Add HTTPRequest/HTTPResponse: real HTTP parsing on the Parser/StreamReader stack

Stage 2 of the APIServer plan. This codebase has only ever sat behind a
CGI gateway - IMFRequest/IMFResponse/IMFMessage predate the
StreamReader/Parser rebuild and are stubs (IMFResponse's header-emission
calls addHeader()/output(), neither of which exist on the class;
IMFRequest is a fragile non-resumable 3-token line split). Nothing in
this repo has ever actually parsed a raw HTTP request off a socket.

HTTPRequest : public Parser fills that gap the same way every other
Parser here is built: REQUEST_LINE -> HEADERS -> BODY -> DONE, chaining
LineParser (request line) and IMFHeader (headers, reused as-is - HTTP
header syntax is IMF-derived, nothing email-specific in IMFHeader
itself) via checkParse() over one shared StreamReader, exactly the
pattern CGIFormData already established for chaining IMFHeader/
IMFBoundary. The BODY state dispatches on Content-Type as CGIFormData
itself does for CGI bodies - application/x-www-form-urlencoded and
multipart/form-data delegate directly to CGIFormData (reused, not
reimplemented), anything else is captured as a raw body (e.g. JSON).
Query string is split off the path before anything else sees it.
getHeader() does a case-insensitive lookup (HTTP header names are
case-insensitive per RFC 7230, unlike IMFHeader's own std::map, which
compares byte-for-byte) since a route handler constantly needs this and
getting it wrong would be a silent framing bug, not a cosmetic one.

The boundary-extraction code mirrors __jet.cpp's own fix exactly (same
cursor-mutation trap: ZString::find() leaves the searched string's
cursor at the match, and CGIFormData's constructor copy-constructs its
own contentType member from whatever's passed in, inheriting that same
moved cursor unless reset() first - would otherwise silently miss its
own startsWith() dispatch checks).

HTTPResponse is a small, new, output-only builder (status line + a real
headers container + body, Content-Length auto-computed unless set
explicitly) - deliberately not patched from IMFResponse, which has
nothing safe to extend.

v1 limitations, flagged not fixed (matching the standard already set by
other Parser subclasses here for known-but-out-of-scope gaps): chunked
Transfer-Encoding isn't supported, and repeated header names last-wins
(inherited from IMFHeader's own std::map).

Verified live via testing/http_test.cpp (34 assertions, new - now wired
into testing/compile): byte-at-a-time GET with a query string and
case-insensitive header lookup; a GET with no query string; a POST with
an urlencoded body deliberately split mid-body (not just mid-header)
across two feed() calls, confirming the CGIFormData delegation resumes
correctly; a raw JSON POST body; testing/http_request.data (a real
fixture, previously unused - normalized from its stored LF line endings
to real CRLF, since HTTP requires CRLF and IMFHeader enforces that
strictly) fed byte-at-a-time; and HTTPResponse::serialize() both with an
auto-computed and an explicitly-set Content-Length (no duplicate
header). Also reran the full existing testing/compile suite
(zstring/mstring/jstring/mfile/jfile) against the rebuilt libCoreUtils.a
to confirm no regression from the two new files.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

testing/compile

@@ -4,3 +4,4 @@ g++ -g -std=c++20 -o mstring_test mstring_test.cpp -I.. -L.. -lCoreUtils -lb64
 g++ -o jstring_test jstring_test.cpp -I.. -L.. -lCoreUtils
 g++ -g -std=c++20 -o mfile_test mfile_test.cpp -I.. -L.. -lCoreUtils
 g++ -g -std=c++20 -o jfile_test jfile_test.cpp -I.. -L.. -lCoreUtils
+g++ -g -std=c++23 -o http_test http_test.cpp -I.. -L.. -lCoreUtils

testing/http_test.cpp

@@ -0,0 +1,179 @@
+#include <iostream>
+#include <cassert>
+#include <cstring>
+#include <fstream>
+#include <sstream>
+#include "../StreamReader.h"
+#include "../HTTPRequest.h"
+#include "../HTTPResponse.h"
+
+using namespace coreutils;
+
+static int checks = 0;
+#define CHECK(cond) do { checks++; if(!(cond)) { std::cerr << "FAIL line " << __LINE__ << ": " #cond << std::endl; return 1; } } while(0)
+
+// Feeds a StreamReader one byte at a time and drives a Parser's doParse()
+// after every single byte - the harshest fragmentation test used elsewhere
+// in this subsystem (see CGIFormData's byte-at-a-time stress test) - to
+// prove HTTPRequest is genuinely resumable across arbitrary split points,
+// not just "works when handed the whole request at once".
+static Parser::Status feedByteAtATime(StreamReader &reader, Parser &parser, const std::string &data) {
+   Parser::Status status = Parser::Status::INCOMPLETE;
+   for (size_t i = 0; i < data.size() && status != Parser::Status::COMPLETE; ++i) {
+      ZString chunk((char *)data.data() + i, 1);
+      reader.feed(chunk);
+      status = parser.doParse();
+   }
+   return status;
+}
+
+int test_get_request_byte_at_a_time() {
+   StreamReader reader;
+   HTTPRequest req(reader);
+   std::string raw =
+      "GET /hello?x=1&y=two HTTP/1.1\r\n"
+      "Host: example.com\r\n"
+      "cookie: session=abc123\r\n"
+      "\r\n";
+   CHECK(feedByteAtATime(reader, req, raw) == Parser::Status::COMPLETE);
+   CHECK(req.getMethod() == "GET");
+   CHECK(req.getPath() == "/hello");
+   CHECK(req.getQueryString() == "x=1&y=two");
+   CHECK(req.getVersion() == "HTTP/1.1");
+   CHECK(req.getHeader("Host") == "example.com");
+   // Case-insensitive lookup: header arrived lowercase ("cookie:"), looked
+   // up with the canonical HTTP capitalization ("Cookie") - real clients
+   // and servers disagree on casing constantly, this must not care.
+   CHECK(req.getHeader("Cookie") == "session=abc123");
+   CHECK(req.getBody().getLength() == 0);
+   CHECK(req.getFormData() == NULL);
+   return 0;
+}
+
+int test_get_request_no_query_string() {
+   StreamReader reader;
+   HTTPRequest req(reader);
+   std::string raw = "GET /plain HTTP/1.1\r\nHost: h\r\n\r\n";
+   CHECK(feedByteAtATime(reader, req, raw) == Parser::Status::COMPLETE);
+   CHECK(req.getPath() == "/plain");
+   CHECK(req.getQueryString().getLength() == 0);
+   return 0;
+}
+
+int test_post_urlencoded_body_split_across_feeds() {
+   StreamReader reader;
+   HTTPRequest req(reader);
+   std::string raw =
+      "POST /submit HTTP/1.1\r\n"
+      "Host: example.com\r\n"
+      "Content-Type: application/x-www-form-urlencoded\r\n"
+      "Content-Length: 14\r\n"
+      "\r\n"
+      "name=A%20B&x=1";
+   // Split mid-body, not just mid-header - the whole point of building this
+   // on the same StreamReader/checkParse() chain CGIFormData already uses.
+   size_t splitAt = raw.size() - 5;
+   ZString part1((char *)raw.data(), splitAt);
+   reader.feed(part1);
+   CHECK(req.doParse() == Parser::Status::INCOMPLETE);
+   ZString part2((char *)raw.data() + splitAt, raw.size() - splitAt);
+   reader.feed(part2);
+   CHECK(req.doParse() == Parser::Status::COMPLETE);
+   CHECK(req.getMethod() == "POST");
+   CHECK(req.getFormData() != NULL);
+   MString name = (*req.getFormData())["name"];
+   CHECK(name == "A B");
+   MString x = (*req.getFormData())["x"];
+   CHECK(x == "1");
+   CHECK(req.getBody().getLength() == 0);   // consumed by the form parser, not the raw-body path
+   return 0;
+}
+
+int test_post_raw_json_body() {
+   StreamReader reader;
+   HTTPRequest req(reader);
+   std::string body = "{\"a\":1}";
+   std::string raw =
+      "POST /api HTTP/1.1\r\n"
+      "Host: example.com\r\n"
+      "Content-Type: application/json\r\n"
+      "Content-Length: " + std::to_string(body.size()) + "\r\n"
+      "\r\n" + body;
+   CHECK(feedByteAtATime(reader, req, raw) == Parser::Status::COMPLETE);
+   CHECK(req.getFormData() == NULL);
+   CHECK(req.getBody() == body.c_str());
+   return 0;
+}
+
+int test_real_fixture_file() {
+   // testing/http_request.data is a real captured request, but stored with
+   // plain LF line endings (a text-editing convenience, confirmed via
+   // `file`/`xxd` - not a literal wire capture); real HTTP requires CRLF
+   // per RFC 7230, and IMFHeader (which HTTPRequest's HEADERS state reuses
+   // as-is) enforces that strictly. Normalize LF -> CRLF before feeding,
+   // same as any real client would actually send it.
+   std::ifstream in("http_request.data");
+   CHECK((bool)in);
+   std::stringstream ss;
+   ss << in.rdbuf();
+   std::string raw = ss.str();
+   std::string crlf;
+   crlf.reserve(raw.size() + 16);
+   for (char c : raw) {
+      if (c == '\n') crlf += '\r';
+      crlf += c;
+   }
+
+   StreamReader reader;
+   HTTPRequest req(reader);
+   CHECK(feedByteAtATime(reader, req, crlf) == Parser::Status::COMPLETE);
+   CHECK(req.getMethod() == "GET");
+   CHECK(req.getPath() == "/");
+   CHECK(req.getHeader("Host") == "localhost:8080");
+   CHECK(req.getHeader("Cookie") == "_ga=GA1.1.1017526412.1628195401");
+   return 0;
+}
+
+int test_http_response_serialize() {
+   HTTPResponse resp;
+   resp.setStatus(200, "OK");
+   resp.setHeader("Content-Type", "text/plain");
+   resp.setBody("hello");
+   MString out = resp.serialize();
+   std::string s(out.getData(), out.getLength());
+   CHECK(s.find("HTTP/1.1 200 OK\r\n") == 0);
+   CHECK(s.find("Content-Type: text/plain\r\n") != std::string::npos);
+   CHECK(s.find("Content-Length: 5\r\n") != std::string::npos);
+   CHECK(s.find("\r\n\r\nhello") != std::string::npos);
+   return 0;
+}
+
+int test_http_response_explicit_content_length_respected() {
+   HTTPResponse resp;
+   resp.setStatus(404, "Not Found");
+   resp.setHeader("Content-Length", "0");
+   resp.setBody("");
+   MString out = resp.serialize();
+   std::string s(out.getData(), out.getLength());
+   CHECK(s.find("HTTP/1.1 404 Not Found\r\n") == 0);
+   int firstCL = s.find("Content-Length:");
+   int lastCL = s.rfind("Content-Length:");
+   CHECK(firstCL == lastCL);   // no duplicate Content-Length header
+   return 0;
+}
+
+int main() {
+   int rc = 0;
+   rc |= test_get_request_byte_at_a_time();
+   rc |= test_get_request_no_query_string();
+   rc |= test_post_urlencoded_body_split_across_feeds();
+   rc |= test_post_raw_json_body();
+   rc |= test_real_fixture_file();
+   rc |= test_http_response_serialize();
+   rc |= test_http_response_explicit_content_length_respected();
+   if (rc == 0)
+      std::cout << "ALL " << checks << " CHECKS PASSED" << std::endl;
+   else
+      std::cout << "SOME CHECKS FAILED (see above)" << std::endl;
+   return rc;
+}