@hupan56/wlkj 2.0.0

package/templates/qoder/scripts/search_index.py

@@ -0,0 +1,352 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+Search knowledge index with platform filtering and PRD search.
+
+Usage:
+  python search_index.py 保险 --platform web
+  python search_index.py 保险 --platform app
+  python search_index.py --prd 保险
+  python search_index.py --style table [--platform web]
+  python search_index.py --field nickName
+  python search_index.py --api salary
+  python search_index.py --components
+  python search_index.py --modules
+  python search_index.py --list
+  python search_index.py --vben
+"""
+import os, json, sys
+
+# UTF-8 stdio (防御性: stdout 被捕获时不崩溃)
+try:
+    sys.stdout.reconfigure(encoding='utf-8', errors='replace')
+except (AttributeError, TypeError, OSError, IOError):
+    try:
+        sys.stdout.reconfigure(encoding='utf-8')
+    except Exception:
+        pass
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+from common.terms import CN_MAP, BUSINESS_PATH_MAP, CN_TO_EN, get_platform_map
+PLATFORM_MAP = get_platform_map()
+
+BASE = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+INDEX_DIR = os.path.join(BASE, 'data', 'index')
+
+# Substring matching against keys shorter than this produces too much noise
+MIN_FUZZY_LEN = 4
+
+
+def load_index(filename, hint=''):
+    path = os.path.join(INDEX_DIR, filename)
+    if not os.path.isfile(path):
+        print('{} not found.{}'.format(filename, ' ' + hint if hint else ''))
+        return None
+    try:
+        with open(path, encoding='utf-8') as f:
+            return json.load(f)
+    except (json.JSONDecodeError, OSError) as e:
+        print('ERROR: failed to read {}: {}'.format(filename, e))
+        return None
+
+
+def filter_by_platform(files, platform=None):
+    if not platform:
+        return files
+    target = PLATFORM_MAP.get(platform.lower(), platform)
+    return [f for f in files if target.lower() in f.lower()]
+
+
+def _match_keyword(query_word, kw):
+    """Match a query word against an index keyword without short-key noise."""
+    w, k = query_word.lower(), kw.lower()
+    if w == k:
+        return True
+    if len(k) >= MIN_FUZZY_LEN and k in w:
+        return True
+    if len(w) >= MIN_FUZZY_LEN and w in k:
+        return True
+    return False
+
+
+def search_keywords(query, platform=None):
+    ki = load_index('keyword-index.json', hint='Run: python git_sync.py')
+    if ki is None:
+        return
+
+    # Expand Chinese query via the shared term maps (same maps used at build time)
+    words = []
+    if query in CN_MAP:
+        words.extend(CN_MAP[query].split())
+    if query in BUSINESS_PATH_MAP:
+        words.extend(p.strip('/-').lower() for p in BUSINESS_PATH_MAP[query])
+    if words:
+        words = sorted(set(w.lower() for w in words))
+        print('Chinese: {} -> English: {}'.format(query, ' '.join(words)))
+    else:
+        words = [query.lower()]
+    if platform:
+        print('Platform: {} -> {}'.format(platform, PLATFORM_MAP.get(platform.lower(), platform)))
+    print()
+
+    # 性能优化 A2: 用倒排索引缩小候选集 (从 O(all_keys) 降到 O(命中数))
+    try:
+        from common.search_engine import get_inverted_index, search_keywords_fast
+        from common.paths import get_repo_root
+        inverted = get_inverted_index(str(get_repo_root()))
+        kw_matches = search_keywords_fast(words, inverted, ki)
+    except Exception:
+        # 回退: 全扫描
+        kw_matches = {}
+        for word in words:
+            for kw, files in ki.items():
+                if _match_keyword(word, kw):
+                    kw_matches[kw] = files
+
+    # 按平台过滤 + 分组
+    matches = {}
+    for kw, files in kw_matches.items():
+        for f in filter_by_platform(files, platform):
+            p = f.split('/')[0] if '/' in f else f.split('\\')[0]
+            matches.setdefault(p, {}).setdefault(kw, [])
+            if f not in matches[p][kw]:
+                matches[p][kw].append(f)
+
+    if not matches:
+        print('No matches for: {}'.format(query))
+        return
+
+    total = 0
+    for proj, kws in sorted(matches.items()):
+        count = sum(len(fs) for fs in kws.values())
+        total += count
+        print('[{}] {} files'.format(proj, count))
+        for kw, fs in sorted(kws.items(), key=lambda x: len(x[1]), reverse=True)[:5]:
+            for f in fs[:3]:
+                parts = f.replace('\\', '/').split('/')
+                short = '/'.join(parts[:2]) + '/.../' + '/'.join(parts[-2:]) if len(parts) > 5 else '/'.join(parts)
+                print('  ' + short)
+        print()
+    print('Total: {} matches'.format(total))
+
+
+def search_prds(query):
+    prd_index = load_index('prd-index.json', hint='Run: python git_sync.py --prd-only')
+    if prd_index is None:
+        return
+    q = query.lower()
+
+    matches = []
+    for prd_file, prd in prd_index.items():
+        score = 0
+        title = prd.get('title', '').lower()
+        keywords = [kw.lower() for kw in prd.get('keywords', [])]
+        cn_terms = [t.lower() for t in prd.get('cn_terms', [])]
+
+        if q in title:
+            score += 10
+        if q in keywords or q in cn_terms:
+            score += 5
+        if q in CN_TO_EN and CN_TO_EN[q] in keywords:
+            score += 3
+        for kw in keywords:
+            if q in kw or kw in q:
+                score += 2
+
+        if score > 0:
+            matches.append((score, prd_file, prd))
+
+    if not matches:
+        print('No PRDs matching: {}'.format(query))
+        return
+
+    matches.sort(key=lambda x: x[0], reverse=True)
+
+    print('PRDs matching: {} ({} found)'.format(query, len(matches)))
+    print()
+    for score, prd_file, prd in matches[:10]:
+        title = prd.get('title', prd_file)[:60]
+        features = len(prd.get('features', []))
+        related = sum(len(fs) for fs in prd.get('related_code', {}).values())
+        print('  {} - {}'.format(prd_file, title))
+        print('    Features: {}, Related code: {} files'.format(features, related))
+        if prd.get('related_code'):
+            for proj, files in list(prd['related_code'].items())[:2]:
+                print('    -> {}: {} files'.format(proj, len(files)))
+        print()
+
+
+def search_style(style_type, platform=None):
+    """Find pages of a given UI type (table/form/detail/modal/dashboard)."""
+    si = load_index('style-index.json', hint='Run: python build_style_index.py')
+    if si is None:
+        return
+    q = style_type.lower()
+    projects = si.get('projects', {})
+    if platform:
+        target = PLATFORM_MAP.get(platform.lower(), platform)
+        projects = {p: v for p, v in projects.items() if p.lower() == target.lower()}
+
+    found_any = False
+    for proj, pdata in sorted(projects.items()):
+        page_types = pdata.get('page_types', {})
+        examples = pdata.get('page_examples', {})
+        hit_types = [pt for pt in page_types if q in pt.lower()]
+        if not hit_types:
+            continue
+        found_any = True
+        for pt in hit_types:
+            print('[{}] {}: {} pages'.format(proj, pt, page_types[pt]))
+            for f in examples.get(pt, [])[:8]:
+                print('  ' + f)
+            if not examples.get(pt):
+                print('  (no examples stored - run: python build_style_index.py)')
+            print()
+
+    if not found_any:
+        known = sorted({pt for v in si.get('projects', {}).values()
+                        for pt in v.get('page_types', {})})
+        print('No style matches for: {}'.format(style_type))
+        print('Known page types: {}'.format(', '.join(known)))
+
+
+def search_field(field_name):
+    """Search field usage in style-index.json field_map."""
+    si = load_index('style-index.json', hint='Run: python build_style_index.py')
+    if si is None:
+        return
+    fm = si.get('field_map', {})
+    matches = {f: i for f, i in fm.items() if field_name.lower() in f.lower()}
+    if not matches:
+        print('No fields matching: {}'.format(field_name))
+        return
+    print('Fields matching: {} ({} found)'.format(field_name, len(matches)))
+    for field, info in sorted(matches.items())[:10]:
+        titles = ', '.join(info.get('titles', []))
+        print('  {} -> {} (used {}x)'.format(field, titles, info.get('count', '?')))
+        for f in info.get('sample_files', [])[:3]:
+            print('    ' + f)
+
+
+def search_api(query):
+    """Search API endpoints in api-index.json."""
+    ai = load_index('api-index.json', hint='Run: python git_sync.py --index-only')
+    if ai is None:
+        return
+    q = query.lower()
+    matches = {ep: f for ep, f in ai.items() if q in ep.lower() or q in f.lower()}
+    if not matches:
+        print('No APIs matching: {}'.format(query))
+        return
+    print('APIs matching: {} ({} found)'.format(query, len(matches)))
+    for ep, f in sorted(matches.items())[:20]:
+        print('  {} -> {}'.format(ep, f))
+    if len(matches) > 20:
+        print('  ... and {} more'.format(len(matches) - 20))
+
+
+def show_modules():
+    """Show project overview from module-map.json."""
+    mm = load_index('module-map.json', hint='Run: python git_sync.py --index-only')
+    if mm is None:
+        return
+    print('=== Project Modules ===')
+    print()
+    for proj, info in sorted(mm.items()):
+        files = info.get('files', 0)
+        files = len(files) if isinstance(files, list) else files
+        print('[{}] files: {}, classes: {}, apis: {}'.format(
+            proj, files, info.get('classes', 0), info.get('apis', 0)))
+
+
+def show_top_keywords():
+    """Show top keywords by file count."""
+    ki = load_index('keyword-index.json', hint='Run: python git_sync.py')
+    if ki is None:
+        return
+    top = sorted(ki.items(), key=lambda x: len(x[1]), reverse=True)[:50]
+    print('=== Top 50 Keywords (by file count) ===')
+    print()
+    for kw, files in top:
+        print('  {} {}'.format(kw.ljust(30), len(files)))
+
+
+def show_components():
+    meta = load_index('style-meta.json', hint='Run: python build_style_index.py')
+    if meta is None:
+        return
+    print('=== UI Components Usage ===')
+    print()
+    print('Ant Design Vue (import counts):')
+    for comp, count in meta.get('common_components', {}).items():
+        print('  {} {}'.format(comp.ljust(20), str(count).ljust(5)))
+    print()
+    print('Form Components (usage in schemas):')
+    for comp, count in sorted(meta.get('form_components', {}).items(), key=lambda x: x[1], reverse=True)[:20]:
+        print('  {} {}'.format(comp.ljust(20), str(count).ljust(5)))
+
+
+def show_vben():
+    data = load_index('vben-style-reference.json')
+    if data is None:
+        return
+    print('=== Vben Admin Design Tokens ===')
+    print()
+    for section, values in data.items():
+        print('[{}]'.format(section))
+        if isinstance(values, dict):
+            for key, val in values.items():
+                print('  {}: {}'.format(key, val))
+        print()
+
+
+def print_usage():
+    print('Usage:')
+    print('  search_index.py <keyword> [--platform web|app]  - Search code')
+    print('  search_index.py --prd <keyword>                - Search PRDs')
+    print('  search_index.py --style <type> [--platform p]  - UI patterns (table/form/detail/modal/dashboard)')
+    print('  search_index.py --field <name>                 - Field usage')
+    print('  search_index.py --api <keyword>                - API endpoints')
+    print('  search_index.py --components                   - UI components')
+    print('  search_index.py --modules                      - Project overview')
+    print('  search_index.py --list                         - Top 50 keywords')
+    print('  search_index.py --vben                         - Vben design tokens')
+    print()
+    print('Platform: web/pc/管理端 -> fywl-ui, app/mobile/h5/移动端 -> Carmg-H5')
+
+
+if __name__ == '__main__':
+    platform = None
+    args = sys.argv[1:]
+    if '--platform' in args:
+        idx = args.index('--platform')
+        if idx + 1 < len(args):
+            platform = args[idx + 1]
+            args = args[:idx] + args[idx+2:]
+        else:
+            args = args[:idx]
+
+    if len(args) < 1:
+        print_usage()
+    elif args[0] == '--prd' and len(args) >= 2:
+        search_prds(args[1])
+    elif args[0] == '--style' and len(args) >= 2:
+        search_style(args[1], platform)
+    elif args[0] == '--field' and len(args) >= 2:
+        search_field(args[1])
+    elif args[0] == '--api' and len(args) >= 2:
+        search_api(args[1])
+    elif args[0] == '--components':
+        show_components()
+    elif args[0] == '--modules':
+        show_modules()
+    elif args[0] == '--list':
+        show_top_keywords()
+    elif args[0] == '--vben':
+        show_vben()
+    elif args[0].startswith('--'):
+        print('Unknown option: {}'.format(args[0]))
+        print()
+        print_usage()
+    else:
+        search_keywords(args[0], platform)