jackwener
f4a36f2358
Bug fixes (P0): - _extract_cursor: both branches now check cursorType=='Bottom', preventing Top cursor from corrupting pagination state - _api_get: add retry loop with exponential backoff for HTTP 429 and JSON error code 88 rate limits Improvements (P1): - _deep_get: support int keys for list index access - Extract BEARER_TOKEN/USER_AGENT into constants.py (was duplicated in auth.py and client.py) Code quality (P2): - fetch_likes: add override_base_variables=True to avoid sending latestControlAvailable/requestContext which Likes endpoint rejects - CLI: extract _fetch_and_display() helper to dedup favorite/likes - _get_client(): remove useless try/except that re-raised same error - formatter: remove no-op tweets_to_json wrapper - _as_int/_as_float: filter.py now imports from config.py Polish (P3): - _parse_tweet_result: split into _extract_media + _extract_author - CLI module docstring: add likes command - verify_cookies: increase timeout from 3s to 10s
94 lines
2.7 KiB
Python
94 lines
2.7 KiB
Python
"""Tweet filtering and engagement scoring.
|
||
|
||
Scores tweets by a weighted engagement formula and filters by
|
||
configurable rules (topN, min score, language, etc.).
|
||
"""
|
||
|
||
from __future__ import annotations
|
||
|
||
from dataclasses import replace
|
||
import math
|
||
|
||
from .config import _as_float, _as_int
|
||
|
||
DEFAULT_WEIGHTS = {
|
||
"likes": 1.0,
|
||
"retweets": 3.0,
|
||
"replies": 2.0,
|
||
"bookmarks": 5.0,
|
||
"views_log": 0.5,
|
||
}
|
||
|
||
|
||
def score_tweet(tweet, weights=None):
|
||
# type: (Tweet, Optional[FilterWeights]) -> float
|
||
"""Calculate engagement score for a single tweet.
|
||
|
||
Formula:
|
||
score = w_likes × likes
|
||
+ w_retweets × retweets
|
||
+ w_replies × replies
|
||
+ w_bookmarks × bookmarks
|
||
+ w_views_log × log10(views)
|
||
"""
|
||
weight_map = _build_weights(weights or {})
|
||
m = tweet.metrics
|
||
return (
|
||
weight_map["likes"] * m.likes
|
||
+ weight_map["retweets"] * m.retweets
|
||
+ weight_map["replies"] * m.replies
|
||
+ weight_map["bookmarks"] * m.bookmarks
|
||
+ weight_map["views_log"] * math.log10(max(m.views, 1))
|
||
)
|
||
|
||
|
||
def filter_tweets(tweets, config):
|
||
# type: (Sequence[Tweet], Mapping[str, Any]) -> List[Tweet]
|
||
"""Filter and rank tweets according to config.
|
||
|
||
Config keys:
|
||
mode: "topN" | "score" | "all"
|
||
topN: int
|
||
minScore: float
|
||
lang: list[str] (empty = no filter)
|
||
excludeRetweets: bool
|
||
weights: dict
|
||
"""
|
||
filtered = list(tweets)
|
||
|
||
# 1. Language filter
|
||
lang_filter = config.get("lang", [])
|
||
if lang_filter:
|
||
lang_set = {str(lang) for lang in lang_filter if str(lang)}
|
||
filtered = [tweet for tweet in filtered if tweet.lang in lang_set]
|
||
|
||
# 2. Exclude retweets
|
||
if config.get("excludeRetweets", False):
|
||
filtered = [tweet for tweet in filtered if not tweet.is_retweet]
|
||
|
||
# 3. Score all tweets
|
||
weights = _build_weights(config.get("weights", {}))
|
||
scored = [replace(tweet, score=round(score_tweet(tweet, weights), 1)) for tweet in filtered]
|
||
|
||
# 4. Sort by score (descending)
|
||
scored.sort(key=lambda tweet: tweet.score, reverse=True)
|
||
|
||
# 5. Apply filter mode
|
||
mode = str(config.get("mode", "topN"))
|
||
if mode == "topN":
|
||
top_n = max(_as_int(config.get("topN"), 20), 1)
|
||
return scored[:top_n]
|
||
if mode == "score":
|
||
min_score = _as_float(config.get("minScore"), 50.0)
|
||
return [tweet for tweet in scored if tweet.score >= min_score]
|
||
return scored
|
||
|
||
|
||
def _build_weights(raw_weights):
|
||
# type: (Mapping[str, Any]) -> Dict[str, float]
|
||
"""Merge custom weights with defaults and coerce to float."""
|
||
merged = {}
|
||
for key, default_value in DEFAULT_WEIGHTS.items():
|
||
merged[key] = _as_float(raw_weights.get(key), default_value)
|
||
return merged
|