support proxypass in web server
C/C++ CI Linux / ${{ matrix.target }} (ubuntu-24.04-arm) (push) Has been cancelled
C/C++ CI Linux / ${{ matrix.target }} (ubuntu-latest) (push) Has been cancelled
C/C++ CI MacOS / ${{ matrix.target }} (macos-15) (push) Has been cancelled
C/C++ CI Windows / ${{ matrix.target }} (windows-2022) (push) Has been cancelled
C/C++ CI cmake / ${{ matrix.target }} (macos-15) (push) Has been cancelled
C/C++ CI cmake / ${{ matrix.target }} (ubuntu-24.04-arm) (push) Has been cancelled
C/C++ CI cmake / ${{ matrix.target }} (ubuntu-latest) (push) Has been cancelled
C/C++ CI cmake / ${{ matrix.target }} (windows-2022) (push) Has been cancelled
C/C++ CI cmake / ubuntu-latest (wolfSSL) (push) Has been cancelled

This commit is contained in:
Vladimir Dubrovin committed 2026-08-27 17:57:51 +03:00
1 parent 710ee8789d
commit 5be794dfd2
15 files changed
+699 -30

No files matched your search

+59
View File
@@ -867,6 +867,65 @@ http file * /** /usr/local/web/blocked.html text/html * * 403
httpsrv -p8080 -i127.0.0.1
</pre>
<p>
<b>Both a site and a proxy.</b> A request may arrive the way it arrives at a
site - a path, with the name in the Host header - or the way it arrives at a
proxy, naming the whole URL, or the host alone with CONNECT. Both are read. A
proxy-form request authenticates with Proxy-Authorization and is refused with
407, the way a proxy refuses one; a site-form request uses Authorization and
401. What answers it is decided by the rules either way.
</p>
<p>
<code>proxypass</code> is the rule which answers by fetching, so one service can
serve what it has and proxy the rest:
</p><pre>
auth iponly
allow *
http file * /local/** "/usr/local/web/$1"
http proxypass * /**
httpsrv -p8080
</pre>
<p>
An access rule redirecting to the local proxy does the same without a rule for
it. The chain with no address is what "the local proxy" is written as, and the
second <code>allow</code> is what the proxy matches on the pass it makes itself,
since a rule carrying the chain is not taken twice:
</p><pre>
auth iponly
allow *
parent 1000 http 0.0.0.0 0
allow *
http file * /local/** "/usr/local/web/$1"
httpsrv -p8080
</pre>
<p>
Authentication happens twice for the same reason - once for the service, once
for the proxy - so a configuration which asks for credentials asks for them the
way a proxy does.
</p>
<p>
The access rules are read from the top on both passes, and the second one is
where the request's destination is known. On the first pass the service is
answering for itself, so an address or a port in a rule is matched against the
address the client connected to; the name from the request is matched on both
passes. On the second the destination is the one the request names, so rules
written with an address, a port or a name decide what the proxy may fetch, and
they decide it before it connects:
</p><pre>
auth iponly
allow *
parent 1000 http 0.0.0.0 0
allow * * * 80,443
deny *
httpsrv -p8080
</pre>
<p>
Everything reaches the rules, only ports 80 and 443 are fetched, and a
<code>deny</code> written before the rule carrying the chain applies on both
passes just the same. The connection to the server is kept for the next request and
closed when that request goes elsewhere, or when the server has closed it in the
meantime.
</p>
<p>
<b>Connections.</b> A client asking in HTTP/1.1 gets a 1.1 answer and the
connection is kept for the next request, unless it sent
<code>Connection: close</code>; a 1.0 client has to ask for keep-alive. The