Spaces:
Running
Running
Update parser/assembler.py
Browse files- parser/assembler.py +39 -36
parser/assembler.py
CHANGED
|
@@ -2,49 +2,56 @@ from typing import List, Dict
|
|
| 2 |
from helpers.utils import normalize_digits, extract_article_number, is_article, is_section
|
| 3 |
|
| 4 |
|
| 5 |
-
def extract_title_and_preamble(
|
| 6 |
"""
|
| 7 |
استخراج عنوان القانون والمقدمة، مع الاحتفاظ بالنصوص غير المصنفة.
|
| 8 |
"""
|
| 9 |
title = ""
|
| 10 |
-
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
if t.lower() != "html" and t != "":
|
| 13 |
title = t
|
| 14 |
break
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
for t in texts:
|
| 20 |
if is_section(t) or is_article(t):
|
| 21 |
-
|
| 22 |
else:
|
| 23 |
-
|
| 24 |
|
| 25 |
-
|
| 26 |
-
|
|
|
|
| 27 |
|
| 28 |
|
| 29 |
-
def
|
| 30 |
"""
|
| 31 |
استخراج المواد مع الاحتفاظ بالنصوص التابعة لها.
|
| 32 |
"""
|
| 33 |
articles = []
|
| 34 |
current = None
|
| 35 |
|
| 36 |
-
for
|
| 37 |
-
t =
|
| 38 |
if is_article(t):
|
| 39 |
if current:
|
| 40 |
current["text"] = current["text"].strip()
|
| 41 |
articles.append(current)
|
| 42 |
-
current = {"number": extract_article_number(t), "text": t}
|
| 43 |
else:
|
| 44 |
if current:
|
| 45 |
current["text"] += "\n" + t
|
|
|
|
| 46 |
else:
|
| 47 |
-
|
|
|
|
| 48 |
|
| 49 |
if current:
|
| 50 |
current["text"] = current["text"].strip()
|
|
@@ -53,30 +60,25 @@ def extract_articles_from_texts(texts: List[str]) -> List[Dict]:
|
|
| 53 |
return articles
|
| 54 |
|
| 55 |
|
| 56 |
-
def extract_sections(
|
| 57 |
"""
|
| 58 |
تقسيم النصوص إلى أقسام وفصول باستخدام الأنماط الذكية.
|
| 59 |
"""
|
| 60 |
sections = []
|
| 61 |
current = {"name": "", "texts": []}
|
| 62 |
|
| 63 |
-
for
|
| 64 |
-
|
| 65 |
-
continue
|
| 66 |
-
|
| 67 |
-
t_norm = normalize_digits(t.strip())
|
| 68 |
|
| 69 |
-
|
| 70 |
-
if is_section(t_norm):
|
| 71 |
# لو يوجد قسم سابق → نحفظه
|
| 72 |
if current["texts"] or current["name"]:
|
| 73 |
sections.append(current)
|
| 74 |
|
| 75 |
# نبدأ قسم/فصل جديد
|
| 76 |
-
current = {"name":
|
| 77 |
else:
|
| 78 |
-
|
| 79 |
-
current["texts"].append(t_norm)
|
| 80 |
|
| 81 |
# إضافة آخر قسم
|
| 82 |
if current["texts"] or current["name"]:
|
|
@@ -90,29 +92,29 @@ def parse_law_from_texts(text_blocks: List[Dict[str, str]]) -> Dict:
|
|
| 90 |
تحويل النصوص إلى هيكل قانوني منظم مع ضمان عدم فقد أي نص.
|
| 91 |
"""
|
| 92 |
# النصوص الخام (منظفة مسبقًا في text_blocks.py)
|
| 93 |
-
|
| 94 |
|
| 95 |
# استخراج العنوان والمقدمة
|
| 96 |
-
title, preamble,
|
| 97 |
|
| 98 |
# استخراج الأقسام
|
| 99 |
-
sections_raw = extract_sections(
|
| 100 |
|
| 101 |
sections = []
|
| 102 |
for sec in sections_raw:
|
| 103 |
-
|
| 104 |
|
| 105 |
-
# المحتوى غير المواد
|
| 106 |
-
content = "\n".join(
|
| 107 |
|
| 108 |
# استخراج المواد
|
| 109 |
-
articles =
|
| 110 |
|
| 111 |
sections.append({
|
| 112 |
"title": sec["name"],
|
| 113 |
"content": content,
|
| 114 |
"articles": [
|
| 115 |
-
{"number": a["number"], "text": a["text"]}
|
| 116 |
for a in articles
|
| 117 |
]
|
| 118 |
})
|
|
@@ -123,6 +125,7 @@ def parse_law_from_texts(text_blocks: List[Dict[str, str]]) -> Dict:
|
|
| 123 |
"law": {
|
| 124 |
"title": title,
|
| 125 |
"preamble": preamble,
|
| 126 |
-
"sections": sections
|
|
|
|
| 127 |
}
|
| 128 |
}
|
|
|
|
| 2 |
from helpers.utils import normalize_digits, extract_article_number, is_article, is_section
|
| 3 |
|
| 4 |
|
| 5 |
+
def extract_title_and_preamble(text_blocks: List[Dict[str, str]]) -> (str, str, List[Dict[str, str]]):
|
| 6 |
"""
|
| 7 |
استخراج عنوان القانون والمقدمة، مع الاحتفاظ بالنصوص غير المصنفة.
|
| 8 |
"""
|
| 9 |
title = ""
|
| 10 |
+
preamble_blocks = []
|
| 11 |
+
remaining_blocks = []
|
| 12 |
+
|
| 13 |
+
# استخراج العنوان
|
| 14 |
+
while text_blocks:
|
| 15 |
+
block = text_blocks.pop(0)
|
| 16 |
+
t = block.get("text", "").strip()
|
| 17 |
if t.lower() != "html" and t != "":
|
| 18 |
title = t
|
| 19 |
break
|
| 20 |
|
| 21 |
+
# استخراج المقدمة وبقية النصوص
|
| 22 |
+
for block in text_blocks:
|
| 23 |
+
t = block.get("text", "").strip()
|
|
|
|
| 24 |
if is_section(t) or is_article(t):
|
| 25 |
+
remaining_blocks.append(block)
|
| 26 |
else:
|
| 27 |
+
preamble_blocks.append(block)
|
| 28 |
|
| 29 |
+
# المقدمة تبقى كـ نصوص منفصلة لكن نعرضها مجمعة عند الإخراج
|
| 30 |
+
preamble = "\n".join([b["text"] for b in preamble_blocks]).strip()
|
| 31 |
+
return title, preamble, remaining_blocks
|
| 32 |
|
| 33 |
|
| 34 |
+
def extract_articles_from_blocks(blocks: List[Dict[str, str]]) -> List[Dict]:
|
| 35 |
"""
|
| 36 |
استخراج المواد مع الاحتفاظ بالنصوص التابعة لها.
|
| 37 |
"""
|
| 38 |
articles = []
|
| 39 |
current = None
|
| 40 |
|
| 41 |
+
for block in blocks:
|
| 42 |
+
t = block.get("text", "").strip()
|
| 43 |
if is_article(t):
|
| 44 |
if current:
|
| 45 |
current["text"] = current["text"].strip()
|
| 46 |
articles.append(current)
|
| 47 |
+
current = {"number": extract_article_number(t), "text": t, "blocks": [block]}
|
| 48 |
else:
|
| 49 |
if current:
|
| 50 |
current["text"] += "\n" + t
|
| 51 |
+
current["blocks"].append(block)
|
| 52 |
else:
|
| 53 |
+
# نص غير مصنف → نحفظه كمادة بدون رقم
|
| 54 |
+
current = {"number": None, "text": t, "blocks": [block]}
|
| 55 |
|
| 56 |
if current:
|
| 57 |
current["text"] = current["text"].strip()
|
|
|
|
| 60 |
return articles
|
| 61 |
|
| 62 |
|
| 63 |
+
def extract_sections(blocks: List[Dict[str, str]]) -> List[Dict]:
|
| 64 |
"""
|
| 65 |
تقسيم النصوص إلى أقسام وفصول باستخدام الأنماط الذكية.
|
| 66 |
"""
|
| 67 |
sections = []
|
| 68 |
current = {"name": "", "texts": []}
|
| 69 |
|
| 70 |
+
for block in blocks:
|
| 71 |
+
t = normalize_digits(block.get("text", "").strip())
|
|
|
|
|
|
|
|
|
|
| 72 |
|
| 73 |
+
if is_section(t):
|
|
|
|
| 74 |
# لو يوجد قسم سابق → نحفظه
|
| 75 |
if current["texts"] or current["name"]:
|
| 76 |
sections.append(current)
|
| 77 |
|
| 78 |
# نبدأ قسم/فصل جديد
|
| 79 |
+
current = {"name": t, "texts": []}
|
| 80 |
else:
|
| 81 |
+
current["texts"].append(block)
|
|
|
|
| 82 |
|
| 83 |
# إضافة آخر قسم
|
| 84 |
if current["texts"] or current["name"]:
|
|
|
|
| 92 |
تحويل النصوص إلى هيكل قانوني منظم مع ضمان عدم فقد أي نص.
|
| 93 |
"""
|
| 94 |
# النصوص الخام (منظفة مسبقًا في text_blocks.py)
|
| 95 |
+
raw_blocks = [b for b in text_blocks if b.get("text")]
|
| 96 |
|
| 97 |
# استخراج العنوان والمقدمة
|
| 98 |
+
title, preamble, remaining_blocks = extract_title_and_preamble(raw_blocks)
|
| 99 |
|
| 100 |
# استخراج الأقسام
|
| 101 |
+
sections_raw = extract_sections(remaining_blocks)
|
| 102 |
|
| 103 |
sections = []
|
| 104 |
for sec in sections_raw:
|
| 105 |
+
raw_blocks = sec["texts"]
|
| 106 |
|
| 107 |
+
# المحتوى غير المواد (��بقى كـ نصوص منفصلة لكن نعرضه مجمّعًا)
|
| 108 |
+
content = "\n".join([b["text"] for b in raw_blocks if not is_article(b.get("text", ""))]).strip()
|
| 109 |
|
| 110 |
# استخراج المواد
|
| 111 |
+
articles = extract_articles_from_blocks(raw_blocks)
|
| 112 |
|
| 113 |
sections.append({
|
| 114 |
"title": sec["name"],
|
| 115 |
"content": content,
|
| 116 |
"articles": [
|
| 117 |
+
{"number": a["number"], "text": a["text"], "blocks": a["blocks"]}
|
| 118 |
for a in articles
|
| 119 |
]
|
| 120 |
})
|
|
|
|
| 125 |
"law": {
|
| 126 |
"title": title,
|
| 127 |
"preamble": preamble,
|
| 128 |
+
"sections": sections,
|
| 129 |
+
"all_blocks": text_blocks # ضمان أن كل النصوص محفوظة
|
| 130 |
}
|
| 131 |
}
|