I am writing a small windows script in javascript/jscript for finding a match for a regexp with a string that i got by manipulating a file.
I know the line number for starting and ending of the search. But I am stuck at this position for a while.
I am writing a small windows script in javascript/jscript for finding a match for a regexp with a string that i got by manipulating a file.
I know the line number for starting and ending of the search. But I am stuck at this position for a while.
You would need to match all the lines preceding the bit you're interested in, then capture the part you want in a capturing group. For example, to extract the fourth and fifth lines, you could use this regex:
/^(?:.*(?:\r\n|[\r\n])){3}((?:.*(?:\r\n|[\r\n])){2})/
The part you want will be available in capturing group #1.
The best way to look at something like this is to not overcomplicate your regular expression - you could make it rather unreadable and confusing, particularly if your regex skills aren't very good. Simplify things first - split the document into an array of lines, pick out your line and then perform your regular expression. Running your regular expression on a smaller string in this fashion could prove to be much faster on very large text files.
// Assuming file is a FileSystemObject TextStream object:
var lines = file.ReadAll().split("\r\n"); // split based on Windows newline
// Arrays have zero-based indexes, so the line you need is your line number -1
var match = lines[4].match(myRegEx); // lines[4] is line 5 of the string
// If you need it to span multiple lines, extract and join the lines you need:
var specificLines = lines.splice(4, 6); // extract the 5th and 6th lines
specificLines.match(myRegEx); // don't forget the regex's /m modifier
splice
returns the array items from the starting index upto and excluding the ending index. You can also provide a limit
parameter to the split
method to stop splitting after a certain number of matches. This should improve speed on a large file where the lines you need aren't near the end.
Glossary: