Rework and improve the decompress stream handler

The read buffer was changed into an iovec to avoid doing
too many binary concatenations and allocations.

Decompression happens transparently: when decoding gzip,
the content-encoding header is removed (we only decode
when "gzip" is the only encoding so nothing remains).

We always add a content_decoded key to the Req object.
This key contains a list of codings that were decoded,
in the reverse order in which they were. Currently it
can only be empty or contain <<"gzip">> but future
improvements or user handlers may see it contain more
values.

The option to disable decompression was renamed to
decompress_enabled and defaults to true.

It is no longer possible to enable/disable decompression
in the middle of reading the body: this ensures that the
data we pass forward is always valid.

Various smaller improvements were made to the code,
tests and manual pages.

1 files changed, 23 insertions, 11 deletions

diff --git a/doc/src/manual/cowboy_decompress_h.asciidoc b/doc/src/manual/cowboy_decompress_h.asciidoc
index 63ee0ee..8598ae4 100644
--- a/doc/src/manual/cowboy_decompress_h.asciidoc
+++ b/doc/src/manual/cowboy_decompress_h.asciidoc
@@ -7,18 +7,26 @@ cowboy_decompress_h - Decompress stream handler
 == Description
 
 The module `cowboy_decompress_h` decompresses request bodies
-automatically when the server supports it. Requests will
-only be decompressed when their compression ratio is lower
-than the configured limit. Mismatch of the content and
-`content-encoding` is rejected with `400 Bad Request`.
+automatically when the server supports it.
+
+The only compression algorithm currently supported is the
+gzip algorithm. Another limitation is that decompression
+is only attempted when gzip is the only content-encoding
+in the request.
+
+This stream handler always adds a field to the Req object
+with the name `content_decoded` which is treated as a
+list of decoded content-encoding values. Currently this
+list may only contain the `<<"gzip">>` binary if content
+was decoded; or be empty otherwise.
 
 == Options
 
 [source,erlang]
 ----
 opts() :: #{
-	decompress_ratio_limit => non_neg_integer(),
-	decompress_ignore => boolean()
+	decompress_enabled => boolean(),
+	decompress_ratio_limit => non_neg_integer()
 }
 ----
 
@@ -28,17 +36,21 @@ The default value is given next to the option name:
 
 decompress_ratio_limit (20)::
 The max ratio of the compressed and decompressed body
-before it is rejected with `413 Payload Too Large`.
+before it is rejected with a `413 Payload Too Large`
+error response.
 +
 This option can be updated at any time using the
 `set_options` stream handler command.
 
-decompress_ignore (false)::
+decompress_enabled (true)::
 
-Whether the handler will be ignored.
+Whether the handler is enabled by default.
 +
-This option can be updated at any time using the
-`set_options` stream handler command.
+This option can be updated using the `set_options`
+stream handler command. This allows disabling
+decompression for the current stream. Attempts
+to enable or disable decompression after starting
+to read the body will be ignored.
 
 == Events