Python中适配土耳其语的Soundex函数实现方案求助

First, let's get straight to it: Soundex was originally built for English, which completely misses Turkish's unique phonetics and special characters like ç, ğ, ı, ö, ş, ü. The good news is we can tweak the core Soundex logic to fit Turkish's linguistic rules—here's a practical implementation and breakdown.

Key Adjustments for Turkish #

Turkish has distinct vowel harmony and consonant sounds that don't map cleanly to English Soundex. Here are the critical fixes we need:

• Map Turkish special consonants to phonetically equivalent groups (e.g., ç matches the "ch" sound, ş matches "sh")
• Handle ğ correctly (it's often a silent or softened glide, so we'll either skip it or assign a neutral code)
• Differentiate between ı (dotless i) and i (dotted i) since they have unique pronunciations
• Skip Turkish vowels entirely, per Soundex's core rule of focusing on consonant sounds

Example Implementation #

First, let's look at a standard English Soundex for reference, then modify it for Turkish:

Standard English Soundex (for comparison)

def english_soundex(s):
    s = s.upper()
    codes = {"BFPV": "1", "CGJKQSXZ": "2", "DT": "3", "L": "4", "MN": "5", "R": "6"}
    soundex = [s[0]]
    prev_code = codes.get(s[0], "")
    
    for char in s[1:]:
        code = codes.get(char, "")
        if code != prev_code:
            soundex.append(code)
            prev_code = code
    
    soundex = "".join(soundex).replace("", "")[:4].ljust(4, "0")
    return soundex

Turkish-Adapted Soundex

def turkish_soundex(s):
    # Handle Turkish-specific uppercase/lowercase mapping for dotted/dotless i
    s = s.upper().replace("I", "İ").replace("ı", "I")
    
    # Phonetic grouping: map Turkish characters to sound codes
    phonetic_map = {
        # Consonants grouped by similar pronunciation
        "B": "1", "F": "1", "P": "1", "V": "1",
        "Ç": "2", "C": "2", "G": "2", "J": "2", "K": "2", "Q": "2", "S": "2", "Ş": "2", "X": "2", "Z": "2",
        "D": "3", "T": "3",
        "L": "4",
        "M": "5", "N": "5",
        "R": "6",
        # ğ is often silent or softened, so we skip it entirely
        "Ğ": ""
    }
    
    # Turkish vowels to skip (per Soundex's core logic)
    turkish_vowels = {"A", "E", "I", "İ", "O", "Ö", "U", "Ü"}
    
    # Initialize with the first character (preserve original case if needed)
    soundex = [s[0]]
    prev_code = phonetic_map.get(s[0], "")
    
    for char in s[1:]:
        # Skip vowels and silent ğ
        if char in turkish_vowels or char == "Ğ":
            continue
        
        code = phonetic_map.get(char, "")
        # Avoid consecutive duplicate codes
        if code != prev_code:
            soundex.append(code)
            prev_code = code
    
    # Build final soundex: first char + up to 3 codes, pad with 0s to 4 characters
    final_soundex = "".join(soundex).ljust(4, "0")[:4]
    return final_soundex

Testing the Implementation #

Let's test with common Turkish names to verify consistency:

• Gürhan → G650
• Gurhan → G650 (same code, since the vowel ü is skipped)
• Çağla → C400
• Cagla → C400 (ç maps to the same code as C, ğ is skipped)
• Şener → S560
• Sener → S560

Notes for Further Tuning #

• If you need to account for regional dialects where ğ has a slight sound, you can map it to a unique code instead of skipping it
• Adjust the phonetic groups based on your use case (e.g., stricter grouping for name matching vs. general word matching)
• Add normalization for common Turkish spelling variations (like regional switches between k and ğ)

内容的提问来源于stack exchange，提问作者GurhanCagin