fix(server): use regex instead of isAbsoluteUrl

[EslamHiko]

• This is a bugfix
• This is a feature
• This is a code refactor
• This is a test update
• This is a docs update
• This is a metadata update

For Bugs and Features; did you add new tests?

Yes

Motivation / Use-Case

fixes: #2190

Breaking Changes

no

Additional Info

If we want to still isAbsoluteUrl use can add an extra check for the second character in the contentBase and replace all isAbsoluteUrl(String(contetBase)) with isAbsoluteUrl(String(contetBase)) && String(contetBase)[1] !== ':' assuming that always local paths will have : as a second character ex:c: , e: , ... but I think reverting the changes to the original regex will be better, what do you think ?

[hiroppy]

Need to uninstall is-absolute-url. And I think it is better to create util for this change.

[codecov]

Codecov Report

Merging #2202 into master will not change coverage.
The diff coverage is 100%.

@@ Coverage Diff @@ ## master #2202 +/- ## ======================================= Coverage 93.89% 93.89% ======================================= Files 34 34 Lines 1278 1278 Branches 366 366 ======================================= Hits 1200 1200 Misses 71 71 Partials 7 7

Impacted Files	Coverage Δ
lib/Server.js	97.33% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update eab6acd...b415ede. Read the comment docs.

[alexander-akait]

I think we need ti wait answer sindresorhus/is-absolute-url#7 before doing something here

[EslamHiko]

@hiroppy done
@evilebottnawi sorry I didn't see the comment, we have now a util file we can use any regex we want.we can modify is-absolute-url regex and use it in checkUrl util. what do you think?

[alexander-akait]

/cc @EslamHiko looks it is solved on dep side, but will be great keep test to avoid future regression, can you rebase?

[EslamHiko]

@evilebottnawi yes, I kept only the tests.

[alexander-akait]

No need this here, if you have problem place let's solve this in other PR

[alexander-akait]

Let's add two tests here c:\absolute\path\to\content-base and C:\absolute\path\to\content-base

[alexander-akait]

Typos

[alexander-akait]

Typo 😄

[hiroppy]

Don't use the port number directly.

[EslamHiko]

if I use port it'll cause listen EADDRINUSE: address already in use 127.0.0.1:8107 error and cause the tests to hang, can I use port+1,port+2 or add new values in ports-map ?

[hiroppy]

Don't add the port number.

diff --git a/test/server/contentBase-option.test.js b/test/server/contentBase-option.test.js index fe120e5..bcde1e6 100644 --- a/test/server/contentBase-option.test.js +++ b/test/server/contentBase-option.test.js @@ -269,10 +269,10 @@ describe('contentBase option', () => { describe('testing single & multiple external paths', () => { afterEach((done) => { - testServer.close(() => { - done(); - }); + testServer.close(done); }); + it('Should throw exception (string)', (done) => { try { // eslint-disable-next-line no-unused-vars @@ -302,71 +302,40 @@ describe('contentBase option', () => { } }); it('Should not throw exception (local path with lower case first character)', (done) => { - try { - // eslint-disable-next-line no-unused-vars - server = new Promise((resolve, reject) => { - testServer.start(config, { - contentBase: - contentBasePublic.charAt(0).toLowerCase() + - contentBasePublic.substring(1), - watchContentBase: true, - port: 2222, - }); - resolve(testServer); - }); - - server.then((testServer) => { - testServer.close(() => { - done(); - }); - }); - } catch (e) { - expect(true).toBe(false); - } + testServer.start( + config, + { + contentBase: + contentBasePublic.charAt(0).toLowerCase() + + contentBasePublic.substring(1), + watchContentBase: true, + port, + }, + done + ); }); it("Should not throw exception (local path with lower case first character & has '-')", (done) => { - try { - // eslint-disable-next-line no-unused-vars - server = new Promise((resolve, reject) => { - testServer.start(config, { - contentBase: 'c:\\absolute\\path\\to\\content-base', - watchContentBase: true, - port: 2223, - }); - resolve(testServer); - }); - - server.then((testServer) => { - testServer.close(() => { - done(); - }); - }); - } catch (e) { - expect(true).toBe(false); - } + testServer.start( + config, + { + contentBase: 'c:\\absolute\\path\\to\\content-base', + watchContentBase: true, + port, + }, + done + ); }); it("Should not throw exception (local path with upper case first character & has '-')", (done) => { - try { - // eslint-disable-next-line no-unused-vars - server = new Promise((resolve, reject) => { - testServer.start(config, { - contentBase: 'C:\\absolute\\path\\to\\content-base', - watchContentBase: true, - port: 2224, - }); - resolve(testServer); - }); - - server.then((testServer) => { - testServer.close(() => { - done(); - }); - }); - } catch (e) { - expect(true).toBe(false); - } + testServer.start( + config, + { + contentBase: 'C:\\absolute\\path\\to\\content-base', + watchContentBase: true, + port, + }, + done + ); }); - it('Should throw exception (array)', (done) => { try { // eslint-disable-next-line no-unused-vars

You can patch this code.

[EslamHiko]

/cc @evilebottnawi @hiroppy
CI is green 😄

[alexander-akait]

Why we add typeof item === 'number'?

[EslamHiko]

@evilebottnawi it's already blocking number values in options.js accepting only strings, in case in future refactors if we change the type of array items to accept any type we still need it to throw an exception if the array item is number type.

[EslamHiko]

@evilebottnawi we can remove number type from options.json

 "contentBase": { "anyOf": [ { "enum": [false] }, { "type": "number" }, { "type": "string" }, { "type": "array", "items": { "type": "string" }, "minItems": 1 } ] }, 

and remove || typeof item === 'number' checks and update tests to check for the exception "should be {String|Array} (https://webpack.js.org/configuration/dev-server/#devservercontentbase)" in number type cases
what do you think ?

[alexander-akait]

Let's do it next version

[alexander-akait]

Good job, one question

[alexander-akait]

/cc @hiroppy