crm-backend-matt/.scratch/customer-rework/bruno_scan_d.py

"""bruno-sync Step 2/3 (本机 D:\ 路径版, 由 bruno_scan.py 改路径而来): scan Spring controllers -> reconcile with docs repo.

Outputs JSON: {apis: [...], report: {...}}
"""
import json
import os
import re
import sys

if sys.stdout.encoding.lower() != "utf-8":
    sys.stdout.reconfigure(encoding="utf-8")

ROOTS = [
    "crm-auth/src/main/java", "crm-file/src/main/java", "crm-lead/src/main/java",
    "crm-opportunity/src/main/java", "crm-rule/src/main/java",
    "crm-preference/src/main/java", "crm-dict/src/main/java",
    "crm-customer/src/main/java",
]
PROJECT = r"d:\code\crm-backend-matt"
DOCS = r"D:\code\crm-api-docs"

def strip_comments(src):
    src = re.sub(r"/\*.*?\*/", "", src, flags=re.S)
    src = re.sub(r"//[^\n]*", "", src)
    return src

def extract_balanced(text, open_idx):
    """text[open_idx] == '('. return inner content, end idx after ')'."""
    depth = 0
    i = open_idx
    in_str = None
    while i < len(text):
        c = text[i]
        if in_str:
            if c == "\\":
                i += 2
                continue
            if c == in_str:
                in_str = None
        else:
            if c in "\"'":
                in_str = c
            elif c == "(":
                depth += 1
            elif c == ")":
                depth -= 1
                if depth == 0:
                    return text[open_idx + 1:i], i + 1
        i += 1
    return text[open_idx + 1:], len(text)

def parse_str_array(body):
    """parse {'a','b'} or "a" -> list of strings; also allow bare string."""
    body = body.strip()
    if body.startswith("{"):
        return re.findall(r'"([^"]*)"', body)
    m = re.match(r'^"([^"]*)"$', body)
    return [m.group(1)] if m else []

def find_annotations(src, name):
    """find all @Name(...) blocks, return list of inner contents."""
    out = []
    for m in re.finditer(r"@" + name + r"\s*\(", src):
        inner, _ = extract_balanced(src, m.end() - 1)
        out.append(inner)
    return out

def scan_source():
    apis = []
    conflicts = []
    for root in ROOTS:
        base = os.path.join(PROJECT, root)
        for dirpath, dirnames, filenames in os.walk(base):
            for fn in filenames:
                if not fn.endswith("Controller.java"):
                    continue
                path = os.path.join(dirpath, fn)
                src = strip_comments(open(path, encoding="utf-8").read())
                if "@RestController" not in src:
                    continue
                cls_m = re.search(r"class\s+(\w+)", src)
                cls = cls_m.group(1) if cls_m else fn[:-5]
                prefixes = []
                for rm in re.finditer(r"@RequestMapping\s*\(", src):
                    inner, end = extract_balanced(src, rm.end() - 1)
                    if end <= len(src) and "class" in src[end:end + 200]:
                        vals = parse_str_array(inner) or re.findall(r'"([^"]*)"', inner)
                        pv = re.search(r'value\s*=\s*', inner)
                        prefixes = parse_str_array(inner[pv.end():]) if pv else vals
                        break
                prefix = prefixes[0] if prefixes else ""
                # class-level tags from @Tag(name=...)
                cls_tags = []
                head = src[:cls_m.start()]
                for tag_inner in find_annotations(head, "Tag"):
                    nm = re.search(r'name\s*=\s*"([^"]*)"', tag_inner)
                    if nm:
                        cls_tags.append(nm.group(1))
                # collect all @Operation blocks first (project style: @Operation BEFORE @Mapping)
                ops = []
                for om in re.finditer(r"@Operation\s*\(", src):
                    op_inner, op_end = extract_balanced(src, om.end() - 1)
                    ops.append({"start": om.start(), "end": op_end, "inner": op_inner})
                # methods: find mapping annotations; assign the NEAREST @Operation BEFORE them
                for mm in re.finditer(r"@(Get|Post|Put|Delete|Request)Mapping\s*(\(|$)", src, re.M):
                    anno_start = mm.start()
                    if mm.group(1) == "Request":
                        inner, end = extract_balanced(src, mm.end() - 1) if mm.group(2) == "(" else ("", mm.end())
                        meth_m = re.search(r"method\s*=\s*RequestMethod\.(\w+)", inner)
                        if not meth_m:
                            continue  # class-level @RequestMapping (already consumed) or no method
                        http = meth_m.group(1)
                        pv = re.search(r'(?:value\s*=\s*)?"([^"]*)"', inner)
                        sub = pv.group(1) if pv else ""
                    else:
                        http = mm.group(1).upper()
                        if mm.group(2) == "(":
                            inner, end = extract_balanced(src, mm.end() - 1)
                            pv = re.search(r'(?:value\s*=\s*)?"([^"]*)"', inner)
                            sub = pv.group(1) if pv else ""
                        else:
                            end = mm.end()
                            sub = ""
                    url = (prefix + sub) if sub.startswith("/") else (prefix + ("/" + sub if sub else ""))
                    # find owning @Operation: project has BOTH styles.
                    # forward style: @Operation ... @PreAuthorize? ... @Mapping (customer/lead/opp)
                    # trailing style: @Mapping ... @PreAuthorize ... @Operation (crm-auth role/resource)
                    op = None
                    for cand in reversed(ops):
                        if cand["end"] <= anno_start:
                            between = src[cand["end"]:anno_start]
                            if "{" in between:
                                continue  # a method body separates them -> belongs to an earlier method
                            op = cand
                            break
                    if op is None:
                        body_brace = src.find("{", end)
                        if body_brace == -1:
                            body_brace = len(src)
                        for cand in ops:
                            if cand["start"] >= end and cand["start"] < body_brace:
                                op = cand
                                break
                    op_tags = None
                    summary = ""
                    if op is not None:
                        sm = re.search(r'summary\s*=\s*"([^"]*)"', op["inner"])
                        if sm:
                            summary = sm.group(1)
                        tm = re.search(r"tags\s*=\s*(\{[^}]*\}|\"[^\"]*\")", op["inner"], re.S)
                        if tm:
                            op_tags = parse_str_array(tm.group(1))
                    method_tags = []
                    if op_tags is None:
                        # method-level @Tag between op end and mapping
                        back = src[anno_start - 800:anno_start]
                        t_m = None
                        for ti in find_annotations(back, "Tag"):
                            t_m = ti
                        if t_m:
                            nm = re.search(r'name\s*=\s*"([^"]*)"', t_m)
                            if nm:
                                method_tags = [nm.group(1)]
                    tags = op_tags if op_tags is not None else (method_tags if method_tags else cls_tags)
                    if not tags:
                        tags = []
                    apis.append({
                        "file": os.path.relpath(path, PROJECT).replace("\\", "/"),
                        "className": cls, "http": http, "url": url,
                        "tags": tags, "summary": summary,
                    })
    # key conflicts (method+url, before tag expansion)
    seen = {}
    for a in apis:
        k = a["http"] + " " + a["url"]
        seen.setdefault(k, []).append(a)
    for k, v in seen.items():
        if len(v) > 1:
            conflicts.append({"key": k, "files": [x["file"] for x in v]})
    return apis, conflicts

def scan_docs():
    """return [{path, key, generated, folder}] for all .bru files."""
    out = []
    skip = {"node_modules", ".git", "environments"}
    for dirpath, dirnames, filenames in os.walk(DOCS):
        dirnames[:] = [d for d in dirnames if d not in skip]
        for fn in filenames:
            if not fn.endswith(".bru") or fn == "collection.bru" or fn == "folder.bru":
                continue
            path = os.path.join(dirpath, fn)
            content = open(path, encoding="utf-8").read()
            gen = bool(re.search(r"tags\s*:\s*\[[^\]]*generated", content, re.S))
            key = None
            km = re.search(r"docs\s*\{.*?`([A-Z]+)\s+(\S+)`", content, re.S)
            if km:
                key = km.group(1) + " " + km.group(2)
            folder = os.path.basename(dirpath)
            out.append({"path": os.path.relpath(path, DOCS).replace("\\", "/"),
                        "key": key, "generated": gen, "folder": folder})
    return out

def main():
    apis, conflicts = scan_source()
    docs = scan_docs()
    # expand source keys: one per (method+url, tag)
    src_keys = {}
    for a in apis:
        tags = a["tags"]
        if tags:
            tl = list(tags)
        else:
            # no tag -> urlPattern fallback dir: module[/resource]
            seg = [s for s in a["url"].split("/") if s]
            if seg and seg[0] == "api":
                seg = seg[1:]
            tl = ["/".join(seg[:2])]
        for t in tl:
            k = (a["http"] + " " + a["url"], t)
            if k in src_keys:
                conflicts.append({"key": k[0] + " @" + str(t), "files": [src_keys[k]["file"], a["file"]]})
            else:
                src_keys[k] = a
    # file keys
    file_keys = {}
    for d in docs:
        if not d["generated"]:
            continue
        if d["key"] is None:
            d["unrecognized"] = True
            continue
        rel = d["path"]
        parts = rel.split("/")
        # tag = full dir path between repo root and file
        tag = "/".join(parts[:-1])
        k = (d["key"], tag)
        file_keys[k] = d
    new_items = []
    del_items = []
    keep = 0
    for k in src_keys:
        if k not in file_keys:
            a = src_keys[k]
            new_items.append({"key": k[0], "tag": k[1], "summary": a["summary"],
                              "file": a["file"], "className": a["className"]})
    for k, d in file_keys.items():
        if k not in src_keys:
            del_items.append({"key": k[0], "tag": k[1], "path": d["path"]})
    for k in src_keys:
        if k in file_keys:
            keep += 1
    report = {
        "source_api_count": len(apis),
        "source_key_count": len(src_keys),
        "generated_file_count": len([d for d in docs if d["generated"]]),
        "private_file_count": len([d for d in docs if not d["generated"]]),
        "unrecognized_count": len([d for d in docs if d.get("unrecognized")]),
        "key_conflicts": conflicts,
        "keep_count": keep,
        "new_count": len(new_items),
        "delete_file_count": len(set(x["path"] for x in del_items)),
        "delete_distinct_url_count": len(set(x["key"] for x in del_items)),
    }
    out = {"report": report, "new": new_items, "delete": del_items, "docs": docs, "apis": apis}
    with open(os.path.join(PROJECT, r".scratch\customer-rework\bruno-recon-d.json"), "w", encoding="utf-8") as f:
        json.dump(out, f, ensure_ascii=False, indent=1)
    print(json.dumps(report, ensure_ascii=False, indent=1))

if __name__ == "__main__":
    main()