Re: Large text file to a CSV file...
From: Louiie Warren (lwarren_at_ticom.com)
Date: 05/25/04
- Next message: Paul: "OK -- who else finds they can no longer import a named range using Access 2000/2002 (it was fine using 97) Any solution?"
- Previous message: Louie Warren: "Re: Large text file to a CSV file..."
- In reply to: John Nurick: "Re: Large text file to a CSV file..."
- Next in thread: Ted Allen: "Re: Large text file to a CSV file..."
- Reply: Ted Allen: "Re: Large text file to a CSV file..."
- Reply: John Nurick: "Re: Large text file to a CSV file..."
- Messages sorted by: [ date ] [ thread ]
Date: Tue, 25 May 2004 06:38:59 -0700
There's the rub... I would have to write a preprocessor
routine because a 3,094,112KB file would be difficult (to
say the least) to edit. WSP does imbed the - when there
is no value. I may be finding that this is more work than
I need to go through.
>-----Original Message-----
>Hi Louie,
>
>Access's text import routine has problems with files like
this, because
>the space is not just being used as a delimiter, it also
appears in the
>data without being "qualified" by enclosing the entire
field in quote
>marks.
>
>It looks as if the problematic field is the datestamp.
Using "-" for an
>unknown username shouldn't be a problem because it
doesn't introduce any
>extra spaces, though if WebSite Pro uses " - " it would.
>
>One approach would be to pre-process the file by simply
replacing the [
>and ] around the datestamp with " and ". That should
convert it into a
>regular space-delimited file, with fields that may
contain a space
>qualified with " " . It would then be possible to import
it to a
>temporary Access table, where you could run a query to
parse the
>datestamp field and convert it to an Access date time
value as you
>append it to your "permanent" table.
>
>The pre-processing could be done in a text editor, or if
it's a regular
>task you could work up a VBA routine using Line Input #
to read the file
>line by line, the Replace() function to change the line,
and Print # to
>write it to a new file. Then import the new file.
>
>Another approach would be to use a regular expression
engine to parse
>the input line into fields and append these directly to
your table.
>
>
>On Fri, 21 May 2004 09:50:16 -0700, "Louie Warren"
<lwarren@ticom.com>
>wrote:
>
>>I have a WebsitePro Access Log Combined (W3C/NCSA)
Format
>>file; and yes, I know there is an ODBC routine out there
>>to do this, but my company doesn't want to spend the
money
>>to get it because we are abandoning WebSite Pro in the
>>next few months. The data is on a classified system so
I
>>will attempt to recreate the problem with generic data.
I
>>already see problems with the blank delimiters...
>>
>>An example from the book (string these together with a
>>space between fields.
>>
>>172.18.96.157 (IP Address)
>>
>>gryphonaire.com (Hostname of server
>>receiving request)
>>
>>Admin (Authentication user
name)
>>
>>[02/Jan/2001:09:44:56 - 0800] (Date timestamp UTC
offset)
>>
>>"GET /members.html HTTP/1.1" (Complete HTTP request)
>>
>>200 (HTTP Response Code)
>>
>>178 (Number of bytes
transfered)
>>
>>"http://gryphonaire.com/index.html" (Complete referring
>>URL)
>>
>>"Mozilla/4.0 (compatible; MSIE 4.01; Windows NT)" (ID
of
>>browser)
>>
>>"Member=WebSiteRules" (Cookie)
>>
>>The problem in the Access import is if the
Authentication
>>username isn't available. WSP puts in a - This is
>>common for most of the requests. When an actual ID is
>>present, it doesn't fit the pattern with the others.
The
>>data is consistantly like the above example (", [, and
>>all).
>>
>>Hopefully that explains it.
>>
>>L
>
>--
>John Nurick [Microsoft Access MVP]
>
>Please respond in the newgroup and not by email.
>.
>
- Next message: Paul: "OK -- who else finds they can no longer import a named range using Access 2000/2002 (it was fine using 97) Any solution?"
- Previous message: Louie Warren: "Re: Large text file to a CSV file..."
- In reply to: John Nurick: "Re: Large text file to a CSV file..."
- Next in thread: Ted Allen: "Re: Large text file to a CSV file..."
- Reply: Ted Allen: "Re: Large text file to a CSV file..."
- Reply: John Nurick: "Re: Large text file to a CSV file..."
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|