Merge pull request #56 from ceedubs/document-limitations

Document limitations

README.md

@@ -59,6 +59,23 @@ Gen.listOfN(3, phraseGen).apply(Gen.Parameters.default, Seed(1046527L))
 // )
 ```
 
+## warnings and limitations
+
+At the moment, irrec does not support the following features:
+
+* **capturing and backreferences** (ex: using `hello, (\w+)` and capturing the string that matched `\w+`)
+* **lookarounds** (ex: `(?=...)` for positive lookahead, `?<=...)` for negative lookahead)
+* **non-greedy matches** (ex: `.*?`)
+* **inline modifiers** (ex: `(?i)` for case-insensitive mode)
+* **anchors** (ex: `\b`, `^`, `$`)
+* some other odds and ends that are probably mostly straightforward to support if desired
+
+Irrec only checks that an entire string matches a regex, as though the regex had a leading `^` and a trailing `$`.
+
+The wildcard `.` matches on newlines (consistent with a `DOTALL` flag being turned on in some regular expression implementations).
+
+Some of these limitations would probably be pretty easy to remove. Others might be tough to address with the current design of irrec.
+
 ## getting irrec
 
 If you are using SBT, you can add irrec as a dependency to your project with:

docs/README.md

@@ -53,6 +53,23 @@ val phraseGen: Gen[String] = regexMatchingStringGen(arbitrary[Char])(phrase)
 Gen.listOfN(3, phraseGen).apply(Gen.Parameters.default, Seed(1046527L))
 ```
 
+## warnings and limitations
+
+At the moment, irrec does not support the following features:
+
+* **capturing and backreferences** (ex: using `hello, (\w+)` and capturing the string that matched `\w+`)
+* **lookarounds** (ex: `(?=...)` for positive lookahead, `?<=...)` for negative lookahead)
+* **non-greedy matches** (ex: `.*?`)
+* **inline modifiers** (ex: `(?i)` for case-insensitive mode)
+* **anchors** (ex: `\b`, `^`, `$`)
+* some other odds and ends that are probably mostly straightforward to support if desired
+
+Irrec only checks that an entire string matches a regex, as though the regex had a leading `^` and a trailing `$`.
+
+The wildcard `.` matches on newlines (consistent with a `DOTALL` flag being turned on in some regular expression implementations).
+
+Some of these limitations would probably be pretty easy to remove. Others might be tough to address with the current design of irrec.
+
 ## getting irrec
 
 If you are using SBT, you can add irrec as a dependency to your project with:

regex/src/main/scala/Glushkov.scala

@@ -38,7 +38,6 @@ object Glushkov {
  case KleeneF.Zero => List.empty
  }
 
- // TODO ceedubs formatting
  def kleeneLocalTransitions[I, A](k: KleeneF[LocalLanguage[I, A]])(
  implicit orderingI: Ordering[I]): SortedMap[I, List[(I, A)]] = {
  implicit val orderI: Order[I] = Order.fromOrdering(orderingI)