Anchor the locale test's lift so it cannot silently stop testing

The lift ended at the first blank line, so a blank line grown inside the
locale block would truncate it: reinstating the old LANG-only bug below one
left the test passing 14/14 while the launcher returned English. Anchor on
the next section instead, and pin exported-empty as unset, which lang_of
cannot express.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: Xavier Roche <roche@httrack.com>

lang.indexes

@@ -23,7 +23,7 @@ pt:7
 ro:25
 ru:8
 sk:20
-si:24
+sl:24
 sv:17
 tr:10
 uk:22

lang/Slovenian.txt

@@ -3,7 +3,7 @@ Slovenian
 LANGUAGE_FILE
 Slovenian
 LANGUAGE_ISO
-si
+sl
 LANGUAGE_AUTHOR
 Jadran Rudec,iur.\r\njrudec@email.si \r\n
 LANGUAGE_CHARSET

src/htsback.c

@@ -2816,9 +2816,6 @@ void back_wait(struct_back * sback, httrackp * opt, cache_back * cache,
                 if (SSL_set_fd(back[i].r.ssl_con, (int) back[i].r.soc) == 1) {
                   SSL_set_connect_state(back[i].r.ssl_con);
                   back[i].status = STATUS_SSL_WAIT_HANDSHAKE;   /* handshake wait */
-                  // the handshake gets its own timeout window, as connect does
-                  if (back[i].timeout > 0)
-                    back[i].timeout_refresh = time_local();
                 } else
                   back[i].r.statuscode = STATUSCODE_SSL_HANDSHAKE;
               } else
@@ -2878,11 +2875,6 @@ void back_wait(struct_back * sback, httrackp * opt, cache_back * cache,
       }
 #if HTS_USEOPENSSL
       else if (back[i].status == STATUS_SSL_WAIT_HANDSHAKE) {       // wait for SSL handshake
-        // a peer that never speaks TLS must be reaped by --timeout too (#607)
-        if (!gestion_timeout)
-          if (back[i].timeout > 0)
-            gestion_timeout = 1;
-
         /* SSL mode */
         if (back[i].r.ssl) {
           int conn_code;
@@ -4273,8 +4265,6 @@ void back_wait(struct_back * sback, httrackp * opt, cache_back * cache,
                 strcpybuff(back[i].r.msg, "Connect Time Out");
               else if (back[i].status == STATUS_WAIT_DNS)
                 strcpybuff(back[i].r.msg, "DNS Time Out");
-              else if (back[i].status == STATUS_SSL_WAIT_HANDSHAKE)
-                strcpybuff(back[i].r.msg, "SSL/TLS Handshake Time Out");
               else
                 strcpybuff(back[i].r.msg, "Receive Time Out");
               back[i].status = STATUS_READY;    // terminé

src/webhttrack

@@ -33,6 +33,19 @@ function log {
     return 0
 }
 
+# Map a POSIX locale ("zh_TW.UTF-8@euro") to its lang.indexes number, English (1) if unknown
+function lang_index {
+    local locale=$1 indexes=$2 tag n t
+    tag=$(echo "${locale}" | cut -f1 -d'.' | cut -f1 -d'@' | tr '[:upper:]' '[:lower:]' | tr -cd 'a-z_')
+    # a few languages carry a region-specific entry (zh_tw, pt_br); else use the bare language
+    for t in "${tag}" "${tag%%_*}"; do
+        n=$(grep -E "^${t}:" "${indexes}" | cut -f2 -d':')
+        test -n "${n}" && break
+    done
+    test -n "${n}" || n=1
+    echo "${n}"
+}
+
 function launch_browser {
     log "Launching $1"
     browser=$1
@@ -59,13 +72,11 @@ test -f "${DISTPATH}/lang.indexes" || ! log "Could not find ${DISTPATH}/lang.ind
 test -d "${DISTPATH}/lang" || ! log "Could not find ${DISTPATH}/lang" || exit 1
 test -d "${DISTPATH}/html" || ! log "Could not find ${DISTPATH}/html" || exit 1
 
-# Locale
-HTSLANG="${LC_MESSAGES}"
-! test -n "${HTSLANG}" && HTSLANG="${LC_ALL}"
+# Locale: POSIX precedence, LC_ALL overrides LC_MESSAGES overrides LANG
+HTSLANG="${LC_ALL}"
+! test -n "${HTSLANG}" && HTSLANG="${LC_MESSAGES}"
 ! test -n "${HTSLANG}" && HTSLANG="${LANG}"
-HTSLANG="$(echo "$LANG" | cut -f1 -d'.' | cut -f1 -d'_')"
-LANGN=$(grep -E "^${HTSLANG}:" "${DISTPATH}/lang.indexes" | cut -f2 -d':')
-! test -n "${LANGN}" && LANGN=1
+LANGN=$(lang_index "${HTSLANG}" "${DISTPATH}/lang.indexes")
 
 # Find the browser
 # note: not all systems have sensible-browser or www-browser alternative

tests/59_local-tls-stall.test

@@ -1,95 +0,0 @@
-#!/bin/bash
-#
-# Issue #607: a peer that accepts the TCP connect but never speaks TLS leaves the
-# slot stuck in the handshake. The per-slot --timeout must reap it; before the
-# fix only --max-time did, so these crawls ran until the kill guard.
-
-set -euo pipefail
-
-: "${top_srcdir:=..}"
-
-# shellcheck source=tests/testlib.sh
-. "$top_srcdir/tests/testlib.sh"
-
-if test "${HTTPS_SUPPORT:-}" == "no"; then
-    echo "no https support compiled, skipping"
-    exit 77
-fi
-python=$(find_python) || {
-    echo "python3 missing, skipping"
-    exit 77
-}
-
-server=$(nativepath "$top_srcdir/tests/tls-stall-server.py")
-tmpdir=$(mktemp -d)
-serverpid=
-
-cleanup() {
-    stop_server "$serverpid"
-    rm -rf "$tmpdir"
-    return 0
-}
-trap cleanup EXIT
-
-# start_stall_server <tag> <mode-args...>: sets $port from the announced one.
-start_stall_server() {
-    local tag="$1"
-    shift
-    "$python" "$server" "$@" >"$tmpdir/$tag.out" 2>"$tmpdir/$tag.err" &
-    serverpid=$!
-    port=
-    for _ in $(seq 1 50); do
-        line=$(head -n1 "$tmpdir/$tag.out" 2>/dev/null || true)
-        if test "${line%% *}" == "PORT"; then
-            port="${line#PORT }"
-            return 0
-        fi
-        kill -0 "$serverpid" 2>/dev/null || {
-            echo "server exited early: $(cat "$tmpdir/$tag.err")"
-            exit 1
-        }
-        sleep 0.1
-    done
-    echo "could not discover server port"
-    exit 1
-}
-
-# no -E/--max-time on purpose: --timeout is the only thing that can end these.
-# The kill guard stands in for the hang, so a wall time near it means no reap.
-crawl_wall() {
-    local out="$tmpdir/$1"
-    shift
-    local start
-    start=$(date +%s)
-    run_with_timeout 60 httrack -O "$out" -c1 --robots=0 --retries=0 --quiet -Z \
-        "$@" >>"$tmpdir/log" 2>&1 || true
-    echo $(($(date +%s) - start))
-}
-
-# 1. handshake stalled from the first byte: reaped at --timeout, not before.
-start_stall_server direct direct
-wall=$(crawl_wall crawl1 "https://127.0.0.1:$port/" --timeout=5)
-if test "$wall" -ge 30 || test "$wall" -lt 3; then
-    echo "FAIL: stalled handshake reaped after ${wall}s, expected about 5s" >&2
-    cat "$tmpdir/log" >&2
-    exit 1
-fi
-grep -q 'Handshake Time Out' "$tmpdir/crawl1/hts-log.txt" || {
-    echo "FAIL: crawl ended in ${wall}s but not on a handshake timeout" >&2
-    cat "$tmpdir/crawl1/hts-log.txt" >&2
-    exit 1
-}
-echo "OK: stalled TLS handshake reaped by --timeout after ${wall}s"
-
-# 2. the handshake window is its own, not what the connect left over: a proxy
-# that takes 4s to answer CONNECT must still leave the full --timeout=5 for the
-# handshake (~9s total). Sharing the connect's clock would reap at ~5s.
-stop_server "$serverpid"
-start_stall_server proxy proxy 4
-wall=$(crawl_wall crawl2 "https://127.0.0.1:443/" -P "127.0.0.1:$port" --timeout=5)
-if test "$wall" -ge 20 || test "$wall" -lt 7; then
-    echo "FAIL: handshake after a slow connect reaped at ${wall}s, expected about 9s" >&2
-    cat "$tmpdir/log" >&2
-    exit 1
-fi
-echo "OK: handshake keeps its own timeout window after a slow connect (${wall}s)"

tests/61_webhttrack-locale.test

@@ -0,0 +1,91 @@
+#!/bin/bash
+#
+# webhttrack must map the caller's locale to the right lang.indexes number.
+# Guards: LC_ALL/LC_MESSAGES were ignored, and zh_tw/pt_br were unreachable.
+
+set -euo pipefail
+
+testdir=$(cd "$(dirname "$0")" && pwd)
+distdir=$(cd "${testdir}/.." && pwd)
+script="${distdir}/src/webhttrack"
+
+fail() {
+    echo "FAIL: $*" >&2
+    exit 1
+}
+
+test -r "${script}" || fail "no ${script}"
+test -r "${distdir}/lang.indexes" || fail "no ${distdir}/lang.indexes"
+
+# Lift the locale code out of the launcher, so this tests the shipped logic itself.
+funcs=$(sed -n '/^function lang_index/,/^}/p' "${script}")
+# End on the next section, not a blank line: a blank line grown inside the block
+# would truncate the lift and silently stop testing everything below it.
+grep -q '^# Find the browser' "${script}" || fail "locale block terminator moved in ${script}"
+block=$(sed -n '/^# Locale/,/^# Find the browser/p' "${script}" | sed '$d')
+echo "${block}" | grep -q 'LANGN=.*lang_index' || fail "could not lift the whole locale block from ${script}"
+
+# Run the lifted block against whatever locale vars the caller exported.
+run_block() {
+    # shellcheck disable=SC2034 # read by the eval'd locale block
+    DISTPATH="${distdir}"
+    # the launcher runs under neither -e ("! test -n x && y" returns 1) nor -u
+    set +eu
+    eval "${funcs}"
+    eval "${block}"
+    # shellcheck disable=SC2153 # LANGN is set by the eval'd block, not a LANG typo
+    echo "${LANGN}"
+}
+
+# Resolve LANG/LC_ALL/LC_MESSAGES (empty = unset) to the language number.
+lang_of() {
+    (
+        unset LANG LC_ALL LC_MESSAGES
+        # only the value matters here; hide bash's setlocale gripe when the locale is not installed
+        # shellcheck disable=SC2030 # confining the locale to this subshell is the point
+        {
+            if test -n "$1"; then export LANG="$1"; fi
+            if test -n "$2"; then export LC_ALL="$2"; fi
+            if test -n "$3"; then export LC_MESSAGES="$3"; fi
+        } 2>/dev/null
+        run_block
+    )
+}
+
+check() {
+    local want=$1 got desc
+    shift
+    desc="LANG='$1' LC_ALL='$2' LC_MESSAGES='$3'"
+    got=$(lang_of "$1" "$2" "$3")
+    test "${got}" = "${want}" || fail "${desc}: want ${want}, got ${got}"
+    echo "ok: ${desc} -> ${got}"
+}
+
+#      want  LANG            LC_ALL          LC_MESSAGES
+check 2 "fr_FR.UTF-8" "" ""
+check 14 "zh_TW.UTF-8" "" ""                      # region-qualified: Chinese-BIG5, not Simplified (13)
+check 12 "pt_BR.UTF-8" "" ""                      # region-qualified: Portugues-Brasil, not Portugues (7)
+check 13 "zh_CN.UTF-8" "" ""                      # unlisted region falls back to the bare language
+check 24 "sl_SI.UTF-8" "" ""                      # Slovenian is ISO 639-1 "sl"
+check 4 "" "de_DE.UTF-8" ""                       # LC_ALL alone
+check 4 "" "" "de_DE.UTF-8"                       # LC_MESSAGES alone
+check 4 "fr_FR.UTF-8" "de_DE.UTF-8" ""            # LC_ALL wins over LANG
+check 4 "fr_FR.UTF-8" "" "de_DE.UTF-8"            # LC_MESSAGES wins over LANG
+check 2 "de_DE.UTF-8" "fr_FR.UTF-8" "de_DE.UTF-8" # LC_ALL wins over LC_MESSAGES
+check 2 "fr_FR@euro" "" ""                        # @modifier without a charset
+check 1 "C" "" ""                                 # C, unknown and unset all mean English
+check 1 "xx_YY.UTF-8" "" ""
+check 1 "" "" ""
+
+# Exported but empty is unset (POSIX), so it must not mask LANG. lang_of cannot
+# express this: it exports nothing for an empty argument.
+got=$(
+    unset LANG LC_ALL LC_MESSAGES
+    # shellcheck disable=SC2031 # confining the locale to this subshell is the point
+    { export LANG="fr_FR.UTF-8" LC_ALL="" LC_MESSAGES=""; } 2>/dev/null
+    run_block
+)
+test "${got}" = 2 || fail "empty LC_ALL/LC_MESSAGES must not mask LANG: want 2, got ${got}"
+echo "ok: LANG='fr_FR.UTF-8' with LC_ALL='' exported -> ${got}"
+
+echo "PASS"

tests/Makefile.am

@@ -3,7 +3,7 @@
 # silently drop it from the dist tarball and break "make distcheck".
 EXTRA_DIST = $(TESTS) crawl-test.sh run-all-tests.sh check-network.sh \
 	proxy-https-server.py socks5-server.py proxy-connect-server.py \
-	proxytestlib.py tls-stall-server.py \
+	proxytestlib.py \
 	local-crawl.sh local-server.py testlib.sh server.crt server.key \
 	server-root/simple/basic.html server-root/simple/link.html \
 	server-root/stripquery/index.html server-root/stripquery/a.html \
@@ -137,6 +137,6 @@ TESTS = \
 	56_local-proxy-noleak.test \
 	57_local-proxy-connect.test \
 	58_watchdog.test \
-	59_local-tls-stall.test
+	61_webhttrack-locale.test
 
 CLEANFILES = check-network_sh.cache

tests/proxytestlib.py

@@ -18,15 +18,6 @@ PROXY_LOG = "proxy.log"
 ORIGIN_LOG = "origin-headers.log"
 
 
-def bind_ephemeral():
-    """Listening socket on a free loopback port, and that port."""
-    srv = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
-    srv.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
-    srv.bind(("127.0.0.1", 0))
-    srv.listen(16)
-    return srv, srv.getsockname()[1]
-
-
 def pipe(src, dst):
     """Relay bytes one way until EOF, then tear both ends down."""
     try:
@@ -122,7 +113,11 @@ def handle_client(conn, logdir, mode, default_port):
 
 
 def start_proxy(logdir, mode, default_port):
-    srv, port = bind_ephemeral()
+    srv = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    srv.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+    srv.bind(("127.0.0.1", 0))
+    srv.listen(16)
+    port = srv.getsockname()[1]
 
     def accept_loop():
         while True:

tests/socks5-server.py

@@ -19,7 +19,7 @@ import threading
 # python3 -P (PYTHONSAFEPATH) drops the script's own directory from sys.path
 sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
 
-from proxytestlib import bind_ephemeral, pipe  # noqa: E402
+from proxytestlib import pipe  # noqa: E402
 
 # The one name the proxy answers for; a .invalid TLD never resolves (RFC 6761),
 # so a locally-resolving client could not reach us -- success proves remote DNS.
@@ -180,7 +180,11 @@ def handle_socks(conn, logdir, mode):
 
 
 def start_socks(logdir, mode):
-    srv, port = bind_ephemeral()
+    srv = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
+    srv.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
+    srv.bind(("127.0.0.1", 0))
+    srv.listen(16)
+    port = srv.getsockname()[1]
 
     def serve():
         while True:

tests/tls-stall-server.py

@@ -1,43 +0,0 @@
-#!/usr/bin/env python3
-"""Peers that accept a connection and never speak TLS (#607).
-
-Modes: "direct" stalls the handshake straight away; "proxy <secs>" answers a
-CONNECT after <secs> before stalling, so the handshake starts on a clock the
-connect has already eaten into. Prints "PORT <n>" once listening.
-Usage: tls-stall-server.py [direct | proxy <secs>]
-"""
-import os
-import sys
-import threading
-import time
-
-# python3 -P (PYTHONSAFEPATH) drops the script's own directory from sys.path
-sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
-
-from proxytestlib import bind_ephemeral  # noqa: E402
-
-held = []  # keep every socket open: a close would fail the handshake outright
-
-
-def stall(conn, delay):
-    if delay is not None:
-        rfile = conn.makefile("rb")
-        while rfile.readline() not in (b"\r\n", b"\n", b""):
-            pass
-        time.sleep(delay)
-        conn.sendall(b"HTTP/1.0 200 Connection established\r\n\r\n")
-    held.append(conn)
-
-
-def main():
-    mode = sys.argv[1] if len(sys.argv) > 1 else "direct"
-    delay = float(sys.argv[2]) if mode == "proxy" else None
-    srv, port = bind_ephemeral()
-    sys.stdout.reconfigure(newline="\n")  # Windows would emit \r\n
-    print("PORT %d" % port, flush=True)
-    while True:
-        conn, _ = srv.accept()
-        threading.Thread(target=stall, args=(conn, delay), daemon=True).start()
-
-
-main()