change html generation

README.rst

@@ -1,4 +1,4 @@
 poems
 -----
 
-All of the poems in here are good, or interesting. There are currently 8,348 poems by 546 poets.
+All of the poems in here are good, or interesting. There are currently 8,396 poems by 548 poets.

poems/author.py

@@ -1,12 +1,13 @@
 from dataclasses import dataclass, field
+from datetime import datetime
 
 @dataclass
 class Author():
     """Author dataclass"""
     tag: str
     name: str
-    birth: str
-    death: str
+    birth: dict
+    death: dict
     gender: str
     religion: str
     nationality: str
@@ -27,24 +28,16 @@ def dates(self):
             if not self.birth:
                 return ""
             else:
-                return f"(born {self.birth})"
-
-        birth_is_circa = True if "~" in self.birth else False
-        death_is_circa = True if "~" in self.death else False
+                return f"(born {self.birth['year']})"
 
-        b_numeric = int(self.birth.strip("~"))
-        d_numeric = int(self.death.strip("~"))
-
-        birth_string, death_string = str(abs(b_numeric)), str(abs(d_numeric))
-
-        birth_string = f'{"c. " if birth_is_circa else ""}{abs(b_numeric)}'
-        death_string = f'{"c. " if death_is_circa else ""}{abs(d_numeric)}'
+        birth_string = f"{'c. ' if 'circa' in self.birth else ''}{abs(self.birth['year'])}"
+        death_string = f"{'c. ' if 'circa' in self.death else ''}{abs(self.death['year'])}"
 
-        if b_numeric < 0: 
-            if d_numeric < 0: 
+        if self.birth["year"] < 0: 
+            if self.death["year"] < 0: 
                 death_string += " BC"
             else: 
                 birth_string += " BC"
                 death_string += " AD"
 
-        return f"({birth_string} -- {death_string})"
+        return f"({birth_string} – {death_string})"

poems/catalog.py

@@ -2,16 +2,18 @@
 
 import json
 import os
+import pathlib
 import warnings
+import yaml
 
 from pandas import DataFrame
 from .context import Context
 from .poem import Poem, Author
+from .utils import make_author_stats
 
 here, this_file = os.path.split(__file__)
 
-with open(f"{here}/weights.json", "r+") as f:
-    CONTEXT_WEIGHTS = json.load(f)
+CONTEXT_WEIGHTS = yaml.safe_load(pathlib.Path(f"{here}/data/weights.yml").read_text())
 
 class Catalog():
 
@@ -88,7 +90,7 @@ def apply_context(self, context: dict, forced=[], verbose=False):
                 if keyword == context[category]: 
                     if keyword in forced:
                         if verbose:
-                            print(f"Forcing context '{keyword}'")
+                            print(f"FORCING CONTEXT: {category}='{keyword}'")
                         multiplier = 1e18
                     else:
                         multiplier = weight
@@ -102,36 +104,49 @@ def apply_context(self, context: dict, forced=[], verbose=False):
         self.df.loc[:, "probability"] = self.df.likelihood / self.df.likelihood.sum()
 
 
-    def apply_history(self, history: DataFrame, latency: int = 30 * 86400, verbose: bool = False):
+    def apply_history(self, history: DataFrame, cooldown: int = 30 * 86400, manage_attrition: bool = False, verbose: bool = False):
 
         timestamp = Context.now().timestamp
 
+        author_stats = make_author_stats(history, self)
+
         last_occurence = DataFrame(columns=["timestamp"], dtype=float)
         for index, entry in history.iterrows():
             last_occurence.loc[entry.author, "timestamp"] = entry.timestamp
 
         last_occurence = last_occurence.sort_values("timestamp")
 
         indices_to_drop = []
+        treated_authors = []
         dropped_authors = []
 
         for _, entry in history.iterrows():
 
-            res = self.df.loc[(self.df.author==entry.author) & (self.df.title==entry.title)]
+            author_mask = self.df.author == entry.author
+
+            res = self.df.loc[author_mask & (self.df.title==entry.title)]
 
             if not len(res):
                 warnings.warn(f"Could not remove poem '{entry.title}' by '{entry.author}'.")
 
             indices_to_drop.extend(res.index)
 
-            if entry.timestamp > timestamp - latency:
-                if entry.author not in dropped_authors:
-                    indices_to_drop.extend(self.df.loc[self.df.author == entry.author].index)
+            # apply to all poems by this author, if not done for this author yet:
+            if entry.author not in dropped_authors:
+
+                if entry.timestamp > timestamp - cooldown:
+                    indices_to_drop.extend(self.df.loc[author_mask].index)
                     dropped_authors.append(entry.author)
 
+            if manage_attrition:
+                if entry.author not in treated_authors:
+                    self.df.loc[author_mask, "likelihood"] *= 1 - author_stats.loc[entry.author, "attrition"]
+                    treated_authors.append(entry.author)
+
         if verbose:
             print(f"Dropped authors {dropped_authors}.")
 
+
         self.df.loc[indices_to_drop, "likelihood"] = 0
 
     def __repr__(self):

poems/context.py

@@ -8,7 +8,7 @@
 
 here, this_filename = os.path.split(__file__)
 
-HOLIDAYS = yaml.safe_load(pathlib.Path(f"{here}/holidays.yml").read_text())
+HOLIDAYS = yaml.safe_load(pathlib.Path(f"{here}/data/holidays.yml").read_text())
 MONTHS = ["january", "february", "march", "april", "may", "june", "july", "august", "september", "october", "november", "december"]
 WEEKDAYS = ["monday", "tuesday", "wednesday", "thursday", "friday", "saturday", "sunday"]
 
@@ -63,9 +63,9 @@ def to_dict(self):
                     "season": self.season, 
                    "liturgy": self.liturgy, 
                    "holiday": self.holiday, 
-                      "year": f"{self.year:04}",
-                     "month": f"{self.month:02}",
-                       "day": f"{self.day:02}",
+                      "year": self.year,
+                     "month": self.month,
+                       "day": self.day,
                   "year_day": self.year_day, 
                    "weekday": self.weekday, 
                "month_epoch": self.month_epoch}

poems/curator.py

@@ -1,18 +1,14 @@
-import json, os
+import os, pathlib, yaml
 import numpy as np
 from .errors import AuthorNotFoundError, PoemNotFoundError
 from .catalog import Catalog
 from .poem import Author, Poem
 
 here, this_file = os.path.split(__file__)
 
-with open(f"{here}/weights.json", "r+") as f:
-    CONTEXT_WEIGHTS = json.load(f)
-
 class Curator():
 
-    def __init__(self, filepath=f"{here}/poems.json"):
-
+    def __init__(self, filepath=f"{here}/data/poems.json"):
         self.catalog = Catalog(filepath=filepath)
 
 

poems/holidays.yml → poems/data/holidays.yml

@@ -108,7 +108,7 @@ dates:
     24: saint_catherine_of_genoa
     25: annunciation
     26: saint_margaret_clitherow
-    27: blessed_francis_faà_di_bruno
+    27: blessed_francis_faa_di_bruno
     28: saint_hesychius_of_jerusalem
     29: saint_berthold
     30: saint_peter_regalado
@@ -284,7 +284,7 @@ dates:
     4: saint_rose_of_viterbo
     5: blessed_teresa_of_kolkata
     6: saint_eleutherius
-    7: blessed_frédéric_ozanam
+    7: blessed_frederic_ozanam
     8: marymas
     9: saint_peter_claver
     10: pedro_de_corpa_and_companions
@@ -312,7 +312,7 @@ dates:
   october:
     1: saint_therese_of_lisieux
     2: guardian_angels
-    3: saint_theodora_guérin
+    3: saint_theodora_guerin
     4: saint_francis_of_assisi
     5: blessed_francis_xavier_seelos
     6: saint_bruno
@@ -334,7 +334,7 @@ dates:
     22: saint_john_paul_ii
     23: saint_john_of_capistrano
     24: saint_anthony_mary_claret
-    25: saint_antônio_de_sant’anna_galvão
+    25: saint_antonio_de_santanna_galvao
     26: saint_bean
     27: saint_frumentius
     28: saints_simon_and_jude