Early-exit validation: stop once enough valid proxies found

- validate_proxies: target= param — once N valid found, remaining
  tasks skip the HTTP call and drain immediately (no wasted time)
- _build_pool: target = max(min_pool_size, chain_length * 3)
- Settings: min_pool_size (default 20), max_candidates default 400->200
- UI: Stop after N valid / Max candidates fields in Settings tab
- README: updated settings table
- 1 new test for early-exit with unreachable target

Made-with: Cursor

README.md

@@ -128,7 +128,8 @@ Hits config sanitization, fetcher smoke, validator timeouts, and a live `get_dir
 | Health check | `180` s | How often the exit **gets re-proven** |
 | Pool refresh | `1800` s | How often lists get **re-fetched** and re-tested |
 | Concurrency | `64` | How hard you **spam** validation |
-| Max candidates | `400` | Cap per cycle before testing |
+| Stop after N valid | `20` | **Early-exit** — stop testing as soon as N good proxies found |
+| Max candidates | `200` | Cap per cycle (shuffled, so random sample) |
 | Timeout | `12` s | Per-proxy **patience** |
 | Kill-switch | ON | **All-or-nothing** outbound (when Admin + firewall engaged) |
 

proxy_chain_manager/app.py

@@ -759,8 +759,10 @@ def main() -> None:
 
     val_sec = _section("Validation")
     _fld(val_sec, "Concurrent validators", "conc", str(s.validation_concurrency))
-    _fld(val_sec, "Max candidates per cycle", "maxc", str(s.max_candidates),
-         "(random sample from fetch)")
+    _fld(val_sec, "Stop after N valid proxies found", "minpool", str(s.min_pool_size),
+         "(early-exit — stop testing once enough found)")
+    _fld(val_sec, "Max candidates to test per cycle", "maxc", str(s.max_candidates),
+         "(shuffled random sample from fetch)")
     _fld(val_sec, "Per-proxy timeout (sec)", "timeout", str(s.validation_timeout_seconds))
 
     sec_sec = _section("Security")
@@ -808,6 +810,7 @@ def main() -> None:
                 health_check_seconds=min(3600, max(10, int(entries["health"].get().strip()))),
                 full_refresh_seconds=min(86400, max(60, int(entries["refresh"].get().strip()))),
                 validation_concurrency=max(1, int(entries["conc"].get().strip())),
+                min_pool_size=max(3, int(entries["minpool"].get().strip())),
                 max_candidates=max(10, int(entries["maxc"].get().strip())),
                 validation_timeout_seconds=min(120.0, max(2.0, float(entries["timeout"].get().strip()))),
                 prefer_elite=bool(elite_var.get()),

proxy_chain_manager/config.py

@@ -126,7 +126,8 @@ class Settings:
 
     # ── validation ────────────────────────────────────────────────────────
     validation_concurrency: int = 64
-    max_candidates: int = 400
+    max_candidates: int = 200          # cap before validation (shuffled, so random sample)
+    min_pool_size: int = 20            # stop validating once this many good proxies found
     validation_timeout_seconds: float = 12.0
     prefer_elite: bool = False
 
@@ -204,7 +205,21 @@ def sanitize_settings(s: Settings) -> tuple[Settings, bool]:
             s.max_candidates = 10
             changed = True
     except (TypeError, ValueError):
-        s.max_candidates = 400
+        s.max_candidates = 200
+        changed = True
+    try:
+        mps = int(s.min_pool_size)
+        if mps < 3:
+            s.min_pool_size = 3
+            changed = True
+        elif mps > 500:
+            s.min_pool_size = 500
+            changed = True
+    except (TypeError, ValueError):
+        s.min_pool_size = 20
+        changed = True
+    if not hasattr(s, "min_pool_size"):
+        s.min_pool_size = 20
         changed = True
     try:
         vt = float(s.validation_timeout_seconds)

proxy_chain_manager/service.py

@@ -579,12 +579,14 @@ class ChainService:
         def on_prog(done: int, total: int) -> None:
             self._notify({"type": "validate_progress", "done": done, "total": total})
 
+        target = max(s.min_pool_size, s.chain_length * 3)
         good = await validate_proxies(
             unique,
             s.ip_check_url,
             s.validation_concurrency,
             s.validation_timeout_seconds,
             on_progress=on_prog,
+            target=target,
         )
         random.shuffle(good)
 

proxy_chain_manager/validator.py

@@ -44,7 +44,13 @@ async def validate_proxies(
     concurrency: int,
     timeout_seconds: float,
     on_progress: Callable[[int, int], None] | None = None,
+    target: int = 0,
 ) -> list[str]:
+    """Validate proxies concurrently.
+
+    Args:
+        target: Stop as soon as this many valid proxies are found (0 = test all).
+    """
     if not proxy_urls:
         return []
     sem = asyncio.Semaphore(max(1, concurrency))
@@ -57,30 +63,46 @@ async def validate_proxies(
     cap = min(600.0, float(waves) * float(timeout_seconds) + 45.0)
     cap = max(90.0, cap)
     log.info(
-        "validate_proxies: n=%d concurrency=%d per_timeout=%.1fs wall_cap~=%.0fs check=%s",
+        "validate_proxies: n=%d concurrency=%d per_timeout=%.1fs wall_cap~=%.0fs target=%s check=%s",
         total,
         c,
         timeout_seconds,
         cap,
+        target if target else "all",
         (check_url[:70] + "…") if len(check_url) > 70 else check_url,
     )
 
     last_prog_t = 0.0
     prog_step = max(1, total // 120)
+    enough = False   # set True once target reached; tasks skip HTTP call and drain fast
 
     async def one(u: str) -> None:
-        nonlocal done, last_prog_t
+        nonlocal done, last_prog_t, enough
+        # Skip cost of the HTTP check once we already have enough
+        if enough:
+            async with lock:
+                done += 1
+                if on_progress and done == total:
+                    on_progress(done, total)
+            return
         async with sem:
-            good = await _check_one(u, check_url, timeout_seconds)
+            if enough:   # re-check after potentially waiting on semaphore
+                good = False
+            else:
+                good = await _check_one(u, check_url, timeout_seconds)
         async with lock:
             done += 1
             if good:
                 ok.append(u)
+                if target and len(ok) >= target:
+                    enough = True
+                    log.info("validate_proxies: target %d reached after %d tested", target, done)
             if on_progress:
                 now = time.monotonic()
                 if (
                     done == total
                     or done == 1
+                    or enough
                     or done % prog_step == 0
                     or (now - last_prog_t) >= 0.1
                 ):
@@ -98,7 +120,7 @@ async def validate_proxies(
             cap,
             total,
         )
-    log.info("validate_proxies: finished ok=%d / %d", len(ok), total)
+    log.info("validate_proxies: finished ok=%d / %d tested", len(ok), done)
     return ok
 
 

tests/test_proxy_god.py

@@ -129,6 +129,20 @@ class TestValidator(unittest.TestCase):
 
         self.assertEqual(asyncio.run(run()), [])
 
+    def test_validate_early_exit_target(self) -> None:
+        """Once target is reached remaining tasks should be skipped (all dead here → no early exit but no crash)."""
+        async def run() -> list[str]:
+            return await validate_proxies(
+                ["http://127.0.0.1:1"] * 10,
+                "https://api.ipify.org?format=json",
+                4,
+                2.0,
+                target=2,   # target unreachable — should still return [] cleanly
+            )
+
+        result = asyncio.run(run())
+        self.assertEqual(result, [])
+
     def test_get_direct_ip_network(self) -> None:
         async def run() -> str | None:
             return await get_direct_ip("https://api.ipify.org?format=json", 12.0)