Satrlar: matn bilan professional ishlash

a = 'bitta qo\'shtirnoq'
b = "qo'sh tirnoq — ichida ' bemalol"
c = """Uch qo'shtirnoq —
bir necha qatorli matn uchun.
Prompt shablonlari uchun juda qulay."""
d = ""                      # bo'sh satr

print("C:\new\table")     # \n va \t yangi qator va tab bo'lib qoladi!
print(r"C:\new\table")    # C:\new\table — aynan shunday

>>> s = "salom\t\n"
>>> print(s)       # ko'rinmaydi
salom
>>> repr(s)
"'salom\\t\\n'"
>>> print(repr(s))
'salom\t\n'

 T   o   s   h   k   e   n   t
 0   1   2   3   4   5   6   7     ← musbat indeks
-8  -7  -6  -5  -4  -3  -2  -1     ← manfiy indeks

s = "Toshkent"
s[0]       # 'T'
s[-1]      # 't' — oxirgi belgi
s[10]      # IndexError: string index out of range

s = "Toshkent"
s[1:4]     # 'osh'   — 1 dan 4 gacha (4 kirmaydi!)
s[:3]      # 'Tos'   — boshidan
s[3:]      # 'hkent' — oxirigacha
s[:]       # 'Toshkent' — to'liq nusxa
s[::2]     # 'Tskn'  — har ikkinchi belgi
s[::-1]    # 'tnekhsoT' — teskari
s[-3:]     # 'ent'   — oxirgi 3 ta
s[2:100]   # 'shkent' — chegaradan chiqsa xato yo'q, shunchaki kesadi

s = "salom"
s[0] = "S"            # TypeError: 'str' object does not support item assignment
s = "S" + s[1:]       # to'g'ri yo'l — yangi satr yaratish

"  salom  \n".strip()        # 'salom' — ikki chetdagi bo'sh joy, \n, \t
"  salom  ".lstrip()         # 'salom  '
"xxsalomxx".strip("x")       # 'salom'
"report.pdf".removesuffix(".pdf")    # 'report'
"Mr. Aziz".removeprefix("Mr. ")      # 'Aziz'

"a,b,,c".split(",")          # ['a', 'b', '', 'c'] — bo'sh element saqlanadi
"a b   c\n".split()          # ['a', 'b', 'c'] — argumentsiz: har qanday bo'sh joy bo'yicha
"k=v=w".split("=", 1)        # ['k', 'v=w'] — faqat birinchi marta
"satr1\nsatr2\r\nsatr3".splitlines()   # ['satr1', 'satr2', 'satr3']
"kalit=qiymat".partition("=")          # ('kalit', '=', 'qiymat')
 
", ".join(["olma", "nok", "uzum"])     # 'olma, nok, uzum'
"\n".join(qatorlar)                    # qatorlarni birlashtirish

"python" in "I love python"      # True — eng o'qiladigan usul
"banana".find("n")               # 2 — birinchi uchragan indeks
"banana".find("x")               # -1 — topilmadi
"banana".index("x")              # ValueError — topilmasa xato
"banana".count("a")              # 3
"report.pdf".endswith((".pdf", ".docx"))   # True — tuple bilan bir nechta variant
"Toshkent".startswith("Tosh")    # True
 
"123".isdigit()      # True
"12.5".isdigit()     # False — nuqta raqam emas
"abc".isalpha()      # True
"   ".isspace()      # True

"salom dunyo".title()        # 'Salom Dunyo'
"Salom".lower()              # 'salom'
"salom".upper()              # 'SALOM'
"salom".capitalize()         # 'Salom'
"banana".replace("a", "o")   # 'bonono'
"banana".replace("a", "o", 1)  # 'bonana' — faqat birinchisi

"Straße".casefold() == "STRASSE".casefold()   # True

"3".zfill(3)           # '003'
"x".center(7, "*")     # '***x***'
"ab" * 3               # 'ababab'

ism = "Aziz"
yosh = 25
print(f"{ism} {yosh} yoshda. Kelasi yil {yosh + 1} bo'ladi.")

x = 3.14159
n = 1234567.891
 
f"{x:.2f}"          # '3.14'          — 2 xona kasr
f"{n:,.2f}"         # '1,234,567.89'  — minglik ajratuvchi
f"{n:_.0f}"         # '1_234_568'
f"{42:05d}"         # '00042'         — 5 xonali, nol bilan to'ldirilgan
f"{0.256:.1%}"      # '25.6%'         — foiz
f"{1e-5:.2e}"       # '1.00e-05'      — ilmiy yozuv
f"{'chap':<8}|"     # 'chap    |'     — chapga tekislash, 8 belgi
f"{'ong':>8}|"      # '     ong|'     — o'ngga
f"{'markaz':^10}|"  # '  markaz  |'   — markazga
f"{255:x}"          # 'ff'            — 16-lik
f"{255:b}"          # '11111111'      — ikkilik

natijalar = [("logreg", 0.8731, 1.2), ("random_forest", 0.9102, 14.8), ("xgboost", 0.9254, 9.1)]
 
print(f"{'Model':<15}{'Accuracy':>10}{'Vaqt, s':>10}")
print("-" * 35)
for nom, acc, vaqt in natijalar:
    print(f"{nom:<15}{acc:>10.2%}{vaqt:>10.1f}")

Model            Accuracy   Vaqt, s
-----------------------------------
logreg             87.31%       1.2
random_forest      91.02%      14.8
xgboost            92.54%       9.1

x = 3.14159
print(f"{x=}")              # x=3.14159
print(f"{x * 2 = :.3f}")    # x * 2 = 6.283

ism = "Ali "
print(f"Ism: {ism}.")       # Ism: Ali .   — oxiridagi bo'sh joy sezilmaydi
print(f"Ism: {ism!r}.")     # Ism: 'Ali '. — darhol ko'rinadi

"{} va {}".format("A", "B")       # str.format()
"%s yoshi %d" % ("Ali", 30)       # % formatlash (C uslubi)

s = "😀"
len(s)                    # 1 — bitta belgi
b = s.encode("utf-8")     # b'\xf0\x9f\x98\x80'
len(b)                    # 4 — to'rt bayt
b.decode("utf-8")         # '😀'
 
"Ш".encode("utf-8")       # b'\xd0\xa8' — kirill harfi 2 bayt
ord("A"), chr(1071)       # (65, 'Я') — belgi ↔ kod nuqtasi

APOSTROFLAR = str.maketrans({"ʻ": "'", "‘": "'", "’": "'", "`": "'", "ʼ": "'"})
 
def normalize_uz(matn: str) -> str:
    return matn.translate(APOSTROFLAR)
 
normalize_uz("Oʻzbekiston gʻalabasi")   # "O'zbekiston g'alabasi"

with open("data.txt", encoding="utf-8") as f:   # encoding'ni HAR DOIM yozing
    text = f.read()

"abc" < "abd"     # True — leksikografik (lug'at tartibi)
"Z" < "a"         # True! — katta harflarning kodi kichikroq (Z=90, a=97)
"10" < "9"        # True! — sonlar emas, belgilar solishtiriladi

PROMPT_TEMPLATE = """\
Quyidagi hujjatlarga asoslanib savolga {til} tilida javob ber.
Agar javob hujjatlarda bo'lmasa, "Bilmayman" de.
 
{hujjatlar}
 
Savol: {savol}
"""
 
 
def build_prompt(savol: str, kontekst: list[str], til: str = "o'zbek") -> str:
    hujjatlar = "\n\n".join(
        f'<hujjat id="{i}">\n{matn.strip()}\n</hujjat>' for i, matn in enumerate(kontekst, 1)
    )
    return PROMPT_TEMPLATE.format(til=til, hujjatlar=hujjatlar, savol=savol.strip())
 
 
print(build_prompt("Poytaxt qaysi?", ["O'zbekiston poytaxti — Toshkent.", "Aholisi 37 mln."]))

Quyidagi hujjatlarga asoslanib savolga o'zbek tilida javob ber.
Agar javob hujjatlarda bo'lmasa, "Bilmayman" de.
 
<hujjat id="1">
O'zbekiston poytaxti — Toshkent.
</hujjat>
 
<hujjat id="2">
Aholisi 37 mln.
</hujjat>
 
Savol: Poytaxt qaysi?

import re
import unicodedata
 
APOSTROFLAR = str.maketrans({"ʻ": "'", "‘": "'", "’": "'", "`": "'", "ʼ": "'"})
 
 
def clean_text(text: str) -> str:
    text = unicodedata.normalize("NFC", text)       # Unicode'ni bitta shaklga
    text = text.translate(APOSTROFLAR)              # apostroflarni birxillashtirish
    text = text.replace("\u00a0", " ")              # bo'linmaydigan bo'sh joy
    text = re.sub(r"[ \t]+", " ", text)             # ketma-ket bo'sh joylar → bitta
    text = re.sub(r"\n{3,}", "\n\n", text)          # 3+ bo'sh qator → 2
    return text.strip()
 
 
print(repr(clean_text("  Oʻzbekiston\u00a0\u00a0 Respublikasi\n\n\n\nPoytaxti:   Toshkent ")))
# "O'zbekiston Respublikasi\n\nPoytaxti: Toshkent"