Library for creating and querying segmented feeds.
+--------------------+
|Segmented OSI Model |
+--------------------+
magic, where XXXX is a feed type (0-12 chars)
+----------------------+
| "seg:XXXX" [char:16] |
+----------------------+
c -- countries
a -- asns
n -- naics
header
+-----------------+
| seg_num: uint64 |
+-----------------+
segments
+-----------------------+-----------------------+
| segment_offset: int32 | segment_length: int32 |
+-----------------------+-----------------------+
segment vector
+-------------+
| key: X, ... |
+-------------+
segment vector
+-------------+
| val: Y, ... |
+-------------+
Vector keys are fixed-length values such as 4-byte IPv4 addresses, 2-byte country codes, or 4-byte NAIC IDs.
All values are little-endian.
Values are other ints or can be null-terminated strings.
Segment 0 must contain a concatenated list of null terminated filenames.
The first filename must be .asif/filenames
.
The n-th filename in the segment 0 describes the n-th segment in the file.
The asif
file may contain contain a named segment .asif/createtime
.
If it does, this segment concatenated Word64
numbers representing the create
time of each file in microseconds since epoch. If the segment does not have a
create time, it's create time entry will be 0
instead.
The asif
file may contain contain a named segment .asif/formats
.
If it does, this segment concatenated null terminated strings representing the
format of the data in the respective segments. The dump
command will use
this information to choose the most appropriate way to print the data in each
segment.