net: extend Sock with methods for robust send & read until terminator

Introduce two high level, convenience methods in the `Sock` class:

* `SendComplete()`: keep trying to send the specified data until either
  successfully sent all of it, timeout or interrupted.

* `RecvUntilTerminator()`: read until a terminator is encountered (never
  after it), timeout or interrupted.

These will be convenient in the I2P SAM implementation.

`SendComplete()` can also be used in the SOCKS5 implementation instead
of calling `send()` directly.

src/compat.h

@@ -44,6 +44,7 @@ typedef unsigned int SOCKET;
 #define WSAEINVAL           EINVAL
 #define WSAEALREADY         EALREADY
 #define WSAEWOULDBLOCK      EWOULDBLOCK
+#define WSAEAGAIN           EAGAIN
 #define WSAEMSGSIZE         EMSGSIZE
 #define WSAEINTR            EINTR
 #define WSAEINPROGRESS      EINPROGRESS
@@ -51,6 +52,14 @@ typedef unsigned int SOCKET;
 #define WSAENOTSOCK         EBADF
 #define INVALID_SOCKET      (SOCKET)(~0)
 #define SOCKET_ERROR        -1
+#else
+#ifndef WSAEAGAIN
+#ifdef EAGAIN
+#define WSAEAGAIN EAGAIN
+#else
+#define WSAEAGAIN WSAEWOULDBLOCK
+#endif
+#endif
 #endif
 
 #ifdef WIN32

src/util/sock.cpp

@@ -4,6 +4,7 @@
 
 #include <compat.h>
 #include <logging.h>
+#include <threadinterrupt.h>
 #include <tinyformat.h>
 #include <util/sock.h>
 #include <util/system.h>
@@ -12,12 +13,18 @@
 #include <codecvt>
 #include <cwchar>
 #include <locale>
+#include <stdexcept>
 #include <string>
 
 #ifdef USE_POLL
 #include <poll.h>
 #endif
 
+static inline bool IOErrorIsPermanent(int err)
+{
+    return err != WSAEAGAIN && err != WSAEINTR && err != WSAEWOULDBLOCK && err != WSAEINPROGRESS;
+}
+
 Sock::Sock() : m_socket(INVALID_SOCKET) {}
 
 Sock::Sock(SOCKET s) : m_socket(s) {}
@@ -125,6 +132,124 @@ bool Sock::Wait(std::chrono::milliseconds timeout, Event requested, Event* occur
 #endif /* USE_POLL */
 }
 
+void Sock::SendComplete(const std::string& data,
+                        std::chrono::milliseconds timeout,
+                        CThreadInterrupt& interrupt) const
+{
+    const auto deadline = GetTime<std::chrono::milliseconds>() + timeout;
+    size_t sent{0};
+
+    for (;;) {
+        const ssize_t ret{Send(data.data() + sent, data.size() - sent, MSG_NOSIGNAL)};
+
+        if (ret > 0) {
+            sent += static_cast<size_t>(ret);
+            if (sent == data.size()) {
+                break;
+            }
+        } else {
+            const int err{WSAGetLastError()};
+            if (IOErrorIsPermanent(err)) {
+                throw std::runtime_error(strprintf("send(): %s", NetworkErrorString(err)));
+            }
+        }
+
+        const auto now = GetTime<std::chrono::milliseconds>();
+
+        if (now >= deadline) {
+            throw std::runtime_error(strprintf(
+                "Send timeout (sent only %u of %u bytes before that)", sent, data.size()));
+        }
+
+        if (interrupt) {
+            throw std::runtime_error(strprintf(
+                "Send interrupted (sent only %u of %u bytes before that)", sent, data.size()));
+        }
+
+        // Wait for a short while (or the socket to become ready for sending) before retrying
+        // if nothing was sent.
+        const auto wait_time = std::min(deadline - now, std::chrono::milliseconds{MAX_WAIT_FOR_IO});
+        Wait(wait_time, SEND);
+    }
+}
+
+std::string Sock::RecvUntilTerminator(uint8_t terminator,
+                                      std::chrono::milliseconds timeout,
+                                      CThreadInterrupt& interrupt) const
+{
+    const auto deadline = GetTime<std::chrono::milliseconds>() + timeout;
+    std::string data;
+    bool terminator_found{false};
+
+    // We must not consume any bytes past the terminator from the socket.
+    // One option is to read one byte at a time and check if we have read a terminator.
+    // However that is very slow. Instead, we peek at what is in the socket and only read
+    // as many bytes as possible without crossing the terminator.
+    // Reading 64 MiB of random data with 262526 terminator chars takes 37 seconds to read
+    // one byte at a time VS 0.71 seconds with the "peek" solution below. Reading one byte
+    // at a time is about 50 times slower.
+
+    for (;;) {
+        char buf[512];
+
+        const ssize_t peek_ret{Recv(buf, sizeof(buf), MSG_PEEK)};
+
+        switch (peek_ret) {
+        case -1: {
+            const int err{WSAGetLastError()};
+            if (IOErrorIsPermanent(err)) {
+                throw std::runtime_error(strprintf("recv(): %s", NetworkErrorString(err)));
+            }
+            break;
+        }
+        case 0:
+            throw std::runtime_error("Connection unexpectedly closed by peer");
+        default:
+            auto end = buf + peek_ret;
+            auto terminator_pos = std::find(buf, end, terminator);
+            terminator_found = terminator_pos != end;
+
+            const size_t try_len{terminator_found ? terminator_pos - buf + 1 :
+                                                    static_cast<size_t>(peek_ret)};
+
+            const ssize_t read_ret{Recv(buf, try_len, 0)};
+
+            if (read_ret < 0 || static_cast<size_t>(read_ret) != try_len) {
+                throw std::runtime_error(
+                    strprintf("recv() returned %u bytes on attempt to read %u bytes but previous "
+                              "peek claimed %u bytes are available",
+                              read_ret, try_len, peek_ret));
+            }
+
+            // Don't include the terminator in the output.
+            const size_t append_len{terminator_found ? try_len - 1 : try_len};
+
+            data.append(buf, buf + append_len);
+
+            if (terminator_found) {
+                return data;
+            }
+        }
+
+        const auto now = GetTime<std::chrono::milliseconds>();
+
+        if (now >= deadline) {
+            throw std::runtime_error(strprintf(
+                "Receive timeout (received %u bytes without terminator before that)", data.size()));
+        }
+
+        if (interrupt) {
+            throw std::runtime_error(strprintf(
+                "Receive interrupted (received %u bytes without terminator before that)",
+                data.size()));
+        }
+
+        // Wait for a short while (or the socket to become ready for reading) before retrying.
+        const auto wait_time = std::min(deadline - now, std::chrono::milliseconds{MAX_WAIT_FOR_IO});
+        Wait(wait_time, RECV);
+    }
+}
+
 #ifdef WIN32
 std::string NetworkErrorString(int err)
 {

src/util/sock.h

@@ -6,6 +6,7 @@
 #define BITCOIN_UTIL_SOCK_H
 
 #include <compat.h>
+#include <threadinterrupt.h>
 #include <util/time.h>
 
 #include <chrono>
@@ -114,6 +115,34 @@ public:
                       Event requested,
                       Event* occurred = nullptr) const;
 
+    /* Higher level, convenience, methods. These may throw. */
+
+    /**
+     * Send the given data, retrying on transient errors.
+     * @param[in] data Data to send.
+     * @param[in] timeout Timeout for the entire operation.
+     * @param[in] interrupt If this is signaled then the operation is canceled.
+     * @throws std::runtime_error if the operation cannot be completed. In this case only some of
+     * the data will be written to the socket.
+     */
+    virtual void SendComplete(const std::string& data,
+                              std::chrono::milliseconds timeout,
+                              CThreadInterrupt& interrupt) const;
+
+    /**
+     * Read from socket until a terminator character is encountered. Will never consume bytes past
+     * the terminator from the socket.
+     * @param[in] terminator Character up to which to read from the socket.
+     * @param[in] timeout Timeout for the entire operation.
+     * @param[in] interrupt If this is signaled then the operation is canceled.
+     * @return The data that has been read, without the terminating character.
+     * @throws std::runtime_error if the operation cannot be completed. In this case some bytes may
+     * have been consumed from the socket.
+     */
+    virtual std::string RecvUntilTerminator(uint8_t terminator,
+                                            std::chrono::milliseconds timeout,
+                                            CThreadInterrupt& interrupt) const;
+
 private:
     /**
      * Contained socket. `INVALID_SOCKET` designates the object is empty.