Ruby regex parsing

Question

I'm trying to parse a text file. Occurrences of the following format are buried within continuous text (so they are never at the start of a line, for example):

"name":"Fred Flintstone","neighborhood":  ...
... "name":"Barney Rubble","address":

I need to find the occurrence of "name":. name appears in other places, so only the word name with the quotes and colon should match. Then I need to print or store the text inside the first pairs of quotes to follow. I'd like to have it clean with just Barney Rubble on one line, Fred Flintstone on another.

This is what I've come up with:

File.open('textfile.txt','r') do |s|
  s.each_line do |eachline|
    wordmatch = eachline.match(/"name":"(.*?)(?=["])/)
    puts wordmatch
  end
end

but it doesn't work. The results appear like:

(lots of space)
"name":"random"
(lots of space)
"name":"Barney Rubble

It prints lots of spaces. It also is not showing all results. I don't see why.

So, apologies if it's confusing. Just to clarify. after the parser finds "name": everything inside the first, immediately following set of quotes needs to be selected/stored/printed. in the first example only Fred Flintsone should be selected, nothing else until the next "name": is encountered. Any characters and amount of space inside the quotes is legitimate.

It looks like you have a JSON string. If it is, I'd rather use a JSON parser for this task. If it is not, you can try using .scan(/"name":"([^"]+)/). The captured texts are the ones you must be looking for. — Wiktor Stribiżew
– Wiktor Stribiżew, Commented Dec 5, 2015 at 9:58

Baldrick · Accepted Answer · 2015-12-05 12:33:30Z

3

You could do it with a non greedy expression:

s = '"name":"Fred Flintstone","neighborhood":"foo","name":"Barney Rubble","address":"bar"'
s.scan(/"name":"(.*?)"/).flatten  #  => ["Fred Flintstone", "Barney Rubble"]

edited Dec 5, 2015 at 12:33

answered Dec 5, 2015 at 10:13

Baldrick

24.4k6 gold badges76 silver badges80 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

boogiewonder Over a year ago

thanks. When I use this with the text file I get an 'Undefined method 'scan'. parse.rb:26:in block in <main>': undefined method scan' for #<File:textfile.txt (closed)> (NoMethodError)

Cary Swoveland Over a year ago

boggle, what do you mean, "When I use this with the text file"? Here s is a string consisting of the contents of the file, e.g., s = File.read(" my_file.txt").

boogiewonder Over a year ago

@CarySwoveland Sorry, just to clarify. I used s= File.open with a "r" and this caused the error for some reason. Using File.read (as you suggested) worked and didn't give the error above but the result is still the same unfortunately

Mark Reed · Accepted Answer · 2015-12-05 15:08:44Z

2

match only finds the first occurrence on a line; it sounds like you may have multiple matches per line, in which case you need to use scan with a loop body:

File.read('textfile.txt').scan(/"name":"([^"]*)"/) do |wordmatch|
  puts wordmatch
end

But that format looks suspisciously JSONlike, and if it's JSON, you should treat it as such:

require 'json'
require 'pp'

obj = JSON.parse(File.read 'textfile.txt')
pp obj

Then look at the structure, which is probably an array of hashes, so what you want is

puts obj.map { |o| o['name'] }

or similar.

answered Dec 5, 2015 at 15:08

Mark Reed

96k17 gold badges149 silver badges189 bronze badges

Comments

Nermin · Accepted Answer · 2015-12-05 09:59:29Z

1

You can use this regex pattern

/(?<="name":")([\w\s]+)/

Meaning:

(?<="name":") will look for occurrences of "name":", but will not include them in the result positive look-behind

([\w\s]+) will match string that contains letters or empty space, in your case until character ", that is the name

You can also check these sites: Rubular, Regex101, this can help you in building your regex

edited Dec 5, 2015 at 9:59

answered Dec 5, 2015 at 9:54

Nermin

6,11016 silver badges23 bronze badges

3 Comments

boogiewonder Over a year ago

unfortunately it's doing exactly the same thing with this regular expression. I'm starting to wonder if there's something wrong with the text file, although I can pick out every mention of 'name:' without problems. It just seems to be problematic when trying to find the following text.

Cary Swoveland Over a year ago

You also need a positive lookahead for a double-quote. If one is not found it's not a match. The question does not specify what characters are permitted between the double quotes, so you are have no basis for limiting the them to word characters and spaces.

boogiewonder Over a year ago

@CarySwoveland I just edited the original question and tagged on a clarification. I've tried everything suggested so far, to varying levels of success but nothing is giving me more than one or two of the values required.

Collectives™ on Stack Overflow

Ruby regex parsing

3 Answers 3

3 Comments

Comments

Meaning:

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

3 Comments

Comments

Meaning:

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related