You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Implement TODO in dlis_index_records "handle situation where explicit bits are not the same for every segment"
Usecase:
File contains many LFs.
Last fdata in LF is not complete (expects successor).
Then FILE-HEADER follows, but it is assumed to be the part of the previous record (relevant to #222 as we do not check predecessor).
'Explicit' marker is overwritten to be as in the last record segment. So the fdata + fileheader were considered to be one "explicit" record.
Then find_fileheaders searches explicit records of type 0 (FHLR is 0), but because FDATA is also 0, 'combined' record was assumed to be the start of a new LF.
Which then caused issues while parsing the record as explicit, and debugging was not straightforward.
The text was updated successfully, but these errors were encountered:
Implement TODO in dlis_index_records
"handle situation where explicit bits are not the same for every segment"
Usecase:
File contains many LFs.
Last fdata in LF is not complete (expects successor).
Then FILE-HEADER follows, but it is assumed to be the part of the previous record (relevant to #222 as we do not check predecessor).
'Explicit' marker is overwritten to be as in the last record segment. So the fdata + fileheader were considered to be one "explicit" record.
Then find_fileheaders searches explicit records of type 0 (FHLR is 0), but because FDATA is also 0, 'combined' record was assumed to be the start of a new LF.
Which then caused issues while parsing the record as explicit, and debugging was not straightforward.
The text was updated successfully, but these errors were encountered: