OpenEuphoria

2. Re: Parsing problem

• Posted by don cole <doncole at pacbell.net> Aug 16, 2005
• 587 views

Patrick Barnes wrote:
> 
> I have some finance data that my bank insists on lumping all together
> into one field. A sample of the data is below:
> 
> Visa Purchase                26DECThe Carss Park Super Carss Par
> Non Stg/Bsa Atm Wdl Fee
> Internet Deposit             05JAN23:11itamoney
> Visa Purchase                03JANWorld Vision Of Aust Burwood E
> Adi Limited28686
> Visa Cash Advance            07JANEur210.00 Banco Di Brescia
> Visa Purchase                18JANOptus Tv/Net Autopay Chatswood
> Internet Deposit             25JAN10:05shopping
> Atm Withdrawal               25JAN11:09Westpaccrlngfrd 2 O/S    Carlingfor2=
>   Au
> Adi Limited28686
> Atm Withdrawal               27JAN07:37St.George Telopea Hair Telopea Nsw \=
> Au
> Atm Withdrawal -Cba           27JAN09:20Cba Atm  Uts B'Way Op   Nsw 228498 =
>   Aus
> Visa Cash Advance            21JANEur120.00 Banco Di Brescia
> Atm Withdrawal -Cba           28JAN08:48Cba Atm  Uts B'Way Op   Nsw 228498 =
>   Aus
> Visa Purchase                25JANWoolworths W1122    Carlingfo
> Visa Purchase                26JANColes Express Dundas Dundas
> Visa Purchase                25JANVodafone            Chatswood
> Internet Deposit             29JAN09:57billls....
> Eftpos Purchase              29JAN10:28N R M A Mcc        North Ryde
> Atm Withdrawal - O/Bank       31JAN12:34Garden Island           Sydney     =
>   Au
> O/Seas Cash Withdrawal Fee
> Visa Purchase                27JANBilo Telopea 4106   Telopea
> Non Stg/Bsa Atm Wdl Fee
> Atm Withdrawal               01FEB19:25St.George Telopea Hair Telopea Nsw \=
> Au
> Visa Purchase                29JANIkea Homebush Bay   Rhodes
> Visa Purchase                28JANWoolworths W1200    Eastwood
> Visa Purchase                29JANPlatinum Communicatn Rhodes
> Visa Purchase                29JANCaci Clinic         Edgecliff
> Visa Purchase                31JANWorld Vision Of Aust Burwood E
> Visa Purchase                28JANVodafone            Chatswood
> Visa Purchase                31JANWoolworths W1200    Eastwood
> Tfr Wdl Bpay Internet         03FEB19:0311174893           Integral Energy
> 
> Each line is a single field in the incoming .csv file.
> Any suggestions on how to parse it? In most but not all cases, the
> second 'column' starts at about the 29th element. Sometimes the date
> is given, sometimes data and time, sometimes nothing.
> 
> --
> MrTrick
> ----------
> 
> 

I would start of with something like:

function splitline(sequence line) integer a a=match(" ",line) return {trim(line[1..a])}&{trim(line[a..length(line)])} end function

---first pass---------- while 1 do line=gets(fn) if match(" ",line) then good=splitline(line) else good=line end if end while

------------second pass------- col=repeat({},2) col=repeat(col,length(good)) for x=1 to length(good) do if length(good)=2 then col[1]=good[1] col[2]=good[2] else col[1]=good[1] col[2]={} end if end for

from there you could put in match("JAN",col[2])

or

for x=1 to length(SHORTMONTHS) dosee my moredates if match(SHORTMONS[x],col[2]) then a=match(SHORTMONTHS[x],col[2] month=col[2][1..3] day=trim(col[2][4..5]) end if end for etc. etc.. you could cut out any information you wanted in this fashsion.

Don Cole, SF }}}

4. Re: Parsing problem

• Posted by DB James <larch at adelphia.net> Aug 16, 2005
• 620 views

Patrick Barnes wrote:
> 
> I have some finance data that my bank insists on lumping all together
> into one field. A sample of the data is below:
> 
> Visa Purchase                26DECThe Carss Park Super Carss Par
> Non Stg/Bsa Atm Wdl Fee
> Internet Deposit             05JAN23:11itamoney
<SNIP>
> Visa Purchase                31JANWorld Vision Of Aust Burwood E
> Visa Purchase                28JANVodafone            Chatswood
> Visa Purchase                31JANWoolworths W1200    Eastwood
> Tfr Wdl Bpay Internet         03FEB19:0311174893           Integral Energy
> 
> Each line is a single field in the incoming .csv file.
> Any suggestions on how to parse it? In most but not all cases, the
> second 'column' starts at about the 29th element. Sometimes the date
> is given, sometimes data and time, sometimes nothing.
> 
> --
> MrTrick
> ----------

Hi.

It appears as if the second set of data in each line always begins with a
number, so you have a basis for separation there, as well as the tabbing or
spacing otherwise.  (Question: are the wide spaces tabs originally?  If so, that
would make parsing easy.)  Also the first part of the second set either ends with
a number, or it ends with a three-letter word for the month, so this can create a
"rule" for separation of the second set.

It would be handy if the bank would provide a copy of their rules for
generationg these lines (some techy deep in the bowels of the bank would know,
but they may not let him/her out of the cage for public communication :^D )

--Quark

6. Re: Parsing problem

• Posted by jacques deschênes <desja at globetrotter.net> Aug 17, 2005
• 592 views

Hi Patrick,
First thing to do,
complain to your bank.

Regards,
jacques Deschênes

8. Re: Parsing problem

• Posted by "Kat" <gertie at visionsix.com> Aug 17, 2005
• 596 views

On 16 Aug 2005, at 22:26, Patrick Barnes wrote:

> 
> I have some finance data that my bank insists on lumping all together
> into one field. A sample of the data is below:
> 
> Visa Purchase                26DECThe Carss Park Super Carss Par
> Non Stg/Bsa Atm Wdl Fee
> Internet Deposit             05JAN23:11itamoney
> Visa Purchase                03JANWorld Vision Of Aust Burwood E
> Adi Limited28686
> Visa Cash Advance            07JANEur210.00 Banco Di Brescia
> Visa Purchase                18JANOptus Tv/Net Autopay Chatswood
> Internet Deposit             25JAN10:05shopping
> Atm Withdrawal               25JAN11:09Westpaccrlngfrd 2 O/S    Carlingfor2 
> Au
> Adi Limited28686 Atm Withdrawal               27JAN07:37St.George Telopea Hair
> Telopea Nsw \Au Atm Withdrawal -Cba           27JAN09:20Cba Atm  Uts B'Way Op 
>
> Nsw 228498   Aus Visa Cash Advance            21JANEur120.00 Banco Di Brescia
> Atm Withdrawal -Cba           28JAN08:48Cba Atm  Uts B'Way Op   Nsw 228498  
> Aus
> Visa Purchase                25JANWoolworths W1122    Carlingfo Visa Purchase 
>
>             26JANColes Express Dundas Dundas Visa Purchase               
> 25JANVodafone            Chatswood Internet Deposit            
> 29JAN09:57billls.... Eftpos Purchase              29JAN10:28N R M A Mcc       
> North Ryde Atm Withdrawal - O/Bank       31JAN12:34Garden Island          
> Sydney       Au O/Seas Cash Withdrawal Fee Visa Purchase               
> 27JANBilo Telopea 4106   Telopea Non Stg/Bsa Atm Wdl Fee Atm Withdrawal       
>
>      01FEB19:25St.George Telopea Hair Telopea Nsw \Au Visa Purchase           
>
>   29JANIkea Homebush Bay   Rhodes Visa Purchase                28JANWoolworths
> W1200    Eastwood Visa Purchase                29JANPlatinum Communicatn
> Rhodes
> Visa Purchase                29JANCaci Clinic         Edgecliff Visa Purchase 
>
>             31JANWorld Vision Of Aust Burwood E Visa Purchase               
> 28JANVodafone            Chatswood Visa Purchase               
> 31JANWoolworths
> W1200    Eastwood Tfr Wdl Bpay Internet         03FEB19:0311174893          
> Integral Energy

Could you send me the file without email linewraps? Or however it's done to 
post a .zip to the euforum? Or something else?

> Each line is a single field in the incoming .csv file.
> Any suggestions on how to parse it? In most but not all cases, the
> second 'column' starts at about the 29th element. Sometimes the date
> is given, sometimes data and time, sometimes nothing.

I have an idea that's easy with strtok v3 (not uploaded yet, not finished
either)
parses(), if i understand what's there. Can you provide any more info on 
what's on each line?

Kat

10. Re: Parsing problem

• Posted by "Kat" <gertie at visionsix.com> Aug 17, 2005
• 593 views

On 17 Aug 2005, at 21:25, Patrick Barnes wrote:

> 
> On 8/16/05, Kat <gertie at visionsix.com> wrote:
> > Could you send me the file without email linewraps? Or however it's done to
> > post a .zip to the euforum? Or something else?
> 
> No. This is my bank transaction history. 

I meant only what you posted to euforum anyhow, nothing additional.

> > I have an idea that's easy with strtok v3 (not uploaded yet, not finished
> > either) parses(), if i understand what's there. Can you provide any more
> > info
> > on what's on each line?
> 
> Thanks, but that's fine now, I've used the suggestions in the forum,
> an 'explode', 'trim', and a func to split at 2-or-more-space
> boundaries... and it works nicely.

Ok, no pressing need for strtok v3 noted.

Kat