Merge pull request #341 from zero-sum-seattle/fix/issue-340-pydantic-float-uiuzmh

Use Pydantic coercion for numeric stat fields

Author: Mattsface

docs/stats.md

@@ -263,6 +263,37 @@ The MLB Stats API publishes the available values directly:
 
 Common stat groups include `hitting`, `pitching`, and `fielding`. Available stat types depend on the group and endpoint. Examples include `season`, `career`, `seasonAdvanced`, `gameLog`, and `playLog`.
 
+## Numeric stat fields are typed as `float`
+
+Rate and average stats such as `avg`, `obp`, `slg`, `ops`, `era`, `whip`, and
+`babip` are typed as `Optional[float]`. The MLB Stats API returns these as
+decimal strings (for example `".287"`), and Pydantic converts them to floats
+automatically:
+
+```python
+split.stat.avg == 0.287        # not ".287"
+split.stat.model_dump()["avg"] # 0.287, not ".287"
+```
+
+This is a behavioral change from earlier releases where these fields were
+`str`. Code relying on string operations (`avg.startswith(".")`) or on
+`avg == ".287"` needs to switch to numeric comparisons.
+
+The MLB Stats API also uses two placeholder strings, `".---"` and `"-.--"`,
+for these rate stats when the underlying value is not applicable (for
+example a caught-stealing percentage when nobody has attempted a steal).
+These two known sentinels are normalized to `None` before conversion, so
+`split.stat.avg` is `None` rather than raising a validation error. Any other
+non-numeric string still raises a `ValidationError`, since it isn't a
+sentinel MLB is known to send.
+
+Fields that use MLB's innings notation remain `str`, because values like
+`"6.2"` mean 6 2/3 innings rather than the decimal 6.2:
+
+- `SimpleFieldingSplit.innings`
+- `SimplePitchingSplit.innings_pitched`
+- `AdvancedPitchingSplit.innings_pitched_per_game`
+
 ## Related documentation
 
 - [Method reference](methods.md)

mlbstatsapi/models/stats/catching.py

@@ -1,9 +1,10 @@
 from typing import Optional, List, ClassVar
-from pydantic import Field
+from pydantic import Field, field_validator
 from mlbstatsapi.models.base import MLBBaseModel
 from mlbstatsapi.models.teams import Team
 from mlbstatsapi.models.game import Game
 from .stats import Split
+from .sentinels import normalize_mlb_float_sentinel
 
 
 class SimpleCatchingSplit(MLBBaseModel):
@@ -30,23 +31,23 @@ class SimpleCatchingSplit(MLBBaseModel):
         The number of hits while catching.
     hit_by_pitch : int
         The number of batters hit by a pitch while catching.
-    avg : str
+    avg : float
         The batting average while catching.
     at_bats : int
         The number of at bats while catching.
-    obp : str
+    obp : float
         The on base percentage while catching.
-    slg : str
+    slg : float
         The slugging percentage while catching.
-    ops : str
+    ops : float
         The on-base slugging while catching.
     caught_stealing : int
         The number of runners caught stealing by the catcher.
-    caught_stealing_percentage : str
+    caught_stealing_percentage : float
         Percentage of runners caught stealing by the catcher.
     stolen_bases : int
         The number of stolen bases while catching.
-    stolen_base_percentage : str
+    stolen_base_percentage : float
         The stolen base percentage against the catcher.
     earned_runs : int
         The earned run amount against the catcher.
@@ -62,7 +63,7 @@ class SimpleCatchingSplit(MLBBaseModel):
         The number of pick offs while catching.
     total_bases : int
         The total number of bases.
-    strikeout_walk_ratio : str
+    strikeout_walk_ratio : float
         The strike out to walk ratio while catching.
     catchers_interference : int
         The number of times catcher interference committed.
@@ -84,29 +85,43 @@ class SimpleCatchingSplit(MLBBaseModel):
     intentional_walks: Optional[int] = Field(default=None, alias="intentionalWalks")
     hits: Optional[int] = None
     hit_by_pitch: Optional[int] = Field(default=None, alias="hitByPitch")
-    avg: Optional[str] = None
+    avg: Optional[float] = None
     at_bats: Optional[int] = Field(default=None, alias="atBats")
-    obp: Optional[str] = None
-    slg: Optional[str] = None
-    ops: Optional[str] = None
+    obp: Optional[float] = None
+    slg: Optional[float] = None
+    ops: Optional[float] = None
     caught_stealing: Optional[int] = Field(default=None, alias="caughtStealing")
-    caught_stealing_percentage: Optional[str] = Field(default=None, alias="caughtStealingPercentage")
+    caught_stealing_percentage: Optional[float] = Field(default=None, alias="caughtStealingPercentage")
     stolen_bases: Optional[int] = Field(default=None, alias="stolenBases")
-    stolen_base_percentage: Optional[str] = Field(default=None, alias="stolenBasePercentage")
+    stolen_base_percentage: Optional[float] = Field(default=None, alias="stolenBasePercentage")
     earned_runs: Optional[int] = Field(default=None, alias="earnedRuns")
     batters_faced: Optional[int] = Field(default=None, alias="battersFaced")
     games_pitched: Optional[int] = Field(default=None, alias="gamesPitched")
     hit_batsmen: Optional[int] = Field(default=None, alias="hitBatsmen")
     wild_pitches: Optional[int] = Field(default=None, alias="wildPitches")
     pickoffs: Optional[int] = None
     total_bases: Optional[int] = Field(default=None, alias="totalBases")
-    strikeout_walk_ratio: Optional[str] = Field(default=None, alias="strikeoutWalkRatio")
+    strikeout_walk_ratio: Optional[float] = Field(default=None, alias="strikeoutWalkRatio")
     catchers_interference: Optional[int] = Field(default=None, alias="catchersInterference")
     sac_bunts: Optional[int] = Field(default=None, alias="sacBunts")
     sac_flies: Optional[int] = Field(default=None, alias="sacFlies")
     passed_ball: Optional[int] = Field(default=None, alias="passedBall")
     pickoff_attempts: Optional[int] = Field(default=None, alias="pickoffAttempts")
 
+    @field_validator(
+        "avg",
+        "obp",
+        "slg",
+        "ops",
+        "caught_stealing_percentage",
+        "stolen_base_percentage",
+        "strikeout_walk_ratio",
+        mode="before",
+    )
+    @classmethod
+    def normalize_float_sentinels(cls, value):
+        return normalize_mlb_float_sentinel(value)
+
 
 class CatchingSeason(Split):
     """

mlbstatsapi/models/stats/fielding.py

@@ -5,6 +5,7 @@
 from mlbstatsapi.models.teams import Team
 from mlbstatsapi.models.game import Game
 from .stats import Split
+from .sentinels import normalize_mlb_float_sentinel
 
 
 class SimpleFieldingSplit(MLBBaseModel):
@@ -21,11 +22,11 @@ class SimpleFieldingSplit(MLBBaseModel):
         The number of games started.
     caught_stealing : int
         The number of runners caught stealing.
-    caught_stealing_percentage : str
+    caught_stealing_percentage : float
         The percentage of runners caught stealing.
     stolen_bases : int
         The number of stolen bases.
-    stolen_base_percentage : str
+    stolen_base_percentage : float
         The stolen base percentage.
     assists : int
         The number of assists.
@@ -35,14 +36,15 @@ class SimpleFieldingSplit(MLBBaseModel):
         The number of errors committed.
     chances : int
         The number of chances.
-    fielding : str
+    fielding : float
         The fielding percentage.
-    range_factor_per_game : str
+    range_factor_per_game : float
         Range rating per game.
-    range_factor_per_9_inn : str
+    range_factor_per_9_inn : float
         Range factor per 9 innings.
     innings : str
-        The number of innings played.
+        The number of innings played. Represented as MLB innings notation
+        (e.g. "6.2" means 6 2/3 innings), not a true decimal value.
     games : int
         The number of games played.
     passed_ball : int
@@ -51,7 +53,7 @@ class SimpleFieldingSplit(MLBBaseModel):
         The number of double plays.
     triple_plays : int
         The number of triple plays.
-    catcher_era : str
+    catcher_era : float
         The catcher ERA of the fielding stat.
     catchers_interference : int
         The number of times catchers interference was committed.
@@ -67,22 +69,22 @@ class SimpleFieldingSplit(MLBBaseModel):
     games_played: Optional[int] = Field(default=None, alias="gamesPlayed")
     games_started: Optional[int] = Field(default=None, alias="gamesStarted")
     caught_stealing: Optional[int] = Field(default=None, alias="caughtStealing")
-    caught_stealing_percentage: Optional[str] = Field(default=None, alias="caughtStealingPercentage")
+    caught_stealing_percentage: Optional[float] = Field(default=None, alias="caughtStealingPercentage")
     stolen_bases: Optional[int] = Field(default=None, alias="stolenBases")
-    stolen_base_percentage: Optional[str] = Field(default=None, alias="stolenBasePercentage")
+    stolen_base_percentage: Optional[float] = Field(default=None, alias="stolenBasePercentage")
     assists: Optional[int] = None
     putouts: Optional[int] = None
     errors: Optional[int] = None
     chances: Optional[int] = None
-    fielding: Optional[str] = None
-    range_factor_per_game: Optional[str] = Field(default=None, alias="rangeFactorPerGame")
-    range_factor_per_9_inn: Optional[str] = Field(default=None, alias="rangeFactorPer9Inn")
+    fielding: Optional[float] = None
+    range_factor_per_game: Optional[float] = Field(default=None, alias="rangeFactorPerGame")
+    range_factor_per_9_inn: Optional[float] = Field(default=None, alias="rangeFactorPer9Inn")
     innings: Optional[str] = None
     games: Optional[int] = None
     passed_ball: Optional[int] = Field(default=None, alias="passedBall")
     double_plays: Optional[int] = Field(default=None, alias="doublePlays")
     triple_plays: Optional[int] = Field(default=None, alias="triplePlays")
-    catcher_era: Optional[str] = Field(default=None, alias="catcherEra")
+    catcher_era: Optional[float] = Field(default=None, alias="catcherEra")
     catchers_interference: Optional[int] = Field(default=None, alias="catchersInterference")
     wild_pitches: Optional[int] = Field(default=None, alias="wildPitches")
     throwing_errors: Optional[int] = Field(default=None, alias="throwingErrors")
@@ -96,6 +98,19 @@ def empty_dict_to_none(cls, v: Any) -> Any:
             return None
         return v
 
+    @field_validator(
+        "caught_stealing_percentage",
+        "stolen_base_percentage",
+        "fielding",
+        "range_factor_per_game",
+        "range_factor_per_9_inn",
+        "catcher_era",
+        mode="before",
+    )
+    @classmethod
+    def normalize_float_sentinels(cls, value: Any) -> Any:
+        return normalize_mlb_float_sentinel(value)
+
 
 class FieldingSeasonAdvanced(Split):
     """