From 1fb78a78f280d2ec0391a787e7716887700a068f Mon Sep 17 00:00:00 2001 From: Simon Rozman Date: Thu, 16 Mar 2023 12:34:28 +0100 Subject: [PATCH] parser: Stabilize HTTP suite Signed-off-by: Simon Rozman --- UnitTests/parser.cpp | 92 ++++++++++++++++++++++++++++++++++++++++ include/stdex/parser.hpp | 28 +++++++++--- 2 files changed, 113 insertions(+), 7 deletions(-) diff --git a/UnitTests/parser.cpp b/UnitTests/parser.cpp index 022f9f0e0..50f865b23 100644 --- a/UnitTests/parser.cpp +++ b/UnitTests/parser.cpp @@ -221,5 +221,97 @@ namespace UnitTests Assert::AreEqual((size_t)31, p.interval.end); } } + + TEST_METHOD(http_test) + { + static const std::locale locale("en_US.UTF-8"); + static const char request[] = + "GET / HTTP/2\r\n" + "Host: stackoverflow.com\r\n" + "User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:109.0) Gecko/20100101 Firefox/110.0\r\n" + "Accept: text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,*/*;q=0.8\r\n" + "Accept-Language: sl,en-US;q=0.8,en;q=0.6,de-DE;q=0.4,de;q=0.2\r\n" + "Accept-Encoding: gzip, deflate, br\r\n" + "DNT: 1\r\n" + "Connection: keep-alive\r\n" + "Cookie: prov=00000000-0000-0000-0000-000000000000; acct=t=00000000000000000%2f%2f0000%2b0000%2b000&s=00000000000000000000000000000000; OptanonConsent=isGpcEnabled=0&datestamp=Fri+Feb+03+2023+11%3A11%3A08+GMT%2B0100+(Srednjeevropski+standardni+%C4%8Das)&version=6.37.0&isIABGlobal=false&hosts=&consentId=00000000-0000-0000-0000-000000000000&interactionCount=1&landingPath=NotLandingPage&groups=00000%3A0%2C00000%3A0%2C00000%3A0%2C00000%3A0; OptanonAlertBoxClosed=2023-02-03T10:11:08.683Z\r\n" + "Upgrade-Insecure-Requests: 1\r\n" + "Sec-Fetch-Dest: document\r\n" + "Sec-Fetch-Mode: navigate\r\n" + "Sec-Fetch-Site: none\r\n" + "Sec-Fetch-User: ?1\r\n" + "Pragma: no-cache\r\n" + "Cache-Control: no-cache\r\n" + "\r\n"; + + { + http_request p(locale); + Assert::IsTrue(p.match(request)); + Assert::AreEqual((size_t)0, p.interval.start); + Assert::AreEqual((size_t)14, p.interval.end); + Assert::AreEqual((size_t)0, p.verb.start); + Assert::AreEqual((size_t)3, p.verb.end); + Assert::AreEqual((size_t)4, p.url.interval.start); + Assert::AreEqual((size_t)5, p.url.interval.end); + Assert::AreEqual((size_t)6, p.protocol.interval.start); + Assert::AreEqual((size_t)12, p.protocol.interval.end); + Assert::AreEqual((uint16_t)0x200, p.protocol.version); + } + + { + std::list hdrs; + size_t offset = 14; + for (;;) { + http_header h; + if (h.match(request, offset)) { + offset = h.interval.end; + hdrs.push_back(std::move(h)); + } + else + break; + } + Assert::AreEqual((size_t)15, hdrs.size()); + http_weighted_collection> langs; + for (const auto& h : hdrs) + if (strnicmp(request + h.name.start, h.name.size(), "Accept-Language", (size_t)-1, locale) == 0) + langs.insert(request, h.value.start, h.value.end); + Assert::IsTrue(!langs.empty()); + { + const vector control = { + "sl", "en-US", "en", "de-DE", "de" + }; + auto c = control.cbegin(); + auto l = langs.cbegin(); + for (; c != control.cend() && l != langs.cend(); ++c, ++l) + Assert::IsTrue(strnicmp(request + l->value.interval.start, l->value.interval.size(), c->c_str(), c->size(), locale) == 0); + Assert::IsTrue(c == control.cend()); + Assert::IsTrue(l == langs.cend()); + } + } + + //static const char response[] = + // "HTTP/2 200 OK\r\n" + // "cache-control: private\r\n" + // "content-type: text/html; charset=utf-8\r\n" + // "content-encoding: gzip\r\n" + // "strict-transport-security: max-age=15552000\r\n" + // "x-frame-options: SAMEORIGIN\r\n" + // "set-cookie: acct=t=00000000000000000%2f%2f0000%2b0000%2b000&s=00000000000000000000000000000000; expires=Sat, 16 Sep 2023 10:23:00 GMT; domain=.stackoverflow.com; path=/; secure; samesite=none; httponly\r\n" + // "set-cookie: prov_tgt=; expires=Tue, 14 Mar 2023 10:23:00 GMT; domain=.stackoverflow.com; path=/; secure; samesite=none; httponly\r\n" + // "x-request-guid: a6536a49-b473-4c6f-b313-c1e7c0d8f600\r\n" + // "feature-policy: microphone 'none'; speaker 'none'\r\n" + // "content-security-policy: upgrade-insecure-requests; frame-ancestors 'self' https://stackexchange.com\r\n" + // "accept-ranges: bytes\r\n" + // "date: Thu, 16 Mar 2023 10:23:00 GMT\r\n" + // "via: 1.1 varnish\r\n" + // "x-served-by: cache-vie6354-VIE\r\n" + // "x-cache: MISS\r\n" + // "x-cache-hits: 0\r\n" + // "x-timer: S1678962181.533907,VS0,VE144\r\n" + // "vary: Accept-Encoding,Fastly-SSL\r\n" + // "x-dns-prefetch-control: off\r\n" + // "X-Firefox-Spdy: h2\r\n" + // "\r\n"; + } }; } diff --git a/include/stdex/parser.hpp b/include/stdex/parser.hpp index 59c816b6c..86650c858 100644 --- a/include/stdex/parser.hpp +++ b/include/stdex/parser.hpp @@ -5457,7 +5457,10 @@ namespace stdex class http_url : public parser { public: - http_url(_In_ const std::locale& locale = std::locale()) : parser(locale) {} + http_url(_In_ const std::locale& locale = std::locale()) : + parser(locale), + port(locale) + {} virtual bool match( _In_reads_or_z_(end) const char* text, @@ -5711,6 +5714,11 @@ namespace stdex class http_weighted_value : public parser { public: + http_weighted_value(_In_ const std::locale& locale = std::locale()) : + parser(locale), + factor(locale) + {} + virtual bool match( _In_reads_or_z_(end) const char* text, _In_ size_t start = 0, @@ -6104,6 +6112,12 @@ namespace stdex class http_request : public parser { public: + http_request(_In_ const std::locale& locale = std::locale()) : + parser(locale), + url(locale), + protocol(locale) + {} + virtual bool match( _In_reads_or_z_(end) const char* text, _In_ size_t start = 0, @@ -6340,14 +6354,14 @@ namespace stdex }; /// - /// Collection of HTTP headers + /// Collection of HTTP values /// template - class http_header_collection : public T + class http_value_collection : public T { public: void insert( - _In_reads_or_z_(end) const T* text, + _In_reads_or_z_(end) const char* text, _In_ size_t start = 0, _In_ size_t end = (size_t)-1, _In_ int flags = match_default) @@ -6378,10 +6392,10 @@ namespace stdex }; /// - /// Collection of weighted HTTP headers + /// Collection of weighted HTTP values /// - template , class _Alloc = std::allocator> - using http_weighted_header_collection = http_header_collection>; + template > + using http_weighted_collection = http_value_collection, _Alloc>>; /// /// Test for JSON string