Cetin

Thanks for that.

>Terry,
>As per being strange it was not at all and you already got plenty of explanations.
>Tilde usage is an indirect 'set exact' implementation. Adding tilde to both sides you introduce a word boundary so if they match they should match up to and including tilde:
>"137~" = "1~"
>would be false for example.

Tell you what, I've never come across that notation - that's why it's strange to me - never used it (I MAY in future). And where are the lots of examples you refer to - on the UT or in this thread? I don't recall seeing either.
>
>Anyway I jumped in to say that Terry's < bg > suggestion is what I had in mind as a solution to your problem. IMHO it's easier and faster to manipulate an intermediate cursor than using fgets() series (at least what I do in my generic importer class).
>Cetin

I don't think it's as simple as that. This is so specialised that it couldn't be a "generic" class. I meant "generic" as in any form of importing.
Below is a small sample of one of the source files, which might run to tens of thousands of lines. The data represented pertains to 5 of my tables, and is read twice, in different modes. IMO it can't be slap-banged into an int. cursor, as the structure changes repeatedly. As such, I'd need the LL routines to get it into a cursor anyway, so what would be the saving? Lower in the file it takes several other forms again, with "note" records, details of bus operators et al. The process is so slow that I'd love to be able to!
...
QSNAN 1535 20020225204912310010000 X193 BUS 5
ZOSCC FP
QN1 Sponsored by Suffolk County Council
QO390060872 1035 T1
QI390060242 10361036B T0
QI390060927 10381038B T0
QI390032950 10401040B T0
QI390060024 10421042B T0
QI390060027 10451045B T1
QI390060054 10471047B T0
QI390060055 10491049B T0
QI390070356 10511051B T0
QI390070354 10531053B T0
QT390070351 1057 T1
QSNAN 1535 20020225204912310010000 X193 BUS 5
ZOSCC FP
QN1 Sponsored by Suffolk County Council
QO390060872 1350 T1
QI390060242 13511351B T0
QI390060927 13531353B T0
QI390032950 13551355B T0
QI390060024 13571357B T0
QI390060027 14001400B T1
QI390060054 14021402B T0
QI390060055 14041404B T0
QI390070356 14061406B T0
QI390070354 14081408B T0
QT390070351 1412 T1
QSNAN 1534 20020225204912311000100 X195 BUS 5
ZOSCC FP
QN1 Sponsored by Suffolk County Council
QO390060031 0908 T1
QI390032950 09100910B T0
QI390060870 09130913B T1
QI390032950 09150915B T0
QI390070348 09180918B T1
QI390032950 09200920B T0
QI390060035 09220922B T0
QI390032950 09240924B T1
QI390060036 09250925B T0
QI390060033 09260926B T1
QI390060222 09280928B T0
QI390060224 09310931B T1
QI390060221 09370937B T1
QI390032950 09390939B T0
QI390060868 09410941B T0
QT390060208 0944 T1
...

Another of the files in in the following format:
...
QLN390030980 Ipswich: Sandyhill Lane, Roebeck Road E0057632
QBN390030980 617249 242417 Ipswich
QLN390030981 Ipswich: Sandyhill Lane, opp Roebeck Road E0057632
QBN390030981 617238 242410 Ipswich
QLN390030982 Ipswich: Tuddenham Road, Millenium Cemetery E0057632
QBN390030982 617982 246751 Ipswich
QLN390030002 Ipswich: St Helens Street, opp Dove Street E0057632
QBN390030002 617125 244541 Ipswich
QLN390030003 Ipswich: St Helens Street, Majors Corner E0057632
QBN390030003 616693 244637 Ipswich
...
and again this changes in the file. If you know of a way of doing this I'd be interested.

Cheers

Terry