From fff42533524d3ee91d298a2873ae19957304be2e Mon Sep 17 00:00:00 2001
From: vahidrezanezhad <vahid631983@gmail.com>
Date: Tue, 3 Feb 2026 20:20:20 +0100
Subject: [PATCH] generate or update list of characters in the case of cnn-rnn
 ocr training

---
 src/eynollah/training/cli.py                  |  2 +
 ...te_or_update_cnn_rnn_ocr_character_list.py | 59 +++++++++++++++++++
 2 files changed, 61 insertions(+)
 create mode 100644 src/eynollah/training/generate_or_update_cnn_rnn_ocr_character_list.py

diff --git a/src/eynollah/training/cli.py b/src/eynollah/training/cli.py
index 3718275..862d212 100644
--- a/src/eynollah/training/cli.py
+++ b/src/eynollah/training/cli.py
@@ -10,6 +10,7 @@ from .inference import main as inference_cli
 from .train import ex
 from .extract_line_gt import linegt_cli
 from .weights_ensembling import main as ensemble_cli
+from .generate_or_update_cnn_rnn_ocr_character_list import main as update_ocr_characters_cli
 
 @click.command(context_settings=dict(
         ignore_unknown_options=True,
@@ -28,3 +29,4 @@ main.add_command(inference_cli, 'inference')
 main.add_command(train_cli, 'train')
 main.add_command(linegt_cli, 'export_textline_images_and_text')
 main.add_command(ensemble_cli, 'ensembling')
+main.add_command(update_ocr_characters_cli, 'generate_or_update_cnn_rnn_ocr_character_list')
diff --git a/src/eynollah/training/generate_or_update_cnn_rnn_ocr_character_list.py b/src/eynollah/training/generate_or_update_cnn_rnn_ocr_character_list.py
new file mode 100644
index 0000000..6d1028c
--- /dev/null
+++ b/src/eynollah/training/generate_or_update_cnn_rnn_ocr_character_list.py
@@ -0,0 +1,59 @@
+import os
+import numpy as np
+import json
+import click
+import logging
+
+
+
+def run_character_list_update(dir_labels, out, current_character_list):
+    ls_labels = os.listdir(dir_labels)
+    ls_labels = [ind for ind in ls_labels if ind.endswith('.txt')]
+    
+    if current_character_list:
+        with open(current_character_list, 'r') as f_name:
+            characters = json.load(f_name)
+            
+        characters = set(characters)
+    else:
+        characters = set()
+
+
+    for ind in ls_labels:
+        label  = open(os.path.join(dir_labels,ind),'r').read().split('\n')[0]
+            
+        for char in label:
+            characters.add(char)
+                
+                
+    characters = sorted(list(set(characters)))
+
+    with open(out, 'w') as f_name:
+        json.dump(characters, f_name)
+        
+
+@click.command()
+@click.option(
+    "--dir_labels",
+    "-dl",
+    help="directory of labels which are txt files",
+    type=click.Path(exists=True, file_okay=False),
+    required=True,
+)
+@click.option(
+    "--current_character_list",
+    "-ccl",
+    help="current exsiting character list which is txt file and wished to be updated with a set of labels",
+    type=click.Path(exists=True, file_okay=True),
+    required=False,
+)
+@click.option(
+    "--out",
+    "-o",
+    help="output file which is a txt file where generated or updated character list will be written",
+    type=click.Path(exists=False, file_okay=True),
+)
+
+def main(dir_labels, out, current_character_list):
+    run_character_list_update(dir_labels, out, current_character_list)
+