Archived
1
0
This repository has been archived on 2024-09-09. You can view files and clone it, but cannot push or open issues or pull requests.
code-server/src/node/http.ts

311 lines
10 KiB
TypeScript
Raw Normal View History

import { field, logger } from "@coder/logger"
2020-10-21 01:05:58 +02:00
import * as express from "express"
import * as expressCore from "express-serve-static-core"
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
import * as http from "http"
import * as net from "net"
import * as qs from "qs"
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
import { Disposable } from "../common/emitter"
import { CookieKeys, HttpCode, HttpError } from "../common/http"
import { normalize } from "../common/util"
import { AuthType, DefaultedArgs } from "./cli"
import { version as codeServerVersion } from "./constants"
import { Heart } from "./heart"
import { CoderSettings, SettingsProvider } from "./settings"
import { UpdateProvider } from "./update"
import { getPasswordMethod, IsCookieValidArgs, isCookieValid, sanitizeString, escapeHtml, escapeJSON } from "./util"
2020-02-04 20:27:46 +01:00
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
/**
* Base options included on every page.
*/
export interface ClientConfiguration {
codeServerVersion: string
/** Relative path from this page to the root. No trailing slash. */
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
base: string
/** Relative path from this page to the static root. No trailing slash. */
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
csStaticBase: string
}
declare global {
// eslint-disable-next-line @typescript-eslint/no-namespace
namespace Express {
export interface Request {
args: DefaultedArgs
heart: Heart
settings: SettingsProvider<CoderSettings>
updater: UpdateProvider
}
}
}
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
export const createClientConfiguration = (req: express.Request): ClientConfiguration => {
const base = relativeRoot(req.originalUrl)
return {
base,
csStaticBase: base + "/_static",
codeServerVersion,
}
}
2020-02-04 20:27:46 +01:00
/**
2020-10-21 01:05:58 +02:00
* Replace common variable strings in HTML templates.
2020-02-04 20:27:46 +01:00
*/
2020-10-21 01:05:58 +02:00
export const replaceTemplates = <T extends object>(
req: express.Request,
content: string,
extraOpts?: Omit<T, "base" | "csStaticBase" | "logLevel">,
): string => {
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
const serverOptions: ClientConfiguration = {
...createClientConfiguration(req),
2020-10-21 01:05:58 +02:00
...extraOpts,
}
2020-10-21 01:05:58 +02:00
return content
.replace(/{{TO}}/g, (typeof req.query.to === "string" && escapeHtml(req.query.to)) || "/")
.replace(/{{BASE}}/g, serverOptions.base)
.replace(/{{CS_STATIC_BASE}}/g, serverOptions.csStaticBase)
.replace("{{OPTIONS}}", () => escapeJSON(serverOptions))
2020-02-04 20:27:46 +01:00
}
/**
2020-11-03 23:45:03 +01:00
* Throw an error if not authorized. Call `next` if provided.
2020-02-04 20:27:46 +01:00
*/
export const ensureAuthenticated = async (
req: express.Request,
_?: express.Response,
next?: express.NextFunction,
): Promise<void> => {
const isAuthenticated = await authenticated(req)
if (!isAuthenticated) {
2020-10-21 01:05:58 +02:00
throw new HttpError("Unauthorized", HttpCode.Unauthorized)
}
2020-11-03 23:45:03 +01:00
if (next) {
next()
}
2020-02-04 20:27:46 +01:00
}
2020-10-21 01:05:58 +02:00
/**
* Return true if authenticated via cookies.
*/
export const authenticated = async (req: express.Request): Promise<boolean> => {
2020-10-21 01:05:58 +02:00
switch (req.args.auth) {
case AuthType.None: {
2020-10-21 01:05:58 +02:00
return true
}
case AuthType.Password: {
2020-10-21 01:05:58 +02:00
// The password is stored in the cookie after being hashed.
const hashedPasswordFromArgs = req.args["hashed-password"]
const passwordMethod = getPasswordMethod(hashedPasswordFromArgs)
const isCookieValidArgs: IsCookieValidArgs = {
passwordMethod,
cookieKey: sanitizeString(req.cookies[CookieKeys.Session]),
passwordFromArgs: req.args.password || "",
hashedPasswordFromArgs: req.args["hashed-password"],
}
return await isCookieValid(isCookieValidArgs)
}
default: {
2020-10-21 01:05:58 +02:00
throw new Error(`Unsupported auth type ${req.args.auth}`)
}
2020-10-21 01:05:58 +02:00
}
2020-02-04 23:55:27 +01:00
}
2020-10-21 01:05:58 +02:00
/**
* Get the relative path that will get us to the root of the page. For each
* slash we need to go up a directory. Will not have a trailing slash.
*
* For example:
*
2020-10-21 01:05:58 +02:00
* / => .
* /foo => .
* /foo/ => ./..
* /foo/bar => ./..
* /foo/bar/ => ./../..
*
* All paths must be relative in order to work behind a reverse proxy since we
* we do not know the base path. Anything that needs to be absolute (for
* example cookies) must get the base path from the frontend.
*
* All relative paths must be prefixed with the relative root to ensure they
* work no matter the depth at which they happen to appear.
*
* For Express `req.originalUrl` should be used as they remove the base from the
* standard `url` property making it impossible to get the true depth.
2020-10-21 01:05:58 +02:00
*/
export const relativeRoot = (originalUrl: string): string => {
const depth = (originalUrl.split("?", 1)[0].match(/\//g) || []).length
2020-10-21 01:05:58 +02:00
return normalize("./" + (depth > 1 ? "../".repeat(depth - 1) : ""))
2020-02-14 22:57:51 +01:00
}
/**
* A helper function to construct a redirect path based on
* an Express Request, query and a path to redirect to.
*
* Redirect path is relative to `/${to}`.
*/
export const constructRedirectPath = (req: express.Request, query: qs.ParsedQs, to: string): string => {
const relativePath = normalize(`${relativeRoot(req.originalUrl)}/${to}`, true)
// %2f or %2F are both equalivent to an encoded slash /
const queryString = qs.stringify(query).replace(/%2[fF]/g, "/")
const redirectPath = `${relativePath}${queryString ? `?${queryString}` : ""}`
return redirectPath
}
2020-10-21 01:05:58 +02:00
/**
* Redirect relatively to `/${to}`. Query variables on the current URI will be
* preserved. `to` should be a simple path without any query parameters
2020-10-21 01:05:58 +02:00
* `override` will merge with the existing query (use `undefined` to unset).
*/
export const redirect = (
req: express.Request,
res: express.Response,
to: string,
override: expressCore.Query = {},
): void => {
const query = Object.assign({}, req.query, override)
Object.keys(override).forEach((key) => {
if (typeof override[key] === "undefined") {
delete query[key]
}
})
const redirectPath = constructRedirectPath(req, query, to)
2020-10-21 01:05:58 +02:00
logger.debug(`redirecting from ${req.originalUrl} to ${redirectPath}`)
res.redirect(redirectPath)
}
2020-02-04 20:27:46 +01:00
/**
2020-10-21 01:05:58 +02:00
* Get the value that should be used for setting a cookie domain. This will
* allow the user to authenticate once no matter what sub-domain they use to log
* in. This will use the highest level proxy domain (e.g. `coder.com` over
* `test.coder.com` if both are specified).
2020-02-04 20:27:46 +01:00
*/
2020-10-21 01:05:58 +02:00
export const getCookieDomain = (host: string, proxyDomains: string[]): string | undefined => {
const idx = host.lastIndexOf(":")
host = idx !== -1 ? host.substring(0, idx) : host
// If any of these are true we will still set cookies but without an explicit
// `Domain` attribute on the cookie.
2020-10-21 01:05:58 +02:00
if (
// The host can be be blank or missing so there's nothing we can set.
2020-10-21 01:05:58 +02:00
!host ||
// IP addresses can't have subdomains so there's no value in setting the
// domain for them. Assume that anything with a : is ipv6 (valid domain name
// characters are alphanumeric or dashes)...
2020-10-21 01:05:58 +02:00
host.includes(":") ||
// ...and that anything entirely numbers and dots is ipv4 (currently tlds
2020-10-21 01:05:58 +02:00
// cannot be entirely numbers).
!/[^0-9.]/.test(host) ||
// localhost subdomains don't seem to work at all (browser bug?). A cookie
// set at dev.localhost cannot be read by 8080.dev.localhost.
2020-10-21 01:05:58 +02:00
host.endsWith(".localhost") ||
// Domains without at least one dot (technically two since domain.tld will
// become .domain.tld) are considered invalid according to the spec so don't
// set the domain for them. In my testing though localhost is the only
// problem (the browser just doesn't store the cookie at all). localhost has
// an additional problem which is that a reverse proxy might give
// code-server localhost even though the domain is really domain.tld (by
// default NGINX does this).
!host.includes(".")
2020-10-21 01:05:58 +02:00
) {
logger.debug("no valid cookie domain", field("host", host))
2020-10-21 01:05:58 +02:00
return undefined
}
proxyDomains.forEach((domain) => {
if (host.endsWith(domain) && domain.length < host.length) {
host = domain
}
})
logger.debug("got cookie domain", field("host", host))
2020-11-03 23:44:08 +01:00
return host || undefined
2020-10-21 01:05:58 +02:00
}
fix(testing): revert change & fix playwright tests (#4310) * fix(testing): revert change & fix playwright tests * fix(constants): add type to import statement * refactor(e2e): delete browser test This test was originally added to ensure playwright was working. At this point, we know it works so removing this test because it doesn't help with anything specific to code-server and only adds unnecessary code to the codebase plus increases the e2e test job duration. * chore(e2e): use 1 worker for e2e test I don't know if it's a resources issue, playwright, or code-server but it seems like the e2e tests choke when multiple workers are used. This change is okay because our CI runner only has 2 cores so it would only use 1 worker anyway, but by specifying it in our playwright config, we ensure more stability in our e2e tests working correctly. See these PRs: - https://github.com/cdr/code-server/pull/3263 - https://github.com/cdr/code-server/pull/4310 * revert(vscode): add missing route with redirect * chore(vscode): update to latest fork * Touch up compilation step. * Bump vendor. * Fix VS Code minification step * Move ClientConfiguration to common Common code must not import Node code as it is imported by the browser. * Ensure lib directory exists before curling cURL errors now because VS Code was moved and the directory does not exist. * Update incorrect e2e test help output Revert workers change as well; this can be overridden when desired. * Add back extension compilation step * Include missing resources in release This includes a favicon, for example. I opted to include the entire directory to make sure we do not miss anything. Some of the other stuff looks potentially useful (like completions). * Set quality property in product configuration When httpWebWorkerExtensionHostIframe.html is fetched it uses the web endpoint template (in which we do not include the commit) but if the quality is not set it prepends the commit to the web endpoint instead. The new static endpoint does not use/handle commits so this 404s. Long-term we might want to make the new static endpoint use commits like the old one but we will also need to update the various other static URLs to include the commit. For now I just fixed this by adding the quality since: 1. Probably faster than trying to find and update all static uses. 2. VS Code probably expects it anyway. 3. Gives us better control over the endpoint. * Update VS Code This fixes several build issues. * Bump vscode. * Bump. * Bump. * Use CLI directly. * Update tests to reflect new upstream behavior. * Move unit tests to after the build Our code has new dependencies on VS Code that are pulled in when the unit tests run. Because of this we need to build VS Code before running the unit tests (as it only pulls built code). * Upgrade proxy-agent dependencies This resolves a security report with one of its dependencies (vm2). * Symlink VS Code output directory before unit tests This is necessary now that we import from the out directory. * Fix issues surrounding persistent processes between tests. * Update VS Code cache directories These were renamed so the cached paths need to be updated. I changed the key as well to force a rebuild. * Move test symlink to script This way it works for local testing as well. I had to use out-build instead of out-vscode-server-min because Jest throws some obscure error about a handlebars haste map. * Fix listening on a socket * Update VS Code It contains fixes for missing files in the build. * Standardize disposals * Dispose HTTP server Shares code with the test HTTP server. For now it is a function but maybe we should make it a class that is extended by tests. * Dispose app on exit * Fix logging link errors Unfortunately the logger currently chokes when provided with error objects. Also for some reason the bracketed text was not displaying... * Update regex used by e2e to extract address The address was recently changed to use URL which seems to add a trailing slash when using toString, causing the regex match to fail. * Log browser console in e2e tests * Add base back to login page This is used to set cookies when using a base path. * Remove login page test The file this was testing no longer exists. * Use path.posix for static base Since this is a web path and not platform-dependent. * Add test for invalid password Co-authored-by: Teffen Ellis <teffen@nirri.us> Co-authored-by: Asher <ash@coder.com>
2021-10-28 22:27:17 +02:00
/**
* Return a function capable of fully disposing an HTTP server.
*/
export function disposer(server: http.Server): Disposable["dispose"] {
const sockets = new Set<net.Socket>()
let cleanupTimeout: undefined | NodeJS.Timeout
server.on("connection", (socket) => {
sockets.add(socket)
socket.on("close", () => {
sockets.delete(socket)
if (cleanupTimeout && sockets.size === 0) {
clearTimeout(cleanupTimeout)
cleanupTimeout = undefined
}
})
})
return () => {
return new Promise<void>((resolve, reject) => {
// The whole reason we need this disposer is because close will not
// actually close anything; it only prevents future connections then waits
// until everything is closed.
server.close((err) => {
if (err) {
return reject(err)
}
resolve()
})
// If there are sockets remaining we might need to force close them or
// this promise might never resolve.
if (sockets.size > 0) {
// Give sockets a chance to close up shop.
cleanupTimeout = setTimeout(() => {
cleanupTimeout = undefined
for (const socket of sockets.values()) {
console.warn("a socket was left hanging")
socket.destroy()
}
}, 1000)
}
})
}
}
/**
* Get the options for setting a cookie. The options must be identical for
* setting and unsetting cookies otherwise they are considered separate.
*/
export const getCookieOptions = (req: express.Request): express.CookieOptions => {
// Normally we set paths relatively. However browsers do not appear to allow
// cookies to be set relatively which means we need an absolute path. We
// cannot be guaranteed we know the path since a reverse proxy might have
// rewritten it. That means we need to get the path from the frontend.
// The reason we need to set the path (as opposed to defaulting to /) is to
// avoid code-server instances on different sub-paths clobbering each other or
// from accessing each other's tokens (and to prevent other services from
// accessing code-server's tokens).
// When logging in or out the request must include the href (the full current
// URL of that page) and the relative path to the root as given to it by the
// backend. Using these two we can determine the true absolute root.
const url = new URL(
req.query.base || req.body.base || "/",
req.query.href || req.body.href || "http://" + (req.headers.host || "localhost"),
)
return {
domain: getCookieDomain(url.host, req.args["proxy-domain"]),
path: normalize(url.pathname) || "/",
sameSite: "lax",
}
}
/**
* Return the full path to the current page, preserving any trailing slash.
*/
export const self = (req: express.Request): string => {
return normalize(`${req.baseUrl}${req.originalUrl.endsWith("/") ? "/" : ""}`, true)
}