Incorporated errata.

Makefile

@@ -25,6 +25,9 @@ reverse-dependency-test :
 		$(OCAMLBUILD) -clean && \
 		$(OCAMLBUILD) $(CFLAGS) dependency.native --
 
+.PHONY : all-tests
+all-tests : uninstall install test reverse-dependency-test
+
 HTML := docs/html
 
 .PHONY : docs

README.md

@@ -102,6 +102,10 @@ perhaps we should add it to Lambda Soup.
 
 Lambda Soup is distributed under the BSD license. See [LICENSE][license].
 
+Lambda Soup includes a modified version of [Ocamlnet][ocamlnet]'s Nethtml HTML
+parser. The license for this parser is also found in the [LICENSE][license]
+file.
+
 [docs]:         http://aantron.github.io/lambda-soup
 [postprocess]:  https://github.com/aantron/lambda-soup/blob/master/docs/postprocess.ml
 [tests]:        https://github.com/aantron/lambda-soup/blob/master/test/test.ml
@@ -114,3 +118,4 @@ Lambda Soup is distributed under the BSD license. See [LICENSE][license].
 [extracss]:     http://aantron.github.io/lambda-soup#VALselect
 [traversals]:   http://aantron.github.io/lambda-soup#2_Elementarytraversals
 [combinators]:  http://aantron.github.io/lambda-soup#2_Combinators
+[ocamlnet]:     http://projects.camlcity.org/projects/ocamlnet.html

docs/CONTRIBUTING.md

@@ -24,13 +24,14 @@ improvement. Here are a few examples:
   of performance improvements to be made.
 - Lambda Soup could extend CSS with parent, ancestor, previous sibling, and
   adjacent previous sibling combinators. These are the natural inverses of the
-  standard combinators.
+  standard combinators. There are also some potentially useful extensions in
+  jQuery, such as `:has`, that Lambda Soup could borrow.
+- Lambda Soup could also support XPath for selection. XPath does seem to be much
+  less familiar to people, on average, however.
 - Lambda Soup relies on [Ocamlnet's parser][nethtml] for HTML. This parser, like
   every other, has quirks, such as the way it handles attributes without a
   value and the fact that it does not resolve entities. Lambda Soup is also
-  completely reliant on it for dealing properly with character encodings. We may
-  need to work around some of these behaviors and/or parametrize Lambda Soup
-  over parsers.
+  completely reliant on it for dealing properly with character encodings.
 
 Of course, this is aside from any bugs that may be found :)
 
@@ -39,14 +40,20 @@ Of course, this is aside from any bugs that may be found :)
 Clone the repository locally. The library source is organized as follows:
 
 - `src/`: main source, i.e. `src/soup.mli` and `src/soup.ml`.
-- `test/`: test cases in `test.ml`.
+- `test/`: unit tests in `test.ml`, a performance test, and a reverse dependency
+  ("integration") test.
 - `docs/`: files related to the `ocamldoc` documentation, such as extra HTML and
   the postprocessing script.
 
-To run tests, run `make test`. This requires the `ounit` package. To generate
-docs, run `make docs`. To install the library from your repo clone, run
-`make install`. This uses OPAM to pin the repository and install a `lambdasoup`
-package. To undo this, run `make uninstall`.
+Tu run tests, run `make all-tests`. This requires the `ounit` package. This
+takes some time, because it reinstalls Lambda Soup. If you want to run only the
+unit and performance tests, run `make test`.
+
+To generate docs, run `make docs`.
+
+To install the library from your repo clone, run `make install`. This uses OPAM
+to pin the repository and install a `lambdasoup` package. To undo this, run
+`make uninstall`.
 
 Make a branch off master, make your changes, rebase over master (if you have to)
 when done, and submit a pull request :)

docs/postprocess.ml

@@ -97,6 +97,14 @@ let () =
   |> filter (fun e -> e $? "br" <> None)
   |> iter (fun e -> e $ "+ .info" |> add_class "multiline-member");
 
+  (* Find every section that has additional text after the title, and add a
+     class to the last element of such text. This is used by CSS to increase
+     spacing. *)
+  soup $$ "h2" |> iter (fun h2 ->
+    let e = h2 $ "~ pre:not(.codepre)" |> R.previous_element in
+    if name e = "h2" then ()
+    else add_class "end-of-section-text" e);
+
   (* Clean up links in the head. *)
   soup $$ "head link:not([rel=stylesheet])" |> iter delete;
 

docs/style.css

@@ -182,6 +182,10 @@ body > pre:first-of-type {
     margin-top: 1.5em;
 }
 
+.end-of-section-text {
+    margin-bottom: 3em;
+}
+
 /* Links. */
 
 a {