How to escape curly braces and backslashes in YAML with Ruby?

Question

My app validates and imports data files for various vendors. One of the validations on the file will be to check whether it's filename is in the right format. I have my regex stored in my YAML file along with various other configurations as follows -

---
filename_regex: !ruby/regexp '/^input_.{1,10}_company_\d{8}.tsv$/'
check_header_row: true
foo: "Bar"
....

Essentially the above enforces that it has to be of format: input_companyName_company_20141127.tst . I'm loading the file with YAML.load(filename) in the app.

I'm running into 2 problems, as you may have already guessed

1. The curly braces don't escape correctly. When I take them out it works fine, so those are almost certainly the problem. Apparently \\ doesn't work for escaping, and neither does ' ?

2. The \\d needs to have it's own backslash escaped, I assume? In that case would I use \\\\d in the file? That didn't seem to work either.

Thanks!

Answer 1

I can't seem to replicate any problem with what you already have.

Don't forget you probably want to be escaping the . at the end for .tsv and as someone already mentioned use \\A and \\Z instead of ^ and $ like '/\\Ainput_.{1,10}_company_\\d{8}\\.tsv\\Z/'

regex.yml

---
filename_regex: !ruby/regexp '/^input_.{1,10}_company_\d{8}.tsv$/'
foo: 'bar'
---

regex.rb

require 'yaml'

cfg = YAML.load_file('regex.yml')

valid_test_name = 'input_ABCD_company_12345678.tsv'
invalid_company_test_name = 'input_CompanyTooLong_company_12345678.tsv'
invalid_date_test_name = 'input_Company_company_12345678901.tsv'

valid_pass = valid_test_name =~ cfg["filename_regex"] ? 'pass' : 'fail'
invalid_company_pass = invalid_date_test_name =~ cfg["filename_regex"] ? 'pass' : 'fail'
invalid_date_pass = invalid_date_test_name =~ cfg["filename_regex"] ? 'pass' : 'fail'

puts "#{valid_test_name}: #{valid_pass}"
puts "#{invalid_company_test_name}: #{invalid_company_pass}"
puts "#{invalid_date_test_name}: #{invalid_date_pass}"

Running it ruby regex.rb outputs

input_ABCD_company_12345678.tsv: pass
input_CompanyTooLong_company_12345678.tsv: fail
input_Company_company_12345678901.tsv: fail