add duration of the scan to JSON

[armijnhemel]

Short Description

currently there is already a start time stamp and end time stamp in the JSON output. It would be nice to also have a "duration" field so I don't have to compute that myself.

Possible Labels

• new feature

Select Category

• Enhancement [X]
• Add License/Copyright []
• Scan Feature []
• Packaging []
• Documentation []
• Expand Support []
• Other []

Describe the Update

How This Feature will help you/your organization

Possible Solution/Implementation Details

end_timestamp - start_timestamp

Example/Links if Any

Can you help with this Feature

[pombredanne]

@armijnhemel Thank you for this: would a duration in seconds work as in 7776045? (as opposed to a human readable and formatted string such as 3:12:45 for 3 hours, 12 minutes and 45 seconds?
Note also this cannot be 100% accurate as the actual time it takes to write the JSON cannot be integrated, and this could be several seconds on large scans.

[armijnhemel]

@armijnhemel Thank you for this: would a duration in seconds work as in 7776045? (as opposed to a human readable and formatted string such as 3:12:45 for 3 hours, 12 minutes and 45 seconds?
Note also this cannot be 100% accurate as the actual time it takes to write the JSON cannot be integrated, and this could be several seconds on large scans.

Having just seconds would be perfect. I could do formatting myself later.

[mnv936]

@pombredanne i would like to work on this feature.
Anything I should know that might help me or I should take care of?

[shivam-bit]

@pombredanne even I would love to work on this feature

[MankaranSingh]

Hi, I have added the requested feature in the following pull requests:
#1939

Now, you can see a 'time_taken' field in the header of the output JSON file,

[MankaranSingh]

i guess i need to edit the test cases too since this feature requires direct change in the output JSON file, alot of test cases.

So how should i go dealing with this ? any help is appreciated

[pombredanne]

@MankaranSingh thank you for this!
@armijnhemel do you see this feature as something always present in the JSON scan outputs? Or as guarded by a command line option?

[pombredanne]

@MankaranSingh let's see if this need to be enabled at all times or not first before updating the tests. There are ways to regenerate test fixtures to help a bit

[mjherzog]

I think that we can avoid yet another command line option.

[armijnhemel]

I would just always include it.

[pombredanne]

@MankaranSingh so you will have indeed to update a good bunch of test scan files since they will now contain your new field... On te other hand, since this field will have a different value on each run, it is best excluded from most of the scan tests assertions, except for a few tests focused only on the tests of that feature. Does it make sense? There are very few places where these checks are done, mostly in this place https://github.com/nexB/scancode-toolkit/blob/develop/src/scancode/cli_test_utils.py

[MankaranSingh]

ok, on it.
So, if i need further help, shall i ask here or on the gitter discussion channel ?

[pombredanne]

@MankaranSingh best is to keep ticket related discussions on the ticket and you can ask for help in the chat linking to the ticket.

[MankaranSingh]

I have opened another pull request #1942
Here are all the changes
@pombredanne ok, so all the tests seem to be passing now.
What i found out was, that the "start_timestamp" and "end_timestamp" were excluded from the JSON file prior to comparing with the expected JSON file. So, it made sense for me to exclude the new "time_taken" field too !

also, i found no other tests that specifically tested the "start_timestamp" and "end_timestamp", so i too didn't write any other test for the "time_taken" feature.

Also, @armijnhemel, i have updated my commit to store time as float instead of string.

[dakshaladia]

I wish to work on this issue. Is it still available?

[armijnhemel]

I wish to work on this issue. Is it still available?

I suppose it is no longer available, but perhaps @pombredanne can clarify?