xGblob 18397#!/usr/bin/env python3 # -*- coding: utf-8 -*- """ 最终 SDK 包 demo 工程清理器 用途： 1. 清理旧的 ../OSFrameworks、../OSResources、../CNFrameworks、../CNResources 引用 （包括 PBXGroup 和 PBXFileSystemSynchronizedRootGroup 两种形式） 2. 清理旧的 SDK 组引用（默认组名：SDK） 3. 清理主 framework / bundle / middleware .a 的旧 fileRef、buildFile、build phase 引用 4. 按块级精确删除 FRAMEWORK_SEARCH_PATHS / HEADER_SEARCH_PATHS / LIBRARY_SEARCH_PATHS 中包含目标路径的条目 此脚本采用 "按块定位 + 按 ID 清除" 策略，避免误删正常行。 """ import os import re import shutil import sys from typing import Iterable, List, Set, Tuple LEGACY_PATHS = { "../OSFrameworks", "../OSResources", "../CNFrameworks", "../CNResources", } SEARCH_PATH_KEYS = [ "FRAMEWORK_SEARCH_PATHS", "HEADER_SEARCH_PATHS", "LIBRARY_SEARCH_PATHS", ] class FinalSDKProjectCleaner: def __init__(self, pbxproj_path: str): self.pbxproj_path = pbxproj_path if not os.path.exists(pbxproj_path): raise FileNotFoundError(f"项目文件不存在: {pbxproj_path}") def clean(self, sdk_name: str, sdk_rel_path: str = "../SDK", group_name: str = "SDK", exported_libs: Iterable[str] = ()): # noqa: E501 print("🧹 开始清理最终 SDK 包中的旧引用...") print(f"📄 pbxproj: {self.pbxproj_path}") print(f"📦 SDK 名称: {sdk_name}") print(f"📂 SDK 相对路径: {sdk_rel_path}") print(f"📁 组名: {group_name}") backup_path = f"{self.pbxproj_path}.backup_final_sdk_clean" shutil.copy2(self.pbxproj_path, backup_path) print(f"✅ 已创建备份: {backup_path}") with open(self.pbxproj_path, "r", encoding="utf-8") as f: content = f.read() try: exported_libs = [name.strip() for name in exported_libs if str(name).strip()] target_files = [f"{sdk_name}.framework", f"{sdk_name}.bundle", *exported_libs] sdk_folder_name = os.path.basename(sdk_rel_path.replace("\\", "/").rstrip("/")) removed_ids: Set[str] = set() content, ids = self._remove_sync_root_groups(content, LEGACY_PATHS) removed_ids |= ids content, ids = self._remove_pbx_groups_by_path(content, LEGACY_PATHS) removed_ids |= ids content, ids = self._remove_pbx_groups_by_name_or_path(content, group_name, sdk_folder_name, sdk_rel_path) removed_ids |= ids # Recursively remove sub-groups whose IDs were collected as children of deleted groups content, ids = self._remove_sub_groups_recursively(content, removed_ids) removed_ids |= ids # Remove orphaned PBXFileReference entries whose IDs were children of deleted groups content, ids = self._remove_file_references_by_ids(content, removed_ids) removed_ids |= ids content, ids = self._remove_file_references(content, target_files) removed_ids |= ids content, ids, build_file_ids_by_ref = self._remove_build_files(content, removed_ids | set()) removed_ids |= ids # Also remove build files for any of the target_files (regardless) content, ids = self._remove_build_files_by_target_names(content, target_files) removed_ids |= ids # Strip references to removed IDs anywhere (children / files / synchronizedGroups / etc.) content = self._strip_id_references(content, removed_ids) content = self._clean_search_paths(content, sdk_folder_name) if not self._validate(content): raise RuntimeError("清理后的 pbxproj 结构校验失败") with open(self.pbxproj_path, "w", encoding="utf-8") as f: f.write(content) print("✅ 旧引用清理完成") return True except Exception as e: print(f"❌ 清理失败: {e}") shutil.copy2(backup_path, self.pbxproj_path) print("✅ 已恢复备份文件") return False # ---------- helpers ---------- @staticmethod def _strip_quotes(value: str): value = value.strip() if len(value) >= 2 and value[0] == '"' and value[-1] == '"': return value[1:-1] return value def _find_block(self, content: str, start_idx: int) -> Tuple[int, int]: """Given index pointing to `{`, return (start, end_after_matching_brace_and_semicolon).""" depth = 0 i = start_idx n = len(content) in_string = False while i < n: ch = content[i] if ch == '"' and (i == 0 or content[i - 1] != "\\"): in_string = not in_string elif not in_string: if ch == "{": depth += 1 elif ch == "}": depth -= 1 if depth == 0: # include trailing `;` j = i + 1 while j < n and content[j] in " \t": j += 1 if j < n and content[j] == ";": j += 1 # include trailing newline while j < n and content[j] == "\n": j += 1 break return start_idx, j i += 1 return start_idx, n def _iter_object_blocks(self, content: str, isa_name: str): """Yield (id, full_match_start, block_start, block_end, block_text) for each top-level object of given isa.""" pattern = re.compile( r"^\s*([A-F0-9]{24}) /\* (.*?) \*/ = \{[^{}]*?isa = " + re.escape(isa_name) + r";", re.MULTILINE, ) for m in pattern.finditer(content): obj_id = m.group(1) comment = m.group(2) # find opening brace position brace_idx = content.index("{", m.start()) # find beginning of line line_start = content.rfind("\n", 0, m.start()) + 1 _, end = self._find_block(content, brace_idx) yield obj_id, comment, line_start, end, content[line_start:end] def _remove_blocks(self, content: str, spans: List[Tuple[int, int]]) -> str: if not spans: return content spans = sorted(set(spans), key=lambda s: s[0], reverse=True) for start, end in spans: content = content[:start] + content[end:] return content # ---------- group / file / build-file removers ---------- def _remove_sync_root_groups(self, content: str, paths: Iterable[str]): print("🗂️ 清理旧的 PBXFileSystemSynchronizedRootGroup...") removed: Set[str] = set() spans: List[Tuple[int, int]] = [] paths = set(paths) for obj_id, comment, s, e, text in self._iter_object_blocks(content, "PBXFileSystemSynchronizedRootGroup"): path_match = re.search(r"path = ([^;]+);", text) if not path_match: continue raw = self._strip_quotes(path_match.group(1)) if raw in paths: print(f" - 删除同步组: {raw} ({comment})") removed.add(obj_id) spans.append((s, e)) content = self._remove_blocks(content, spans) return content, removed def _remove_pbx_groups_by_path(self, content: str, paths: Iterable[str]): print("📂 清理按路径匹配的旧 PBXGroup...") removed: Set[str] = set() spans: List[Tuple[int, int]] = [] paths = set(paths) for obj_id, comment, s, e, text in self._iter_object_blocks(content, "PBXGroup"): path_match = re.search(r"path = ([^;]+);", text) if not path_match: continue raw = self._strip_quotes(path_match.group(1)) if raw in paths: print(f" - 删除 PBXGroup: {raw} ({comment})") removed.add(obj_id) spans.append((s, e)) # also collect child ids to prune later (children lines) for child_id in re.findall(r"([A-F0-9]{24}) /\*", text): if child_id != obj_id: removed.add(child_id) content = self._remove_blocks(content, spans) return content, removed def _remove_pbx_groups_by_name_or_path(self, content: str, group_name: str, folder_name: str, sdk_rel_path: str): print("📂 清理旧的 SDK PBXGroup (按名称或路径)...") removed: Set[str] = set() spans: List[Tuple[int, int]] = [] candidates = {group_name, folder_name, sdk_rel_path} for obj_id, comment, s, e, text in self._iter_object_blocks(content, "PBXGroup"): name_match = re.search(r"name = ([^;]+);", text) path_match = re.search(r"path = ([^;]+);", text) raw_name = self._strip_quotes(name_match.group(1)) if name_match else "" raw_path = self._strip_quotes(path_match.group(1)) if path_match else "" comment_name = self._strip_quotes(comment) hits = {raw_name, raw_path, comment_name} & candidates if hits: print(f" - 删除 SDK 组: name={raw_name}, path={raw_path}, comment={comment_name}") removed.add(obj_id) spans.append((s, e)) for child_id in re.findall(r"([A-F0-9]{24}) /\*", text): if child_id != obj_id: removed.add(child_id) content = self._remove_blocks(content, spans) return content, removed def _remove_sub_groups_recursively(self, content: str, parent_ids: Set[str]): """Recursively remove PBXGroup entries whose ID is in parent_ids, collecting all descendant IDs.""" print("📂 递归清理子组...") all_removed: Set[str] = set() pending_ids = set(parent_ids) while True: found_spans: List[Tuple[int, int]] = [] new_child_ids: Set[str] = set() for obj_id, comment, s, e, text in self._iter_object_blocks(content, "PBXGroup"): if obj_id in pending_ids: print(f" - 删除子组: {comment} ({obj_id})") found_spans.append((s, e)) all_removed.add(obj_id) for child_id in re.findall(r"([A-F0-9]{24}) /\*", text): if child_id != obj_id and child_id not in all_removed: new_child_ids.add(child_id) all_removed.add(child_id) content = self._remove_blocks(content, found_spans) if not new_child_ids: break pending_ids = new_child_ids return content, all_removed def _remove_file_references_by_ids(self, content: str, ids: Set[str]): """Remove PBXFileReference entries whose ID is in the given set (orphaned children of deleted groups).""" if not ids: return content, set() print("📄 清理已删除组的子项 PBXFileReference（防止 Recovered References）...") removed: Set[str] = set() pattern = re.compile( r"^(\s*)([A-F0-9]{24}) /\* ([^*]+?) \*/ = \{isa = PBXFileReference;[^}]*\};\s*$", re.MULTILINE, ) def replacer(match): ref_id = match.group(2) name = match.group(3).strip() if ref_id in ids: removed.add(ref_id) print(f" - 删除孤儿文件引用: {name} ({ref_id})") return "" return match.group(0) content = pattern.sub(replacer, content) return content, removed def _remove_file_references(self, content: str, file_names: List[str]): print("📄 清理旧的文件引用...") removed: Set[str] = set() if not file_names: return content, removed name_set = set(file_names) pattern = re.compile( r"^(\s*)([A-F0-9]{24}) /\* ([^*]+?) \*/ = \{isa = PBXFileReference;[^}]*\};\s*$", re.MULTILINE, ) def replacer(match): name = match.group(3).strip() if name in name_set: removed.add(match.group(2)) print(f" - 删除文件引用: {name}") return "" return match.group(0) content = pattern.sub(replacer, content) return content, removed def _remove_build_files(self, content: str, file_ref_ids: Set[str]): """Remove PBXBuildFile entries whose fileRef is in the given set.""" print("🔧 清理与已删除 fileRef 关联的构建文件...") removed: Set[str] = set() mapping = {} if not file_ref_ids: return content, removed, mapping pattern = re.compile( r"^\s*([A-F0-9]{24}) /\* ([^*]+?) \*/ = \{isa = PBXBuildFile; fileRef = ([A-F0-9]{24}) [^}]*\};\s*$", re.MULTILINE, ) def replacer(match): build_id = match.group(1) desc = match.group(2).strip() ref_id = match.group(3) if ref_id in file_ref_ids: removed.add(build_id) mapping[ref_id] = build_id print(f" - 删除构建文件: {desc}") return "" return match.group(0) content = pattern.sub(replacer, content) return content, removed, mapping def _remove_build_files_by_target_names(self, content: str, file_names: List[str]): print("🔧 再次按文件名清理残留的 PBXBuildFile...") removed: Set[str] = set() if not file_names: return content, removed name_re = "|".join(re.escape(n) for n in file_names) pattern = re.compile( rf"^\s*([A-F0-9]{{24}}) /\* ({name_re}) in (Frameworks|Resources|Sources) \*/ = \{{isa = PBXBuildFile;[^}}]*\}};\s*$", re.MULTILINE, ) def replacer(match): removed.add(match.group(1)) print(f" - 删除构建文件: {match.group(2)} in {match.group(3)}") return "" content = pattern.sub(replacer, content) return content, removed def _strip_id_references(self, content: str, ids: Set[str]): if not ids: return content print("🧽 清理 children / build phases / synchronizedGroups 中的旧引用...") for item_id in ids: pattern = re.compile(rf"^\s*{re.escape(item_id)} /\* [^*]* \*/,\s*$", re.MULTILINE) content = pattern.sub("", content) return content # ---------- search path block cleaner ---------- def _clean_search_paths(self, content: str, sdk_folder_name: str): print("🔍 清理搜索路径（按块精确处理）...") bad_tokens = [ "../OSFrameworks", "../OSResources", "../CNFrameworks", "../CNResources", f"/{sdk_folder_name}", ] for key in SEARCH_PATH_KEYS: # Multi-line: KEY = ( ... ); multi_pattern = re.compile(rf"(\s*){re.escape(key)} = \((.*?)\);", re.DOTALL) def multi_replacer(m): indent = m.group(1) body = m.group(2) items = [line for line in body.split("\n")] kept = [] for line in items: stripped = line.strip().rstrip(",").strip() if not stripped: kept.append(line) continue if any(tok in stripped for tok in bad_tokens): print(f" - 删除 {key} 条目: {stripped}") continue kept.append(line) new_body = "\n".join(kept) return f"{indent}{key} = ({new_body});" content = multi_pattern.sub(multi_replacer, content) # Single-line: KEY = "..."; single_pattern = re.compile(rf"(\s*){re.escape(key)} = \"([^\"]*)\";\s*\n") def single_replacer(m): value = m.group(2) if any(tok in value for tok in bad_tokens): print(f" - 删除 {key} 单值: {value}") return "" return m.group(0) content = single_pattern.sub(single_replacer, content) return content def _validate(self, content: str): print("🔍 校验 pbxproj 结构...") required_sections = [ "/* Begin PBXFileReference section */", "/* End PBXFileReference section */", "/* Begin PBXGroup section */", "/* End PBXGroup section */", "/* Begin XCBuildConfiguration section */", "/* End XCBuildConfiguration section */", ] for section in required_sections: if section not in content: print(f"❌ 缺少必要节: {section}") return False if content.count("{") != content.count("}"): print("❌ 花括号数量不匹配") return False if not re.search(r"rootObject = [A-F0-9]{24}", content): print("❌ 缺少 rootObject") return False print("✅ pbxproj 结构校验通过") return True def main(): if len(sys.argv) < 3: print("用法: python custom_pbxproj_clean.py [SDK相对路径] [组名] [导出.a逗号列表]") sys.exit(1) pbxproj_path = sys.argv[1] sdk_name = sys.argv[2] sdk_rel_path = sys.argv[3] if len(sys.argv) > 3 else "../SDK" group_name = sys.argv[4] if len(sys.argv) > 4 else "SDK" exported_libs_csv = sys.argv[5] if len(sys.argv) > 5 else "" exported_libs = [item.strip() for item in exported_libs_csv.split(",") if item.strip()] cleaner = FinalSDKProjectCleaner(pbxproj_path) ok = cleaner.clean(sdk_name, sdk_rel_path=sdk_rel_path, group_name=group_name, exported_libs=exported_libs) sys.exit(0 if ok else 1) if __name__ == "__main__": main() o