All unpacking is performed by a parser class unique to the data type being unpacked (GRIF-16 fragment, GRIF-16 scalar, PPG or GRIF-4G fragment). Simultaneously, the parser checks for inconsistencies in the data, via the following pattern
-
Each parser contains an
assessComposition
method, which checks that a valid number of words appear in a valid order in the event; any problems are recorded in thedataStore
in an array of error messages. -
For each word type, the parser contains a function to unpack that word and raise warnings about problems aparent in that word.
-
Finally, each parser contains a last-pass function
postProcessingFlags
to check for any inconsistencies apparent after unpacking is complete.
- must have at least 10 words
- first word must start with 0x8
- second word must start with 0xD
- third word must start with 00
- fourth word must start with 0
- type V must start with 0x9
- type VI must start with 0xA
- type VII must start with 0xB
- allow a run of type VIIa words
- type VIII must start with 0
- type IX must start with 0
- type X must start with 0xE
- type X word must be last word in event
- Type I
- Must have valid module code
- Number of words reported in header must match actual number of words (excluding type VIIa)
- Must have valid detector code
- Type III
- Make sure there are no type VIIa words if the waveform bit is unset, or the converse.
- Type VI & IX
- Timestamp low bits and CFD time should be within 10k units of each other.
Analogous to GRIF-16 Fragments.
- must have 7 words
- each word must have the correct type flag, in order
- Type I
- Must have valid module code
- Must have valid detector code
- must have at least 5 words
- each word must have the correct type flag, in order
- Type I
- Must have valid module code