse.test.js.snapshot b/tests/McpResponse.test.js.snapshot
index 52652269d..d03588984 100644
--- a/tests/McpResponse.test.js.snapshot
+++ b/tests/McpResponse.test.js.snapshot
@@ -324,6 +324,7 @@ exports[`McpResponse > list pages 2`] = `
{
"id": 1,
"url": "about:blank",
+ ... | true |
ChromeDevTools/chrome-devtools-mcp | 2,166 | comment_to_fix | feat: include page title in list_pages output | Please add a character limit as requested in https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/2156#issuecomment-4610053690
It would be good to extract this to a separate function since you'll need to do this in three places. | a97c642d43d19fc5198038a7544ff41528ddc316 | 403465713614e9ef9875e9667437357ad7229235 | diff --git a/src/McpResponse.ts b/src/McpResponse.ts
index b2abf7660..d0de129a8 100644
--- a/src/McpResponse.ts
+++ b/src/McpResponse.ts
@@ -759,7 +759,7 @@ export class McpResponse implements Response {
);
}
- format(
+ async format(
toolName: string,
context: McpContext,
data: {
@@ -777,7 +... | [
"src/McpResponse.ts"
] | [
{
"comment": "Please add a character limit as requested in https://github.com/ChromeDevTools/chrome-devtools-mcp/issues/2156#issuecomment-4610053690\nIt would be good to extract this to a separate function since you'll need to do this in three places.",
"path": "src/McpResponse.ts",
"hunk": "@@ -849,10 ... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,189 | issue_to_patch | docs(skill): guide agent to prompt for --categoryExtensions | ...when extension tools are missing. I've run into the situation that the coding agent would try different esoteric approach to install the extension when asked to debug it in chrome DevTools. | aa33bff19fcbf87949eec152c4b49d74e0a9330d | 80b7b4f3db785080e4e46d77018e6eee436a2475 | diff --git a/skills/chrome-devtools/SKILL.md b/skills/chrome-devtools/SKILL.md
index 622619bdd..d62f66eca 100644
--- a/skills/chrome-devtools/SKILL.md
+++ b/skills/chrome-devtools/SKILL.md
@@ -41,6 +41,21 @@ You can send multiple tool calls in parallel, but maintain correct order: naviga
### Testing an extension
+... | [
"skills/chrome-devtools/SKILL.md"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,042 | issue_to_patch | feat: add experimental TOON support for structured content output | Adds `--experimentalToonFormat` boolean flag, controlling how structured content is formatted in text response.
By default, custom shorthand format is used. With this flag, TOON format (see https://github.com/toon-format/toon) is used instead.
TOON format is supposed to be more token-efficient and less error-prone ... | 29e389848fd0cae1116620d44fbce508b4404137 | ea498ffa226e4a174517d0caf6914dbf3ffd9535 | diff --git a/package-lock.json b/package-lock.json
index 55723ae9c..09fdbf3ae 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -20,6 +20,7 @@
"@rollup/plugin-json": "^6.1.0",
"@rollup/plugin-node-resolve": "^16.0.3",
"@stylistic/eslint-plugin": "^5.4.0",
+ "@toon-format/toon":... | [
"package-lock.json",
"package.json",
"src/McpResponse.ts",
"src/ToolHandler.ts",
"src/bin/chrome-devtools-mcp-cli-options.ts",
"src/telemetry/flag_usage_metrics.json",
"src/third_party/index.ts",
"tests/third_party_notices.test.js.snapshot"
] | [] | diff --git a/tests/third_party_notices.test.js.snapshot b/tests/third_party_notices.test.js.snapshot
index 4b9412642..4cce074ad 100644
--- a/tests/third_party_notices.test.js.snapshot
+++ b/tests/third_party_notices.test.js.snapshot
@@ -907,6 +907,36 @@ IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF ... | true | |
ChromeDevTools/chrome-devtools-mcp | 2,037 | issue_to_patch | feat: support allowedUrlPattern & blockedUrlPattern Options | ## Support for Network Blocklists and Allowlists (`--blocked-url-pattern` & `--allowed-url-pattern` arguments)
This PR adds support for CLI options to restrict network access in the browser session via URL patterns.
### Key Features & How It Works
- **Pattern Matching:** Utilizes the [URLPattern Standard](http... | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/README.md b/README.md
index cfe7dd3ea..ea3f4e962 100644
--- a/README.md
+++ b/README.md
@@ -637,6 +637,14 @@ The Chrome DevTools MCP server supports the following configuration option:
Additional arguments for Chrome. Only applies when Chrome is launched by chrome-devtools-mcp.
- **Type:** array
+-... | [
"README.md",
"src/DevtoolsUtils.ts",
"src/McpContext.ts",
"src/bin/chrome-devtools-mcp-cli-options.ts",
"src/browser.ts",
"src/index.ts",
"src/telemetry/flag_usage_metrics.json",
"tests/browser.test.ts",
"tests/cli.test.ts",
"tests/network_blocking.test.ts",
"tests/utils.ts"
] | [
{
"comment": "Please document CLI args in details, what will be blocked and what will not be blocked and starting from which Chrome version.",
"path": "src/bin/chrome-devtools-mcp-cli-options.ts",
"hunk": "@@ -210,6 +210,17 @@ export const cliOptions = {\n describe:\n 'Additional arguments for... | diff --git a/tests/browser.test.ts b/tests/browser.test.ts
index 85e9c592f..3c4ed5976 100644
--- a/tests/browser.test.ts
+++ b/tests/browser.test.ts
@@ -13,6 +13,8 @@ import {executablePath} from 'puppeteer';
import {detectDisplay, ensureBrowserConnected, launch} from '../src/browser.js';
+import {serverHooks} fro... | true | |
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Please document CLI args in details, what will be blocked and what will not be blocked and starting from which Chrome version. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/src/bin/chrome-devtools-mcp-cli-options.ts b/src/bin/chrome-devtools-mcp-cli-options.ts
index b65929e12..0cfb3c2b8 100644
--- a/src/bin/chrome-devtools-mcp-cli-options.ts
+++ b/src/bin/chrome-devtools-mcp-cli-options.ts
@@ -205,6 +205,18 @@ export const cliOptions = {
describe:
'Additional argu... | [
"src/bin/chrome-devtools-mcp-cli-options.ts"
] | [
{
"comment": "Please document CLI args in details, what will be blocked and what will not be blocked and starting from which Chrome version.",
"path": "src/bin/chrome-devtools-mcp-cli-options.ts",
"hunk": "@@ -210,6 +210,17 @@ export const cliOptions = {\n describe:\n 'Additional arguments for... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | We do not need to mention `Cannot be used with` as 'conflicts' already would handle this. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/src/bin/chrome-devtools-mcp-cli-options.ts b/src/bin/chrome-devtools-mcp-cli-options.ts
index b65929e12..0cfb3c2b8 100644
--- a/src/bin/chrome-devtools-mcp-cli-options.ts
+++ b/src/bin/chrome-devtools-mcp-cli-options.ts
@@ -205,6 +205,18 @@ export const cliOptions = {
describe:
'Additional argu... | [
"src/bin/chrome-devtools-mcp-cli-options.ts"
] | [
{
"comment": "We do not need to mention `Cannot be used with` as 'conflicts' already would handle this.",
"path": "src/bin/chrome-devtools-mcp-cli-options.ts",
"hunk": "@@ -210,6 +210,17 @@ export const cliOptions = {\n describe:\n 'Additional arguments for Chrome. Only applies when Chrome is ... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | how does it compare to using a string type? Does `array` support comma separated input? I think if we go with the array we should rename the argument from blocklist/allowlist to `blocked-url-pattern`, `allowed-url-pattern`. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/src/bin/chrome-devtools-mcp-cli-options.ts b/src/bin/chrome-devtools-mcp-cli-options.ts
index b65929e12..0cfb3c2b8 100644
--- a/src/bin/chrome-devtools-mcp-cli-options.ts
+++ b/src/bin/chrome-devtools-mcp-cli-options.ts
@@ -205,6 +205,18 @@ export const cliOptions = {
describe:
'Additional argu... | [
"src/bin/chrome-devtools-mcp-cli-options.ts"
] | [
{
"comment": "how does it compare to using a string type? Does `array` support comma separated input? I think if we go with the array we should rename the argument from blocklist/allowlist to `blocked-url-pattern`, `allowed-url-pattern`.",
"path": "src/bin/chrome-devtools-mcp-cli-options.ts",
"hunk": "@... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | block list should not need M149, right? | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/network_blocking.test.ts b/tests/network_blocking.test.ts
new file mode 100644
index 000000000..fbc2279b3
--- /dev/null
+++ b/tests/network_blocking.test.ts
@@ -0,0 +1,283 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+import assert from 'node:asser... | [
"tests/network_blocking.test.ts"
] | [
{
"comment": "block list should not need M149, right?",
"path": "tests/network_blocking.test.ts",
"hunk": "@@ -0,0 +1,241 @@\n+/**\n+ * @license\n+ * Copyright 2026 Google LLC\n+ * SPDX-License-Identifier: Apache-2.0\n+ */\n+\n+import assert from 'node:assert/strict';\n+import {describe, it} from 'node:... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | blocklist should not need M149? | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/network_blocking.test.ts b/tests/network_blocking.test.ts
new file mode 100644
index 000000000..fbc2279b3
--- /dev/null
+++ b/tests/network_blocking.test.ts
@@ -0,0 +1,283 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+import assert from 'node:asser... | [
"tests/network_blocking.test.ts"
] | [
{
"comment": "blocklist should not need M149?",
"path": "tests/network_blocking.test.ts",
"hunk": "@@ -0,0 +1,241 @@\n+/**\n+ * @license\n+ * Copyright 2026 Google LLC\n+ * SPDX-License-Identifier: Apache-2.0\n+ */\n+\n+import assert from 'node:assert/strict';\n+import {describe, it} from 'node:test';\n... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Should we throw if we try to set emulation? Reason is I try to emulate network conditions and this just silently ignores it, an LLM most likely will be confused. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/src/McpContext.ts b/src/McpContext.ts
index 3e7e410a8..4346070d2 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -60,6 +60,8 @@ interface McpContextOptions {
experimentalIncludeAllPages?: boolean;
// Whether CrUX data should be fetched.
performanceCrux: boolean;
+ // Whether allowlist/b... | [
"src/McpContext.ts"
] | [
{
"comment": "Should we throw if we try to set emulation? Reason is I try to emulate network conditions and this just silently ignores it, an LLM most likely will be confused. ",
"path": "src/McpContext.ts",
"hunk": "@@ -345,24 +347,27 @@ export class McpContext implements Context {\n const mcpPage ... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Nit: We can just return true or false as the message may change. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/browser.test.ts b/tests/browser.test.ts
index 85e9c592f..3c4ed5976 100644
--- a/tests/browser.test.ts
+++ b/tests/browser.test.ts
@@ -13,6 +13,8 @@ import {executablePath} from 'puppeteer';
import {detectDisplay, ensureBrowserConnected, launch} from '../src/browser.js';
+import {serverHooks} fro... | [
"tests/browser.test.ts"
] | [
{
"comment": "Nit: We can just return true or false as the message may change.",
"path": "tests/browser.test.ts",
"hunk": "@@ -100,4 +102,88 @@ describe('browser', () => {\n await browser.close();\n }\n });\n+\n+ describe('Blocking', () => {\n+ const server = serverHooks();\n+\n+ it('... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Is this also still needed? | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/browser.test.ts b/tests/browser.test.ts
index 85e9c592f..3c4ed5976 100644
--- a/tests/browser.test.ts
+++ b/tests/browser.test.ts
@@ -13,6 +13,8 @@ import {executablePath} from 'puppeteer';
import {detectDisplay, ensureBrowserConnected, launch} from '../src/browser.js';
+import {serverHooks} fro... | [
"tests/browser.test.ts"
] | [
{
"comment": "Is this also still needed?",
"path": "tests/browser.test.ts",
"hunk": "@@ -100,4 +102,88 @@ describe('browser', () => {\n await browser.close();\n }\n });\n+\n+ describe('Blocking', () => {\n+ const server = serverHooks();\n+\n+ it('blocks URLs in blocklist', async () =>... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Can you double check if ` '--blocked-url-pattern', '<pattern-1>', <pattern-2>` works? I think that is also valid CLI args. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/cli.test.ts b/tests/cli.test.ts
index 05f2d3b06..25f07d18e 100644
--- a/tests/cli.test.ts
+++ b/tests/cli.test.ts
@@ -345,4 +345,94 @@ describe('cli args parsing', () => {
);
assert.strictEqual(disabledArgs.performanceCrux, false);
});
+
+ it('parses blocked-url-pattern flags as array',... | [
"tests/cli.test.ts"
] | [
{
"comment": "Can you double check if ` '--blocked-url-pattern', '<pattern-1>', <pattern-2>` works? I think that is also valid CLI args.",
"path": "tests/cli.test.ts",
"hunk": "@@ -345,4 +345,62 @@ describe('cli args parsing', () => {\n );\n assert.strictEqual(disabledArgs.performanceCrux, false... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Same here. we can use true/false. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/network_blocking.test.ts b/tests/network_blocking.test.ts
new file mode 100644
index 000000000..fbc2279b3
--- /dev/null
+++ b/tests/network_blocking.test.ts
@@ -0,0 +1,283 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+import assert from 'node:asser... | [
"tests/network_blocking.test.ts"
] | [
{
"comment": "Same here. we can use true/false.",
"path": "tests/network_blocking.test.ts",
"hunk": "@@ -0,0 +1,240 @@\n+/**\n+ * @license\n+ * Copyright 2026 Google LLC\n+ * SPDX-License-Identifier: Apache-2.0\n+ */\n+\n+import assert from 'node:assert/strict';\n+import {describe, it} from 'node:test';... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | This was confusingly at the end we should move it to the start, some where easier to find. | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/network_blocking.test.ts b/tests/network_blocking.test.ts
new file mode 100644
index 000000000..fbc2279b3
--- /dev/null
+++ b/tests/network_blocking.test.ts
@@ -0,0 +1,283 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+import assert from 'node:asser... | [
"tests/network_blocking.test.ts"
] | [
{
"comment": "This was confusingly at the end we should move it to the start, some where easier to find.",
"path": "tests/network_blocking.test.ts",
"hunk": "@@ -0,0 +1,240 @@\n+/**\n+ * @license\n+ * Copyright 2026 Google LLC\n+ * SPDX-License-Identifier: Apache-2.0\n+ */\n+\n+import assert from 'node:... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,037 | comment_to_fix | feat: support allowedUrlPattern & blockedUrlPattern Options | Do we need this? | f90f863d4b9d643ab880c9dc8cdd78b6c88ae38d | 86eeb4f80670dcfe52170288e8ecbd657c2a38f6 | diff --git a/tests/network_blocking.test.ts b/tests/network_blocking.test.ts
new file mode 100644
index 000000000..fbc2279b3
--- /dev/null
+++ b/tests/network_blocking.test.ts
@@ -0,0 +1,283 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+import assert from 'node:asser... | [
"tests/network_blocking.test.ts"
] | [
{
"comment": "Do we need this?",
"path": "tests/network_blocking.test.ts",
"hunk": "@@ -0,0 +1,240 @@\n+/**\n+ * @license\n+ * Copyright 2026 Google LLC\n+ * SPDX-License-Identifier: Apache-2.0\n+ */\n+\n+import assert from 'node:assert/strict';\n+import {describe, it} from 'node:test';\n+\n+import {lig... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,173 | issue_to_patch | fix: Cursor plugin homepage setting field | Fix Cursor plugin homepage setting field as per #2172 | 30d59a78727c31ec9d70d2bd6d9310e78f1888b3 | 2d8a46bfdae9cf6ec5803740dc749c1350f1d9ec | diff --git a/.cursor-plugin/plugin.json b/.cursor-plugin/plugin.json
index e88830433..fc0aad514 100644
--- a/.cursor-plugin/plugin.json
+++ b/.cursor-plugin/plugin.json
@@ -6,7 +6,7 @@
"name": "Google Chrome"
},
"logo": "https://github.com/ChromeDevTools/devtools-logo/blob/master/logos/svg/chrome-devtools-sq... | [
".cursor-plugin/plugin.json"
] | [
{
"comment": "Is this part of the diff needed / intended? If not, let’s keep it as-is:\n\n\n```suggestion\n \"description\": \"Help your agent build, debug, and verify your code correctly. With Chrome DevTools for agents, your AI agent can interact with the Chrome browser to test code, emulate users, and catch... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,173 | comment_to_fix | fix: Cursor plugin homepage setting field | Is this part of the diff needed / intended? If not, let’s keep it as-is:
```suggestion
"description": "Help your agent build, debug, and verify your code correctly. With Chrome DevTools for agents, your AI agent can interact with the Chrome browser to test code, emulate users, and catch bugs using Chrome DevTools’ ... | 30d59a78727c31ec9d70d2bd6d9310e78f1888b3 | 2d8a46bfdae9cf6ec5803740dc749c1350f1d9ec | diff --git a/.cursor-plugin/plugin.json b/.cursor-plugin/plugin.json
index e88830433..fc0aad514 100644
--- a/.cursor-plugin/plugin.json
+++ b/.cursor-plugin/plugin.json
@@ -6,7 +6,7 @@
"name": "Google Chrome"
},
"logo": "https://github.com/ChromeDevTools/devtools-logo/blob/master/logos/svg/chrome-devtools-sq... | [
".cursor-plugin/plugin.json"
] | [
{
"comment": "Is this part of the diff needed / intended? If not, let’s keep it as-is:\n\n\n```suggestion\n \"description\": \"Help your agent build, debug, and verify your code correctly. With Chrome DevTools for agents, your AI agent can interact with the Chrome browser to test code, emulate users, and catch... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,164 | issue_to_patch | chore(deps-dev): bump the dev-dependencies group across 1 directory with 4 updates | Bumps the dev-dependencies group with 4 updates in the / directory: [@google/genai](https://github.com/googleapis/js-genai), [@rollup/plugin-commonjs](https://github.com/rollup/plugins/tree/HEAD/packages/commonjs), [eslint-import-resolver-typescript](https://github.com/import-js/eslint-import-resolver-typescript) and [... | 63304ef89624f1f89a6a2aad14bbc9a7ae6dde58 | a574d57c4618667a39ab0fca27c6728348e53f0b | diff --git a/package-lock.json b/package-lock.json
index 505103376..dfa81ca7f 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -38,7 +38,7 @@
"lighthouse": "13.3.0",
"prettier": "^3.6.2",
"puppeteer": "25.1.0",
- "rollup": "4.60.4",
+ "rollup": "4.61.0",
"rollu... | [
"package-lock.json",
"package.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,110 | issue_to_patch | chore(deps): bump qs from 6.14.2 to 6.15.2 | Bumps [qs](https://github.com/ljharb/qs) from 6.14.2 to 6.15.2.
<details>
<summary>Changelog</summary>
<p><em>Sourced from <a href="https://github.com/ljharb/qs/blob/main/CHANGELOG.md">qs's changelog</a>.</em></p>
<blockquote>
<h2><strong>6.15.2</strong></h2>
<ul>
<li>[Fix] <code>stringify</code>: skip null/undefined e... | 57f32b0cd4afe1775b96ba35c27f25d6f0770331 | 4173bcb181e702ac9e6c8949d6ae22f433694b4e | diff --git a/package-lock.json b/package-lock.json
index 2d3adfc60..b1b9d1b73 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -7260,9 +7260,9 @@
}
},
"node_modules/qs": {
- "version": "6.14.2",
- "resolved": "https://registry.npmjs.org/qs/-/qs-6.14.2.tgz",
- "integrity": "sha51... | [
"package-lock.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,111 | issue_to_patch | chore(deps-dev): bump the dev-dependencies group with 6 updates | Bumps the dev-dependencies group with 6 updates:
| Package | From | To |
| --- | --- | --- |
| [@google/genai](https://github.com/googleapis/js-genai) | `2.4.0` | `2.6.0` |
| [@types/node](https://github.com/DefinitelyTyped/DefinitelyTyped/tree/HEAD/types/node) | `25.8.0` | `25.9.1` |
| [@typescript-eslint/eslint-plug... | 57f32b0cd4afe1775b96ba35c27f25d6f0770331 | b4bc25c79b8d7b3e62178b70afe658f497fed8a4 | diff --git a/package-lock.json b/package-lock.json
index 2d3adfc60..50b01a9da 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -383,9 +383,9 @@
}
},
"node_modules/@google/genai": {
- "version": "2.4.0",
- "resolved": "https://registry.npmjs.org/@google/genai/-/genai-2.4.0.tgz",
- ... | [
"package-lock.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,112 | issue_to_patch | chore(deps-dev): bump chrome-devtools-frontend from 1.0.1631386 to 1.0.1632065 in the bundled-devtools group | Bumps the bundled-devtools group with 1 update: [chrome-devtools-frontend](https://github.com/ChromeDevTools/devtools-frontend).
Updates `chrome-devtools-frontend` from 1.0.1631386 to 1.0.1632065
<details>
<summary>Commits</summary>
<ul>
<li><a href="https://github.com/ChromeDevTools/devtools-frontend/commit/53df97e01... | 57f32b0cd4afe1775b96ba35c27f25d6f0770331 | e11a5d0a1481e1a65029edefaafcda64e740e7a9 | diff --git a/package-lock.json b/package-lock.json
index 2d3adfc60..96a46e700 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -28,7 +28,7 @@
"@types/yargs": "^17.0.33",
"@typescript-eslint/eslint-plugin": "^8.43.0",
"@typescript-eslint/parser": "^8.43.0",
- "chrome-devtools-f... | [
"package-lock.json",
"package.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,124 | issue_to_patch | fix(cli): address pid file creation issues | 57f32b0cd4afe1775b96ba35c27f25d6f0770331 | 0020ad809c8192b7225ea941807782d53510a6f3 | diff --git a/src/daemon/daemon.ts b/src/daemon/daemon.ts
index e6cb8bd08..6708f31f5 100644
--- a/src/daemon/daemon.ts
+++ b/src/daemon/daemon.ts
@@ -6,8 +6,9 @@
* SPDX-License-Identifier: Apache-2.0
*/
-import fs from 'node:fs';
+import fs, {constants, openSync, writeSync, closeSync} from 'node:fs';
import {crea... | [
"src/daemon/daemon.ts",
"tests/daemon/symlink.test.ts"
] | [] | diff --git a/tests/daemon/symlink.test.ts b/tests/daemon/symlink.test.ts
new file mode 100644
index 000000000..3726464f9
--- /dev/null
+++ b/tests/daemon/symlink.test.ts
@@ -0,0 +1,126 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+
+import assert from 'node:assert';
+i... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,127 | issue_to_patch | fix: use realpath for MCP roots validation | 2eab50938093e848ca9786f2ce0c5bf371e63619 | b68617198aacdd566015187168bff37f4dbdadf0 | diff --git a/src/McpContext.ts b/src/McpContext.ts
index f8b5f3cd8..2a4e1da9d 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -5,6 +5,7 @@
*/
import fs from 'node:fs/promises';
+import fsPromises from 'node:fs/promises';
import os from 'node:os';
import path from 'node:path';
import {fileURLToPath, pa... | [
"src/McpContext.ts",
"src/tools/ToolDefinition.ts",
"src/tools/input.ts",
"src/tools/lighthouse.ts",
"src/tools/memory.ts",
"src/tools/network.ts",
"src/tools/performance.ts",
"src/tools/screencast.ts",
"src/tools/screenshot.ts",
"src/tools/script.ts",
"src/tools/snapshot.ts",
"src/utils/files... | [
{
"comment": "Could there be cases where a directory structure of 2 or more directories need to be created?\n`/existing/first_new/second_new/foo.bar`",
"path": "src/McpContext.ts",
"hunk": "@@ -175,27 +176,83 @@ export class McpContext implements Context {\n this.#roots = roots;\n }\n \n- validat... | diff --git a/tests/McpContext.test.ts b/tests/McpContext.test.ts
index 487e59561..1d5d2cdcf 100644
--- a/tests/McpContext.test.ts
+++ b/tests/McpContext.test.ts
@@ -5,6 +5,7 @@
*/
import assert from 'node:assert';
+import fs from 'node:fs/promises';
import os from 'node:os';
import path from 'node:path';
import... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,127 | comment_to_fix | fix: use realpath for MCP roots validation | Could there be cases where a directory structure of 2 or more directories need to be created?
`/existing/first_new/second_new/foo.bar` | 2eab50938093e848ca9786f2ce0c5bf371e63619 | b68617198aacdd566015187168bff37f4dbdadf0 | diff --git a/src/McpContext.ts b/src/McpContext.ts
index f8b5f3cd8..2a4e1da9d 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -5,6 +5,7 @@
*/
import fs from 'node:fs/promises';
+import fsPromises from 'node:fs/promises';
import os from 'node:os';
import path from 'node:path';
import {fileURLToPath, pa... | [
"src/McpContext.ts"
] | [
{
"comment": "Could there be cases where a directory structure of 2 or more directories need to be created?\n`/existing/first_new/second_new/foo.bar`",
"path": "src/McpContext.ts",
"hunk": "@@ -175,27 +176,83 @@ export class McpContext implements Context {\n this.#roots = roots;\n }\n \n- validat... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,134 | issue_to_patch | fix: Set viewport after updating timeouts when setting emulation | Tentatively addresses #2115
| 521c388624d448c8c55ee3b5415971ebc35b1ec3 | f921ed89d0642d0d4bc37934a5d954cdca77e2e9 | diff --git a/src/McpContext.ts b/src/McpContext.ts
index fe288c6cc..3855c02f4 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -413,7 +413,6 @@ export class McpContext implements Context {
}
if (!options.viewport) {
- await page.setViewport(null);
delete newSettings.viewport;
} els... | [
"src/McpContext.ts"
] | [
{
"comment": "I think we should avoid running setViewport if it was not changed in the current update.",
"path": "src/McpContext.ts",
"hunk": "@@ -404,6 +401,10 @@ export class McpContext implements Context {\n : {};\n \n this.#updateSelectedPageTimeouts();\n+\n+ // This should happen after... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,134 | comment_to_fix | fix: Set viewport after updating timeouts when setting emulation | I think we should avoid running setViewport if it was not changed in the current update. | 521c388624d448c8c55ee3b5415971ebc35b1ec3 | f921ed89d0642d0d4bc37934a5d954cdca77e2e9 | diff --git a/src/McpContext.ts b/src/McpContext.ts
index fe288c6cc..3855c02f4 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -413,7 +413,6 @@ export class McpContext implements Context {
}
if (!options.viewport) {
- await page.setViewport(null);
delete newSettings.viewport;
} els... | [
"src/McpContext.ts"
] | [
{
"comment": "I think we should avoid running setViewport if it was not changed in the current update.",
"path": "src/McpContext.ts",
"hunk": "@@ -404,6 +401,10 @@ export class McpContext implements Context {\n : {};\n \n this.#updateSelectedPageTimeouts();\n+\n+ // This should happen after... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,130 | issue_to_patch | chore(deps-dev): bump puppeteer from 25.0.4 to 25.1.0 in the bundled group | Bumps the bundled group with 1 update: [puppeteer](https://github.com/puppeteer/puppeteer).
Updates `puppeteer` from 25.0.4 to 25.1.0
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a href="https://github.com/puppeteer/puppeteer/releases">puppeteer's releases</a>.</em></p>
<blockquote>
<h2>puppeteer-co... | 8ea5f098ef7e8d8ae4f4bbaea5291cef84b8f15f | bcb62b86580590c6508ef314b9ff7e7b7f99ddd1 | diff --git a/package-lock.json b/package-lock.json
index ebb296e8c..1ab7d8ec1 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -37,7 +37,7 @@
"globals": "^17.0.0",
"lighthouse": "13.3.0",
"prettier": "^3.6.2",
- "puppeteer": "25.0.4",
+ "puppeteer": "25.1.0",
"... | [
"package-lock.json",
"package.json",
"tests/McpResponse.test.js.snapshot",
"tests/tools/input.test.ts"
] | [] | diff --git a/tests/McpResponse.test.js.snapshot b/tests/McpResponse.test.js.snapshot
index 9b5597d12..b98a0aef4 100644
--- a/tests/McpResponse.test.js.snapshot
+++ b/tests/McpResponse.test.js.snapshot
@@ -1286,7 +1286,7 @@ exports[`webmcp > includes webmcp tools in list_pages response 1`] = `
## Pages
1: about:blank ... | true | |
ChromeDevTools/chrome-devtools-mcp | 2,129 | issue_to_patch | chore(deps-dev): bump chrome-devtools-frontend from 1.0.1632065 to 1.0.1635648 in the bundled-devtools group | Bumps the bundled-devtools group with 1 update: [chrome-devtools-frontend](https://github.com/ChromeDevTools/devtools-frontend).
Updates `chrome-devtools-frontend` from 1.0.1632065 to 1.0.1635648
<details>
<summary>Commits</summary>
<ul>
<li><a href="https://github.com/ChromeDevTools/devtools-frontend/commit/baad1f976... | 3ba70d350a135f5b444826f204724d08aaa9b924 | 44d499c2f4ce36fdca3093134ba2bbc9a0719e38 | diff --git a/package-lock.json b/package-lock.json
index 4e64ad44f..127cb22c5 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -28,7 +28,7 @@
"@types/yargs": "^17.0.33",
"@typescript-eslint/eslint-plugin": "^8.43.0",
"@typescript-eslint/parser": "^8.43.0",
- "chrome-devtools-f... | [
"package-lock.json",
"package.json",
"src/tools/performance.ts"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,128 | issue_to_patch | chore(deps-dev): bump the dev-dependencies group across 1 directory with 3 updates | Bumps the dev-dependencies group with 3 updates in the / directory: [@typescript-eslint/eslint-plugin](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/eslint-plugin), [@typescript-eslint/parser](https://github.com/typescript-eslint/typescript-eslint/tree/HEAD/packages/parser) and [typescript-e... | b39131172b2b0bbb6735f5a3c490dad6c38552e7 | 0810f2d04c1beeea0afe6d8a593cd85fa13e4401 | diff --git a/package-lock.json b/package-lock.json
index 127cb22c5..a06a97544 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -2227,17 +2227,17 @@
"license": "MIT"
},
"node_modules/@typescript-eslint/eslint-plugin": {
- "version": "8.59.4",
- "resolved": "https://registry.npmjs.org/... | [
"package-lock.json"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,123 | issue_to_patch | chore: fix two typos in issue templates | ## Summary
Two small typo fixes in `.github/ISSUE_TEMPLATE/`:
- `02-feature.yml`: "Suggest an idea for for chrome-devtools-mcp" had a duplicate "for".
- `03-task.yml`: "mainainers only" was missing the first `t`.
Both strings appear on the GitHub issue picker when a user clicks "New issue", so they're user-facing. | 57f32b0cd4afe1775b96ba35c27f25d6f0770331 | ce5332eea1a649dbcc013b7496bcc0b974677637 | diff --git a/.github/ISSUE_TEMPLATE/02-feature.yml b/.github/ISSUE_TEMPLATE/02-feature.yml
index 784433690..0bd648bf4 100644
--- a/.github/ISSUE_TEMPLATE/02-feature.yml
+++ b/.github/ISSUE_TEMPLATE/02-feature.yml
@@ -1,5 +1,5 @@
name: Feature
-description: Suggest an idea for for chrome-devtools-mcp
+description: Sugg... | [
".github/ISSUE_TEMPLATE/02-feature.yml",
".github/ISSUE_TEMPLATE/03-task.yml"
] | [] | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,168 | issue_to_patch | feat: Handle multiple providers of third-party developer tools | This allows a page to have multiple providers of third-party developer tools, which each respond to the `devtoolstooldiscovery` event.
- Multiple `ToolGroup`s
- MCP tool responses only mention third-party developer tools, if there are any. Otherwise this part of the output is skipped.
| bf0574da8ce0986beb5dd2f026a2332a7e63b791 | db0e80c095d1984a1b6d1ec1545aaba654bbc83a | diff --git a/src/McpPage.ts b/src/McpPage.ts
index 0a32c2de8..7dd2955d7 100644
--- a/src/McpPage.ts
+++ b/src/McpPage.ts
@@ -14,7 +14,7 @@ import type {
WebMCPTool,
} from './third_party/index.js';
import {takeSnapshot} from './tools/snapshot.js';
-import type {ToolGroup, ToolDefinition} from './tools/thirdPartyDe... | [
"src/McpPage.ts",
"src/McpResponse.ts",
"src/tools/ToolDefinition.ts",
"src/tools/thirdPartyDeveloper.ts",
"tests/McpResponse.test.js.snapshot",
"tests/McpResponse.test.ts",
"tests/tools/thirdPartyDeveloper.test.ts"
] | [
{
"comment": "Nit: Maybe we can extract a type as I see it used very often.",
"path": "src/McpPage.ts",
"hunk": "@@ -59,7 +59,7 @@ export class McpPage implements ContextPage {\n #dialog?: Dialog;\n #dialogHandler: (dialog: Dialog) => void;\n \n- thirdPartyDeveloperTools: ToolGroup<ToolDefinition> ... | diff --git a/tests/McpResponse.test.js.snapshot b/tests/McpResponse.test.js.snapshot
index b98a0aef4..52652269d 100644
--- a/tests/McpResponse.test.js.snapshot
+++ b/tests/McpResponse.test.js.snapshot
@@ -1261,24 +1261,26 @@ name="myTool", description="Does something", inputSchema={"type":"object","prope
exports[`th... | true | |
ChromeDevTools/chrome-devtools-mcp | 2,168 | comment_to_fix | feat: Handle multiple providers of third-party developer tools | Nit: Maybe we can extract a type as I see it used very often. | bf0574da8ce0986beb5dd2f026a2332a7e63b791 | db0e80c095d1984a1b6d1ec1545aaba654bbc83a | diff --git a/src/McpPage.ts b/src/McpPage.ts
index 0a32c2de8..7dd2955d7 100644
--- a/src/McpPage.ts
+++ b/src/McpPage.ts
@@ -14,7 +14,7 @@ import type {
WebMCPTool,
} from './third_party/index.js';
import {takeSnapshot} from './tools/snapshot.js';
-import type {ToolGroup, ToolDefinition} from './tools/thirdPartyDe... | [
"src/McpPage.ts"
] | [
{
"comment": "Nit: Maybe we can extract a type as I see it used very often.",
"path": "src/McpPage.ts",
"hunk": "@@ -59,7 +59,7 @@ export class McpPage implements ContextPage {\n #dialog?: Dialog;\n #dialogHandler: (dialog: Dialog) => void;\n \n- thirdPartyDeveloperTools: ToolGroup<ToolDefinition> ... | true | ||
ChromeDevTools/chrome-devtools-mcp | 2,096 | issue_to_patch | fix: Fix throttling info in performance trace output | This addresses #1955.
Throttling information needs to be passed to the parser for the output to be correct. | ce31594d6c9614c63a93cd7abddf4522a4c4a053 | 19451ed328b414f4e673b8ea5db7e4f723389537 | diff --git a/src/McpContext.ts b/src/McpContext.ts
index a77bc06bc..2916163cd 100644
--- a/src/McpContext.ts
+++ b/src/McpContext.ts
@@ -481,7 +481,7 @@ export class McpContext implements Context {
page.networkConditions,
);
page.pptrPage.setDefaultNavigationTimeout(
- NAVIGATION_TIMEOUT * network... | [
"src/McpContext.ts",
"src/tools/ToolDefinition.ts",
"src/tools/performance.ts",
"src/trace-processing/parse.ts",
"tests/e2e/chrome-devtools-commands.test.ts"
] | [] | diff --git a/tests/e2e/chrome-devtools-commands.test.ts b/tests/e2e/chrome-devtools-commands.test.ts
index 4884d4f82..97d23fb7d 100644
--- a/tests/e2e/chrome-devtools-commands.test.ts
+++ b/tests/e2e/chrome-devtools-commands.test.ts
@@ -106,4 +106,41 @@ describe('chrome-devtools', () => {
'restart command sugges... | true | |
ClickHouse/ClickHouse | 100,479 | issue_to_patch | Occurs error when (I)LIKE doesn't support search with-non-constant needles in constant haystack when left side is constant
### Company or project name
_No response_
### Describe what's wrong
When execution query like
```
select * from sample_tables where 'foo' ilike concat('%', sample_column, '%');
```
Occurs er... | Support constant haystack with non-constant needle in LIKE/ILIKE/match | ### Changelog category (leave one):
- Bug Fix (user-visible misbehavior in an official stable release)
### Changelog entry (a [user-readable short description](https://github.com/ClickHouse/ClickHouse/blob/master/docs/changelog_entry_guidelines.md) of the changes that goes into CHANGELOG.md):
Functions `like`, `ilike... | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h",
"src/Functions/match.cpp",
"tests/queries/0_stateless/03278_enum_string_functions.reference",
"tests/queries/0_stateless/03278_enum_string_functions.sql",
"tests/queries/0_stateless/04327_like_constant_haystack_nonconstant_needle.reference",
"tests/queries/0_stateless/04327_li... | [
{
"comment": "`constantVector` no longer throws `ILLEGAL_COLUMN`, so `ErrorCodes::ILLEGAL_COLUMN` at the top of this file became unused. This is currently failing style CI (`ErrorCodes::ILLEGAL_COLUMN is defined but not used`). Please remove the unused declaration from `namespace ErrorCodes`.\n",
"path": "s... | diff --git a/tests/queries/0_stateless/03278_enum_string_functions.reference b/tests/queries/0_stateless/03278_enum_string_functions.reference
index c7198e2de762..88ef7fc1ba5f 100644
--- a/tests/queries/0_stateless/03278_enum_string_functions.reference
+++ b/tests/queries/0_stateless/03278_enum_string_functions.referen... | true |
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | `constantVector` no longer throws `ILLEGAL_COLUMN`, so `ErrorCodes::ILLEGAL_COLUMN` at the top of this file became unused. This is currently failing style CI (`ErrorCodes::ILLEGAL_COLUMN is defined but not used`). Please remove the unused declaration from `namespace ErrorCodes`.
| bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "`constantVector` no longer throws `ILLEGAL_COLUMN`, so `ErrorCodes::ILLEGAL_COLUMN` at the top of this file became unused. This is currently failing style CI (`ErrorCodes::ILLEGAL_COLUMN is defined but not used`). Please remove the unused declaration from `namespace ErrorCodes`.\n",
"path": "s... | true | ||
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | `constantVector` is missing the trivial-pattern handling that `vectorConstant` already has for `needle == '%'` / `needle == '%%'`.
With the current code, `'' LIKE '%%'` enters `likePatternIsSubstring`, produces an empty `required_substr`, then the `Searcher` path sets result to `0` for an empty haystack (`search` retu... | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "`constantVector` is missing the trivial-pattern handling that `vectorConstant` already has for `needle == '%'` / `needle == '%%'`.\n\nWith the current code, `'' LIKE '%%'` enters `likePatternIsSubstring`, produces an empty `required_substr`, then the `Searcher` path sets result to `0` for an empty... | true | ||
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | `constantVector` now short-circuits `needle == '%' || needle == '%%'`, but the same `LIKE` edge case is still live in `vectorVector` (and similarly in `vectorFixedVector`).
Both paths still go through `likePatternIsSubstring` for `needle == '%%'`, which produces an empty `required_substr`; for an empty haystack row th... | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "`constantVector` now short-circuits `needle == '%' || needle == '%%'`, but the same `LIKE` edge case is still live in `vectorVector` (and similarly in `vectorFixedVector`).\n\nBoth paths still go through `likePatternIsSubstring` for `needle == '%%'`, which produces an empty `required_substr`; for ... | true | ||
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | Cannot we just check if the needle contains only `%`? Then even `'%%%%%%'` would be trivial. | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "Cannot we just check if the needle contains only `%`? Then even `'%%%%%%'` would be trivial.",
"path": "src/Functions/MatchImpl.h",
"hunk": "@@ -494,7 +489,15 @@ struct MatchImpl\n reinterpret_cast<const char *>(cur_needle_data),\n cur_needle_length)... | true | ||
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | Cannot we just use a `string_view` over the needle? I don't see why would we need to copy the needle.
```suggestion
const auto & needle = std::string_view(
``` | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "Cannot we just use a `string_view` over the needle? I don't see why would we need to copy the needle. \n```suggestion\n const auto & needle = std::string_view(\n```",
"path": "src/Functions/MatchImpl.h",
"hunk": "@@ -668,10 +679,115 @@ struct MatchImpl\n }\n }\n \n- ... | true | ||
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | This is the third case where this logic is written. Let's unify it so it will be in sync. | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "This is the third case where this logic is written. Let's unify it so it will be in sync.",
"path": "src/Functions/MatchImpl.h",
"hunk": "@@ -668,10 +679,115 @@ struct MatchImpl\n }\n }\n \n- template <typename... Args>\n- static void constantVector(Args &&...)\n+ stat... | true | ||
ClickHouse/ClickHouse | 100,479 | comment_to_fix | Support constant haystack with non-constant needle in LIKE/ILIKE/match | Since this path now accepts a non-constant needle for `match` as well, the source docs should stop advertising the second argument as `const String`: `src/Functions/match.cpp` still has `{"pattern", ..., {"const String"}}`. Otherwise generated docs will continue to tell users that the newly supported `match('foo', patt... | bf2051399e17d9e2b620d7d2c02568ff125011a6 | 38d0e3535d5d16ae471b8f753284fd142a5ca3cb | diff --git a/src/Functions/MatchImpl.h b/src/Functions/MatchImpl.h
index a35e6a7f4650..26f47a3b63ec 100644
--- a/src/Functions/MatchImpl.h
+++ b/src/Functions/MatchImpl.h
@@ -11,13 +11,14 @@
namespace DB
{
-namespace ErrorCodes
+namespace impl
{
- extern const int ILLEGAL_COLUMN;
-}
-namespace impl
+/// Does ... | [
"src/Functions/MatchImpl.h"
] | [
{
"comment": "Since this path now accepts a non-constant needle for `match` as well, the source docs should stop advertising the second argument as `const String`: `src/Functions/match.cpp` still has `{\"pattern\", ..., {\"const String\"}}`. Otherwise generated docs will continue to tell users that the newly su... | true | ||
ClickHouse/ClickHouse | 105,543 | issue_to_patch | `variant_throw_on_type_mismatch=false` fails for comparison functions: `try_build` only wraps `build`, not `execute`
_Found via ClickGap automated review. Please close or comment if this is incorrect or needs adjustment._
_Retrospective finding from a historical scan of [PR #95811](https://github.com/ClickHouse/Click... | Fix `variant_throw_on_type_mismatch/dynamic_throw_on_type_mismatch=false` not catching exceptions during function execution |
Closes https://github.com/ClickHouse/ClickHouse/issues/103484
### Changelog category (leave one):
- Bug Fix (user-visible misbehavior in an official stable release)
### Changelog entry (a [user-readable short description](https://github.com/ClickHouse/ClickHouse/blob/master/docs/changelog_entry_guidelines.md... | ef937ef552415585853632049b17591e0c7c37f4 | 41b943bafc47fe9e7b467ca712c94a6eccd9270b | diff --git a/src/Functions/FunctionDynamicAdaptor.cpp b/src/Functions/FunctionDynamicAdaptor.cpp
index 8d0400a3e2a1..7cd00eddb9c5 100644
--- a/src/Functions/FunctionDynamicAdaptor.cpp
+++ b/src/Functions/FunctionDynamicAdaptor.cpp
@@ -85,6 +85,33 @@ ColumnPtr ExecutableFunctionDynamicAdaptor::executeImpl(const ColumnsW... | [
"src/Functions/FunctionDynamicAdaptor.cpp",
"src/Functions/FunctionVariantAdaptor.cpp",
"tests/queries/0_stateless/04264_variant_throw_on_type_mismatch_execute.reference",
"tests/queries/0_stateless/04264_variant_throw_on_type_mismatch_execute.sql",
"tests/queries/0_stateless/04264_variant_throw_on_type_mis... | [
{
"comment": "This catch is now around `execute`, not just overload resolution, so `CANNOT_CONVERT_TYPE` is too broad here. For example, `accurateCast(v, 'UInt8')` on a `Variant(UInt16, String)` value `300` can build successfully and then throw `CANNOT_CONVERT_TYPE` from the per-value overflow check during exec... | diff --git a/tests/queries/0_stateless/04264_variant_throw_on_type_mismatch_execute.reference b/tests/queries/0_stateless/04264_variant_throw_on_type_mismatch_execute.reference
new file mode 100644
index 000000000000..c220e7d9b316
--- /dev/null
+++ b/tests/queries/0_stateless/04264_variant_throw_on_type_mismatch_execut... | true |
ClickHouse/ClickHouse | 105,543 | comment_to_fix | Fix `variant_throw_on_type_mismatch/dynamic_throw_on_type_mismatch=false` not catching exceptions during function execution | This catch is now around `execute`, not just overload resolution, so `CANNOT_CONVERT_TYPE` is too broad here. For example, `accurateCast(v, 'UInt8')` on a `Variant(UInt16, String)` value `300` can build successfully and then throw `CANNOT_CONVERT_TYPE` from the per-value overflow check during execution; with `variant_t... | ef937ef552415585853632049b17591e0c7c37f4 | 41b943bafc47fe9e7b467ca712c94a6eccd9270b | diff --git a/src/Functions/FunctionVariantAdaptor.cpp b/src/Functions/FunctionVariantAdaptor.cpp
index 8797545fcc24..f5fd599ab77e 100644
--- a/src/Functions/FunctionVariantAdaptor.cpp
+++ b/src/Functions/FunctionVariantAdaptor.cpp
@@ -90,6 +90,33 @@ ColumnPtr ExecutableFunctionVariantAdaptor::executeImpl(
}
... | [
"src/Functions/FunctionVariantAdaptor.cpp"
] | [
{
"comment": "This catch is now around `execute`, not just overload resolution, so `CANNOT_CONVERT_TYPE` is too broad here. For example, `accurateCast(v, 'UInt8')` on a `Variant(UInt16, String)` value `300` can build successfully and then throw `CANNOT_CONVERT_TYPE` from the per-value overflow check during exec... | true | ||
ClickHouse/ClickHouse | 107,389 | issue_to_patch | Fix `arrayResize` with a `Decimal` size | <!--
Related: https://github.com/ClickHouse/ClickHouse/pull/107319
-->
`arrayResize` read its size argument via `IColumn::getInt`, which for a `Decimal` column returns the raw unscaled value rather than the logical one. As a result `arrayResize([1, 2, 3], 1.5::Decimal(2, 1))` resized to 15 elements instead of 1, incon... | a09d30d9e0c46faad1b63a07583ec10f1501397f | b0af15f8c38ffd32582a0a931f6fcab8b5b6baec | diff --git a/src/Functions/array/arrayResize.cpp b/src/Functions/array/arrayResize.cpp
index bbf0f2f18472..8513b4444295 100644
--- a/src/Functions/array/arrayResize.cpp
+++ b/src/Functions/array/arrayResize.cpp
@@ -4,6 +4,7 @@
#include <Functions/GatherUtils/GatherUtils.h>
#include <DataTypes/DataTypeArray.h>
#inclu... | [
"src/Functions/array/arrayResize.cpp",
"src/Functions/tests/gtest_functions_stress.cpp",
"tests/queries/0_stateless/04327_array_resize_decimal_size.reference",
"tests/queries/0_stateless/04327_array_resize_decimal_size.sql"
] | [] | diff --git a/src/Functions/tests/gtest_functions_stress.cpp b/src/Functions/tests/gtest_functions_stress.cpp
index e594c7124f71..bbc2accb6d41 100644
--- a/src/Functions/tests/gtest_functions_stress.cpp
+++ b/src/Functions/tests/gtest_functions_stress.cpp
@@ -1565,16 +1565,26 @@ struct FunctionsStressTestThread
... | true | |
ClickHouse/ClickHouse | 107,430 | issue_to_patch | Do not rewrite NULL-preserving aggregates with optimize_rewrite_aggregate_function_with_if | <!--
Linked issues and pull requests. Use full GitHub URLs, one relationship per line; delete the lines you don't need.
Closes: https://github.com/ClickHouse/ClickHouse/issues/107421
-->
### Changelog category (leave one):
- Bug Fix (user-visible misbehavior in an official stable release)
### Changelog entry (a [u... | 0303010ba9899c1667080ad1b98daeaf326ab2aa | fd2eda719fe3db1cbec94af2409829dcfe8b0a9c | diff --git a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
index 4e1dec0a9b27..bbc16d0571f0 100644
--- a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
+++ b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
@@ -2,8 +2,10 @@
... | [
"src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp",
"tests/queries/0_stateless/04337_rewrite_aggregate_function_with_if_respect_nulls.reference",
"tests/queries/0_stateless/04337_rewrite_aggregate_function_with_if_respect_nulls.sql"
] | [
{
"comment": "And what if it has a custom Null Adapter, which is non-default?",
"path": "src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp",
"hunk": "@@ -44,6 +45,21 @@ class RewriteAggregateFunctionWithIfVisitor : public InDepthQueryTreeVisitorWith\n if (lower_name.ends_with(\"if\"))\n... | diff --git a/tests/queries/0_stateless/04337_rewrite_aggregate_function_with_if_respect_nulls.reference b/tests/queries/0_stateless/04337_rewrite_aggregate_function_with_if_respect_nulls.reference
new file mode 100644
index 000000000000..cfe2bffaba38
--- /dev/null
+++ b/tests/queries/0_stateless/04337_rewrite_aggregate... | true | |
ClickHouse/ClickHouse | 107,430 | comment_to_fix | Do not rewrite NULL-preserving aggregates with optimize_rewrite_aggregate_function_with_if | This guard runs before the pass has proved that the aggregate has one `if` argument. That makes ordinary zero-argument `count` hit `AggregateFunctionCount::getOwnNullAdapter` with an empty `DataTypes`; the implementation indexes `types[0]`. `NormalizeCountVariantsPass` runs earlier and also rewrites `count(1)` / `sum(1... | 0303010ba9899c1667080ad1b98daeaf326ab2aa | fd2eda719fe3db1cbec94af2409829dcfe8b0a9c | diff --git a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
index 4e1dec0a9b27..bbc16d0571f0 100644
--- a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
+++ b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
@@ -2,8 +2,10 @@
... | [
"src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp"
] | [
{
"comment": "This guard runs before the pass has proved that the aggregate has one `if` argument. That makes ordinary zero-argument `count` hit `AggregateFunctionCount::getOwnNullAdapter` with an empty `DataTypes`; the implementation indexes `types[0]`. `NormalizeCountVariantsPass` runs earlier and also rewrit... | true | ||
ClickHouse/ClickHouse | 107,430 | comment_to_fix | Do not rewrite NULL-preserving aggregates with optimize_rewrite_aggregate_function_with_if | The guard only checks the outer aggregate function, so wrappers around a NULL-preserving aggregate still get rewritten. For example, `first_value_respect_nullsOrNull` is supported by `AggregateFunctionOrFill`, but that wrapper's `getOwnNullAdapter` returns `nullptr`, so `first_value_respect_nullsOrNull(if(number % 2 = ... | 0303010ba9899c1667080ad1b98daeaf326ab2aa | fd2eda719fe3db1cbec94af2409829dcfe8b0a9c | diff --git a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
index 4e1dec0a9b27..bbc16d0571f0 100644
--- a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
+++ b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
@@ -2,8 +2,10 @@
... | [
"src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp"
] | [
{
"comment": "The guard only checks the outer aggregate function, so wrappers around a NULL-preserving aggregate still get rewritten. For example, `first_value_respect_nullsOrNull` is supported by `AggregateFunctionOrFill`, but that wrapper's `getOwnNullAdapter` returns `nullptr`, so `first_value_respect_nullsO... | true | ||
ClickHouse/ClickHouse | 107,430 | comment_to_fix | Do not rewrite NULL-preserving aggregates with optimize_rewrite_aggregate_function_with_if | This early return still lets the rewrite fire for types that can contain `NULL` without being `Nullable`, notably `Variant` and `Dynamic`. For example, `Variant` has a `NULL` discriminator and `canContainNull` explicitly treats it as nullable-capable, but `makeNullableSafe(Variant(...))` returns the original non-Nullab... | 0303010ba9899c1667080ad1b98daeaf326ab2aa | fd2eda719fe3db1cbec94af2409829dcfe8b0a9c | diff --git a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
index 4e1dec0a9b27..bbc16d0571f0 100644
--- a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
+++ b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
@@ -2,8 +2,10 @@
... | [
"src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp"
] | [
{
"comment": "This early return still lets the rewrite fire for types that can contain `NULL` without being `Nullable`, notably `Variant` and `Dynamic`. For example, `Variant` has a `NULL` discriminator and `canContainNull` explicitly treats it as nullable-capable, but `makeNullableSafe(Variant(...))` returns t... | true | ||
ClickHouse/ClickHouse | 107,430 | comment_to_fix | Do not rewrite NULL-preserving aggregates with optimize_rewrite_aggregate_function_with_if | This still lets the rewrite change results for `Variant`/`Dynamic` arguments whose `NULL` is carried by the type discriminator. The probe here asks what the aggregate would do if the argument were `Nullable`, but `AggregateFunctionFactory` only applies the internal `Null` combinator when `type->isNullable()`. For a non... | 0303010ba9899c1667080ad1b98daeaf326ab2aa | fd2eda719fe3db1cbec94af2409829dcfe8b0a9c | diff --git a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
index 4e1dec0a9b27..bbc16d0571f0 100644
--- a/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
+++ b/src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp
@@ -2,8 +2,10 @@
... | [
"src/Analyzer/Passes/RewriteAggregateFunctionWithIfPass.cpp"
] | [
{
"comment": "This still lets the rewrite change results for `Variant`/`Dynamic` arguments whose `NULL` is carried by the type discriminator. The probe here asks what the aggregate would do if the argument were `Nullable`, but `AggregateFunctionFactory` only applies the internal `Null` combinator when `type->is... | true | ||
ClickHouse/ClickHouse | 107,426 | issue_to_patch | Creating materialized view over MaterializedPostgreSQL results in DB::Exception: Too large size passed to allocator.
Hi, I am trying to set up replication from Postgres to ClickHouse, and create some materialized view (call it `my_view`) on top of mirrored data.
Replication (MaterializedPostgreSQL) without `my_view`... | Add regression test for materialized view over MaterializedPostgreSQL (#39805) | <!--
Closes: https://github.com/ClickHouse/ClickHouse/issues/39805
Related: https://github.com/ClickHouse/ClickHouse/issues/49045
-->
[#39805](https://github.com/ClickHouse/ClickHouse/issues/39805) reported that creating a materialized view (an aggregation: total orders per customer) on top of a `MaterializedPostgreSQ... | bf2051399e17d9e2b620d7d2c02568ff125011a6 | f59a07b28557024983de71e756070386139fd5e5 | [
"tests/integration/test_postgresql_replica_database_engine/test_3.py"
] | [] | diff --git a/tests/integration/test_postgresql_replica_database_engine/test_3.py b/tests/integration/test_postgresql_replica_database_engine/test_3.py
index 34dd16499005..b8895f1d182e 100644
--- a/tests/integration/test_postgresql_replica_database_engine/test_3.py
+++ b/tests/integration/test_postgresql_replica_databas... | true | |
ClickHouse/ClickHouse | 107,387 | issue_to_patch | Revert "Revert "Fix session_timezone ignored on LowCardinality(DateTime)"" | Reverts ClickHouse/ClickHouse#107315 | 19059120f5e793db617f4370f6493365992fe4df | 8b605e0806f52e43e61ca75c35d88903e1e508f1 | diff --git a/src/DataTypes/Serializations/SerializationLowCardinality.cpp b/src/DataTypes/Serializations/SerializationLowCardinality.cpp
index cb47cc7537d9..5dcdbb2c55c4 100644
--- a/src/DataTypes/Serializations/SerializationLowCardinality.cpp
+++ b/src/DataTypes/Serializations/SerializationLowCardinality.cpp
@@ -44,11... | [
"src/DataTypes/Serializations/SerializationLowCardinality.cpp",
"tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.reference",
"tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.sh"
] | [
{
"comment": "This restores the same shared-server shape that made `#106634` flaky and led to `#107315`: the test warms and checks the process-global `SerializationObjectPool` through `${CLICKHOUSE_CLIENT}`, so other parallel stateless tests using the same server can affect which `LowCardinality(DateTime)` seri... | diff --git a/tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.reference b/tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.reference
new file mode 100644
index 000000000000..99524392ed97
--- /dev/null
+++ b/tests/queries/0_statele... | true | |
ClickHouse/ClickHouse | 107,387 | comment_to_fix | Revert "Revert "Fix session_timezone ignored on LowCardinality(DateTime)"" | This restores the same shared-server shape that made `#106634` flaky and led to `#107315`: the test warms and checks the process-global `SerializationObjectPool` through `${CLICKHOUSE_CLIENT}`, so other parallel stateless tests using the same server can affect which `LowCardinality(DateTime)` serialization state exists... | 19059120f5e793db617f4370f6493365992fe4df | 8b605e0806f52e43e61ca75c35d88903e1e508f1 | diff --git a/tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.sh b/tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.sh
new file mode 100755
index 000000000000..6eb69bf9f177
--- /dev/null
+++ b/tests/queries/0_stateless/04322_10647... | [
"tests/queries/0_stateless/04322_106474_session_timezone_lowcardinality_datetime_csv_roundtrip.sh"
] | [
{
"comment": "This restores the same shared-server shape that made `#106634` flaky and led to `#107315`: the test warms and checks the process-global `SerializationObjectPool` through `${CLICKHOUSE_CLIENT}`, so other parallel stateless tests using the same server can affect which `LowCardinality(DateTime)` seri... | true | ||
ClickHouse/ClickHouse | 107,124 | issue_to_patch | Preserve constant columns of `ActionsDAG` input nodes in serialization | `ActionsDAG::serialize` skipped the column for all `INPUT` nodes:
```cpp
const bool has_column = (node.type != ActionType::INPUT && node.column != nullptr);
```
However, `INPUT` nodes can legitimately carry a constant column: lambda capture DAGs reference constants from the enclosing DAG as inputs with a pre-set cons... | 0f3fdcf807d5dc4e712c66d6cca0fdc30ad4a5a3 | c0aeba1961af0305083fca9cf933476eee3536a6 | diff --git a/src/Interpreters/ActionsDAG.cpp b/src/Interpreters/ActionsDAG.cpp
index a88b14751d5b..ce7921950bc8 100644
--- a/src/Interpreters/ActionsDAG.cpp
+++ b/src/Interpreters/ActionsDAG.cpp
@@ -3886,8 +3886,13 @@ void ActionsDAG::serialize(WriteBuffer & out, SerializedSetsRegistry & registry)
writeVar... | [
"src/Interpreters/ActionsDAG.cpp",
"tests/queries/0_stateless/04338_actions_dag_input_constant_serialization.reference",
"tests/queries/0_stateless/04338_actions_dag_input_constant_serialization.sql"
] | [] | diff --git a/tests/queries/0_stateless/04338_actions_dag_input_constant_serialization.reference b/tests/queries/0_stateless/04338_actions_dag_input_constant_serialization.reference
new file mode 100644
index 000000000000..5a2682b30b06
--- /dev/null
+++ b/tests/queries/0_stateless/04338_actions_dag_input_constant_serial... | true | |
ClickHouse/ClickHouse | 74,691 | issue_to_patch | Add `FORMAT PNG` support | <!---
A technical comment, you are free to remove or leave it as it is when PR is created
The following categories are used in the next scripts, update them accordingly
utils/changelog/changelog.py
tests/ci/cancel_and_rerun_workflow_lambda/app.py
-->
### Changelog category (leave one):
- New Feature
### Chang... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/.gitmodules b/.gitmodules
index 038e3ac05c35..33373dbc389e 100644
--- a/.gitmodules
+++ b/.gitmodules
@@ -357,6 +357,9 @@
[submodule "contrib/SHA3IUF"]
path = contrib/SHA3IUF
url = https://github.com/brainhub/SHA3IUF.git
+[submodule "contrib/libpng"]
+ path = contrib/libpng
+ url = https://github.com/... | [
".gitmodules",
"ci/jobs/fast_test.py",
"contrib/CMakeLists.txt",
"contrib/libpng",
"contrib/libpng-cmake/CMakeLists.txt",
"docs/en/interfaces/formats/PNG.md",
"src/CMakeLists.txt",
"src/Common/config.h.in",
"src/Core/FormatFactorySettings.h",
"src/Core/SettingsChangesHistory.cpp",
"src/Formats/F... | [
{
"comment": "Not making libpng dependent on ENABLE_LIBRARIES will compile it unconditionally in all builds. This isn't good, FastTest does a minimal build with as few as possible library dependencies. Let's restore l. 1-6.",
"path": "contrib/libpng-cmake/CMakeLists.txt",
"hunk": "@@ -0,0 +1,38 @@\n+# o... | diff --git a/ci/jobs/fast_test.py b/ci/jobs/fast_test.py
index 30821f31c019..4332de09e854 100644
--- a/ci/jobs/fast_test.py
+++ b/ci/jobs/fast_test.py
@@ -65,6 +65,7 @@ def clone_submodules():
"contrib/StringZilla",
"contrib/rust_vendor",
"contrib/clickstack",
+ "contrib/libpng",
... | true | |
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Not making libpng dependent on ENABLE_LIBRARIES will compile it unconditionally in all builds. This isn't good, FastTest does a minimal build with as few as possible library dependencies. Let's restore l. 1-6. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/contrib/libpng-cmake/CMakeLists.txt b/contrib/libpng-cmake/CMakeLists.txt
new file mode 100644
index 000000000000..08a1cf8b5687
--- /dev/null
+++ b/contrib/libpng-cmake/CMakeLists.txt
@@ -0,0 +1,57 @@
+option(ENABLE_LIBPNG "Enable libpng" ${ENABLE_LIBRARIES})
+
+if (NOT ENABLE_LIBPNG)
+ message(STATUS "... | [
"contrib/libpng-cmake/CMakeLists.txt"
] | [
{
"comment": "Not making libpng dependent on ENABLE_LIBRARIES will compile it unconditionally in all builds. This isn't good, FastTest does a minimal build with as few as possible library dependencies. Let's restore l. 1-6.",
"path": "contrib/libpng-cmake/CMakeLists.txt",
"hunk": "@@ -0,0 +1,38 @@\n+# o... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | We name these settings `output_format_...` everywhere, so here it should be `output_format_png_...`, not `output_png_...` | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Core/FormatFactorySettings.h b/src/Core/FormatFactorySettings.h
index a43e2d1e78b5..b93a5cb50024 100644
--- a/src/Core/FormatFactorySettings.h
+++ b/src/Core/FormatFactorySettings.h
@@ -1540,6 +1540,28 @@ Set the quoting rule for identifiers in SHOW CREATE query
)", 0) \
DECLARE(IdentifierQuoting... | [
"src/Core/FormatFactorySettings.h"
] | [
{
"comment": "We name these settings `output_format_...` everywhere, so here it should be `output_format_png_...`, not `output_png_...`",
"path": "src/Core/FormatFactorySettings.h",
"hunk": "@@ -1311,6 +1311,27 @@ Set the quoting rule for identifiers in SHOW CREATE query\n )\", 0) \\\n DECLARE(Ident... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | The documentation (here) should list the possible options. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Core/FormatFactorySettings.h b/src/Core/FormatFactorySettings.h
index a43e2d1e78b5..b93a5cb50024 100644
--- a/src/Core/FormatFactorySettings.h
+++ b/src/Core/FormatFactorySettings.h
@@ -1540,6 +1540,28 @@ Set the quoting rule for identifiers in SHOW CREATE query
)", 0) \
DECLARE(IdentifierQuoting... | [
"src/Core/FormatFactorySettings.h"
] | [
{
"comment": "The documentation (here) should list the possible options.",
"path": "src/Core/FormatFactorySettings.h",
"hunk": "@@ -1311,6 +1311,27 @@ Set the quoting rule for identifiers in SHOW CREATE query\n )\", 0) \\\n DECLARE(IdentifierQuotingStyle, show_create_query_identifier_quoting_style, ... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | ### PNG format missing TTY-unfriendly and content-type registration
**Medium Severity**
<!-- DESCRIPTION START -->
The `registerOutputFormatPNG` function doesn't call `factory.markOutputFormatNotTTYFriendly("PNG")` or `factory.setContentType("PNG", "image/png")`. Every other binary output format (Arrow, Avro, BSON, N... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Processors/Formats/Impl/PNGOutputFormat.cpp b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
new file mode 100644
index 000000000000..8e3f6c8dbf7c
--- /dev/null
+++ b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
@@ -0,0 +1,110 @@
+#include <Processors/Formats/Impl/PNGOutputFormat.h>
+
+#if USE_LIBP... | [
"src/Processors/Formats/Impl/PNGOutputFormat.cpp"
] | [
{
"comment": "### PNG format missing TTY-unfriendly and content-type registration\n\n**Medium Severity**\n\n<!-- DESCRIPTION START -->\nThe `registerOutputFormatPNG` function doesn't call `factory.markOutputFormatNotTTYFriendly(\"PNG\")` or `factory.setContentType(\"PNG\", \"image/png\")`. Every other binary ou... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | ### Unchecked overflow in pixel buffer size computation
**Medium Severity**
<!-- DESCRIPTION START -->
`pixels.assign(channels * width * height, 0)` has no overflow guard. `width` and `height` are `UInt64` settings with no upper bound beyond zero. If their product overflows `size_t`, the buffer will be undersized and... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "### Unchecked overflow in pixel buffer size computation\n\n**Medium Severity**\n\n<!-- DESCRIPTION START -->\n`pixels.assign(channels * width * height, 0)` has no overflow guard. `width` and `height` are `UInt64` settings with no upper bound beyond zero. If their product overflows `size_t`, the bu... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Move to 26.6. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Core/SettingsChangesHistory.cpp b/src/Core/SettingsChangesHistory.cpp
index 01fadf7eb3cb..c63e8102f026 100644
--- a/src/Core/SettingsChangesHistory.cpp
+++ b/src/Core/SettingsChangesHistory.cpp
@@ -41,6 +41,9 @@ const VersionToSettingsChangesMap & getSettingsChangesHistory()
/// Note: please c... | [
"src/Core/SettingsChangesHistory.cpp"
] | [
{
"comment": "Move to 26.6.",
"path": "src/Core/SettingsChangesHistory.cpp",
"hunk": "@@ -41,6 +41,8 @@ const VersionToSettingsChangesMap & getSettingsChangesHistory()\n /// Note: please check if the key already exists to prevent duplicate entries.\n addSettingsChanges(settings_changes_h... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Why not in the constructor? | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGWriter.h b/src/Formats/PNGWriter.h
new file mode 100644
index 000000000000..7c741c182d29
--- /dev/null
+++ b/src/Formats/PNGWriter.h
@@ -0,0 +1,75 @@
+#pragma once
+
+#include "config.h"
+
+#if USE_LIBPNG
+
+#include <exception>
+#include <string>
+
+#include <png.h>
+#include <boost/noncopy... | [
"src/Formats/PNGWriter.h"
] | [
{
"comment": "Why not in the constructor?",
"path": "src/Formats/PNGWriter.h",
"hunk": "@@ -0,0 +1,54 @@\n+#pragma once\n+\n+#include <png.h>\n+#include <boost/noncopyable.hpp>\n+\n+#include <Common/Exception.h>\n+\n+namespace DB\n+{\n+\n+class WriteBuffer;\n+\n+/** Thin libpng wrapper that writes an 8-... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Just `PNG`. We don't include the full class name. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Processors/Formats/Impl/PNGOutputFormat.h b/src/Processors/Formats/Impl/PNGOutputFormat.h
new file mode 100644
index 000000000000..8e2105d17a27
--- /dev/null
+++ b/src/Processors/Formats/Impl/PNGOutputFormat.h
@@ -0,0 +1,36 @@
+#pragma once
+
+#include "config.h"
+
+#if USE_LIBPNG
+
+#include <Core/Blo... | [
"src/Processors/Formats/Impl/PNGOutputFormat.h"
] | [
{
"comment": "Just `PNG`. We don't include the full class name.",
"path": "src/Processors/Formats/Impl/PNGOutputFormat.h",
"hunk": "@@ -0,0 +1,30 @@\n+#pragma once\n+\n+#include <Core/Block_fwd.h>\n+#include <Formats/FormatSettings.h>\n+#include <Processors/Chunk.h>\n+#include <Processors/Formats/IOutpu... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | I don't like it. When do we have warnings? If we don't have them right now, make every warning an exception in ClickHouse. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGWriter.cpp b/src/Formats/PNGWriter.cpp
new file mode 100644
index 000000000000..2ba8fd736e4b
--- /dev/null
+++ b/src/Formats/PNGWriter.cpp
@@ -0,0 +1,169 @@
+#include <Formats/PNGWriter.h>
+
+#if USE_LIBPNG
+
+#include <png.h>
+
+#include <algorithm>
+#include <cstring>
+
+#include <IO/Write... | [
"src/Formats/PNGWriter.cpp"
] | [
{
"comment": "I don't like it. When do we have warnings? If we don't have them right now, make every warning an exception in ClickHouse.",
"path": "src/Formats/PNGWriter.cpp",
"hunk": "@@ -0,0 +1,141 @@\n+#include <Formats/PNGWriter.h>\n+\n+#include <png.h>\n+#include <vector>\n+\n+#include <IO/WriteBuf... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Just to clarify - is throwing through C ok? | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGWriter.cpp b/src/Formats/PNGWriter.cpp
new file mode 100644
index 000000000000..2ba8fd736e4b
--- /dev/null
+++ b/src/Formats/PNGWriter.cpp
@@ -0,0 +1,169 @@
+#include <Formats/PNGWriter.h>
+
+#if USE_LIBPNG
+
+#include <png.h>
+
+#include <algorithm>
+#include <cstring>
+
+#include <IO/Write... | [
"src/Formats/PNGWriter.cpp"
] | [
{
"comment": "Just to clarify - is throwing through C ok?",
"path": "src/Formats/PNGWriter.cpp",
"hunk": "@@ -0,0 +1,141 @@\n+#include <Formats/PNGWriter.h>\n+\n+#include <png.h>\n+#include <vector>\n+\n+#include <IO/WriteBuffer.h>\n+#include <Common/Exception.h>\n+\n+namespace DB\n+{\n+\n+namespace Err... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | This is suboptimal. Do all dispatch outside of the loop, so it is not done for each row again and again. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "This is suboptimal. Do all dispatch outside of the loop, so it is not done for each row again and again.",
"path": "src/Formats/PNGSerializer.cpp",
"hunk": "@@ -0,0 +1,417 @@\n+#include <Formats/PNGSerializer.h>\n+\n+#include <cstring>\n+#include <algorithm>\n+#include <optional>\n+\n+#inc... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Use ClickHouse functions from NaNUtils. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "Use ClickHouse functions from NaNUtils.",
"path": "src/Formats/PNGSerializer.cpp",
"hunk": "@@ -0,0 +1,417 @@\n+#include <Formats/PNGSerializer.h>\n+\n+#include <cstring>\n+#include <algorithm>\n+#include <optional>\n+\n+#include <Columns/ColumnNullable.h>\n+#include <DataTypes/IDataType.h... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Use ClickHouse functions from ASCIIUtils | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "Use ClickHouse functions from ASCIIUtils",
"path": "src/Formats/PNGSerializer.cpp",
"hunk": "@@ -0,0 +1,417 @@\n+#include <Formats/PNGSerializer.h>\n+\n+#include <cstring>\n+#include <algorithm>\n+#include <optional>\n+\n+#include <Columns/ColumnNullable.h>\n+#include <DataTypes/IDataType.... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | Suboptimal. Calculate x and y incrementally. | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "Suboptimal. Calculate x and y incrementally.",
"path": "src/Formats/PNGSerializer.cpp",
"hunk": "@@ -0,0 +1,417 @@\n+#include <Formats/PNGSerializer.h>\n+\n+#include <cstring>\n+#include <algorithm>\n+#include <optional>\n+\n+#include <Columns/ColumnNullable.h>\n+#include <DataTypes/IDataT... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `PNG` produces a complete PNG datastream, but the format is still treated as appendable because registration only marks it TTY-unfriendly and sets the content type. `StorageFile` relies on `FormatFactory::checkIfFormatSupportAppend` before appending to an existing file; with the default append support, a second insert ... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Processors/Formats/Impl/PNGOutputFormat.cpp b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
new file mode 100644
index 000000000000..8e3f6c8dbf7c
--- /dev/null
+++ b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
@@ -0,0 +1,110 @@
+#include <Processors/Formats/Impl/PNGOutputFormat.h>
+
+#if USE_LIBP... | [
"src/Processors/Formats/Impl/PNGOutputFormat.cpp"
] | [
{
"comment": "`PNG` produces a complete PNG datastream, but the format is still treated as appendable because registration only marks it TTY-unfriendly and sets the content type. `StorageFile` relies on `FormatFactory::checkIfFormatSupportAppend` before appending to an existing file; with the default append sup... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `output_format_image_width` and `output_format_image_height` are `UInt64`, but `png_set_IHDR` receives 32-bit dimensions here. Values above the PNG/libpng limit are truncated by `static_cast<png_uint_32>` after the serializer has already allocated the larger buffer, so `output_format_image_width = 4294967297` can be en... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGWriter.cpp b/src/Formats/PNGWriter.cpp
new file mode 100644
index 000000000000..2ba8fd736e4b
--- /dev/null
+++ b/src/Formats/PNGWriter.cpp
@@ -0,0 +1,169 @@
+#include <Formats/PNGWriter.h>
+
+#if USE_LIBPNG
+
+#include <png.h>
+
+#include <algorithm>
+#include <cstring>
+
+#include <IO/Write... | [
"src/Formats/PNGWriter.cpp"
] | [
{
"comment": "`output_format_image_width` and `output_format_image_height` are `UInt64`, but `png_set_IHDR` receives 32-bit dimensions here. Values above the PNG/libpng limit are truncated by `static_cast<png_uint_32>` after the serializer has already allocated the larger buffer, so `output_format_image_width =... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | The test only verifies the `IHDR` metadata, so a regression that writes every pixel as zero, ignores explicit `x`/`y`, or breaks clamping/float scaling would still pass. Since `FORMAT PNG`'s contract is mostly in the pixel mapping, please add a small decoder/check that validates a few pixel values for implicit `RGB`, e... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/tests/queries/0_stateless/04326_png_format.sh b/tests/queries/0_stateless/04326_png_format.sh
new file mode 100755
index 000000000000..4e90aa2de9dd
--- /dev/null
+++ b/tests/queries/0_stateless/04326_png_format.sh
@@ -0,0 +1,194 @@
+#!/usr/bin/env bash
+
+CUR_DIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pw... | [
"tests/queries/0_stateless/04326_png_format.sh"
] | [
{
"comment": "The test only verifies the `IHDR` metadata, so a regression that writes every pixel as zero, ignores explicit `x`/`y`, or breaks clamping/float scaling would still pass. Since `FORMAT PNG`'s contract is mostly in the pixel mapping, please add a small decoder/check that validates a few pixel values... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | This new format page is missing the frontmatter keys that the other `docs/en/interfaces/formats/*.md` pages use for routing and reference metadata. Please add at least `slug: /interfaces/formats/PNG` and `doc_type: 'reference'` next to `title`; right now `PNG.md` is the only format page without them.
| 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/docs/en/interfaces/formats/PNG.md b/docs/en/interfaces/formats/PNG.md
new file mode 100644
index 000000000000..41c18bd7f67f
--- /dev/null
+++ b/docs/en/interfaces/formats/PNG.md
@@ -0,0 +1,110 @@
+---
+alias: []
+description: 'Documentation for the PNG image output format'
+input_format: false
+keywords: [... | [
"docs/en/interfaces/formats/PNG.md"
] | [
{
"comment": "This new format page is missing the frontmatter keys that the other `docs/en/interfaces/formats/*.md` pages use for routing and reference metadata. Please add at least `slug: /interfaces/formats/PNG` and `doc_type: 'reference'` next to `title`; right now `PNG.md` is the only format page without th... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | New docs files under `docs/` must include the standard frontmatter keys `description`, `sidebar_label`, `sidebar_position`, `slug`, `title`, and `doc_type`. This page now has `description`, `slug`, `title`, and `doc_type`, but it still lacks `sidebar_label` and `sidebar_position`; please add them before the first headi... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/docs/en/interfaces/formats/PNG.md b/docs/en/interfaces/formats/PNG.md
new file mode 100644
index 000000000000..41c18bd7f67f
--- /dev/null
+++ b/docs/en/interfaces/formats/PNG.md
@@ -0,0 +1,110 @@
+---
+alias: []
+description: 'Documentation for the PNG image output format'
+input_format: false
+keywords: [... | [
"docs/en/interfaces/formats/PNG.md"
] | [
{
"comment": "New docs files under `docs/` must include the standard frontmatter keys `description`, `sidebar_label`, `sidebar_position`, `slug`, `title`, and `doc_type`. This page now has `description`, `slug`, `title`, and `doc_type`, but it still lacks `sidebar_label` and `sidebar_position`; please add them ... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `FORMAT PNG` should treat wrappers around supported numeric channels the same way other output paths do, but this helper only unwraps `Nullable`. A result column such as `v LowCardinality(UInt8)` or `r LowCardinality(UInt8)` is semantically a supported pixel channel and `ColumnLowCardinality` can read it through `getUI... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "`FORMAT PNG` should treat wrappers around supported numeric channels the same way other output paths do, but this helper only unwraps `Nullable`. A result column such as `v LowCardinality(UInt8)` or `r LowCardinality(UInt8)` is semantically a supported pixel channel and `ColumnLowCardinality` can ... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | This top-level `*.c` glob pulls `example.c` and `pngtest.c` into `ch_contrib::libpng`; upstream keeps those in separate tool/test targets and only puts the actual library sources into `libpng_sources`. Both extra files define `main` and add test/example-only code to the contrib target, making the build sensitive to lin... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/contrib/libpng-cmake/CMakeLists.txt b/contrib/libpng-cmake/CMakeLists.txt
new file mode 100644
index 000000000000..08a1cf8b5687
--- /dev/null
+++ b/contrib/libpng-cmake/CMakeLists.txt
@@ -0,0 +1,57 @@
+option(ENABLE_LIBPNG "Enable libpng" ${ENABLE_LIBRARIES})
+
+if (NOT ENABLE_LIBPNG)
+ message(STATUS "... | [
"contrib/libpng-cmake/CMakeLists.txt"
] | [
{
"comment": "This top-level `*.c` glob pulls `example.c` and `pngtest.c` into `ch_contrib::libpng`; upstream keeps those in separate tool/test targets and only puts the actual library sources into `libpng_sources`. Both extra files define `main` and add test/example-only code to the contrib target, making the ... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `ENABLE_LIBPNG` can leave `ch_contrib::libpng` undefined, but `PNG` is still declared/registered unconditionally and the new `PNG*` sources include `<png.h>` unconditionally. That makes `-DENABLE_LIBPNG=0` or a generic `-DENABLE_LIBRARIES=0` build still compile the `PNG` implementation without the contrib include dirs/... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/registerFormats.cpp b/src/Formats/registerFormats.cpp
index 2b6af47d6ec9..e4e09678871e 100644
--- a/src/Formats/registerFormats.cpp
+++ b/src/Formats/registerFormats.cpp
@@ -96,6 +96,7 @@ void registerOutputFormatMarkdown(FormatFactory & factory);
void registerOutputFormatPostgreSQLWire(Format... | [
"src/Formats/registerFormats.cpp"
] | [
{
"comment": "`ENABLE_LIBPNG` can leave `ch_contrib::libpng` undefined, but `PNG` is still declared/registered unconditionally and the new `PNG*` sources include `<png.h>` unconditionally. That makes `-DENABLE_LIBPNG=0` or a generic `-DENABLE_LIBRARIES=0` build still compile the `PNG` implementation without the... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `output_format_image_terminal_mode` changes `FORMAT PNG` from raw `image/png` bytes into terminal escape text, but the format metadata here is still unconditional. In the CLI path, `markOutputFormatNotTTYFriendly` still triggers the binary-output confirmation for the intended `auto`/`kitty`/`iterm`/`sixel` terminal ren... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Processors/Formats/Impl/PNGOutputFormat.cpp b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
new file mode 100644
index 000000000000..8e3f6c8dbf7c
--- /dev/null
+++ b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
@@ -0,0 +1,110 @@
+#include <Processors/Formats/Impl/PNGOutputFormat.h>
+
+#if USE_LIBP... | [
"src/Processors/Formats/Impl/PNGOutputFormat.cpp"
] | [
{
"comment": "`output_format_image_terminal_mode` changes `FORMAT PNG` from raw `image/png` bytes into terminal escape text, but the format metadata here is still unconditional. In the CLI path, `markOutputFormatNotTTYFriendly` still triggers the binary-output confirmation for the intended `auto`/`kitty`/`iterm... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | In implicit coordinate mode this silently ignores every row after `width * height` pixels have been filled. That makes `FORMAT PNG` truncate the result set without an exception, even though the documented implicit contract says each record maps to a pixel in scanline order and only explicit out-of-range coordinates are... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "In implicit coordinate mode this silently ignores every row after `width * height` pixels have been filled. That makes `FORMAT PNG` truncate the result set without an exception, even though the documented implicit contract says each record maps to a pixel in scanline order and only explicit out-of... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `output_format_image_width`/`output_format_image_height` are user-controlled, but this allocates the whole image with `std::vector` after only the PNG 31-bit bound. A query can request a huge but valid dimension such as `100000 x 100000` and force a multi-GB allocation before any rows are processed; implicit `operator ... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGSerializer.cpp b/src/Formats/PNGSerializer.cpp
new file mode 100644
index 000000000000..3e9cfae73bd1
--- /dev/null
+++ b/src/Formats/PNGSerializer.cpp
@@ -0,0 +1,489 @@
+#include <Formats/PNGSerializer.h>
+
+#include <cstring>
+#include <algorithm>
+#include <optional>
+
+#include <Columns/C... | [
"src/Formats/PNGSerializer.cpp"
] | [
{
"comment": "`output_format_image_width`/`output_format_image_height` are user-controlled, but this allocates the whole image with `std::vector` after only the PNG 31-bit bound. A query can request a huge but valid dimension such as `100000 x 100000` and force a multi-GB allocation before any rows are processe... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `PNGOutputFormat` keeps rendered pixels and the implicit coordinate cursor in this `serializer`, but the class does not override `resetFormatterImpl`. Reusable output paths such as `MessageQueueSink` call `finalize`, restart the buffer, and then call `resetFormatter` for the next produced message; with the current impl... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Processors/Formats/Impl/PNGOutputFormat.h b/src/Processors/Formats/Impl/PNGOutputFormat.h
new file mode 100644
index 000000000000..8e2105d17a27
--- /dev/null
+++ b/src/Processors/Formats/Impl/PNGOutputFormat.h
@@ -0,0 +1,36 @@
+#pragma once
+
+#include "config.h"
+
+#if USE_LIBPNG
+
+#include <Core/Blo... | [
"src/Processors/Formats/Impl/PNGOutputFormat.h"
] | [
{
"comment": "`PNGOutputFormat` keeps rendered pixels and the implicit coordinate cursor in this `serializer`, but the class does not override `resetFormatterImpl`. Reusable output paths such as `MessageQueueSink` call `finalize`, restart the buffer, and then call `resetFormatter` for the next produced message;... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `output_format_image_height` is user-controlled up to the PNG 31-bit limit, but this allocates one raw pointer per row with `std::vector`, outside the tracked ClickHouse allocator path used for the pixel buffer. For example, `output_format_image_width = 1` with a very large height can pass the tracked pixel allocation ... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Formats/PNGWriter.cpp b/src/Formats/PNGWriter.cpp
new file mode 100644
index 000000000000..2ba8fd736e4b
--- /dev/null
+++ b/src/Formats/PNGWriter.cpp
@@ -0,0 +1,169 @@
+#include <Formats/PNGWriter.h>
+
+#if USE_LIBPNG
+
+#include <png.h>
+
+#include <algorithm>
+#include <cstring>
+
+#include <IO/Write... | [
"src/Formats/PNGWriter.cpp"
] | [
{
"comment": "`output_format_image_height` is user-controlled up to the PNG 31-bit limit, but this allocates one raw pointer per row with `std::vector`, outside the tracked ClickHouse allocator path used for the pixel buffer. For example, `output_format_image_width = 1` with a very large height can pass the tra... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | `output_format_image_terminal_mode` should keep memory derived from user-controlled image dimensions tracked or bounded, but these terminal paths allocate extra buffers outside the tracked allocator after the `PODArray` pixel buffer has already passed the memory-limit check. `ITerm`/`Kitty` materialize a full encoded `... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/src/Processors/Formats/Impl/PNGOutputFormat.cpp b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
new file mode 100644
index 000000000000..8e3f6c8dbf7c
--- /dev/null
+++ b/src/Processors/Formats/Impl/PNGOutputFormat.cpp
@@ -0,0 +1,110 @@
+#include <Processors/Formats/Impl/PNGOutputFormat.h>
+
+#if USE_LIBP... | [
"src/Processors/Formats/Impl/PNGOutputFormat.cpp"
] | [
{
"comment": "`output_format_image_terminal_mode` should keep memory derived from user-controlled image dimensions tracked or bounded, but these terminal paths allocate extra buffers outside the tracked allocator after the `PODArray` pixel buffer has already passed the memory-limit check. `ITerm`/`Kitty` materi... | true | ||
ClickHouse/ClickHouse | 74,691 | comment_to_fix | Add `FORMAT PNG` support | This `no-fasttest` tag no longer matches the PR state: `ci/jobs/fast_test.py` now checks out `contrib/libpng` and configures Fast Test with `-DENABLE_LIBPNG=1`, so these `PNG` coverage tests are skipped from the fastest CI signal even though their stated blocker is gone. Please remove the `no-fasttest` tags from the ne... | 1a5e24c889955eaff7e2e29e19e0074f30bf1322 | 09b2afe7949a3181110e32075ef51c49474b53e9 | diff --git a/tests/queries/0_stateless/04326_png_format.sh b/tests/queries/0_stateless/04326_png_format.sh
new file mode 100755
index 000000000000..4e90aa2de9dd
--- /dev/null
+++ b/tests/queries/0_stateless/04326_png_format.sh
@@ -0,0 +1,194 @@
+#!/usr/bin/env bash
+
+CUR_DIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pw... | [
"tests/queries/0_stateless/04326_png_format.sh"
] | [
{
"comment": "This `no-fasttest` tag no longer matches the PR state: `ci/jobs/fast_test.py` now checks out `contrib/libpng` and configures Fast Test with `-DENABLE_LIBPNG=1`, so these `PNG` coverage tests are skipped from the fastest CI signal even though their stated blocker is gone. Please remove the `no-fast... | true | ||
ClickHouse/ClickHouse | 102,904 | issue_to_patch | Disable lazy materialization for patch parts | ### Changelog category (leave one):
- Not for changelog (changelog entry is not required)
### Changelog entry (a [user-readable short description](https://github.com/ClickHouse/ClickHouse/blob/master/docs/changelog_entry_guidelines.md) of the changes that goes into CHANGELOG.md):
Lazy materialization is not work... | 1f08686bb837d2b91e23cad6c25dad138bb6c2f3 | 315a58477683b8b311dfb7957838d89cd730afa0 | diff --git a/src/Processors/QueryPlan/Optimizations/optimizeLazyMaterialization.cpp b/src/Processors/QueryPlan/Optimizations/optimizeLazyMaterialization.cpp
index d11c77b038aa..8347d10264d5 100644
--- a/src/Processors/QueryPlan/Optimizations/optimizeLazyMaterialization.cpp
+++ b/src/Processors/QueryPlan/Optimizations/o... | [
"src/Processors/QueryPlan/Optimizations/optimizeLazyMaterialization.cpp",
"tests/queries/0_stateless/04102_lazy_materialization_with_patch_parts.reference",
"tests/queries/0_stateless/04102_lazy_materialization_with_patch_parts.sql"
] | [
{
"comment": "Typo in test comment: \"where for error\" reads oddly. Please rephrase to something like \"Test reproduces regression with error\".\n",
"path": "tests/queries/0_stateless/04102_lazy_materialization_with_patch_parts.sql",
"hunk": "@@ -0,0 +1,17 @@\n+-- Tags: no-shared-merge-tree\n+-- Test r... | diff --git a/tests/queries/0_stateless/04102_lazy_materialization_with_patch_parts.reference b/tests/queries/0_stateless/04102_lazy_materialization_with_patch_parts.reference
new file mode 100644
index 000000000000..6b605f667865
--- /dev/null
+++ b/tests/queries/0_stateless/04102_lazy_materialization_with_patch_parts.r... | true | |
ClickHouse/ClickHouse | 98,827 | issue_to_patch | Implement http header validation for DataLakeCatalog | ### Changelog category (leave one):
- Bug Fix (user-visible misbehavior in an official stable release)
### Changelog entry (a [user-readable short description](https://github.com/ClickHouse/ClickHouse/blob/master/docs/changelog_entry_guidelines.md) of the changes that goes into CHANGELOG.md):
`DataLakeCatalog`... | b0db88ca1cf6084fb5495911ec0a0ca3f53cc9f2 | 5198f3225264c04399828edefa93aec04b3ab11a | diff --git a/src/Databases/DataLake/DatabaseDataLake.cpp b/src/Databases/DataLake/DatabaseDataLake.cpp
index 4215f58cea1a..9744da4455e9 100644
--- a/src/Databases/DataLake/DatabaseDataLake.cpp
+++ b/src/Databases/DataLake/DatabaseDataLake.cpp
@@ -47,6 +47,7 @@
#include <Parsers/ASTFunction.h>
#include <Parsers/ASTDat... | [
"src/Databases/DataLake/DatabaseDataLake.cpp",
"tests/integration/test_database_iceberg/test.py",
"tests/integration/test_database_iceberg_lakekeeper_catalog/test.py",
"tests/integration/test_database_iceberg_nessie_catalog/test.py",
"tests/queries/0_stateless/03913_datalake_restful_catalog_bad_format.refer... | [
{
"comment": "### Header validation runs during ATTACH blocking server restart\n\n**High Severity**\n\n<!-- DESCRIPTION START -->\nThe new forbidden-header validation runs unconditionally in the factory's `create_fn`, which is invoked for both `CREATE DATABASE` and `ATTACH DATABASE` (server restart). The commen... | diff --git a/tests/integration/test_database_iceberg/test.py b/tests/integration/test_database_iceberg/test.py
index 58d0e2d1d181..d1fe0912540e 100644
--- a/tests/integration/test_database_iceberg/test.py
+++ b/tests/integration/test_database_iceberg/test.py
@@ -385,7 +385,7 @@ def test_hide_sensitive_info(started_clus... | true | |
ClickHouse/ClickHouse | 98,827 | comment_to_fix | Implement http header validation for DataLakeCatalog | ### Test still uses `wrong.header` against reviewer feedback
**Low Severity**
<!-- DESCRIPTION START -->
The PR reviewer (@kssenii) explicitly requested adjusting `03913_datalake_restful_catalog_bad_format` to use correct auth header format instead of `auth_header = 'wrong.header'`. The test still uses `'wrong.header... | b0db88ca1cf6084fb5495911ec0a0ca3f53cc9f2 | 5198f3225264c04399828edefa93aec04b3ab11a | diff --git a/tests/queries/0_stateless/03913_datalake_restful_catalog_bad_format.sh b/tests/queries/0_stateless/03913_datalake_restful_catalog_bad_format.sh
index b7f78f733d34..fb32a9b0c162 100755
--- a/tests/queries/0_stateless/03913_datalake_restful_catalog_bad_format.sh
+++ b/tests/queries/0_stateless/03913_datalake... | [
"tests/queries/0_stateless/03913_datalake_restful_catalog_bad_format.sh"
] | [
{
"comment": "### Test still uses `wrong.header` against reviewer feedback\n\n**Low Severity**\n\n<!-- DESCRIPTION START -->\nThe PR reviewer (@kssenii) explicitly requested adjusting `03913_datalake_restful_catalog_bad_format` to use correct auth header format instead of `auth_header = 'wrong.header'`. The tes... | true |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.