adss 1.33__tar.gz → 1.34__tar.gz

{adss-1.33 → adss-1.34}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: adss
-Version: 1.33
+Version: 1.34
 Summary: Astronomical Data Smart System
 Author-email: Gustavo Schwarz <gustavo.b.schwarz@gmail.com>
 Project-URL: Homepage, https://github.com/schwarzam/adss

{adss-1.33 → adss-1.34}/adss/auth.py

@@ -20,6 +20,63 @@ _BACKOFF_FACTOR  = float(os.getenv("ADSS_RETRY_BACKOFF", "0.5"))
 _TRUST_ENV       = os.getenv("ADSS_TRUST_ENV", "1").lower() not in ("0", "false", "no")
 _FORCE_CLOSE_STREAMS = os.getenv("ADSS_FORCE_CLOSE_STREAMS", "0").lower() in ("1", "true", "yes")
 
+def _read_all_bytes(resp: httpx.Response,
+                    chunk_size: int = 1024 * 1024,
+                    total_timeout: Optional[float] = None) -> bytes:
+    """
+    Stream the response body to memory and return bytes.
+    - Respects httpx read timeout between chunks (via iter_bytes()).
+    - Optionally enforces an overall time budget (total_timeout).
+    - Validates Content-Length when present.
+    - Always closes the response.
+    """
+    import io, time
+    from httpx import ReadTimeout, RemoteProtocolError, TransportError
+
+    # If httpx has already cached content, return it.
+    if hasattr(resp, "_content"):
+        return resp._content  # type: ignore[attr-defined]
+
+    buf = io.BytesIO()
+    bytes_read = 0
+    start = time.monotonic()
+
+    expected = None
+    cl = resp.headers.get("Content-Length")
+    if cl:
+        try:
+            expected = int(cl)
+        except ValueError:
+            expected = None
+
+    try:
+        for chunk in resp.iter_bytes(chunk_size=chunk_size):
+            if not chunk:
+                break
+            buf.write(chunk)
+            bytes_read += len(chunk)
+            if total_timeout is not None and (time.monotonic() - start) > total_timeout:
+                raise ReadTimeout("overall read timeout exceeded")
+    except (ReadTimeout, RemoteProtocolError, TransportError):
+        # ensure socket cleanup before propagating
+        try:
+            resp.close()
+        finally:
+            raise
+
+    data = buf.getvalue()
+    if expected is not None and bytes_read != expected:
+        try:
+            resp.close()
+        finally:
+            raise RemoteProtocolError(
+                f"Incomplete body: got {bytes_read} bytes, expected {expected}"
+            )
+
+    # cache like httpx does; then close
+    resp._content = data  # type: ignore[attr-defined]
+    resp.close()
+    return data
 
 def _to_httpx_timeout(t):
     """Map (connect, read) tuple or scalar into httpx.Timeout."""
@@ -31,26 +88,87 @@ def _to_httpx_timeout(t):
     return httpx.Timeout(connect=_CONNECT_TIMEOUT, read=_READ_TIMEOUT, write=_READ_TIMEOUT, pool=_CONNECT_TIMEOUT)
 
 
-def _attach_requests_compat(resp: httpx.Response) -> httpx.Response:
+def _attach_requests_compat(resp: httpx.Response):
     """
-    Attach requests-like helpers so existing code doesn't break:
-    - iter_content(chunk_size) -> yields bytes
-    - raw.read([n]) -> bytes (returns remaining body)
+    Give httpx.Response a requests-like surface and a SAFE .read():
+    - resp.iter_content(chunk_size) -> yields bytes
+    - resp.raw.read()               -> returns remaining bytes
+    - resp.read()                   -> safe, streaming-based, idempotent
     """
+    import io, time
+    from httpx import ReadTimeout, RemoteProtocolError, TransportError
+
+    # requests-like streaming
     if not hasattr(resp, "iter_content"):
         def iter_content(chunk_size: int = 1024 * 1024):
             return resp.iter_bytes(chunk_size=chunk_size)
         setattr(resp, "iter_content", iter_content)
 
+    # requests-like raw.read()
     if not hasattr(resp, "raw"):
         class _RawAdapter:
             def __init__(self, r: httpx.Response):
                 self._r = r
             def read(self, amt: Optional[int] = None) -> bytes:
-                # httpx sync API doesn't expose partial read; return remaining.
-                return self._r.read()
+                # Use the same safe read under the hood
+                return getattr(self._r, "read")( )
         setattr(resp, "raw", _RawAdapter(resp))
 
+    # ---- SAFE .read(): stream to memory, cache, and close ----
+    # Only replace if httpx hasn't already cached content
+    def _safe_read(self, *, chunk_size: int = 1024 * 1024,
+                   total_timeout: Optional[float] = None) -> bytes:
+        # If httpx already cached, return it (idempotent)
+        if hasattr(self, "_content"):
+            return self._content
+
+        buf = io.BytesIO()
+        bytes_read = 0
+        start = time.monotonic()
+
+        # If server provided length, we can validate
+        expected = None
+        cl = self.headers.get("Content-Length")
+        if cl:
+            try:
+                expected = int(cl)
+            except ValueError:
+                expected = None
+
+        try:
+            for chunk in self.iter_bytes(chunk_size=chunk_size):
+                if not chunk:
+                    break
+                buf.write(chunk)
+                bytes_read += len(chunk)
+                if total_timeout is not None and (time.monotonic() - start) > total_timeout:
+                    raise ReadTimeout("overall read timeout exceeded")
+        except (ReadTimeout, RemoteProtocolError, TransportError) as e:
+            # Ensure the socket is cleaned up
+            try:
+                self.close()
+            finally:
+                raise
+
+        data = buf.getvalue()
+        # Validate length if known
+        if expected is not None and bytes_read != expected:
+            try:
+                self.close()
+            finally:
+                raise RemoteProtocolError(
+                    f"Incomplete body: got {bytes_read} bytes, expected {expected}"
+                )
+
+        # Cache like httpx normally does, then close the stream
+        self._content = data
+        self.close()
+        return data
+
+    # Bind as a method (so `resp.read()` calls _safe_read)
+    import types
+    resp.read = types.MethodType(_safe_read, resp)  # type: ignore[attr-defined]
+
     return resp
 
 
@@ -66,7 +184,11 @@ class Auth:
         self.verify_ssl = verify_ssl
 
         # Single keep-alive client; set verify at construction.
-        self._client = httpx.Client(trust_env=_TRUST_ENV, verify=self.verify_ssl)
+        self._client = httpx.Client(
+            trust_env=_TRUST_ENV, 
+            verify=self.verify_ssl, 
+            limits=httpx.Limits(max_keepalive_connections=0, max_connections=10)
+        )
 
     def login(self, username: str, password: str, **kwargs) -> Tuple[str, User]:
         """
@@ -300,6 +422,7 @@ class Auth:
         url: str,
         headers: Optional[Dict[str, str]] = None,
         auth_required: bool = False,
+        timeout: Optional[float] = None,
         **kwargs
     ) -> requests.Response:
         """
@@ -322,7 +445,11 @@ class Auth:
         if _FORCE_CLOSE_STREAMS:
             final_headers.setdefault("Connection", "close")
 
-        timeout   = _to_httpx_timeout(kwargs.pop('timeout', _DEFAULT_TIMEOUT))
+        if timeout is None:
+            timeout   = _to_httpx_timeout(kwargs.pop('timeout', _DEFAULT_TIMEOUT))
+        else:
+            timeout = _to_httpx_timeout(timeout)
+        
         follow_redirects = kwargs.pop('allow_redirects', True)
         kwargs.pop('verify', None)  # verify is fixed on client
 
@@ -343,4 +470,61 @@ class Auth:
             follow_redirects=follow_redirects,
         )
         handle_response_errors(resp)  # fail fast on HTTP errors
-        return _attach_requests_compat(resp)
+        return _attach_requests_compat(resp)
+
+    def download_bytes(
+        self,
+        method: str,
+        url: str,
+        headers: Optional[Dict[str, str]] = None,
+        auth_required: bool = False,
+        timeout: Optional[float | Tuple[float, float]] = None,
+        total_timeout: Optional[float] = None,
+        **kwargs
+    ) -> bytes:
+        """
+        Stream a large body and return bytes.
+        Safe replacement for patterns that do `resp = download(...); resp.read()`.
+        """
+        if auth_required and not self.is_authenticated():
+            raise AuthenticationError("Authentication required for this request")
+
+        url = self._full_url(url)
+
+        final_headers = self._get_auth_headers()
+        if headers:
+            final_headers.update(headers)
+        # avoid gzip surprises on binaries; optionally force close via env
+        final_headers.setdefault("Accept-Encoding", "identity")
+        if _FORCE_CLOSE_STREAMS:
+            final_headers.setdefault("Connection", "close")
+
+        # timeouts
+        if timeout is None:
+            timeout = _to_httpx_timeout(_DEFAULT_TIMEOUT)
+        else:
+            timeout = _to_httpx_timeout(timeout)
+        follow_redirects = kwargs.pop('allow_redirects', True)
+        kwargs.pop('verify', None)
+
+        params = kwargs.pop('params', None)
+        data   = kwargs.pop('data', None)
+        json_  = kwargs.pop('json', None)
+        files  = kwargs.pop('files', None)
+
+        # open the stream
+        resp = self._request_with_retries_stream(
+            method=method,
+            url=url,
+            headers=final_headers,
+            params=params,
+            data=data,
+            json=json_,
+            files=files,
+            timeout=timeout,
+            follow_redirects=follow_redirects,
+        )
+        handle_response_errors(resp)  # raise for HTTP errors
+
+        # read it all safely and return
+        return _read_all_bytes(resp, total_timeout=total_timeout)

{adss-1.33 → adss-1.34}/adss/client.py

@@ -294,6 +294,24 @@ class ADSSClient:
         """
         return self.metadata.get_database_metadata(**kwargs)
     
+    def pretty_print_db_metadata(self, dbmeta: Optional[DatabaseMetadata] = None) -> None:
+        """
+        Pretty print the database metadata in a hierarchical format.
+        
+        Args:
+            dbmeta: Optional DatabaseMetadata object. If not provided, fetches current metadata.
+        """
+        if dbmeta is None:
+            dbmeta = self.get_database_metadata()
+        
+        for schema in dbmeta.schemas:
+            print(f"Schema: {schema.name}")
+            for table in schema.tables:
+                print(f"  Table: {table.name}")
+                for column in table.columns:
+                    nullable = "NULL" if column.is_nullable else "NOT NULL"
+                    print(f"    Column: {column.name} ({column.data_type}, {nullable})")
+    
     def update_profile(self, 
                      email: Optional[str] = None, 
                      full_name: Optional[str] = None,

{adss-1.33 → adss-1.34}/adss/endpoints/images.py

@@ -7,6 +7,7 @@ import os
 from adss.exceptions import ResourceNotFoundError
 from adss.utils import handle_response_errors
 
+import re
 
 class ImagesEndpoint:
     """
@@ -573,24 +574,42 @@ class StampImagesEndpoint:
         except Exception as e:
             raise ResourceNotFoundError(f"Failed to create stamp by coordinates: {e}")
 
-    def create_stamp_by_object(self,
-                               collection_id: int, object_name: str,
-                               filter_name: str, ra: float, dec: float, size: float,
-                               size_unit: str = "arcmin", format: str = "fits",
-                               zmin: Optional[float] = None, zmax: Optional[float] = None,
-                               pattern: Optional[str] = None,
-                               output_path: Optional[str] = None,
-                               **kwargs) -> Union[bytes, str]:
+    # TODO: Apply the same pattern of this functions to all download functions in this file
+    def create_stamp_by_object(
+        self,
+        collection_id: int,
+        object_name: str,
+        filter_name: str,
+        ra: float,
+        dec: float,
+        size: float,
+        size_unit: str = "arcmin",
+        format: str = "fits",
+        zmin: Optional[float] = None,
+        zmax: Optional[float] = None,
+        pattern: Optional[str] = None,
+        output_path: Optional[str] = None,
+        **kwargs
+    ) -> Union[bytes, str]:
         url = f"{self.base_url}/adss/v1/images/collections/{collection_id}/stamp_by_object"
+
+        # Build headers (auth if available), prefer identity for big binaries
         try:
             headers = self.auth_manager._get_auth_headers()
-        except:
-            headers = {"Accept": "image/png" if format == "png" else "application/fits"}
+        except Exception:
+            headers = {}
+        headers.setdefault("Accept", "image/png" if format == "png" else "application/fits")
+        headers.setdefault("Accept-Encoding", "identity")
 
+        # Payload
         payload: Dict[str, Any] = {
-            "object_name": object_name, "filter_name": filter_name,
-            "ra": ra, "dec": dec, "size": size,
-            "size_unit": size_unit, "format": format
+            "object_name": object_name,
+            "filter_name": filter_name,
+            "ra": ra,
+            "dec": dec,
+            "size": size,
+            "size_unit": size_unit,
+            "format": format,
         }
         if zmin is not None:
             payload["zmin"] = zmin
@@ -599,8 +618,9 @@ class StampImagesEndpoint:
         if pattern:
             payload["pattern"] = pattern
 
+        # Download bytes in one go (no Response object leaked to callers)
         try:
-            resp = self.auth_manager.download(
+            data = self.auth_manager.download_bytes(
                 method="POST",
                 url=url,
                 headers=headers,
@@ -608,22 +628,39 @@ class StampImagesEndpoint:
                 auth_required=False,
                 **kwargs
             )
-            handle_response_errors(resp)
+        except Exception as e:
+            raise ResourceNotFoundError(f"Failed to create stamp by object: {e}")
 
-            cd = resp.headers.get('Content-Disposition', '')
-            ext = "fits" if format == "fits" else "png"
-            filename = cd.split('filename=')[1].strip('"') if 'filename=' in cd else f"stamp.{ext}"
+        # If no output_path => return bytes
+        if not output_path:
+            return data
 
-            if output_path and os.path.isdir(output_path):
-                output_path = os.path.join(output_path, filename)
-            if output_path:
-                with open(output_path, 'wb') as f:
-                    f.write(resp.read())
-                return resp.read()
-            return resp.read()
+        # If writing to disk, synthesize a stable filename
+        ext = "fits" if format == "fits" else "png"
 
-        except Exception as e:
-            raise ResourceNotFoundError(f"Failed to create stamp by object: {e}")
+        # sanitize components for filesystem safety
+        def _safe(s: str) -> str:
+            s = s.strip()
+            s = re.sub(r"\s+", "_", s)           # spaces -> underscores
+            s = re.sub(r"[^A-Za-z0-9._\-+]", "", s)  # drop weird chars
+            return s or "unknown"
+
+        obj = _safe(object_name)
+        filt = _safe(filter_name)
+        size_str = f"{size:g}{size_unit}"
+
+        filename = f"stamp_{obj}_{filt}_{size_str}.{ext}"
+
+        # If output_path is a dir, append filename; otherwise treat as full path
+        final_path = output_path
+        if os.path.isdir(final_path):
+            final_path = os.path.join(final_path, filename)
+
+        os.makedirs(os.path.dirname(final_path) or ".", exist_ok=True)
+        with open(final_path, "wb") as f:
+            f.write(data)
+
+        return final_path
 
 
 class TrilogyImagesEndpoint:

{adss-1.33 → adss-1.34}/adss.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: adss
-Version: 1.33
+Version: 1.34
 Summary: Astronomical Data Smart System
 Author-email: Gustavo Schwarz <gustavo.b.schwarz@gmail.com>
 Project-URL: Homepage, https://github.com/schwarzam/adss

{adss-1.33 → adss-1.34}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "adss"
-version = "1.33"
+version = "1.34"
 description = "Astronomical Data Smart System"
 readme = "README.md"
 authors = [{ name = "Gustavo Schwarz", email = "gustavo.b.schwarz@gmail.com" }]