eugenetyc / pe

0 stars 0 forks source link

Inconsistency in error message #5

Open eugenetyc opened 4 years ago

eugenetyc commented 4 years ago

for addc,

the following command gives this error message for p/

image.png

However, for a/, there is no specific error message for address.

image.png

This makes it inconvenient for the user, who expects address to also have a specific message so they know what the problem in their command is.

The same goes for the other parameters other than for the phone prefix p/

nus-pe-bot commented 4 years ago

Team's Response

Screenshot 2020-04-18 at 12.13.56 PM.png As seen here, the format for address is given in the error message when the address format is wrong.

Screenshot 2020-04-18 at 12.15.38 PM.png Same goes for email

Items for the Tester to Verify

:question: Issue response

Team chose [response.Rejected]

Reason for disagreement: The issue here is that the command's error message is inconsistent in being corresponding to the given prefix.

While true, the team shows that addc n/jackson p/111 a/ gives the corresponding error message for address a/,

image.png

it does not follow when there are only 2 prefixes, one n/ and another being any of them other than p/. That is, only p/ will give a corresponding error message when it is the 2nd and only other prefix other than n/.

My issue is justified as follows, with variations to the 2nd prefix (apart from n/):

  1. If I use p/, an error message specific to phone number appears.

image.png

  1. However, if I use e/, the general error message appears.

image.png

  1. It is similar to the issue in the Point 2 if I use a/, a general error message.

image.png

  1. Similarly, for t/.

image.png

Therefore, as you can see, the error messages shown are INCONSISTENT when there are 2 prefixes.

Since this brings inconvenience to the user, who may not understand where he/she is going wrong in their usage of the command, it is marked as a feature flaw. Moreover, I marked this as medium because there are many prefixes, and it is quite common for users to type in wrong prefixes - therefore a need to know exactly which prefix they are getting wrong, rather than trial and error all of them. I believe this follows the following PE guidelines, and therefore my bug issue is justified.

image.png