ticket #116 attempt 1

devplacepy/avatar.py

@@ -15,19 +15,27 @@ def avatar_seed(user) -> str:
     return user.get("avatar_seed") or user.get("username") or ""
 
 
-def generate_avatar_svg(seed: str) -> str:
+def _svg_with_size(svg: str, size: int) -> str:
+    size_attr = f' width="{size}" height="{size}"'
+    if svg.startswith("<svg"):
+        tag_end = svg.index(">")
+        return svg[:tag_end] + size_attr + svg[tag_end:]
+    return svg
+
+
+def generate_avatar_svg(seed: str, size: int = 128) -> str:
     try:
         from multiavatar.multiavatar import multiavatar
 
         svg = multiavatar(seed, None, None)
         if svg and svg.strip().startswith("<svg"):
-            return svg
+            return _svg_with_size(svg, size)
     except Exception as e:
         logger.exception(e)
         logger.warning(f"Avatar generation failed for {seed}: {e}")
     initial = seed[:1].upper() if seed else "?"
     return (
-        '<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 100 100">'
+        f'<svg xmlns="http://www.w3.org/2000/svg" width="{size}" height="{size}" viewBox="0 0 100 100">'
         f'<rect width="100" height="100" rx="50" fill="#ff6b35"/>'
         f'<text x="50" y="65" text-anchor="middle" fill="white" font-size="40" font-weight="700" font-family="sans-serif">{initial}</text></svg>'
     )

devplacepy/routers/avatar.py

@@ -17,14 +17,16 @@ _CACHE_CONTROL = f"public, max-age={SECONDS_PER_DAY}, immutable"
 
 @router.get("/{style}/{seed}")
 async def avatar_proxy(request: Request, style: str, seed: str, size: int = 128):
-    etag = '"' + hashlib.md5(f"{seed}:{size}".encode("utf-8")).hexdigest() + '"'
+    size = max(16, min(512, size))
+    cache_key = f"{seed}:{size}"
+    etag = '"' + hashlib.md5(cache_key.encode("utf-8")).hexdigest() + '"'
     headers = {"ETag": etag, "Cache-Control": _CACHE_CONTROL}
 
     if request.headers.get("if-none-match") == etag:
         return Response(status_code=304, headers=headers)
 
-    svg = _cache.get(seed)
+    svg = _cache.get(cache_key)
     if svg is None:
-        svg = generate_avatar_svg(seed)
-        _cache.set(seed, svg)
+        svg = generate_avatar_svg(seed, size)
+        _cache.set(cache_key, svg)
     return Response(content=svg, media_type="image/svg+xml", headers=headers)

devplacepy/services/jobs/deepsearch/crawl.py

@@ -22,8 +22,6 @@ from .pdf import MAX_PDF_BYTES, extract_pdf_text, is_pdf
 
 logger = logging.getLogger(__name__)
 
-_pw_lock = asyncio.Lock()
-
 RSEARCH_URL = "https://rsearch.app.molodetz.nl"
 RSEARCH_TIMEOUT_SECONDS = 45.0
 FETCH_TIMEOUT_SECONDS = 20.0
@@ -174,7 +172,13 @@ async def _render_with_playwright(
 ) -> tuple[str, str, int, list[tuple[str, str]]]:
     from playwright.async_api import async_playwright
 
-    async def _render(browser) -> tuple[str, str, int, list[tuple[str, str]]]:
+    own_browser = browser is None
+    if own_browser:
+        pw = await async_playwright().__aenter__()
+        browser = await pw.chromium.launch(
+            headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
+        )
+    try:
         context = await browser.new_context(user_agent=USER_AGENT)
         page = await context.new_page()
         response = await page.goto(url, wait_until="load", timeout=30000)
@@ -185,18 +189,10 @@ async def _render_with_playwright(
         await context.close()
         extracted = extract_html(content, base_url=url)
         return extracted.title, extracted.text, status, extracted.links
-
-    if browser is None:
-        async with async_playwright() as pw:
-            browser = await pw.chromium.launch(
-                headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
-            )
-            try:
-                return await _render(browser)
-            finally:
-                await browser.close()
-    else:
-        return await _render(browser)
+    finally:
+        if own_browser:
+            await browser.close()
+            await pw.__aexit__(None, None, None)
 
 
 async def fetch_page(url: str, depth: int, browser=None) -> CrawledPage | None:
@@ -246,20 +242,18 @@ async def fetch_page(url: str, depth: int, browser=None) -> CrawledPage | None:
         except (LookupError, ValueError) as exc:
             logger.info("deepsearch decode failed for %s: %s", url, exc)
     if len(text) < MIN_PAGE_CHARS:
-        async with _pw_lock:
-            try:
-                r_title, r_text, r_status, r_links = await _render_with_playwright(url, browser)
-            except Exception as exc:
-                logger.info("deepsearch render failed for %s: %s", url, exc)
-                r_title, r_text, r_status, r_links = "", "", 0, []
-        if len(r_text) > len(text):
-            title, text, status, source, links = (
-                r_title or title,
-                r_text,
-                r_status or status,
-                "playwright",
-                r_links,
-            )
+        try:
+            r_title, r_text, r_status, r_links = await _render_with_playwright(url, browser)
+            if len(r_text) > len(text):
+                title, text, status, source, links = (
+                    r_title or title,
+                    r_text,
+                    r_status or status,
+                    "playwright",
+                    r_links,
+                )
+        except Exception as exc:
+            logger.info("deepsearch render failed for %s: %s", url, exc)
     if len(text) < MIN_PAGE_CHARS:
         return None
     return CrawledPage(
@@ -302,99 +296,102 @@ async def crawl(
     total = min(len(level_candidates), max_pages)
     cancelled = False
 
+    pw = None
     browser = None
-    async with async_playwright() as pw:
-        try:
-            browser = await pw.chromium.launch(
-                headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
-            )
-        except Exception as exc:
-            logger.warning("deepsearch playwright launch failed, pages will use httpx only: %s", exc)
+    try:
+        pw = await async_playwright().__aenter__()
+        browser = await pw.chromium.launch(
+            headless=True, args=["--no-sandbox", "--disable-dev-shm-usage"]
+        )
+    except Exception as exc:
+        logger.warning("deepsearch playwright launch failed, pages will use httpx only: %s", exc)
 
-        try:
-            for level in range(max(1, depth)):
-                if cancelled or fetched >= max_pages or not level_candidates:
+    try:
+        for level in range(max(1, depth)):
+            if cancelled or fetched >= max_pages or not level_candidates:
+                break
+            next_candidates: list[dict] = []
+            for start in range(0, len(level_candidates), CRAWL_CONCURRENCY):
+                if fetched >= max_pages:
                     break
-                next_candidates: list[dict] = []
-                for start in range(0, len(level_candidates), CRAWL_CONCURRENCY):
+                if await should_stop():
+                    emit({"type": "stage", "stage": "cancelled", "message": "Crawl cancelled"})
+                    cancelled = True
+                    break
+                batch = level_candidates[start : start + CRAWL_CONCURRENCY][: max_pages - fetched]
+                for candidate in batch:
+                    emit(
+                        {
+                            "type": "progress",
+                            "done": fetched,
+                            "total": total,
+                            "url": candidate["url"],
+                            "depth": level,
+                            "message": f"Reading {candidate['url']}",
+                        }
+                    )
+                    if is_cached(candidate["url"]):
+                        emit({"type": "page_cached", "url": candidate["url"], "reason": "seen in a prior run"})
+                fetch_start = time.perf_counter()
+                results = await asyncio.gather(
+                    *(_resolve_candidate(candidate, level, browser) for candidate in batch),
+                    return_exceptions=True,
+                )
+                elapsed_ms = int((time.perf_counter() - fetch_start) * 1000)
+                for candidate, page in zip(batch, results):
+                    url = candidate["url"]
+                    if isinstance(page, BaseException):
+                        logger.info("deepsearch fetch crashed for %s: %s", url, page)
+                        page = None
+                    if page is None:
+                        emit(
+                            {
+                                "type": "page_skipped",
+                                "url": url,
+                                "reason": "no readable content",
+                                "elapsed_ms": elapsed_ms,
+                            }
+                        )
+                        continue
                     if fetched >= max_pages:
                         break
-                    if await should_stop():
-                        emit({"type": "stage", "stage": "cancelled", "message": "Crawl cancelled"})
-                        cancelled = True
-                        break
-                    batch = level_candidates[start : start + CRAWL_CONCURRENCY][: max_pages - fetched]
-                    for candidate in batch:
+                    digest = content_hash(page.text)
+                    if digest in outcome.seen_hashes:
                         emit(
                             {
-                                "type": "progress",
-                                "done": fetched,
-                                "total": total,
-                                "url": candidate["url"],
-                                "depth": level,
-                                "message": f"Reading {candidate['url']}",
-                            }
-                        )
-                        if is_cached(candidate["url"]):
-                            emit({"type": "page_cached", "url": candidate["url"], "reason": "seen in a prior run"})
-                    fetch_start = time.perf_counter()
-                    results = await asyncio.gather(
-                        *(_resolve_candidate(candidate, level, browser) for candidate in batch),
-                        return_exceptions=True,
-                    )
-                    elapsed_ms = int((time.perf_counter() - fetch_start) * 1000)
-                    for candidate, page in zip(batch, results):
-                        url = candidate["url"]
-                        if isinstance(page, BaseException):
-                            logger.info("deepsearch fetch crashed for %s: %s", url, page)
-                            page = None
-                        if page is None:
-                            emit(
-                                {
-                                    "type": "page_skipped",
-                                    "url": url,
-                                    "reason": "no readable content",
-                                    "elapsed_ms": elapsed_ms,
-                                }
-                            )
-                            continue
-                        if fetched >= max_pages:
-                            break
-                        digest = content_hash(page.text)
-                        if digest in outcome.seen_hashes:
-                            emit(
-                                {
-                                    "type": "page_duplicate",
-                                    "url": url,
-                                    "reason": "duplicate content",
-                                    "elapsed_ms": elapsed_ms,
-                                }
-                            )
-                            continue
-                        outcome.seen_hashes.add(digest)
-                        outcome.pages.append(page)
-                        fetched += 1
-                        emit(
-                            {
-                                "type": "page_loaded",
-                                "url": page.url,
-                                "title": page.title,
-                                "source": page.source,
-                                "depth": level,
-                                "render": page.source == "playwright",
+                                "type": "page_duplicate",
+                                "url": url,
+                                "reason": "duplicate content",
                                 "elapsed_ms": elapsed_ms,
-                                "done": fetched,
-                                "total": total,
                             }
                         )
-                        if level + 1 < depth:
-                            for link in relevant_links(page.links, query, LINKS_PER_PAGE):
-                                if link not in seen_urls:
-                                    seen_urls.add(link)
-                                    next_candidates.append({"url": link})
-                level_candidates = next_candidates
-                total = min(total + len(next_candidates), max_pages)
-        finally:
-            if browser is not None:
-                await browser.close()
+                        continue
+                    outcome.seen_hashes.add(digest)
+                    outcome.pages.append(page)
+                    fetched += 1
+                    emit(
+                        {
+                            "type": "page_loaded",
+                            "url": page.url,
+                            "title": page.title,
+                            "source": page.source,
+                            "depth": level,
+                            "render": page.source == "playwright",
+                            "elapsed_ms": elapsed_ms,
+                            "done": fetched,
+                            "total": total,
+                        }
+                    )
+                    if level + 1 < depth:
+                        for link in relevant_links(page.links, query, LINKS_PER_PAGE):
+                            if link not in seen_urls:
+                                seen_urls.add(link)
+                                next_candidates.append({"url": link})
+            level_candidates = next_candidates
+            total = min(total + len(next_candidates), max_pages)
+    finally:
+        if browser is not None:
+            await browser.close()
+        if pw is not None:
+            await pw.__aexit__(None, None, None)
     return outcome

tests/api/avatar.py

@@ -1,7 +1,6 @@
 # retoor <retoor@molodetz.nl>
 
 import requests
-from devplacepy.avatar import avatar_url, generate_avatar_svg
 from tests.conftest import BASE_URL
 
 
@@ -10,3 +9,32 @@ def test_avatar_endpoint_serves_svg(app_server):
     assert r.status_code == 200
     assert "svg" in r.headers.get("content-type", "").lower()
     assert "<svg" in r.text
+
+
+def test_avatar_respects_size(app_server):
+    r = requests.get(f"{BASE_URL}/avatar/multiavatar/alice_test?size=32")
+    assert r.status_code == 200
+    assert 'width="32"' in r.text
+    assert 'height="32"' in r.text
+
+
+def test_avatar_different_sizes_different_response(app_server):
+    r32 = requests.get(f"{BASE_URL}/avatar/multiavatar/bob_test?size=32")
+    r128 = requests.get(f"{BASE_URL}/avatar/multiavatar/bob_test?size=128")
+    assert r32.status_code == 200
+    assert r128.status_code == 200
+    etag32 = r32.headers.get("etag", "")
+    etag128 = r128.headers.get("etag", "")
+    assert etag32 != etag128, "same seed with different sizes must produce different ETags"
+
+
+def test_avatar_size_clamped_low(app_server):
+    r = requests.get(f"{BASE_URL}/avatar/multiavatar/clamp_low?size=1")
+    assert r.status_code == 200
+    assert 'width="16"' in r.text
+
+
+def test_avatar_size_clamped_high(app_server):
+    r = requests.get(f"{BASE_URL}/avatar/multiavatar/clamp_high?size=9999")
+    assert r.status_code == 200
+    assert 'width="512"' in r.text