sergeyglazyrindev
commented
2 years ago Hello guys! I forked this package cause owner disappeared. Hope, he will be back, but it would be easier to merge these changes back if he is back Link to my repo: create issue there and we'll discuss it.
Are you fixing a bug or providing a failing unit test to demonstrate a bug? Yes, For checking a valid URL
How do you reproduce it? By updating the regular expression format for the URL, by correcting the misplaced brackets and updating the character to '+'.The char in regular expression mean the means "0 or more occurrences of the preceding item" and + char means "1 or more occurrences of the preceding item" Also there was a misplaced bracket of the form: ( (exp1... ) OR exp2).... instead it should have been of the form: ( (exp1...) ) OR (exp2....) The specified after the form (.[a-zA-Z])* means it can have nil (.com/.eu) but updating it to '+' ensures (.com/.in/.xyz) be specified in the url to make it a valid url
What was the previous behavior? Earlier it used to return true for http://abcd or xyz/qw which infact is not a valid URL as the domain name is not specified(.com/.eu/.in should be in the URL) which in turn makes the 'govalidator package' work similar to the net/url package with a bug.
What is the current behavior? After modifying it, it returns false for http://abcd or wxe/123 or nmhih and in turn returns true for http://abcd.eu or xyz.ab or jkl.com
Explain why the changes are necessary To check for a valid URL, there are two packages net/url and the govalidator package.But if this package does not return the correct output, there is no difference between the faulty package(net/url) and this one. So to be at a higher level and simplify the complexities of the user this commit needs to be merged.Also in various projects the checking of a valid URL is a crucial part, hence forth moving forward.