Spaces:
Running
Running
Update parser/assembler.py
Browse files- parser/assembler.py +25 -52
parser/assembler.py
CHANGED
|
@@ -1,65 +1,34 @@
|
|
| 1 |
# parser/assembler.py
|
| 2 |
-
|
| 3 |
-
from typing import List, Dict, Optional, Tuple
|
| 4 |
from helpers.utils import normalize_digits, extract_article_number, is_article, is_section
|
| 5 |
|
|
|
|
| 6 |
def extract_title_and_preamble(
|
| 7 |
blocks: List[Dict[str, str]],
|
| 8 |
-
*,
|
| 9 |
-
default_title: Optional[str] = None,
|
| 10 |
-
default_preamble: Optional[str] = None
|
| 11 |
) -> Tuple[str, str, List[Dict[str, str]]]:
|
| 12 |
"""
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
إذا وُجد type يُستخدم مباشرة، وإلا تُستخدم الطريقة التقليدية.
|
| 16 |
-
يمكن أيضاً تمرير عنوان أو مقدمة افتراضية لتُستخدم بدلاً من الاستخراج التلقائي.
|
| 17 |
"""
|
| 18 |
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
body_blocks = [b for b in blocks if b.get("type") != "title" and b.get("type") != "preamble"]
|
| 22 |
-
return default_title, default_preamble, body_blocks
|
| 23 |
-
|
| 24 |
-
# هل القائمة تحتوي على حقل type مُحدد مسبقاً؟
|
| 25 |
-
if any(b.get("type") in {"title", "preamble", "body"} for b in blocks):
|
| 26 |
-
title_blocks = [b for b in blocks if b.get("type") == "title"]
|
| 27 |
-
preamble_blocks = [b for b in blocks if b.get("type") == "preamble"]
|
| 28 |
-
body_blocks = [b for b in blocks if b.get("type") == "body"]
|
| 29 |
-
|
| 30 |
-
title = "\n".join([b["text"].strip() for b in title_blocks]).strip()
|
| 31 |
-
preamble = "\n".join([b["text"].strip() for b in preamble_blocks]).strip()
|
| 32 |
-
return title, preamble, body_blocks
|
| 33 |
-
|
| 34 |
-
# الطريقة التقليدية (المنطق القديم دون تغيير)
|
| 35 |
-
title = ""
|
| 36 |
-
preamble_blocks = []
|
| 37 |
-
|
| 38 |
-
# استخراج العنوان
|
| 39 |
-
while blocks:
|
| 40 |
-
block = blocks.pop(0)
|
| 41 |
-
t = block.get("text", "").strip()
|
| 42 |
-
if t.lower() != "html" and t:
|
| 43 |
-
title = t
|
| 44 |
-
break
|
| 45 |
-
|
| 46 |
-
# استخراج المقدمة حتى أول قسم أو مادة
|
| 47 |
-
while blocks:
|
| 48 |
-
t = blocks[0].get("text", "").strip()
|
| 49 |
-
if is_section(t) or is_article(t):
|
| 50 |
-
break
|
| 51 |
-
preamble_blocks.append(blocks.pop(0))
|
| 52 |
|
|
|
|
| 53 |
preamble = "\n".join([b["text"] for b in preamble_blocks]).strip()
|
| 54 |
-
|
|
|
|
| 55 |
|
| 56 |
|
| 57 |
def extract_articles_from_blocks(blocks: List[Dict[str, str]]) -> List[Dict]:
|
|
|
|
|
|
|
|
|
|
| 58 |
articles = []
|
| 59 |
current = None
|
| 60 |
|
| 61 |
for block in blocks:
|
| 62 |
-
t = block
|
| 63 |
if is_article(t):
|
| 64 |
if current:
|
| 65 |
current["text"] = current["text"].strip()
|
|
@@ -79,11 +48,14 @@ def extract_articles_from_blocks(blocks: List[Dict[str, str]]) -> List[Dict]:
|
|
| 79 |
|
| 80 |
|
| 81 |
def extract_sections(blocks: List[Dict[str, str]]) -> List[Dict]:
|
|
|
|
|
|
|
|
|
|
| 82 |
sections = []
|
| 83 |
current = {"name": "", "texts": []}
|
| 84 |
|
| 85 |
for block in blocks:
|
| 86 |
-
t = normalize_digits(block
|
| 87 |
|
| 88 |
if is_section(t):
|
| 89 |
if current["texts"] or current["name"]:
|
|
@@ -99,16 +71,17 @@ def extract_sections(blocks: List[Dict[str, str]]) -> List[Dict]:
|
|
| 99 |
|
| 100 |
|
| 101 |
def parse_law_from_texts(text_blocks: List[Dict[str, str]]) -> Dict:
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
|
|
|
| 105 |
sections_raw = extract_sections(remaining_blocks)
|
| 106 |
|
| 107 |
sections = []
|
| 108 |
for sec in sections_raw:
|
| 109 |
raw_blocks = sec["texts"]
|
| 110 |
|
| 111 |
-
content = "\n".join([b["text"] for b in raw_blocks if not is_article(b
|
| 112 |
articles = extract_articles_from_blocks(raw_blocks)
|
| 113 |
|
| 114 |
sections.append({
|
|
@@ -121,10 +94,10 @@ def parse_law_from_texts(text_blocks: List[Dict[str, str]]) -> Dict:
|
|
| 121 |
})
|
| 122 |
|
| 123 |
return {
|
| 124 |
-
"message": "success",
|
| 125 |
"all_blocks": {
|
| 126 |
"count": len(text_blocks),
|
| 127 |
-
#"items": text_blocks
|
| 128 |
},
|
| 129 |
"law": {
|
| 130 |
"title": title,
|
|
|
|
| 1 |
# parser/assembler.py
|
| 2 |
+
from typing import List, Dict, Tuple
|
|
|
|
| 3 |
from helpers.utils import normalize_digits, extract_article_number, is_article, is_section
|
| 4 |
|
| 5 |
+
|
| 6 |
def extract_title_and_preamble(
|
| 7 |
blocks: List[Dict[str, str]],
|
|
|
|
|
|
|
|
|
|
| 8 |
) -> Tuple[str, str, List[Dict[str, str]]]:
|
| 9 |
"""
|
| 10 |
+
استخراج العنوان والمقدمة وبقية النصوص (body) من الكتل.
|
| 11 |
+
كل كتلة تحتوي على {"text": ..., "type": ...}، جميع العناصر مستخدمة.
|
|
|
|
|
|
|
| 12 |
"""
|
| 13 |
+
title_blocks = [b for b in blocks if b["type"] == "title"]
|
| 14 |
+
preamble_blocks = [b for b in blocks if b["type"] == "preamble"]
|
| 15 |
+
body_blocks = [b for b in blocks if b["type"] == "body"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
|
| 17 |
+
title = "\n".join([b["text"] for b in title_blocks]).strip()
|
| 18 |
preamble = "\n".join([b["text"] for b in preamble_blocks]).strip()
|
| 19 |
+
|
| 20 |
+
return title, preamble, body_blocks
|
| 21 |
|
| 22 |
|
| 23 |
def extract_articles_from_blocks(blocks: List[Dict[str, str]]) -> List[Dict]:
|
| 24 |
+
"""
|
| 25 |
+
تقسيم النصوص إلى مقالات حسب is_article.
|
| 26 |
+
"""
|
| 27 |
articles = []
|
| 28 |
current = None
|
| 29 |
|
| 30 |
for block in blocks:
|
| 31 |
+
t = block["text"].strip()
|
| 32 |
if is_article(t):
|
| 33 |
if current:
|
| 34 |
current["text"] = current["text"].strip()
|
|
|
|
| 48 |
|
| 49 |
|
| 50 |
def extract_sections(blocks: List[Dict[str, str]]) -> List[Dict]:
|
| 51 |
+
"""
|
| 52 |
+
تقسيم النصوص إلى أقسام حسب is_section.
|
| 53 |
+
"""
|
| 54 |
sections = []
|
| 55 |
current = {"name": "", "texts": []}
|
| 56 |
|
| 57 |
for block in blocks:
|
| 58 |
+
t = normalize_digits(block["text"].strip())
|
| 59 |
|
| 60 |
if is_section(t):
|
| 61 |
if current["texts"] or current["name"]:
|
|
|
|
| 71 |
|
| 72 |
|
| 73 |
def parse_law_from_texts(text_blocks: List[Dict[str, str]]) -> Dict:
|
| 74 |
+
"""
|
| 75 |
+
تحويل قائمة النصوص المفهرسة إلى هيكل القانون النهائي مع أقسامه ومقالاته.
|
| 76 |
+
"""
|
| 77 |
+
title, preamble, remaining_blocks = extract_title_and_preamble(text_blocks)
|
| 78 |
sections_raw = extract_sections(remaining_blocks)
|
| 79 |
|
| 80 |
sections = []
|
| 81 |
for sec in sections_raw:
|
| 82 |
raw_blocks = sec["texts"]
|
| 83 |
|
| 84 |
+
content = "\n".join([b["text"] for b in raw_blocks if not is_article(b["text"])]).strip()
|
| 85 |
articles = extract_articles_from_blocks(raw_blocks)
|
| 86 |
|
| 87 |
sections.append({
|
|
|
|
| 94 |
})
|
| 95 |
|
| 96 |
return {
|
| 97 |
+
"message": "success",
|
| 98 |
"all_blocks": {
|
| 99 |
"count": len(text_blocks),
|
| 100 |
+
#"items": text_blocks # يمكن إلغاء التعليق إذا أردت كل العناصر
|
| 101 |
},
|
| 102 |
"law": {
|
| 103 |
"title": title,
|