Accept a list of credentials in `httpCredentials` of `New Context` and `New Persistent Context`
I maintainer di solito rispondono entro 1 giorno
Nessuno ha ancora preso questa issue.
Valutazione
- Difficoltà
- 3/5
- Tempo stimato
- 1-2 giorni
- Idoneità per principianti
- 76/100
Direzione di ricerca
Read Browser/utils/data_types.py around convert_typed_dict, then inspect Browser/keywords/playwright_state.py around _set_context_options and its New Context and New Persistent Context annotations. Add list handling while preserving per-entry secret resolution and direct-value checks; start by extending utest/test_secrets.py with resolved and rejected list cases. Done means both context keywords accept a list, a single dictionary still works, and the acceptance and logging cases described in the issue pass.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Descrizione
Use case
A test that talks to more than one server behind HTTP authentication needs different credentials per server. Typical cases: the application and its API on different hosts, or a staging site whose static assets come from a second protected host.
New Context httpCredentials= takes one username and password. With origin set, they are sent only to that origin; without it, to any server that asks. A second server with other credentials cannot be covered in the same context today. The only workaround is a second context, which does not help when one page loads from both servers.
Playwright 1.63 accepts an array of credentials, each with its own optional origin. The first entry whose origin matches the request is used, and an entry without origin matches any request.
Proposed keyword / arguments
No new keyword. httpCredentials of New Context and New Persistent Context accepts either one HttpCredentials dictionary, as today, or a list of them.
*** Test Cases ***
Log In To Application And API With Different Credentials
VAR &{app} = username=$APP_USER password=$APP_PASSWORD origin=https://app.example.com
VAR &{api} = username=$API_USER password=$API_PASSWORD origin=https://api.example.com
VAR @{credentials} = ${app} ${api}
New Context httpCredentials=${credentials}
New Page https://app.example.com
Get Text id=status == Data loaded from API
Secret resolution works per entry, the same way it works for a single dictionary today: $var and %ENV placeholders, Robot Framework 7.4 Secret values and crypt: strings from CryptoLibrary. The rule that direct values are refused applies to every entry, so a list that contains one entry with plain values fails with the existing "Direct assignment of values or variables as 'httpCredentials' is not allowed" error.
New Page does not take context options, so it is not affected.
browserContext.setHTTPCredentials() changes the credentials of an open context. Playwright deprecated it earlier and un-deprecated it in 1.63, and it accepts the same list. This issue does not propose a keyword for it. Playwright notes that browsers may cache credentials per origin after a successful authentication, so changing them mid-test may have no effect; a new context is the reliable way to switch users. A setter can be reconsidered if someone shows a case that a new context does not cover.
Playwright API
browser.newContext({ httpCredentials }):httpCredentialsis an object or, since 1.63, an array of objects withusername,password, optionaloriginand optionalsend. Intypes.d.tsof 1.63.0 the type changes from{...}to{...}|Array<{...}>; 1.62.0 has only the object form.browserType.launchPersistentContext(userDataDir, { httpCredentials }): same change in 1.63.browserContext.setHTTPCredentials(httpCredentials): the@deprecatedtag is gone in 1.63 and the parameter isnull|{...}|ReadonlyArray<{...}>.
send applies only to requests from an APIRequestContext, not to requests from the browser. Our HttpCredentials TypedDict does not have it today and does not need it for this change.
Implementation notes
All references are to upstream main (5e350998).
Python:
Browser/utils/data_types.py:804-840:_HttpCredentials(requiredusername,password, bothstr | Secret) andHttpCredentials(optionalorigin). The TypedDict stays as it is. Its docstring gets a list example and the "first matching origin wins" rule.Browser/keywords/playwright_state.py:815(New Context) and:1028(New Persistent Context): annotationHttpCredentials | NonebecomesHttpCredentials | list[HttpCredentials] | None. That stays within the narrow-annotation rule: it names the TypedDict, notdictorAny. ThehttpCredentialslines of both docstrings (:880-882, and the Secret paragraph at:961-971) get the list form.Browser/utils/data_types.py:114-147convert_typed_dict: for a union it picks the TypedDict member and then raisesTypeErrorwhen the value is not adict(:126-129). It needs to convert each item of a list value with the same TypedDict.Browser/keywords/playwright_state.py:1218-1220in_set_context_optionscallsself.resolve_secret(httpCredentials, "httpCredentials")on the raw keyword argument, not on the convertedparamsvalue.resolve_secret(Browser/base/librarycomponent.py:248-257) goes through_replace_placeholder_variables(:297-302), which resolves the values of a dict but returns any other value unchanged. A list would therefore come back equal to the input and hit the "Direct assignment"ValueErrorat:251-256. The fix is to resolve each entry separately in_set_context_options(oneresolve_secretcall per entry), so the equality check stays per entry.resolve_secretitself does not need to know about lists._mask_credentials(playwright_state.py:1272-1275) replaces the whole value with"XXX", so a list is masked in the log without change.
Node: none. newContext (node/playwright-wrapper/playwright-state.ts:1024-1042) and newPersistentContext (:1106-1115) pass the parsed rawOptions to Playwright as they are. No proto change.
Tests:
utest/test_secrets.py:123-130already checks the resolved single dict in the raw options sent to Node; add the same for a list of two entries (both resolved), and a list where one entry has direct values (fails).- Acceptance: the Test App (
node/dynamic-test-app/src/server.ts) has no route that answers with401andWWW-Authenticate. Add one that checks Basic credentials, and reach it throughlocalhostand127.0.0.1as two origins, the same trickatest/test/01_Browser_Management/storage_state.robot:7uses (${OTHER_ORIGIN_URL}). One test with a list of two entries, each origin getting its own credentials. Extendatest/test/01_Browser_Management/no_httpCredentials_logging.robotwith a list case to keep the log masking covered.
Backwards compatibility
Additive. A single dictionary keeps working and is passed to Playwright unchanged. The list form needs Playwright 1.63 or later; the library already ships 1.65.
- Lingua principale
- Python
- Stelle
- 662
- Fork
- 149
- Merge medio
- 6h 13m
- PR unite (30g)
- 121
Preparare l'ambiente
- Nessun Dockerfile né file Docker Compose
- Nessun modello di pull request
- Leggi la guida per i contributori
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di MarketSquare/robotframework-browser
-
priority: high
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
MarketSquare/robotframework-browser#5306 · 2 commenti ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
MarketSquare/robotframework-browser#5153 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 3-5 giorni Idoneità per principianti 48/100
MarketSquare/robotframework-browser#5503 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 3/5 1-2 giorni Idoneità per principianti 78/100
MarketSquare/robotframework-browser#5502 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 4/5 3-5 giorni Idoneità per principianti 54/100
MarketSquare/robotframework-browser#5501 ·
I maintainer di solito rispondono entro 1 giorno
Tutte le issue di MarketSquare/robotframework-browser
Issue simili
-
bug
Difficoltà 2/5 1-3 ore Idoneità per principianti 74/100
igormcsouza/tether#278 ·
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 91/100
aws-samples/sample-per-model-bedrock#803 ·
I maintainer di solito rispondono entro 1 giorno
-
documentation p1
Difficoltà 2/5 1-3 ore Idoneità per principianti 75/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
huggingface/accelerate#4511 ·
I maintainer di solito rispondono entro 1 giorno
-
follow-up
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
I maintainer di solito rispondono entro 1 giorno