streamlit-octostar-utils 2.11a4__tar.gz → 2.11a6__tar.gz

{streamlit_octostar_utils-2.11a4 → streamlit_octostar_utils-2.11a6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: streamlit-octostar-utils
-Version: 2.11a4
+Version: 2.11a6
 Summary: 
 License: MIT
 License-File: LICENSE

{streamlit_octostar_utils-2.11a4 → streamlit_octostar_utils-2.11a6}/pyproject.toml

@@ -5,7 +5,7 @@ include = '\.pyi?$'
 
 [tool.poetry]
 name = "streamlit-octostar-utils"
-version = "2.11a4"
+version = "2.11a6"
 description = ""
 license = "MIT"
 authors = ["Octostar"]

{streamlit_octostar_utils-2.11a4 → streamlit_octostar_utils-2.11a6}/streamlit_octostar_utils/nlp/language.py

@@ -1,20 +1,39 @@
 import re
 import py3langid as langid
-import iso639 as languages
+
+from iso639 import Lang
+
+FLAIR_MODELS = {
+    "en": "flair/ner-english-large",
+    "es": "flair/ner-spanish-large",
+    "de": "flair/ner-german-large",
+    "nl": "flair/ner-dutch-large",
+    "multi": "flair/ner-multi",                     # English, German, French, Spanish
+    "multi-fast": "flair/ner-multi-fast",           # English, German, Dutch, Spanish
+}
+
+SPACY_MODELS = {
+    "en": "en_core_web_sm",
+    "es": "es_core_news_sm",
+    "fr": "fr_core_news_sm",
+    "de": "de_core_news_sm",
+    "it": "it_core_news_sm"
+}
 
 
 def alpha2_to_language(alpha2: str) -> str:
     if not alpha2:
-        return None
-    code = alpha2.strip().lower()
-    return languages.to_name(code)
+        raise ValueError("Language code must be a non-empty string.")
+    return Lang(alpha2).name
+
 
 def language_to_alpha2(language_name: str) -> str:
     if not language_name:
-        return None
-    name = language_name.strip().lower()
-    data = languages.find(name)
-    return data["iso639_1"]
+        raise ValueError("Language name must be a non-empty string.")
+
+    name = re.sub(r'\b\w+', lambda m: m.group(0).capitalize(), name)
+    return Lang(name).pt1
+
 
 def detect_language(text, min_confidence=None):
     detector = langid.langid.LanguageIdentifier.from_pickled_model(
@@ -23,33 +42,23 @@ def detect_language(text, min_confidence=None):
     detected_lang, confidence = detector.classify(text)
     if min_confidence and confidence < min_confidence:
         return None, confidence
-    detected_lang = re.sub("[^A-Za-z]", "", detected_lang).lower()
-    detected_lang = languages.to_name(detected_lang).lower()
+    detected_lang = alpha2_to_language(detected_lang)
     return detected_lang, confidence
 
-FLAIR_MODELS = {
-    "en": "flair/ner-english-large",
-    "es": "flair/ner-spanish-large",
-    "de": "flair/ner-german-large",
-    "nl": "flair/ner-dutch-large",
-    "multi": "flair/ner-multi",
-    "multi-fast": "flair/ner-multi-fast",
-}
-
-SPACY_MODELS = {
-    "en": 'en_core_web_sm',
-}
 
 def load_language_model(language, type):
     from flair.models import SequenceTagger
     from spacy_download import load_spacy
 
     model = None
+
     match type:
         case "spacy":
-            model_name = SPACY_MODELS.get(language, SPACY_MODELS["en"])
+            model_name = SPACY_MODELS.get(language_to_alpha2(language), SPACY_MODELS["en"])
             model = load_spacy(model_name)
+
         case "flair":
             model_name = FLAIR_MODELS.get(language, "flair/ner-multi")
             model = SequenceTagger.load(model_name)
-    return model
+
+    return model

{streamlit_octostar_utils-2.11a4 → streamlit_octostar_utils-2.11a6}/streamlit_octostar_utils/nlp/ner.py

@@ -83,15 +83,6 @@ class FlairRecognizer(EntityRecognizer):
         ({"ORG"}, {"ORG", "ORGANIZATION"}),
     ]
 
-    MODEL_LANGUAGES = {
-        "en": "flair/ner-english-large",
-        "es": "flair/ner-spanish-large",
-        "de": "flair/ner-german-large",
-        "nl": "flair/ner-dutch-large",
-        "multi": "flair/ner-multi",
-        "multi-fast": "flair/ner-multi-fast",
-    }
-
     PRESIDIO_EQUIVALENCES = {
         "PER": "PERSON",
         "LOC": "LOC",
@@ -407,7 +398,7 @@ def analyze_dataframe_optimized(df: pd.DataFrame, analyzer: AnalyzerEngine, lang
 
 
 def compute_ner_presidio(
-        text,
+        text_or_df,
         language,
         analyzer,
         entities=None,
@@ -418,15 +409,15 @@ def compute_ner_presidio(
         batch_size=32,
         n_process=4
 ):
-    if isinstance(text, pd.DataFrame):
-        if len(text) >= 100:
-            return analyze_dataframe_optimized(text, analyzer, language, entities, score_threshold)
+    if isinstance(text_or_df, pd.DataFrame):
+        if len(text_or_df) >= 100:
+            return analyze_dataframe_optimized(text_or_df, analyzer, language, entities, score_threshold)
 
         else:
             texts = []
 
-            for col in text.columns:
-                for idx, value in text[col].dropna().items():
+            for col in text_or_df.columns:
+                for idx, value in text_or_df[col].dropna().items():
                     text_value = str(value).strip()
 
                     if text_value:
@@ -434,7 +425,8 @@ def compute_ner_presidio(
 
             text = "\n".join(texts)
 
-    elif isinstance(text, list):
+    elif isinstance(text_or_df, list):
+        text = text_or_df
         batch_analyzer = BatchAnalyzerEngine(analyzer_engine=analyzer)
 
         results_generator = batch_analyzer.analyze_iterator(
@@ -543,12 +535,13 @@ def get_extractive_summary(text, language, max_chars, fast=False, with_scores=Fa
 
 
 def ner_pipe(
-        text,
+        text_or_df,
         language,
         model,
         engine_type="spacy",
         fast=False,
         compression_ratio="auto",
+        with_scores=False,
         with_comentions=True,
         with_context=True,
         entities=None,
@@ -556,15 +549,18 @@ def ner_pipe(
         batch_size=32,
         n_process=4
 ):
+    if with_scores:
+        raise NotImplementedError("with_scores functionality is not implemented yet")
+
     analyzer = build_presidio_analyzer(
         language=language,
         engine_type=engine_type,
         model=model,
     )
 
-    if isinstance(text, pd.DataFrame):
+    if isinstance(text_or_df, pd.DataFrame):
         ner = compute_ner_presidio(
-            text,
+            text_or_df,
             language,
             analyzer,
             entities,
@@ -575,6 +571,8 @@ def ner_pipe(
             n_process=n_process
         )
     else:
+        text = text_or_df
+
         if compression_ratio == "auto":
             compression_ratio = max(1.0, len(text) / 15000) if fast else 1.0
 
@@ -613,13 +611,14 @@ def get_ner_handler(
     except LookupError:
         language = "en"
 
-    return lambda text, compression_ratio="auto", with_comentions=True, with_context=True: ner_pipe(
-        text,
+    return lambda text_or_df, compression_ratio="auto", with_scores=False, with_comentions=True, with_context=True: ner_pipe(
+        text_or_df,
         language,
         model,
         engine_type,
         fast,
         compression_ratio,
+        with_scores,
         with_comentions,
         with_context,
         entities,