Fix RequestHeaders to proper concat repeats

[BigBlueHat]

Previously RequestHeaders only provided the last instance of a repeated header.

Sadly, writing a unit test for this has proved...daunting. Python libraries from urllib's 1-3, Requests, and even httplib do not allow one to construct a header sequence like the one below (from the Web Annotation Protocol spec):

GET /annotations/ HTTP/1.1
Host: example.org
Accept: application/ld+json; profile="http://www.w3.org/ns/anno.jsonld"
Prefer: return=representation;include="http://www.w3.org/ns/ldp#PreferMinimalContainer"
Prefer: return=representation;include="http://www.w3.org/ns/oa#PreferContainedIRIs"

Most of these libraries use a dictionary style interface, and when given multiple values for a single header, they send them as a comma separated list (as they should! re: RFC 7230, RFC 2616, etc). However, because of that writing a test that would match the exact request seen above is...non-trivial (with the most likely suggestions being to "just use the sockets library")--which goes beyond what I care to attempt (or have time to...) in order to prove that this works. 😖

So. Instead, here's one loverly curl line for your copy/paste enjoyment:

curl http://localhost:8080 \
-H 'Accept: application/ld+json; profile="http://www.w3.org/ns/anno.jsonld"' \
-H 'Prefer: return=representation;include="http://www.w3.org/ns/ldp#PreferMinimalContainer"' \
-H 'Prefer: return=representation;include="http://www.w3.org/ns/oa#PreferContainedIRIs"'

And a simple server to test it against:

import json

import wptserve

port = 8080
doc_root = './docroot/'

@wptserve.handlers.handler
def multi(request, response):
    return json.dumps(request.headers, indent=4)

routes = [
    ("GET", "*", multi)
]

print 'http://localhost:{0}/'.format(port)

httpd = wptserve.server.WebTestHttpd(port=port, doc_root=doc_root,
                                     routes=routes)
httpd.start(block=True)

Which, if run against each other, should result in this JSON:

{
    "host": "localhost:8080",
    "prefer": "return=representation;include=\"http://www.w3.org/ns/ldp#PreferMinimalContainer\", return=representation;include=\"http://www.w3.org/ns/ldp#PreferMinimalContainer\"",
    "accept": "application/ld+json; profile=\"http://www.w3.org/ns/anno.jsonld\"",
    "user-agent": "curl/7.41.0"
}

Cheers.
🎩

---

This change is

[BigBlueHat]

@halindrome your input here would be most welcome. 😄

[halindrome]

Review status: 0 of 1 files reviewed at latest revision, 1 unresolved discussion.

---

wptserve/request.py, line 355 [r1] (raw file):

            # if there are other headers with this name, we need them
            values = items.getallmatchingheaders(header)
            if len(values) > 1:

Silly question - but won't there always be at least 1 if there was a key in items?

---

Comments from Reviewable

[BigBlueHat]

@halindrome right. This is checking for situations where there's more than 1 header returned. Happy Monday! ☕

[halindrome]

Review status: 0 of 1 files reviewed at latest revision, 1 unresolved discussion.

---

wptserve/request.py, line 355 [r1] (raw file):

Previously, halindrome (Shane McCarron) wrote…

Silly question - but won't there always be at least 1 if there was a key in items?

Wow - never mind. I was thinking >= 1.... MORE COFFEE.

---

Comments from Reviewable

[halindrome]

LGTM

---

Review status: 0 of 1 files reviewed at latest revision, 1 unresolved discussion.

---

Comments from Reviewable

[jgraham]

Review status: 0 of 1 files reviewed at latest revision, 2 unresolved discussions.

---

wptserve/request.py, line 360 [r1] (raw file):

                # loop through the values from getallmatchingheaders
                for value in values:
                    if ':' in value:

This seems technically wrong, since you don't do anything for continuation lines. I would be happy if you raised a 400 error in such cases i.e. if there is no : in the line or if the line starts with whitespace.

---

Comments from Reviewable

[BigBlueHat]

@jgraham sorry. yeah. that was left over from an earlier approach. I'll clean it out and resend shortly. Appologies!