Autofill behaving oddly for 3 inputs
Reported by
evolso...@gmail.com,
Nov 9
|
|||||||
Issue descriptionUserAgent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.77 Safari/537.36 Steps to reproduce the problem: Save the attached/following code as a .html file: <form> <label for="firstname">First Name</label><input type="text" maxlength="75" style="width:300px;" class="Question" name="firstname" id="firstname" value=""><br> <label for="lastname">Last Name</label><input type="text" maxlength="75" style="width:300px;" class="Question" name="lastname" id="lastname" value=""><br> <label for="q">q34wregsr rq rwq</label><input type="text" maxlength="75" style="width:300px;" class="Question" name="q_" id="q" value=""><br> </form> Open the html file and auto fill it. What is the expected behavior? Auto fill shouldn't fill out seemingly random fields with seemingly random data from the contact. What went wrong? The auto fill, for whatever reason, fills out #q with the value stored in state. I also have another example where the 5th field/ 4th input on the page is populating as Country, but I can't reproduce this. If I add another label to the example code, the auto populate does NOT populate #q or the new field (where i just replaced "q" with "a"). This does not happen in Safari. Did this work before? N/A Does this work in other browsers? N/A Chrome version: 70.0.3538.77 Channel: n/a OS Version: OS X 10.14.0 Flash Version:
,
Nov 10
,
Nov 13
Able to reproduce the issue on Mac 10.13.6, Win-10 and Ubuntu 17.10 using chrome reported version #70.0.3538.77 but the same is not reproducible in the latest canary #72.0.3608.0. Reverse bisect information will be provided very soon from our end. Hence adding Needs-Bisect label and marking it as Untriaged. Thanks.
,
Nov 14
Able to reproduce the issue on Mac 10.13.6, Win-10 and Ubuntu 17.10 using chrome reported version #70.0.3538.77 but the same is not reproducible in the latest canary #72.0.3608.0. Reverse Bisect Information: ===================== Good build: 71.0.3544.0 Bad Build : 71.0.3543.0 Change Log URL: https://chromium.googlesource.com/chromium/src/+log/75419b8570502363cc81aa2619407e994830f557..7300efea63d93b082e780c02b953bd3ed32e9b09 From the above change log suspecting below change Change-Id: Ic03afbec002e3cf005d22ac25fd96e90b0be1273 Reviewed-on: https://chromium-review.googlesource.com/c/1257184 rogerm@ - Could you please check and merge the fix to M-70 if it is a valid candidate. Note: Adding label RBS as it seems to be a recent regression. Issue is seen from M-60 builds but fixed on latest canary . Hence, not adding RegressedIn label. Thanks...!!
,
Nov 14
The fix identified by the bisect doesn't directly address the OP concern, it just stops autofill from triggering on "file://" URLs. Putting the same test file onto a web-server and browsing to it via "http://" or "https://" may yield the same of different predictions, depending on how user's have interacted with that form/field previously. I would suggest that http://crrev.com/c/1257184 *NOT* be merged to M-70 due to it's size and the indirect nature by which it fixes the OP' concern. --- Digging a little deeper into the crowdsourced predictions for form these fields shows the predictions are based on very little data and should probably be discarded. evolsoulx@gmail.com can you go into chrome://flags and enable the --show-autofill-type-predictions flag and restart? When you hover over an input field you should see a pop-up that displays the type of data autofill thinks the field accepts as well as the form and field signatures. Please paste the form and field signatures for the problem field into this bug. This will allow me to confirm which predictions to discard. Thanks.
,
Nov 14
krajshree@, I don't think this issue should be a release blocker and it's not a regression.
,
Nov 15
Removing RBS label as per comment #5 - "I would suggest that http://crrev.com/c/1257184 *NOT* be merged to M-70 due to it's size and the indirect nature by which it fixes the OP' concern". Thanks...!!
,
Nov 15
Removing 'RBS' label per c#6. |
|||||||
►
Sign in to add a comment |
|||||||
Comment 1 by dtapu...@chromium.org
, Nov 9