Question

I have a CSV file that I want to parse using PHP (don’t want to use php csv functions since they’re too slow for my case). I use preg_replace to select only the columns that I need and write output back to standard out. I have a pattern that look like that

preg_replace("/^\"([^\"]*)\",\"([^\"]*)\"(.*)$/m", "$1;$2", $content);

There are some lines in the CSV that are malformed. Is there a way to skip output for lines that don’t match the pattern above?

Answer 1

You could do it in two steps:

First remove all lines that don’t match your regex:

preg_replace("/^(?!\"([^\"]*)\",\"([^\"]*)\".*$).*\r?\n/m", "", $content);

Then do the original regex replace.

However, you need to be very careful. As soon as you get linebreaks and/or escaped quotes in your CSV fields, these regexes may break and destroy your file.

Or imagine a malformed line like

"foo","bar (missing quote!)
"baz", "bam" (correct line)

The regex would now delete both the incomplete and the correct line because [^\"]* also eats the newlines.