Add named "additional data" support #121
Conversation
This allows representing KV pairs of data in e.g. chrome profiles more faithfully.
nagisa
force-pushed
the
nagisa/kv-additional-data
branch
from
356671f to
48c55c4
Compare
|
Note: this doesn’t actually add the API to measureme to output such data. Ideas on what API would make most sense here would be appreciated. |
| /// <argument> = ['\x1D' <argument_name>] '\x1E' <argument_value> | ||
| /// <argument_name> = <text> | ||
| /// <argument_value> = <text> |
There was a problem hiding this comment.
I wonder if it would be simpler to make the grammar:
<argument> = '\x1E' (<argument_name_value_pair> | <argument_value>)
<argument_name_value_pair> = <text> '\x1D' <text>
<argument_value> = <text>
Then we could always parse the leading \x1E and then do a string split on the \x1D if it exists.
There was a problem hiding this comment.
I think that would also make the format compatible with the current format. It's fine if we need to break that but we'll have to bump the format number and coordinate redeploying new versions on perf.rlo.
There was a problem hiding this comment.
It sounds like there might be another breaking change in the works from another contributor so if you think it's simpler to use the current strategy rather than this suggestion, that's fine with me!
|
I’m probably not going to pursue this further, but if anybody wants feel free to pick this up. |
This allows representing KV pairs of data in e.g. chrome profiles more
faithfully. Previously we would automatically generate arg0... as a key
for all such data.