Issue #19402 has been reported by jamie_ca (Jamie Macey). ---------------------------------------- Bug #19402: CSV skip_lines option not behaving as documented https://bugs.ruby-lang.org/issues/19402 * Author: jamie_ca (Jamie Macey) * Status: Open * Priority: Normal * ruby -v: ruby 3.2.0 (2022-12-25 revision a528908271) [x86_64-darwin21] * Backport: 2.7: UNKNOWN, 3.0: UNKNOWN, 3.1: UNKNOWN, 3.2: UNKNOWN ---------------------------------------- The [CSV documentation](https://ruby-doc.org/stdlib-3.1.0/libdoc/csv/rdoc/CSV.html#class-CSV-label-O...) for the `skip_lines` parser option says "If a String, converts it to a Regexp, ignores lines that match it." Application behaviour as well as [the source](https://github.com/ruby/csv/blob/master/lib/csv/parser.rb#L909-L919) appears to be normalizing the string encoding and running a simple substring check instead. Given the existing behaviour, this might just want a documentation update to describe it accurately? I stumbled across this on a project still on ruby 2.6.9 ([2.6 docs](https://ruby-doc.org/stdlib-2.6.1/libdoc/csv/rdoc/CSV.html#method-c-new)), but it's applicable still at 3.2.0. Reproduction script: ```ruby require 'csv' data = <<CSV data,data test,data data,test CSV puts "Parsing with regexp skip_lines /^test/, expect 2 rows" CSV.parse(data, skip_lines: /^test/).each { |row| pp row } puts puts "Parsing with text skip_lines \"^test\", expect 2 rows" CSV.parse(data, skip_lines: "^test").each { |row| pp row } puts puts "Parsing with unanchored text skip_lines \"test\", expect 1 row" CSV.parse(data, skip_lines: "test").each { |row| pp row } puts ``` ``` $ ruby csv_test.rb Parsing with regexp skip_lines /^test/, expect 2 rows ["data", "data"] ["data", "test"] Parsing with text skip_lines "^test", expect 2 rows ["data", "data"] ["test", "data"] ["data", "test"] Parsing with unanchored text skip_lines "test", expect 1 row ["data", "data"] ``` -- https://bugs.ruby-lang.org/