Write better URL regular expressions for our needs.

2 files changed, 133 insertions, 22 deletions

diff --git a/Hotline/Utility/FoundationExtensions.swift b/Hotline/Utility/FoundationExtensions.swift
index c154dbd..6270ac7 100644
--- a/Hotline/Utility/FoundationExtensions.swift
+++ b/Hotline/Utility/FoundationExtensions.swift
@@ -1,4 +1,5 @@
 import Foundation
+import SwiftUI
 
 enum Endianness {
   case big
@@ -6,23 +7,13 @@ enum Endianness {
 }
 
 extension String {
-  func convertToAttributedStringWithLinks(relaxed: Bool = false) -> AttributedString {
+  func convertToAttributedStringWithLinks() -> AttributedString {
     let attributedString: NSMutableAttributedString = NSMutableAttributedString(string: self)
-    let urlPattern = relaxed ?
-      #"(hotline|http|https)?(://)?[\w-]+(\.[\w-]+)+([\w.,@?^=%&:/~+#-]*[\w@?^=%&/~+#-])?"# :
-      #"(hotline|http|https)://[\w-]+(\.[\w-]+)+([\w.,@?^=%&:/~+#-]*[\w@?^=%&/~+#-])?"#
-    
-    guard let regex = try? NSRegularExpression(pattern: urlPattern, options: .caseInsensitive) else {
-      return AttributedString(attributedString)
-    }
-
-    regex.enumerateMatches(in: self, range: NSMakeRange(0, self.count)) { (result: NSTextCheckingResult!, _, _) in
-      if let newRange = Range(result.range, in: self) {
-        let str = String(self[newRange])
-        attributedString.addAttribute(.link, value: str, range: result.range)
-      }
+    let matches = self.ranges(of: RegularExpressions.relaxedLink)
+    for match in matches {
+      attributedString.addAttribute(.link, value: self[match], range: NSRange(match, in: self))
+//      attributedString.addAttribute(.underlineStyle, value: 1, range: NSRange(match, in: self))
     }
-    
     return AttributedString(attributedString)
   }
   
@@ -51,14 +42,18 @@ extension String {
     }
   }
   
-  func convertLinksToMarkdown() -> String {
-    let urlPattern = #"(hotline|http|https)?(://)?[\w-]+(\.[\w-]+)+([\w.,@?^=%&:/~+#-]*[\w@?^=%&/~+#-])?"#
-    
-    guard let regex = try? NSRegularExpression(pattern: urlPattern, options: .caseInsensitive) else {
-      return self
+  func convertingLinksToMarkdown() -> String {
+    var cp = String(self)
+    cp.replace(RegularExpressions.relaxedLink) { match -> String in
+      let linkText = self[match.range]
+      var injectedScheme = "https://"
+      if let _ = try? RegularExpressions.supportedLinkScheme.prefixMatch(in: linkText) {
+        injectedScheme = ""
+      }
+      
+      return "[\(linkText)](\(injectedScheme)\(linkText))"
     }
-    
-    return regex.stringByReplacingMatches(in: self, range: NSRange(location: 0, length: self.count), withTemplate: "[$0]($0)")
+    return cp
   }
 }
 
diff --git a/Hotline/Utility/RegularExpressions.swift b/Hotline/Utility/RegularExpressions.swift
index 0314dd7..0a0dda9 100644
--- a/Hotline/Utility/RegularExpressions.swift
+++ b/Hotline/Utility/RegularExpressions.swift
@@ -20,4 +20,120 @@ struct RegularExpressions {
       }
     }
   }
+  
+  static let supportedLinkScheme = Regex {
+    Anchor.startOfLine
+    ChoiceOf {
+      "hotline"
+      "http"
+      "https"
+    }
+    "://"
+  }.ignoresCase().anchorsMatchLineEndings()
+  
+  static let relaxedLink = Regex {
+    ChoiceOf {
+      Anchor.startOfLine
+      Anchor.wordBoundary
+    }
+    Capture {
+      // scheme (optional)
+      Optionally {
+        ChoiceOf {
+          "hotline://"
+          "http://"
+          "https://"
+        }
+      }
+      // domain name
+      OneOrMore {
+        CharacterClass(
+          .anyOf(".-"),
+          ("a"..."z"),
+          ("0"..."9")
+        )
+      }
+      // top-level domain name
+      "."
+      ChoiceOf {
+        "com"
+        "net"
+        "org"
+        "edu"
+        "gov"
+        "mil"
+        "aero"
+        "asia"
+        "biz"
+        "cat"
+        "coop"
+        "info"
+        "int"
+        "jobs"
+        "mobi"
+        "museum"
+        "name"
+        "pizza"
+        "post"
+        "pro"
+        "red"
+        "tel"
+        "today"
+        "travel"
+        "ai"
+        "be"
+        "by"
+        "ca"
+        "co"
+        "de"
+        "er"
+        "es"
+        "fr"
+        "gs"
+        "ie"
+        "im"
+        "in"
+        "io"
+        "is"
+        "it"
+        "jp"
+        "la"
+        "ly"
+        "ma"
+        "md"
+        "me"
+        "my"
+        "nl"
+        "ps"
+        "pt"
+        "ja"
+        "st"
+        "to"
+        "tv"
+        "uk"
+        "ws"
+      }
+      // Port
+      Optionally {
+        ":"
+        OneOrMore {
+          CharacterClass(.digit)
+        }
+      }
+      // path
+      ZeroOrMore {
+        CharacterClass(
+          .anyOf("#_-/.?=&%\\()[]"),
+          ("a"..."z"),
+          ("0"..."9")
+        )
+      }
+    }
+    ChoiceOf {
+      Anchor.endOfLine
+      Anchor.wordBoundary
+    }
+  }
+  .anchorsMatchLineEndings()
+  .ignoresCase()
 }