Ajout de plusieurs méthodes pour filtrer Les messages

million/analyze/word_finder.py

@@ -1,11 +1,51 @@
+from datetime import date
 import re
+import million.analyze.message_evaluation as msg_val
 from typing import List
 from million.model.message import Message
 
 
-def _wordFilter(msg: Message, words: List[str]) -> bool:
-    rgx = r"(\b"+ r"\b|\b".join(words) + r"\b)"
-    return msg.content and re.search(rgx, msg.content, re.I)
+def filter_words(messages: List[Message], words: List[str]) -> List[Message]:
+    """
+    Return every message containg the given words, you can use regex syntax inside your words
+    i.e. find_words(messages, ["dogs?","m(ous|ic)e"])
+    will search for : dog, dogs, mouse, mice
+    """
+    r_words = [rf"\b{w}\b" for w in words]
+    rgx = "(" + "|".join(r_words) + ")"
+    return [m for m in messages if m.content and re.search(rgx, m.content, re.I)]
+
+
+def filter_value(messages: List[Message], val: int) -> List[Message]:
+    """
+    Return every message whose value is evaluated to the given val
+    """
+    return [m for m in messages if msg_val.get(m) == val]
+
+
+def filter_date(messages: List[Message], other: date) -> List[Message]:
+    """
+    Return every message posted on the given date
+    """
+    return [m for m in messages if m.date_time.date() == other]
+
+
+def filter_neighbours(
+    messages: List[Message], msg: Message, amplitude: int = 10
+) -> List[Message]:
+    """
+    Return the messages posted just before and after the given one.
+    amplitude will indicate how many messages to look for in each direction
+    """
+    idx = messages.index(msg)
+    start_index = max(0, idx - amplitude)
+    end_index = min(len(messages), idx + amplitude + 1)
+    return messages[start_index:end_index]
+
+
+def retain_counts(messages: List[Message]) -> List[Message]:
+    """
+    Retain only the messages for which are considered having a counted value
+    """
+    return [msg for msg in messages if msg_val.get(msg)]
 
-def findWords(messages: List[Message], words: List[str]) -> List[Message]:
-    return filter(lambda m: _wordFilter(m, words), messages)

scripts/find_gromots.py

@@ -1,5 +1,4 @@
-from datetime import datetime
-from million.analyze.word_finder import find_words
+from million.analyze.word_finder import filter_words
 import million.parse.fb_exports as fb
 
 
@@ -23,7 +22,7 @@ gros_mots = [
     ]
 
 export = fb.parse_dirfiles(DATA_PATH)
-msg_gros_mots = find_words(export.messages, gros_mots)
+msg_gros_mots = filter_words(export.messages, gros_mots)
 
 msg_gros_mots_grp = {}