pyrox-client 0.0.2__tar.gz → 0.0.4__tar.gz

pyrox_client-0.0.4/.gitignore

@@ -0,0 +1,3 @@
+*.pyc
+
+/src/pyrox/__pycache__/

pyrox_client-0.0.4/PKG-INFO

@@ -0,0 +1,30 @@
+Metadata-Version: 2.4
+Name: pyrox-client
+Version: 0.0.4
+Summary: HYROX race data client – retrieve full race results via Python
+Author: Vlad Matei
+License: MIT
+Requires-Python: >=3.9
+Requires-Dist: fastparquet>=2024.5.0
+Requires-Dist: fsspec>=2024.2.0
+Requires-Dist: pandas>=2.2
+Requires-Dist: platformdirs>=4.2.0
+Requires-Dist: requests>=2.31
+Requires-Dist: s3fs>=2024.2.0
+Provides-Extra: duckdb
+Requires-Dist: duckdb>=1.0.0; extra == 'duckdb'
+Description-Content-Type: text/markdown
+
+Package to retrieve Hyrox race data programmatically
+
+At the moment, only 3 endpoints -- examples below
+
+
+
+```commandline
+s6races_info = list_races(season=6)  #  list available races of season 6
+all_races_info = list_races()  #  without passing in a season parameter - see all available races in the data warehouse
+s6races_subset = get_season(season=6, locations=['london', 'hamburg'])  # get a subset of races from a specified season
+london_season6 = get_race(season=6, location="london")   #  get a specific race
+hamburg_season6 = get_race(season=6, location="hamburg", division="open", gender="m")  #  get a specific race, with extra-filtering applied
+```

pyrox_client-0.0.4/README.md

@@ -0,0 +1,13 @@
+Package to retrieve Hyrox race data programmatically
+
+At the moment, only 3 endpoints -- examples below
+
+
+
+```commandline
+s6races_info = list_races(season=6)  #  list available races of season 6
+all_races_info = list_races()  #  without passing in a season parameter - see all available races in the data warehouse
+s6races_subset = get_season(season=6, locations=['london', 'hamburg'])  # get a subset of races from a specified season
+london_season6 = get_race(season=6, location="london")   #  get a specific race
+hamburg_season6 = get_race(season=6, location="hamburg", division="open", gender="m")  #  get a specific race, with extra-filtering applied
+```

{pyrox_client-0.0.2 → pyrox_client-0.0.4}/pyproject.toml

@@ -1,10 +1,10 @@
 [build-system]
-requires = ["hatchling>=1.20"]
+requires = ["hatchling>=1.20", "hatch-vcs>=0.4"]
 build-backend = "hatchling.build"
 
 [project]
 name = "pyrox-client"
-version = "0.0.2"
+dynamic = ["version"]
 description = "HYROX race data client – retrieve full race results via Python"
 authors = [{ name = "Vlad Matei" }]
 readme = "README.md"
@@ -34,3 +34,6 @@ include = [
   "README.md",
   "pyproject.toml"
 ]
+
+[tool.hatch.version]
+path = "src/pyrox/__init__.py"

{pyrox_client-0.0.2 → pyrox_client-0.0.4}/src/pyrox/__init__.py

@@ -1,5 +1,6 @@
 from .core import get_race, list_races
 from .config import set_bucket, set_manifest_path, get_config
 
+__version__ = "0.0.4"
+
 __all__ = ["get_race", "list_races", "set_bucket", "set_manifest_path", "get_config"]
-__version__ = "0.0.2"

{pyrox_client-0.0.2 → pyrox_client-0.0.4}/src/pyrox/config.py

@@ -21,7 +21,6 @@ _config = PyroxConfig()
 def set_bucket(uri:str) -> None:
     _config.bucket = uri.rsplit("/")
 
-
 def set_manifest_path(path: str) -> None:
     _config.manifest_path = path
 

pyrox_client-0.0.4/src/pyrox/core.py

@@ -0,0 +1,74 @@
+from __future__ import annotations
+from concurrent.futures import ThreadPoolExecutor, as_completed
+from typing import Iterable, Optional
+import os
+import pandas as pd
+import httpx
+
+DEFAULT_API_URL = "https://pyrox-api-proud-surf-3131.fly.dev"
+DEFAULT_API_KEY = os.getenv("PYROX_API_KEY")  #  optional for now as not used on api side
+#  DEFAULT_API_URL = "http://localhost:8000"  --> used for testing when running api in docker container
+class ApiError(RuntimeError): ...
+class RaceNotFound(RuntimeError): ...
+
+def _client(base_url: str | None = None, api_key: str | None = None) -> httpx.Client:
+    headers = {}
+    if api_key:
+        headers["Authorization"] = f"Bearer {api_key}"
+    return httpx.Client(base_url=base_url or DEFAULT_API_URL, headers=headers, timeout=30)
+
+def list_races(season: int | None = None, base_url: str | None = None, api_key: str | None = DEFAULT_API_KEY) -> pd.DataFrame:
+    with _client(base_url, api_key) as c:
+        r = c.get("/v1/manifest")
+        if r.status_code != 200:
+            raise ApiError(f"manifest failed: {r.status_code} {r.text}")
+        rows = r.json()  # [{season, location, path}, ...]
+    df = pd.DataFrame(rows)
+    if season is not None:
+        df = df[df["season"] == int(season)]
+    return df[["season", "location"]].drop_duplicates().sort_values(["season", "location"]).reset_index(drop=True)
+
+def get_race(*, season: int, location: str, gender: Optional[str] = None, division: Optional[str] = None,
+             base_url: str | None = None, api_key: str | None = DEFAULT_API_KEY) -> pd.DataFrame:
+    params = {}
+    if gender: params["gender"] = gender
+    if division: params["division"] = division
+    with _client(base_url, api_key) as c:
+        r = c.get(f"/v1/race/{int(season)}/{location}", params=params)
+        if r.status_code == 404:
+            raise RaceNotFound(f"No race found for season={season}, location='{location}'")
+        if r.status_code != 200:
+            raise ApiError(f"race fetch failed: {r.status_code} {r.text}")
+        rows = r.json()  # list[dict]
+    df = pd.DataFrame(rows)
+    return df
+
+
+def get_season(season: int, locations: Optional[Iterable[str]] = None,
+               max_workers: int = 8, base_url: str | None = None, api_key: str | None = DEFAULT_API_KEY,
+               gender: Optional[str] = None, division: Optional[str] = None) -> pd.DataFrame:
+    # discover available races first (avoids 404 spam)
+    m = list_races(season=season, base_url=base_url, api_key=api_key)
+    if locations:
+        want = {loc.casefold() for loc in locations}
+        m = m[m["location"].str.casefold().isin(want)]
+    if m.empty:
+        raise RaceNotFound(f"No races found for season={season}")
+    #  function to get one data frame for a specifc race
+    def _one(loc: str) -> pd.DataFrame:
+        return get_race(season=season, location=loc, gender=gender, division=division, base_url=base_url, api_key=api_key)
+
+    frames: list[pd.DataFrame] = []
+    #  split the work to get the dataframe across multiple workers
+    with ThreadPoolExecutor(max_workers=max_workers) as ex:
+        futs = [ex.submit(_one, loc) for loc in m["location"].tolist()]
+        for f in as_completed(futs):
+            frames.append(f.result())
+    #  join and return the data!
+    if not frames:
+        return pd.DataFrame()
+    out = pd.concat(frames, ignore_index=True)
+
+    return out
+
+

{pyrox_client-0.0.2 → pyrox_client-0.0.4}/src/pyrox/manifest.py

@@ -7,14 +7,15 @@ from pathlib import Path
 import fsspec
 import pandas as pd
 
-from .config import get_config
-from .errors import ManifestUnavailable
-
+from pyrox.config import get_config
+from pyrox.errors import ManifestUnavailable
+import logging
+logger = logging.getLogger(__name__)
+logging.basicConfig(level=logging.INFO)
 
 def _s3_open(path: str, mode: str = "rb", anon: bool = True):
     """
     Unified opener for S3 (via fsspec/s3fs).
-
     Parameters
     ----------
     path : str
@@ -23,7 +24,6 @@ def _s3_open(path: str, mode: str = "rb", anon: bool = True):
         File mode, usually "rb" for binary reads.
     anon : bool
         True => anonymous/public read (no AWS creds required).
-
     Returns
     -------
     contextmanager
@@ -36,7 +36,6 @@ def _s3_open(path: str, mode: str = "rb", anon: bool = True):
 def _cache_paths(name: str) -> Tuple[Path, Path]:
     """
     Compute cache locations for the manifest CSV and its ETag sidecar.
-
     We cache:
       - the CSV bytes as ~/.cache/pyrox/manifest.latest.csv
       - the ETag string as ~/.cache/pyrox/manifest.latest.csv.etag
@@ -122,8 +121,10 @@ def load_manifest(refresh: bool = True) -> pd.DataFrame:
             # Some backends wrap ETag in quotes; we keep it as-is because
             # we compare exact strings on both sides.
             current_etag = info.get("ETag")
-        except Exception:
+        except Exception as e:
             # HEAD can fail (e.g., transient network); we'll fall back to cache or GET.
+
+            logger.info(f"Have not been able to retrieve head_s3 from the manifest uri: {s3_manifest_uri}. Encountered exception: {e}. Setting etag as none and looking for cache.")
             current_etag = None
 
     # 2) If we have a cached file and either:
@@ -135,7 +136,6 @@ def load_manifest(refresh: bool = True) -> pd.DataFrame:
         try:
             return pd.read_csv(cache_file)
         except Exception:
-            # Corrupted cache? Continue to re-download below.
             pass
 
     # 3) Re-download from S3 and update cache.
@@ -164,7 +164,6 @@ def load_manifest(refresh: bool = True) -> pd.DataFrame:
         return df
 
     except Exception as e:
-        # 4) If network failed, try to limp with the existing cache (even if ETag mismatch).
         if cache_file.exists():
             try:
                 return pd.read_csv(cache_file)
@@ -172,3 +171,5 @@ def load_manifest(refresh: bool = True) -> pd.DataFrame:
                 pass
         # No working cache → bail with a clear error
         raise ManifestUnavailable(f"Could not load manifest from {s3_manifest_uri}: {e}")
+
+

pyrox_client-0.0.2/PKG-INFO

@@ -1,19 +0,0 @@
-Metadata-Version: 2.4
-Name: pyrox-client
-Version: 0.0.2
-Summary: HYROX race data client – retrieve full race results via Python
-Author: Vlad Matei
-License: MIT
-Requires-Python: >=3.9
-Requires-Dist: fastparquet>=2024.5.0
-Requires-Dist: fsspec>=2024.2.0
-Requires-Dist: pandas>=2.2
-Requires-Dist: platformdirs>=4.2.0
-Requires-Dist: requests>=2.31
-Requires-Dist: s3fs>=2024.2.0
-Provides-Extra: duckdb
-Requires-Dist: duckdb>=1.0.0; extra == 'duckdb'
-Description-Content-Type: text/markdown
-
-Package to retrieve Hyrox race data programatically.
-

pyrox_client-0.0.2/README.md

@@ -1,2 +0,0 @@
-Package to retrieve Hyrox race data programatically.
-

pyrox_client-0.0.2/src/pyrox/core.py

@@ -1,102 +0,0 @@
-from __future__ import annotations
-
-import io
-from pathlib import Path
-import pandas as pd
-
-from .manifest import load_manifest, _s3_open   # re-use your manifest loader + fsspec opener
-from .config import get_config
-from .errors import RaceNotFound, ParquetReadError
-
-
-def list_races(season: int | None = None) -> pd.DataFrame:
-    """
-    Return available (season, location) pairs from the manifest.
-    """
-    df = load_manifest(refresh=True)
-    if season is not None:
-        try:
-            df = df[df["season"] == int(season)]
-        except ValueError as e:
-            raise (ValueError(f"Input {season} is not of expected type. Please make sure you request an integer."))
-    return (
-        df[["season", "location"]]
-        .drop_duplicates()
-        .sort_values(["season", "location"])
-        .reset_index(drop=True)
-    )
-
-
-def _local_parquet_cache_path(s3_key: str) -> Path:
-    """
-    Map an S3 key to a deterministic local cache path under ~/.cache/pyrox/parquet/.
-    """
-    cfg = get_config()
-    p = cfg.cache_dir / "parquet" / s3_key.replace("/", "__")
-    p.parent.mkdir(parents=True, exist_ok=True)
-    return p
-
-
-def _read_parquet_from_s3(s3_uri: str) -> pd.DataFrame:
-    """
-    Download the parquet bytes anonymously from S3 and read with fastparquet.
-    """
-    try:
-        with _s3_open(s3_uri, "rb", anon=True) as f:
-            data = f.read()
-        return pd.read_parquet(io.BytesIO(data), engine="fastparquet")
-    except Exception as e:
-        raise ParquetReadError(f"Failed to read parquet from {s3_uri}: {e}")
-# core.py
-def _to_s3_uri(bucket: str, key: str) -> str:
-    """Join bucket + key safely. If key is already a full s3:// URI, use it as-is."""
-    key = key.strip()
-    if key.startswith("s3://"):
-        return key
-    # allow bucket to be "s3://bucket" or just "bucket"
-    if not bucket.startswith("s3://"):
-        bucket = f"s3://{bucket}"
-    return f"{bucket.rstrip('/')}/{key.lstrip('/')}"
-
-
-def get_race(*, season: int, location: str) -> pd.DataFrame:
-    """
-    Resolve (season, location) -> s3_key via the manifest, then return the race DataFrame.
-
-    Uses a simple local file cache. If the cached parquet fails to read,
-    it is discarded and we re-fetch from S3.
-    """
-    manifest = load_manifest(refresh=True)
-
-    # case-insensitive match on location
-    row = manifest[
-        (manifest["season"] == int(season)) &
-        (manifest["location"].str.casefold() == location.casefold())
-    ]
-
-    if row.empty:
-        raise RaceNotFound(f"No race found for season={season}, location='{location}'. "
-                           f"Try: pyrox.list_races({season})")
-
-    s3_key = row.iloc[0]["path"]
-    cfg = get_config()
-    s3_uri = _to_s3_uri(cfg.bucket, s3_key)
-
-    # check local cache first
-    local_path = _local_parquet_cache_path(s3_key)
-    if local_path.exists():
-        try:
-            return pd.read_parquet(local_path, engine="fastparquet")
-        except Exception:
-            local_path.unlink(missing_ok=True)  # bad cache → remove and refetch
-
-    # fetch from S3 and cache
-    df = _read_parquet_from_s3(s3_uri)
-    try:
-        df.to_parquet(local_path, index=False, engine="fastparquet")
-    except Exception:
-        # cache failures should not fail the API
-        pass
-    return df
-
-breakheere = 0