![]() | |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
Flimport to check record length, it shows a result of 32500. Very odd. Any thoughts on this issue? |
#3
| |||
| |||
|
|
JoeSpareBedroom wrote: Flimport to check record length, it shows a result of 32500. Very odd. Any thoughts on this issue? Is there a CR/LF (hex 0D 0A) pair at the end of each line? |
#4
| |||
| |||
|
|
Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. |
#5
| |||
| |||
|
|
JoeSpareBedroom wrote: Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. One other thought... some programs that create text files put an end-of-file marker (non-visible) after the last CR/LF. My experience is that Paradox doesn't like it. I suspect that when you copy/paste you are not bringing along that EOF byte. Try deleting it with a hex editor or even using the backspace key after opening the file in Notepad. Then re-save and re-import. |
#6
| |||
| |||
|
|
As Ed has covered, it's most likely that each line is not ending with the CR/LF pair (Hex 0D/0A). You should run a HEX viewer to check the original file - any other "editor" may lie to you - some will automatically repair a lone CR, or a lone LF, or a reversed LF/CR, to a "proper" CR/LF pair, and then you'll not be any wiser at the end. If it's a CR/LF issue, then you'll need to trace back to the source of this file - I expect someone there took a wrong option... or prepared the file for a UNIX platform... or... Ed has also covered the EOF issue... Finally, all the PDox "Import" routines - that I'm familiar with - are not very smart on CSV files, where the field separator char (usually ,) and/or the field delimiter char (usually ") are embedded within string fields. - Mike "Ed Nash" <ask.me (AT) specifically (DOT) pls> wrote in message news:45ad19a3 (AT) pnews (DOT) thedbcommunity.com... JoeSpareBedroom wrote: Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. One other thought... some programs that create text files put an end-of-file marker (non-visible) after the last CR/LF. My experience is that Paradox doesn't like it. I suspect that when you copy/paste you are not bringing along that EOF byte. Try deleting it with a hex editor or even using the backspace key after opening the file in Notepad. Then re-save and re-import. |
#7
| |||
| |||
|
|
This was a fixed length file, so delimiters weren't the problem, although I know what you mean. The files I get often have commas in the product descriptions. Since they serve no useful purpose, in terms of legibility, I do a search/replace and turn them into spaces. As far the source of the file, you're probably correct about the "wrong turn". I have no control over what the vendors send us, unfortunately. A few of the companies have had consistent procedures in place for 10 years, as evidenced by the age of my flimport spec files. Others think it's fun to change their formats every 3 months. "Michael Kennedy" <Info (AT) KennedySoftware (DOT) ie> wrote in message news:cWarh.17508$j7.339594 (AT) news (DOT) indigo.ie... As Ed has covered, it's most likely that each line is not ending with the CR/LF pair (Hex 0D/0A). You should run a HEX viewer to check the original file - any other "editor" may lie to you - some will automatically repair a lone CR, or a lone LF, or a reversed LF/CR, to a "proper" CR/LF pair, and then you'll not be any wiser at the end. If it's a CR/LF issue, then you'll need to trace back to the source of this file - I expect someone there took a wrong option... or prepared the file for a UNIX platform... or... Ed has also covered the EOF issue... Finally, all the PDox "Import" routines - that I'm familiar with - are not very smart on CSV files, where the field separator char (usually ,) and/or the field delimiter char (usually ") are embedded within string fields. - Mike "Ed Nash" <ask.me (AT) specifically (DOT) pls> wrote in message news:45ad19a3 (AT) pnews (DOT) thedbcommunity.com... JoeSpareBedroom wrote: Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. One other thought... some programs that create text files put an end-of-file marker (non-visible) after the last CR/LF. My experience is that Paradox doesn't like it. I suspect that when you copy/paste you are not bringing along that EOF byte. Try deleting it with a hex editor or even using the backspace key after opening the file in Notepad. Then re-save and re-import. |
#8
| |||
| |||
|
|
Just a thought; Can it be linked to a smallint limit? Régis "JoeSpareBedroom" <dishborealis (AT) yahoo (DOT) com> a écrit dans le message de news: 82brh.725$ya1.389 (AT) news02 (DOT) roc.ny... This was a fixed length file, so delimiters weren't the problem, although I know what you mean. The files I get often have commas in the product descriptions. Since they serve no useful purpose, in terms of legibility, I do a search/replace and turn them into spaces. As far the source of the file, you're probably correct about the "wrong turn". I have no control over what the vendors send us, unfortunately. A few of the companies have had consistent procedures in place for 10 years, as evidenced by the age of my flimport spec files. Others think it's fun to change their formats every 3 months. "Michael Kennedy" <Info (AT) KennedySoftware (DOT) ie> wrote in message news:cWarh.17508$j7.339594 (AT) news (DOT) indigo.ie... As Ed has covered, it's most likely that each line is not ending with the CR/LF pair (Hex 0D/0A). You should run a HEX viewer to check the original file - any other "editor" may lie to you - some will automatically repair a lone CR, or a lone LF, or a reversed LF/CR, to a "proper" CR/LF pair, and then you'll not be any wiser at the end. If it's a CR/LF issue, then you'll need to trace back to the source of this file - I expect someone there took a wrong option... or prepared the file for a UNIX platform... or... Ed has also covered the EOF issue... Finally, all the PDox "Import" routines - that I'm familiar with - are not very smart on CSV files, where the field separator char (usually ,) and/or the field delimiter char (usually ") are embedded within string fields. - Mike "Ed Nash" <ask.me (AT) specifically (DOT) pls> wrote in message news:45ad19a3 (AT) pnews (DOT) thedbcommunity.com... JoeSpareBedroom wrote: Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. One other thought... some programs that create text files put an end-of-file marker (non-visible) after the last CR/LF. My experience is that Paradox doesn't like it. I suspect that when you copy/paste you are not bringing along that EOF byte. Try deleting it with a hex editor or even using the backspace key after opening the file in Notepad. Then re-save and re-import. |
#9
| |||
| |||
|
|
This was a fixed length file, so delimiters weren't the problem, although I know what you mean. The files I get often have commas in the product descriptions. Since they serve no useful purpose, in terms of legibility, I do a search/replace and turn them into spaces. As far the source of the file, you're probably correct about the "wrong turn". I have no control over what the vendors send us, unfortunately. A few of the companies have had consistent procedures in place for 10 years, as evidenced by the age of my flimport spec files. Others think it's fun to change their formats every 3 months. "Michael Kennedy" <Info (AT) KennedySoftware (DOT) ie> wrote in message news:cWarh.17508$j7.339594 (AT) news (DOT) indigo.ie... As Ed has covered, it's most likely that each line is not ending with the CR/LF pair (Hex 0D/0A). You should run a HEX viewer to check the original file - any other "editor" may lie to you - some will automatically repair a lone CR, or a lone LF, or a reversed LF/CR, to a "proper" CR/LF pair, and then you'll not be any wiser at the end. If it's a CR/LF issue, then you'll need to trace back to the source of this file - I expect someone there took a wrong option... or prepared the file for a UNIX platform... or... Ed has also covered the EOF issue... Finally, all the PDox "Import" routines - that I'm familiar with - are not very smart on CSV files, where the field separator char (usually ,) and/or the field delimiter char (usually ") are embedded within string fields. - Mike "Ed Nash" <ask.me (AT) specifically (DOT) pls> wrote in message news:45ad19a3 (AT) pnews (DOT) thedbcommunity.com... JoeSpareBedroom wrote: Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. One other thought... some programs that create text files put an end-of-file marker (non-visible) after the last CR/LF. My experience is that Paradox doesn't like it. I suspect that when you copy/paste you are not bringing along that EOF byte. Try deleting it with a hex editor or even using the backspace key after opening the file in Notepad. Then re-save and re-import. |
#10
| |||
| |||
|
|
the only problem with flimport is that it's functionality is so limited, if the file and/or the results aren't straight-forward.. but in your case they seem to be.. I never used it regularly.. it was too easy to build an import shell that can eat *any* file.. -- Steven Green - Waldorf Maryland USA Diamond Software Group http://www.diamondsg.com/main.htm Paradox Support & Sales Diamond Sports Gems http://www.diamondsg.com/gemsmain.htm Sports Memorabilia and Trading Cards "JoeSpareBedroom" <dishborealis (AT) yahoo (DOT) com> wrote in message news:82brh.725$ya1.389 (AT) news02 (DOT) roc.ny... This was a fixed length file, so delimiters weren't the problem, although I know what you mean. The files I get often have commas in the product descriptions. Since they serve no useful purpose, in terms of legibility, I do a search/replace and turn them into spaces. As far the source of the file, you're probably correct about the "wrong turn". I have no control over what the vendors send us, unfortunately. A few of the companies have had consistent procedures in place for 10 years, as evidenced by the age of my flimport spec files. Others think it's fun to change their formats every 3 months. "Michael Kennedy" <Info (AT) KennedySoftware (DOT) ie> wrote in message news:cWarh.17508$j7.339594 (AT) news (DOT) indigo.ie... As Ed has covered, it's most likely that each line is not ending with the CR/LF pair (Hex 0D/0A). You should run a HEX viewer to check the original file - any other "editor" may lie to you - some will automatically repair a lone CR, or a lone LF, or a reversed LF/CR, to a "proper" CR/LF pair, and then you'll not be any wiser at the end. If it's a CR/LF issue, then you'll need to trace back to the source of this file - I expect someone there took a wrong option... or prepared the file for a UNIX platform... or... Ed has also covered the EOF issue... Finally, all the PDox "Import" routines - that I'm familiar with - are not very smart on CSV files, where the field separator char (usually ,) and/or the field delimiter char (usually ") are embedded within string fields. - Mike "Ed Nash" <ask.me (AT) specifically (DOT) pls> wrote in message news:45ad19a3 (AT) pnews (DOT) thedbcommunity.com... JoeSpareBedroom wrote: Not that I can see in WinEdit. However: - If I open the text file in Word using the "Show All" option, there's a paragraph mark at the end of each line. - If I copy the text from the offending file into a new file, the import works. It's a minor extra step, but still, I'm curious. One other thought... some programs that create text files put an end-of-file marker (non-visible) after the last CR/LF. My experience is that Paradox doesn't like it. I suspect that when you copy/paste you are not bringing along that EOF byte. Try deleting it with a hex editor or even using the backspace key after opening the file in Notepad. Then re-save and re-import. |
![]() |
| Thread Tools | |
| Display Modes | |
| |