PARSE$() techniques

[Mike Trader]

I am writing a little utility to convert data. I want to be able to use it on different formats. I have most of the code written to handle different fields (date and time formats etc).

The thing that is giving me a little trouble is the first step, PARSEing out the various fields in a string:

hFile = FREEFILE
OPEN FilePathStr+FileNameStr FOR INPUT AS hFile LEN = 32768
INPUT #hFile, InputStr ' Read a line of Data
DateStr = PARSE$(InputStr, any ", ", 1)
TimeStr = PARSE$(InputStr, any ", ", 2)

Now this works for comma delimited fields like:
19900118, 1610, 340.7
or
19900118,1610,340.7

But I often get data with just spaces delimiting it:
19900118 1610 340.7
or
19900118 1610 340.7
or
19900118 1610 340.7

PARSEing on the space will only work if there is 1 space between each field. that is not allways the case

So how do you PARSE out the fields without getting into a cumbersome loop and checking each character for a space etc etc

Do you wizards have a neat little trick?


------------------
Kind Regards
Mike

[Mike Burns]

Mike,

You can use the Replace function to make the 2 spaces = 1 space
before doing the parse ... I had to do this once and it worked fine.
I knew none of the data fields would have a space and that all fields
would have a non-space character.

Replace " " with " " in InputStr

Mike




------------------

[Eric Pearson]

It sounds like you need a simple routine to "conform" the incoming data before you begin processing it...

REPLACE " " WITH "," IN sTarget$
DO
    IF INSTR(sTarget$,",,") = 0 THEN EXIT LOOP
    REPLACE ",," WITH "," IN sTarget$
LOOP

You could change the first REPLACE to REPLACE ANY if you also wanted to support semicolons, etc.

-- Eric


------------------
Perfect Sync Development Tools
Perfect Sync Web Site
Contact Us: mailto:[email protected][email protected]</A>

[This message has been edited by Eric Pearson (edited June 16, 2001).]

[Mike Trader]

This is waht I am trying to avoid:

hFile = FREEFILE
OPEN FilePathStr+FileNameStr FOR INPUT AS hFile LEN = 32768 ' Open data file for reading
'LINE INPUT #hFile, InputStr ' Discard first line of data
INPUT #hFile, InputStr ' Read first line of Data
IF InputStr = "" THEN EXIT FUNCTION ' end of data or a hole in data
DateStr = PARSE$(InputStr, any ", ", 1)
Done = 0
Count = 2
TimeStr = ""
WHILE TimeStr = "" AND Done = 0 ' figure out spaces B4 Time
TimeStr = PARSE$(InputStr, any ", ", Count)
INCR Count
DIALOG DOEVENTS : DIALOG DOEVENTS
WEND
TimeSpaces = Count - 1
PriceStr = ""
WHILE PriceStr = "" AND Done = 0 ' figure out spaces B4 Price
PriceStr = PARSE$(InputStr, any ", ", Count)
INCR Count
DIALOG DOEVENTS : DIALOG DOEVENTS
WEND
PriceSpaces = Count - 1


WHILE NOT EOF(hFile) AND Done = 0 ' go through the rest of the file
INPUT #hFile, InputStr ' Read subsequent lines of data
IF InputStr = "" THEN EXIT FUNCTION ' end of data or a hole in data
DateStr = PARSE$(InputStr, any ", ", 1)
TimeStr = PARSE$(InputStr, any ", ", TimeSpaces)
PriceStr = PARSE$(InputStr, any ", ", PriceSpaces)
Price = VAL(PriceStr)
'Process Date, Time Price
WEND



------------------
Kind Regards
Mike

[Eric Pearson]

If you use the code I posted to process your InputStr string, you can simplify that code a lot!

It will produce a string that can be used with PARSECOUNT and PARSE$ very easily. Nothing but data separated by commas.

-- Eric

------------------
Perfect Sync Development Tools
Perfect Sync Web Site
Contact Us: mailto:[email protected][email protected]</A>

[Trevor Lane]

Mike,

You must know something about the data, if you know it is always a date
followed by a number followed by say an email address then you have some
knowledge about what needs to be taken out. If the only thing you are
unsure about is the delimiter and you do not want to use the technique
that Eric suggested then you can look at regular expressions. Take a look
at REGEXP and REGREP see if you can do something with that. Unless the
data is quite complex though, you would be better off following Eric's advice.

Regards
Trevor

------------------

[Mike Trader]

Perfect!
Thank you Mike, Eric and Trevor. I learned two new functions that are exactly what I need.

This forum is so cool!

------------------
Kind Regards
Mike