Variable length data

[MarkJeronimus]

I need an option to designate one or more bytes as a length, and then parse the following N bytes as the data.

For example:
30 00 05 aa bb cc dd ee 4f
30 00 01 ff d8
->
ID:30 LEN:0005 DATA:aabbccddeeff CRC:4f
ID:30 LEN:0001 DATA:ff CRC:d8

[zuckschwerdt]

That will take some thinking. It would mean to feed some bits back to the format string to determine length.
Perhaps something like "ID:8h LEN:16h DATA:{16@16}h CRC:8h" with {} meaning variable field length and 16@16 meaning 16 bits starting from 16?

[MarkJeronimus]

Before you do something that can't be undone (for backwards compatiblility) Let me first make a 'preparation' enhancement #9 that this enhancement can make use of.

{16} is just the same as 16: the 16bit value starting at the current position.
{16@24} means 16bit value starting at bit 24 (as in #9)

Here's my enhancement proposal:
{{16@24}*8}h : The length in bytes is the 16bit value starting bit 24.

Here the result of the inner evaluation gets substituted in the outer {} so if the 16bit value starting at bit 24 happens to be 0005h, it becomes {5*8}, which becomes {40}. In case of my example, this reads the next 40 bits because the length field is 5.

The *8 doesn't need to utilize a full formula parser, just a multiplication by a positive literal.

==Extras:==
{<{16@24}*8}h : The length in bytes is the 16bit value starting bit 24, interpreted little endian.

[zuckschwerdt]

That looks recursive… Maybe moving-the-pointer and referencing-as-field-length should use different syntax. […] vs {…}?

[MarkJeronimus]

Recursive was the idea, although I left it up to you whether you want to go deeper or limit to one nesting.

Alternate syntax is also fine with me.