static const char sep[256] = { [' ']=1, ['.']=1, ['-']=1, ['_']=1 };

最新推荐文章于 2025-01-30 23:05:55 发布
转载 最新推荐文章于 2025-01-30 23:05:55 发布 · 489 阅读 · 0

本文介绍了一种使用GCC编译器的特殊初始化数组的方法，这种方法允许使用字符对应的ASCII码作为数组元素的下标进行初始化。然而，需要注意的是并非所有编译器都支持这种方式，例如Visual Studio就不支持。

static const char sep[256] = { [' ']=1, ['.']=1, ['-']=1, ['_']=1 };//gcc能编译通过，vs不支持此格式

意思是定义1个数组，并进行初始化赋值，被赋值元素的下标为字符所对应的ASCII码。

static const char sep[256] = { [' ']=1, ['.']=1, ['-']=1, ['_']=1 }
coobeedongzhou的专栏
05-17 719
static const char sep[256] = { [' ']=1, ['.']=1, ['-']=1, ['_']=1 };//gcc能编译通过，
Linux 内核路由发现过程
12-11 197
路由查找   //这个函数负责从fib_table中查找fn_key相匹配的fib_node，然后找到相应的fib_info, //将结果存放在fib_result结构中   //传入参数fz_divisor是linux kernel 2.6的变化,在2.4.x中是直接传入key的，而flp参数里面包含了数据包的全部地址信息，传入之后再用目的地址来计算key. static int
c语言中的static 和const
热门推荐
带你飞
05-18 2万+
C语言中static 和 const使用 static static在C语言中主要是两种用法 1、修饰局部变量 2、修饰函数和全局变量 局部变量按照存储形式来分，分为auto，static，register 首先从内存四区的角度去看，auto即为普通的局部变量，存储在栈上，当函数结束时，随之释放。 register为寄存器变量，存放在寄存器里面，调用速度快。 在C语言中regist...
Qt Creator编译时出错error: cannot convert 'const char*' to 'LPCWSTR
ziyou飞翔的专栏
07-29 1万+
用Qt5.3.1写的一个windows服务客理程序，就是在图形界面上能过点击按钮，实现
llama.cpp LLM_CHAT_TEMPLATE_DEEPSEEK_3
程永强
01-30 1404
llama.cpp LLM_CHAT_TEMPLATE_DEEPSEEK_3
int tm_to_sec(const char *date_time, time_t *utc_time) { /* 带时区的日期时间字符串（格式：YYYY-MM-DD HH:MM:SS +HHMM）*/ struct tm tm_time = {0}; char timezone_sign; int tz_hour, tz_min; int offset_seconds; /* 解析字符串 */ if (sscanf(date_time, "%d-%d-%d %d:%d:%d %c%02d%02d", &tm_time.tm_year, &tm_time.tm_mon, &tm_time.tm_mday, &tm_time.tm_hour, &tm_time.tm_min, &tm_time.tm_sec, &timezone_sign, &tz_hour, &tz_min) != 9) { RELAY_ERROR_LOG("解析失败\n"); return -1; } /* 调整tm结构：年从1900起计，月从0起计 */ tm_time.tm_year -= 1900; tm_time.tm_mon -= 1; tm_time.tm_isdst = -1; // 自动判断夏令时 /* 计算时区偏移秒数（东区为正，西区为负） */ offset_seconds = (timezone_sign == '+' ? 1 : -1) * (tz_hour * 3600 + tz_min * 60); *utc_time = mktime(&tm_time) - offset_seconds; return 0; } tm_to_sec - 将日期转换为1970.1.1以来的秒数 date_time:日期字符串（格式：YYYY-MM-DD HH:MM:SS +HHMM） utc_time:输出，保存转换的UTC秒数 成功返回0，失败返回-1，该函数有没有问题
09-20
int tm_to_sec(const char *date_time, time_t *utc_time) { struct tm tm_time = {0}; char timezone_sign; int tz_hour, tz_min; int offset_seconds; char *old_tz = NULL; /* 解析字符串 */ if (sscanf...
# === Step 2: 更新 ENUM === if valid_blocks: enum_pattern = rf'(enum\s+{re.escape(self.enum_name)}\s*\{{)([^}}]*)\}}\s*;' match = re.search(enum_pattern, block_content, re.DOTALL | re.IGNORECASE) if not match: raise ValueError(f"未找到枚举定义: {self.enum_name}") header = match.group(1) body_content = match.group(2) lines = [ln for ln in body_content.splitlines() if ln.strip()] # 提取 target_eq_col 用于对齐（保留原逻辑） last_line = lines[-1] if lines else "" clean_last = re.sub(r'//.*|/\*.*?\*/', '', last_line).strip() first_macro_match = re.search(r'RATE_SET_[A-Z0-9_]+', clean_last) eq_match = re.search(r'=\s*\d+', clean_last) target_eq_col = 30 if first_macro_match and eq_match: raw_before = last_line[:first_macro_match.start() + eq_match.start()] expanded_before = raw_before.expandtabs(4) target_eq_col = len(expanded_before) # 构造新宏 existing_enum = parsed.get('existing_enum', {}) next_enum_value = max(existing_enum.values(), default=-1) + 1 new_macros = [] for block in valid_blocks: name = block["enum_name"] padding = max(1, target_eq_col - len(name.replace('\t', ' '))) formatted = f"{name}{' ' * padding}= {next_enum_value}" new_macros.append(formatted) next_enum_value += 1 # 使用通用函数格式化 new_body = self._format_with_inline_fallback( lines=lines, new_items=new_macros, max_per_line=self.MAX_ENUM_PER_LINE, indent_marker=" ", item_separator=" ", line_suffix=",", extract_func=lambda s: re.findall(r'RATE_SET_[A-Z0-9_]+', s), ) new_enum_code = f"{header}{new_body}\n}};" replacements.append((match.start(), match.end(), new_enum_code)) self.logger.debug(f"更新 enum: 添加 {len(valid_blocks)} 项") # === Step 3: 更新 DATA 数组 === if valid_blocks: data_pattern = rf'(static const unsigned char {re.escape(self.data_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(data_pattern, block_content, re.DOTALL) if not match: raise ValueError(f"未找到 data 数组: {self.data_array_name}") header = match.group(1) body_content = match.group(2).strip() footer = match.group(3) lines = body_content.splitlines() last_line = lines[-1] if lines else "" indent_match = re.match(r'^(\s*)', last_line) line_indent = indent_match.group(1) if indent_match else " " new_body = body_content.rstrip() if not new_body.endswith(','): new_body += ',' for block in valid_blocks: # 插入头行（如 /*...*/ 4,） new_body += f"\n{line_indent}{block['raw_header'].strip()}" # 插入每行 body（保持原始缩进） for raw_line in block["raw_body"]: new_body += f"\n{line_indent}{raw_line}" new_data_code = f"{header}{new_body}\n{footer}" replacements.append((match.start(), match.end(), new_data_code)) self.logger.debug(f"计划更新 data 数组: 添加 {len(valid_blocks)} 个原始块") # === Step 4: 更新 INDEX 数组 === if valid_blocks: index_pattern = rf'(static const unsigned short {re.escape(self.index_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(index_pattern, block_content, re.DOTALL) if not match: raise ValueError(f"未找到 index 数组: {self.index_array_name}") header = match.group(1) body_content = match.group(2).strip() footer = match.group(3) lines = body_content.splitlines() if body_content.strip() else [] # 计算 index 值 index_values = [] for block in valid_blocks: index_val = base_data_offset + current_new_data_size index_values.append(str(index_val)) current_new_data_size += 1 + block["count"] # 使用通用函数格式化 new_body = self._format_with_inline_fallback( lines=lines, new_items=index_values, max_per_line=self.MAX_INDEX_ITEMS_PER_LINE, indent_marker=" ", item_separator=", ", line_suffix=",", extract_func=lambda s: re.findall(r'\d+', s), # 提取数字 ) new_index_code = f"{header}{new_body}\n{footer}" replacements.append((match.start(), match.end(), new_index_code)) self.logger.debug( f"更新 index 数组: 同行追加部分数据 → [{index_values[0]} ...]" ) 怎么改
10-26
data_pattern = rf'(static const unsigned char {re.escape(self.data_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(data_pattern, block_content, re.DOTALL) if not match: raise ...
# rate_set/rate_sync.py import json import os import re import logging import sys from pathlib import Path from utils import resource_path from datetime import datetime from typing import Dict, List, Tuple, Any # ------------------------------- # 日志配置 # ------------------------------- PROJECT_ROOT = Path(__file__).parent.parent.resolve() LOG_DIR = PROJECT_ROOT / "output" / "log" LOG_DIR.mkdir(parents=True, exist_ok=True) LOG_FILE = LOG_DIR / f"rate_sync_{datetime.now().strftime('%Y%m%d_%H%M%S')}.log" class RateSetSynchronizer: MAX_ENUM_PER_LINE = 4 # enum 每行最多几个宏 MAX_DATA_ITEMS_PER_LINE = 4 # data 数组每行最多几个值 MAX_INDEX_ITEMS_PER_LINE = 15 # index 数组每行最多几个值 def __init__(self, c_file_path=None, dry_run=False, config_path="config/config.json"): self.logger = logging.getLogger(f"{__name__}.RateSetSynchronizer") # 加载配置 self.config_file_path = resource_path(config_path) if not os.path.exists(self.config_file_path): raise FileNotFoundError(f"配置文件不存在: {self.config_file_path}") with open(self.config_file_path, 'r', encoding='utf-8') as f: self.config = json.load(f) self.dry_run = dry_run # C 文件路径 if c_file_path is None: internal_c_path = self.config["target_c_file"] self.c_file_path = resource_path(internal_c_path) else: self.c_file_path = Path(c_file_path) if not self.c_file_path.exists(): raise FileNotFoundError(f"找不到 C 源文件: {self.c_file_path}") # === 单一锚点标记 === self.block_start = self.config["STR_RATE_SET_DATA"] self.block_end = self.config["END_RATE_SET_DATA"] # 数组与枚举名 self.data_array_name = "rate_sets_2g_20m" self.index_array_name = "rate_sets_index_2g_20m" self.enum_name = "rate_set_2g_20m" # 扫描所有子目录中的 .c 文件（排除自身） self.rate_set_dir = Path(__file__).parent self.rate_files = [ f for f in self.rate_set_dir.rglob("*.c") # 递归匹配所有 .c 文件 if f.is_file() and f.name != "rate_sync.py" ] # 加载文件名和结构映射 self.target_map = self.config.get("rate_set_map") if not isinstance(self.target_map, dict): raise ValueError("config.json 中缺少 'rate_set_map' 字段或格式错误") # 变更摘要 self.summary = { "total_new_sets": 0, "per_band": {"2g": 0, "5g": 0, "6g": 0}, "per_type": {}, # 如 20M, 40M, 80M... "new_macros": [], # RATE_SET_xxx "new_locales": [], # 文件中的 /*NAME*/ 块 "tables_updated": set() } self._validate_target_map() # ← 添加一致性校验 def _validate_target_map(self): """验证 rate_set_map 是否一致，防止多个 full_key 映射到同一数组""" seen_data = {} seen_index = {} seen_enum = {} for key, cfg in self.target_map.items(): d = cfg["data"] i = cfg["index"] e = cfg["enum"] if d in seen_data: raise ValueError(f"data 数组冲突: '{d}' 被 '{seen_data[d]}' 和 '{key}' 同时使用") if i in seen_index: raise ValueError(f"index 数组冲突: '{i}' 被 '{seen_index[i]}' 和 '{key}' 同时使用") if e in seen_enum: raise ValueError(f"enum 名称冲突: '{e}' 被 '{seen_enum[e]}' 和 '{key}' 同时使用") seen_data[d] = key seen_index[i] = key seen_enum[e] = key def parse_filename(self, filename: str) -> str: """ 从文件名提取 band_bw_ext 类型键，用于查找 rate_set_map 示例： 2G_20M_rate_set.c → 2G_20M_BASE 2G_20M_EXT_rate_set.c → 2G_20M_EXT 2G_20M_EXT4_NSS2_rate_set.c → 2G_20M_EXT4 5G_80M_NSS1_rate_set.c → 5G_80M_BASE 6G_320M_EXT4_NSS4_rate_set.c → 6G_320M_EXT4 """ # 增强版正则：允许中间有 _NSS\d+ pattern = r'^([A-Z0-9]+)_([0-9]+M)(?:_(EXT\d*))?(_NSS\d+)?_rate_set\.c$' match = re.match(pattern, filename, re.I) if not match: raise ValueError(f"无法识别的文件名格式: {filename}") band, bw, ext, nss_part = match.groups() ext_type = ext.upper() if ext else "BASE" return f"{band.upper()}_{bw.upper()}_{ext_type}" def extract_sub_rate_sets(self, content: str) -> List[Dict[str, Any]]: """ 提取 /*NAME*/ N, 后续多行 WL_RATE_xxx 列表 支持跨行、缩进、逗号、空行、注释干扰等 使用“按行扫描 + 状态机”方式，避免正则越界 """ self.logger.info("开始提取速率集...") self.logger.info("...") sub_sets = [] lines = [line.rstrip('\r\n') for line in content.splitlines()] # 保留原始行尾 i = 0 # 匹配 /*NAME*/ N, 的开头 header_pattern = re.compile(r'/\*\s*([A-Za-z0-9_]+)\s*\*/\s*(\d+)\s*,?') while i < len(lines): line = lines[i] stripped = line.strip() # 跳过空行和纯注释 if not stripped or stripped.startswith("//"): i += 1 continue # 查找头: /*NAME*/ N, match = header_pattern.search(stripped) if not match: i += 1 continue name = match.group(1) try: count = int(match.group(2)) except ValueError: self.logger.warning(f"⚠️ 计数无效，跳过: {name} = '{match.group(2)}'") i += 1 continue self.logger.info(f"🔍 发现块: {name}, 预期数量={count}") # 开始收集 body 内容（保留原始带缩进的行） body_lines = [] j = i + 1 max_lines_to_read = 200 while j < len(lines) and len(body_lines) < max_lines_to_read: ln = lines[j].strip() # 终止条件：遇到新 block / 结构结束 if ln.startswith("/*") or ln.startswith("}") or ln.startswith("enum"): break if ln and not ln.startswith("//"): body_lines.append(lines[j]) # ← 原样保存（含缩进） else: body_lines.append(lines[j]) # 也保留注释或空行（保持格式一致） j += 1 # 提取宏名用于校验（但不再用于生成数据） body_text = "\n".join(body_lines) all_macros = re.findall(r'WL_RATE_[A-Za-z0-9_]+', body_text) rate_items = all_macros[:count] if len(rate_items) < count: self.logger.warning(f"[{name}] 条目不足: 需要 {count}, 实际 {len(rate_items)}") # 构建结果：增加 raw_header 和 raw_body（关键改动） sub_sets.append({ "name": name, "count": count, "rates": rate_items, "raw_header": line, # ← 原始头行（如 /*...*/ 4,） "raw_body": body_lines, # ← 原始 body 行列表 "start_line": i, "end_line": j - 1 }) self.logger.debug(f"✅ 提取成功: {name} → {len(rate_items)} 个速率") i = j # 跳到下一个 block self.logger.info(f" 共提取 {len(sub_sets)} 个有效子集") return sub_sets def parse_all_structures(self, full_content: str) -> Dict: """ 直接从完整 C 文件中解析 enum/data/index 结构 """ self.logger.info("开始解析所有结构...") self.logger.info("...") result = { 'existing_enum': {}, 'data_entries': [], 'index_values': [], 'data_len': 0 } # === 解析 enum === enum_pattern = rf'enum\s+{re.escape(self.enum_name)}\s*\{{([^}}]+)\}};' enum_match = re.search(enum_pattern, full_content, re.DOTALL) if enum_match: body = enum_match.group(1) entries = re.findall(r'(RATE_SET_[^=,\s]+)\s*=\s*(\d+)', body) result['existing_enum'] = {k: int(v) for k, v in entries} self.logger.info(f"解析出 {len(entries)} 个已有枚举项") else: self.logger.warning(f"未找到 enum 定义: {self.enum_name}") # === 解析 data 数组 === data_pattern = rf'static const unsigned char {re.escape(self.data_array_name)}\[\] = \{{([^}}]+)\}};' data_match = re.search(data_pattern, full_content, re.DOTALL) if not data_match: raise ValueError(f"未找到 data 数组: {self.data_array_name}") data_code = data_match.group(1) result['data_entries'] = [item.strip() for item in re.split(r'[,\n]+', data_code) if item.strip()] result['data_len'] = len(result['data_entries']) # === 解析 index 数组 === index_pattern = rf'static const unsigned short {re.escape(self.index_array_name)}\[\] = \{{([^}}]+)\}};' index_match = re.search(index_pattern, full_content, re.DOTALL) if not index_match: raise ValueError(f"未找到 index 数组: {self.index_array_name}") index_code = index_match.group(1) result['index_values'] = [int(x.strip()) for x in re.split(r'[,\n]+', index_code) if x.strip()] return result def build_injection_with_format(self, new_subsets: List[Dict], existing_enum: Dict[str, int]) -> List[Dict]: """ 返回需要注入的原始块列表（包含 raw_header + raw_body） 不再返回 new_data/new_indices/new_enums """ valid_blocks = [] next_enum_value = max(existing_enum.values(), default=-1) + 1 self.logger.info(f"开始构建注入内容，当前最大枚举值 = {next_enum_value}") for subset in new_subsets: enum_name = subset["name"] if enum_name in existing_enum: self.logger.info(f"跳过已存在的枚举项: {enum_name} = {existing_enum[enum_name]}") continue # 只保存必要信息，不计算偏移 valid_blocks.append({ "enum_name": enum_name, "raw_header": subset["raw_header"], "raw_body": subset["raw_body"], "count": subset["count"], # 用于计算 data 占用空间 "enum_value": next_enum_value, # ✅ 必须存在！ }) self.logger.debug(f" 准备注入: {enum_name}") self.logger.info(f"新增条目: {enum_name} enum={next_enum_value}") next_enum_value += 1 self.logger.info(f"构建完成：共 {len(valid_blocks)} 个新条目（保留原始格式）") return valid_blocks def format_list(self, items: List[str], indent: str = " ", width: int = 8) -> str: """格式化数组为多行字符串""" lines = [] for i in range(0, len(items), width): chunk = items[i:i + width] lines.append(indent + ", ".join(chunk) + ",") return "\n".join(lines).rstrip(",") def _safe_write_back(self, old_content: str, new_content: str) -> bool: """安全写回文件，带备份""" if old_content == new_content: self.logger.info("主文件内容无变化，无需写入") return False if self.dry_run: self.logger.info("DRY-RUN 模式启用，跳过实际写入") print("[DRY RUN] 差异预览（前 20 行）:") diff = new_content.splitlines()[:20] for line in diff: print(f" {line}") return True # 创建备份 timestamp = datetime.now().strftime("%Y%m%d_%H%M%S") backup = self.c_file_path.with_name(f"{self.c_file_path.stem}_{timestamp}.c.bak") try: self.c_file_path.rename(backup) self.logger.info(f"原文件已备份为: {backup.name}") except Exception as e: self.logger.error(f"备份失败: {e}") raise # 写入新内容 try: self.c_file_path.write_text(new_content, encoding='utf-8') self.logger.info(f"✅ 成功写入更新后的文件: {self.c_file_path.name}") return True except Exception as e: self.logger.error(f"写入失败: {e}", exc_info=True) raise def inject_new_data(self) -> bool: try: full_content = self.c_file_path.read_text(encoding='utf-8') except Exception as e: self.logger.error(f"读取主 C 文件失败: {e}") raise self.logger.info(f"正在处理 C 文件: {self.c_file_path.name}") start_pos = full_content.find(self.block_start) end_pos = full_content.find(self.block_end) if start_pos == -1: raise ValueError(f"未找到起始锚点: {self.block_start}") if end_pos == -1: raise ValueError(f"未找到结束锚点: {self.block_end}") if end_pos <= start_pos: raise ValueError("结束锚点位于起始锚点之前") inner_start = start_pos + len(self.block_start) block_content = full_content[inner_start:end_pos].strip() all_changes_made = False # === 遍历每一个 rate set 子文件 === for file_path in self.rate_files: try: self.logger.info(f"→ 处理子文件: {file_path.name}") # --- 1. 解析文件名得到 full_key --- try: full_key = self.parse_filename(file_path.name) self.logger.debug(f" ├─ 解析出 key: {full_key}") except ValueError as ve: self.logger.warning(f" └─ 跳过无效文件名: {ve}") continue # --- 2. 查找 rate_set_map 映射 --- target = self.target_map.get(full_key) if not target: self.logger.warning(f" └─ 未在 config.json 中定义映射关系: {full_key}，跳过") continue # --- 3. 动态设置当前注入目标 --- self.data_array_name = target["data"] self.index_array_name = target["index"] self.enum_name = target["enum"] self.logger.debug(f" ├─ 绑定目标:") self.logger.debug(f" data: {self.data_array_name}") self.logger.debug(f" index: {self.index_array_name}") self.logger.debug(f" enum: {self.enum_name}") # --- 4. 解析主文件中的当前结构 --- try: parsed = self.parse_all_structures(full_content) except Exception as e: self.logger.error(f" └─ 解析主文件结构失败: {e}") continue # --- 5. 提取该子文件中的 rate sets --- file_content = file_path.read_text(encoding='utf-8') subsets = self.extract_sub_rate_sets(file_content) if not subsets: self.logger.info(f" └─ 无有效子集数据") continue # --- 6. 构建要注入的内容 --- valid_blocks = self.build_injection_with_format( subsets, existing_enum=parsed['existing_enum'] ) if not valid_blocks: self.logger.info(f" └─ 无需更新") continue # --- 7. 写回新内容（精准插入）--- updated_content = self._write_back_in_blocks( full_content, parsed, valid_blocks ) if updated_content != full_content: all_changes_made = True full_content = updated_content # 更新内存内容供后续文件使用 # === 统计变更 === count = len(valid_blocks) enum_name_base = self.enum_name.replace("rate_set_", "").upper() self.summary["total_new_sets"] += count self.summary["tables_updated"].add(self.data_array_name) self.summary["tables_updated"].add(self.index_array_name) self.summary["tables_updated"].add(self.enum_name) # 提取 band 和 type (如 2G_20M_EXT) band_key = "unknown" bw_type = "unknown" if "2g" in full_key.lower(): band_key = "2g" elif "5g" in full_key.lower(): band_key = "5g" elif "6g" in full_key.lower(): band_key = "6g" bw_match = re.search(r'_(\d+M)', full_key) bw_type = bw_match.group(1) if bw_match else "unknown" self.summary["per_band"][band_key] += count self.summary["per_type"][bw_type] = self.summary["per_type"].get(bw_type, 0) + count # 收集新宏和 locale 名称 for block in valid_blocks: macro_name = f"RATE_SET_{block['enum_name']}" self.summary["new_macros"].append(macro_name) self.summary["new_locales"].append(f"{block['enum_name']} ({full_key})") self.logger.info(f"✅ 成功注入 {count} 条目 → {self.enum_name}") except Exception as e: self.logger.warning(f"❌ 处理文件失败 [{file_path.name}]: {e}") if self.logger.isEnabledFor(logging.DEBUG): self.logger.debug("详细堆栈:", exc_info=True) continue # 最终写回磁盘 if all_changes_made: try: return self._safe_write_back(self.c_file_path.read_text(encoding='utf-8'), full_content) except Exception as e: self.logger.error(f"写入最终文件失败: {e}") raise else: self.logger.info("没有需要更新的内容") return False def _format_with_inline_fallback( self, lines: List[str], new_items: List[str], max_per_line: int, indent_marker: str = " ", item_separator: str = ", ", line_suffix: str = "", # 注意：现在我们不在这里加逗号！ extract_func=None, align_eq_col: bool = False, detect_spacing_from_last_line: bool = True, ) -> str: if not lines: lines = [""] last_line = lines[-1].rstrip() indent_match = re.match(r'^(\s*)', last_line) line_indent = indent_match.group(1) if indent_match else indent_marker clean_last = re.sub(r'//.*|/\*.*?\*/', '', last_line).strip() existing_items = extract_func(clean_last) if extract_func else re.findall(r'\w+', clean_last) current_count = len(existing_items) space_left = max(0, max_per_line - current_count) to_append_inline = new_items[:space_left] to_append_newline = new_items[space_left:] # === 检测真实分隔符 === actual_sep = item_separator if detect_spacing_from_last_line and len(existing_items) >= 2: first = re.escape(existing_items[0]) second = re.escape(existing_items[1]) match = re.search(f"({first})(\\s+)({second})", last_line) if match: actual_sep = match.group(2) # === 对齐等号列：关键修复 → 对齐后再加逗号 === formatted_new_items = [] for item in new_items: raw_item = item.rstrip(',') # 去掉可能已有的逗号避免重复 if align_eq_col: m = re.match(r'(\w+)\s*=\s*(\d+)', raw_item) if m: name, val = m.groups() # 计算目标列位置 target_eq_col = None for i in range(len(lines) - 1, -1, -1): ln = lines[i] eq_match = re.search(r'=\s*\d+', ln) if eq_match: raw_before = ln[:eq_match.start()] expanded_before = raw_before.expandtabs(4) target_eq_col = len(expanded_before) break if target_eq_col is None: target_eq_col = 30 padding = max(1, target_eq_col - len(name.replace('\t', ' ').expandtabs(4))) spaces = ' ' * padding aligned_item = f"{name}{spaces}= {val}" formatted_new_items.append(aligned_item) else: formatted_new_items.append(raw_item) else: formatted_new_items.append(raw_item) # 现在统一加逗号：每个 item 都要加！ # 注意：是否加逗号应该由调用者或此函数控制，不要混合 final_formatted_items = [f"{item}," for item in formatted_new_items] to_append_inline = final_formatted_items[:space_left] to_append_newline = final_formatted_items[space_left:] # === 构建结果 === result_lines = lines[:-1] # 保留前面所有行 final_main_line = lines[-1].rstrip() # 添加 inline 项 if to_append_inline: joined_inline = actual_sep.join(to_append_inline) if final_main_line.strip(): final_main_line += actual_sep + joined_inline else: final_main_line = joined_inline result_lines.append(final_main_line) # 添加新行（每行最多 max_per_line 个） if to_append_newline: for i in range(0, len(to_append_newline), max_per_line): chunk = to_append_newline[i:i + max_per_line] joined = actual_sep.join(chunk) result_lines.append(f"{line_indent}{joined}") return '\n'.join(result_lines) def _write_back_in_blocks(self, full_content: str, parsed: Dict, valid_blocks: List[Dict]) -> str: """ 使用局部块操作策略：只在 /* START */ ... /* END */ 范围内修改内容 关键改进：直接插入 raw_header + raw_body，保留原始格式 """ self.logger.info("开始执行局部块写入操作...") self.logger.info("...") # 在 _write_back_in_blocks 最上方添加： base_data_offset = parsed['data_len'] current_new_data_size = 0 # 动态记录已写入的新 data 大小 start_pos = full_content.find(self.block_start) end_pos = full_content.find(self.block_end) if start_pos == -1 or end_pos == -1: raise ValueError(f"未找到锚点标记: {self.block_start} 或 {self.block_end}") if end_pos <= start_pos: raise ValueError("结束锚点位于起始锚点之前") inner_start = start_pos + len(self.block_start) block_content = full_content[inner_start:end_pos] replacements = [] # (start_in_block, end_in_block, replacement) # === Step 2: 更新 ENUM === if valid_blocks: # 提取函数：从字符串中提取 RATE_SET_xxx extract_enum = lambda s: re.findall(r'RATE_SET_[A-Z0-9_]+', s) enum_pattern = rf'(enum\s+{re.escape(self.enum_name)}\s*\{{)([^}}]*)\}}\s*;' match = re.search(enum_pattern, block_content, re.DOTALL | re.IGNORECASE) if not match: raise ValueError(f"未找到枚举定义: {self.enum_name}") header = match.group(1) # "enum rate_set_2g_20m {" body_content = match.group(2) lines = [ln.rstrip() for ln in body_content.splitlines() if ln.strip()] # 计算新值 new_macros = [] for block in valid_blocks: name = block["enum_name"] value = block["enum_value"] # ✅ 来自 build_injection_with_format 的正确值 new_macros.append(f"{name} = {value}") # === 关键：获取标准缩进 === indent_match = re.match(r'^(\s*)', lines[0] if lines else "") standard_indent = indent_match.group(1) if indent_match else " " # 格式化新 body new_body = self._format_with_inline_fallback( lines=lines, new_items=new_macros, max_per_line=self.MAX_ENUM_PER_LINE, indent_marker=standard_indent, item_separator=" ", line_suffix="", extract_func=extract_enum, align_eq_col=True, detect_spacing_from_last_line=True, ) # 关键修复：确保每行都有缩进（包括第一行） formatted_lines = [] for line in new_body.splitlines(): stripped = line.strip() if stripped: formatted_lines.append(f"{standard_indent}{stripped}") else: formatted_lines.append(line) final_body = '\n'.join(formatted_lines) # 关键：header 单独占一行，新 body 换行开始 new_enum_code = f"{header}\n{final_body}\n}};" replacements.append((match.start(), match.end(), new_enum_code)) self.logger.debug(f"更新 enum: 添加 {len(valid_blocks)} 项") # === Step 3: 更新 DATA 数组 === if valid_blocks: data_pattern = rf'(static const unsigned char {re.escape(self.data_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(data_pattern, block_content, re.DOTALL) if not match: raise ValueError(f"未找到 data 数组: {self.data_array_name}") header = match.group(1) body_content = match.group(2).strip() footer = match.group(3) lines = body_content.splitlines() last_line = lines[-1] if lines else "" indent_match = re.match(r'^(\s*)', last_line) line_indent = indent_match.group(1) if indent_match else " " new_body = body_content.rstrip() if not new_body.endswith(','): new_body += ',' for block in valid_blocks: # 插入头行（如 /*...*/ 4,） new_body += f"\n{line_indent}{block['raw_header'].strip()}" # 插入每行 body（保持原始缩进） for raw_line in block["raw_body"]: new_body += f"\n{line_indent}{raw_line}" new_data_code = f"{header}{new_body}\n{footer}" replacements.append((match.start(), match.end(), new_data_code)) self.logger.debug(f"计划更新 data 数组: 添加 {len(valid_blocks)} 个原始块") # === Step 2: 更新 INDEX 数组 === if valid_blocks: index_pattern = rf'(static const unsigned short {re.escape(self.index_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(index_pattern, block_content, re.DOTALL) if not match: raise ValueError(f"未找到 index 数组: {self.index_array_name}") header = match.group(1) body_content = match.group(2) footer = match.group(3).strip() lines = [ln.rstrip() for ln in body_content.splitlines()] non_empty_lines = [ln for ln in lines if ln.strip()] # 获取标准缩进（与 enum 一致） if non_empty_lines: indent_match = re.match(r'^(\s*)', non_empty_lines[0]) standard_indent = indent_match.group(1) if indent_match else " " else: standard_indent = " " # 生成新索引值 # 正确计算 index values：基于 data 偏移 + 每个 block 的实际大小 current_offset = parsed['data_len'] # 初始偏移 = 原 data 长度 new_index_values = [] for block in valid_blocks: # 添加当前 block 的起始偏移 new_index_values.append(str(current_offset)) # 偏移 += 当前 block 的数据条数 current_offset += block["count"]+1 # ← 使用 block 自带的 count！ self.logger.info(f"生成新的 index values: {new_index_values}") # 格式化 index body formatted_body = self._format_with_inline_fallback( lines=non_empty_lines, new_items=new_index_values, max_per_line=self.MAX_INDEX_ITEMS_PER_LINE, indent_marker=standard_indent, item_separator=" ", line_suffix="", extract_func=lambda s: re.findall(r'\d+', s), detect_spacing_from_last_line=True, align_eq_col=False, ) # 统一添加缩进 final_lines = [] for line in formatted_body.splitlines(): stripped = line.strip() if stripped: final_lines.append(f"{standard_indent}{stripped}") else: final_lines.append("") final_body = '\n'.join(final_lines) new_index_code = f"{header}\n{final_body}\n{footer}" replacements.append((match.start(), match.end(), new_index_code)) self.logger.debug(f"更新 index 数组: 添加 {len(valid_blocks)} 个索引") # === Step 5: 倒序应用所有替换 === if not replacements: self.logger.info("无任何变更需要写入") return full_content for start, end, r in sorted(replacements, key=lambda x: x[0], reverse=True): block_content = block_content[:start] + r + block_content[end:] # === Step 6: 拼接回完整文件 === final_content = ( full_content[:inner_start] + block_content + full_content[end_pos:] ) self.logger.info(f"成功构建新内容，总长度变化: {len(full_content)} → {len(final_content)}") return final_content def _print_summary(self): """打印最终变更总结""" s = self.summary self.logger.info("=" * 60) self.logger.info("📋 R A T E S E T 变 更 总 结") self.logger.info("=" * 60) if s["total_new_sets"] == 0: self.logger.info("🟢 无新增速率集") else: self.logger.info(f"🟢 新增速率集总数: {s['total_new_sets']} 个") # 按频段分布 self.logger.info(f"📡 按频段分布:") for band in ["2g", "5g", "6g"]: cnt = s["per_band"][band] if cnt > 0: self.logger.info(f" • {band.upper()}: {cnt} 个") # 按带宽类型分布 if s["per_type"]: self.logger.info(f"📏 按带宽分布:") for bw, cnt in s["per_type"].items(): self.logger.info(f" • {bw}: {cnt} 个") # 新增的宏 self.logger.info(f"🔧 新增 ENUM 宏:") for macro in sorted(s["new_macros"])[:20]: # 限制显示前20个 self.logger.info(f" • {macro}") if len(s["new_macros"]) > 20: self.logger.info(f" • ... 还有 {len(s['new_macros']) - 20} 个") # 新增的 Locale 块 self.logger.info(f"📦 新增数据块 (/*NAME*/):") for loc in sorted(s["new_locales"])[:10]: self.logger.info(f" • {loc}") if len(s["new_locales"]) > 10: self.logger.info(f" • ... 还有 {len(s['new_locales']) - 10} 个") # 更新的表 updated_tables = sorted(s["tables_updated"]) self.logger.info(f"📝 更新的数据结构:") for tbl in updated_tables: prefix = "rate_sets_" if "rate_sets_" in tbl else "rate_set_" desc = "ENUM" if tbl.startswith("rate_set_") and not "_" + "idx" in tbl else \ "INDEX" if "_index_" in tbl else "DATA" self.logger.info(f" • {tbl:<30} [{desc}]") self.logger.info("=" * 60) status = "✅ 同步完成" if s["total_new_sets"] > 0 else "ℹ️ 无需更新" self.logger.info(status) def run(self): self.logger.info("开始同步 RATE_SET 数据...") try: changed = self.inject_new_data() # === 打印变更总结 === self._print_summary() if changed: print("🔄 同步完成：已更新 C 文件") else: print("✅ 无新数据，无需更新") return { "success": True, "changed": changed, "file": str(self.c_file_path), "backup": f"{self.c_file_path.stem}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.c.bak" if changed and not self.dry_run else None } except Exception as e: self.logger.error(f"同步失败: {e}", exc_info=True) print("❌ 同步失败，详见日志。") return {"success": False, "error": str(e)} def main(): logging.basicConfig( level=logging.INFO, format='%(asctime)s [%(levelname)s] %(name)s: %(message)s', handlers=[ logging.FileHandler(LOG_FILE, encoding='utf-8'), logging.StreamHandler(sys.stdout) ], force=True ) dry_run = False # 设置为 True 可进行试运行 try: sync = RateSetSynchronizer(dry_run=dry_run) sync.run() print("同步完成！") except FileNotFoundError as e: logging.error(f"文件未找到: {e}") print(" 文件错误，请检查路径。") sys.exit(1) except PermissionError as e: logging.error(f"权限错误: {e}") print(" 权限不足，请关闭编辑器或以管理员运行。") sys.exit(1) except Exception as e: logging.error(f"程序异常退出: {e}", exc_info=True) print(" 同步失败，详见日志。") sys.exit(1) if __name__ == '__main__': main() 这个呢
11-10
我们来**全面分析 `rate_sync.py` 中的日志路径问题，尤其是在 PyInstaller 打包后的表现是否正确。** --- ## 你的日志配置代码： ```python # ------------------------------- # 日志配置 ...
/*************************************************************************** * _ _ ____ _ * Project ___| | | | _ \| | * / __| | | | |_) | | * | (__| |_| | _ <| |___ * \___|\___/|_| \_\_____| * * Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al. * * This software is licensed as described in the file COPYING, which * you should have received as part of this distribution. The terms * are also available at https://curl.se/docs/copyright.html. * * You may opt to use, copy, modify, merge, publish, distribute and/or sell * copies of the Software, and permit persons to whom the Software is * furnished to do so, under the terms of the COPYING file. * * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY * KIND, either express or implied. * * SPDX-License-Identifier: curl * ***************************************************************************/ #include "curl_setup.h" #include "urldata.h" #include <curl/curl.h> #include <stddef.h> #ifdef HAVE_LIBZ #include <zlib.h> #endif #ifdef HAVE_BROTLI #include <brotli/decode.h> #endif #ifdef HAVE_ZSTD #include <zstd.h> #endif #include "sendf.h" #include "http.h" #include "content_encoding.h" #include "strdup.h" #include "strcase.h" #include "curl_memory.h" #include "memdebug.h" #define CONTENT_ENCODING_DEFAULT "identity" #ifndef CURL_DISABLE_HTTP #define DSIZ CURL_MAX_WRITE_SIZE /* buffer size for decompressed data */ #ifdef HAVE_LIBZ /* Comment this out if zlib is always going to be at least ver. 1.2.0.4 (doing so will reduce code size slightly). */ #define OLD_ZLIB_SUPPORT 1 #define GZIP_MAGIC_0 0x1f #define GZIP_MAGIC_1 0x8b /* gzip flag byte */ #define ASCII_FLAG 0x01 /* bit 0 set: file probably ascii text */ #define HEAD_CRC 0x02 /* bit 1 set: header CRC present */ #define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ #define ORIG_NAME 0x08 /* bit 3 set: original file name present */ #define COMMENT 0x10 /* bit 4 set: file comment present */ #define RESERVED 0xE0 /* bits 5..7: reserved */ typedef enum { ZLIB_UNINIT, /* uninitialized */ ZLIB_INIT, /* initialized */ ZLIB_INFLATING, /* inflating started. */ ZLIB_EXTERNAL_TRAILER, /* reading external trailer */ ZLIB_GZIP_HEADER, /* reading gzip header */ ZLIB_GZIP_INFLATING, /* inflating gzip stream */ ZLIB_INIT_GZIP /* initialized in transparent gzip mode */ } zlibInitState; /* Deflate and gzip writer. */ struct zlib_writer { struct contenc_writer super; zlibInitState zlib_init; /* zlib init state */ uInt trailerlen; /* Remaining trailer byte count. */ z_stream z; /* State structure for zlib. */ }; static voidpf zalloc_cb(voidpf opaque, unsigned int items, unsigned int size) { (void) opaque; /* not a typo, keep it calloc() */ return (voidpf) calloc(items, size); } static void zfree_cb(voidpf opaque, voidpf ptr) { (void) opaque; free(ptr); } static CURLcode process_zlib_error(struct Curl_easy *data, z_stream *z) { if(z->msg) failf(data, "Error while processing content unencoding: %s", z->msg); else failf(data, "Error while processing content unencoding: " "Unknown failure within decompression software."); return CURLE_BAD_CONTENT_ENCODING; } static CURLcode exit_zlib(struct Curl_easy *data, z_stream *z, zlibInitState *zlib_init, CURLcode result) { if(*zlib_init == ZLIB_GZIP_HEADER) Curl_safefree(z->next_in); if(*zlib_init != ZLIB_UNINIT) { if(inflateEnd(z) != Z_OK && result == CURLE_OK) result = process_zlib_error(data, z); *zlib_init = ZLIB_UNINIT; } return result; } static CURLcode process_trailer(struct Curl_easy *data, struct zlib_writer *zp) { z_stream *z = &zp->z; CURLcode result = CURLE_OK; uInt len = z->avail_in < zp->trailerlen? z->avail_in: zp->trailerlen; /* Consume expected trailer bytes. Terminate stream if exhausted. Issue an error if unexpected bytes follow. */ zp->trailerlen -= len; z->avail_in -= len; z->next_in += len; if(z->avail_in) result = CURLE_WRITE_ERROR; if(result || !zp->trailerlen) result = exit_zlib(data, z, &zp->zlib_init, result); else { /* Only occurs for gzip with zlib < 1.2.0.4 or raw deflate. */ zp->zlib_init = ZLIB_EXTERNAL_TRAILER; } return result; } static CURLcode inflate_stream(struct Curl_easy *data, struct contenc_writer *writer, zlibInitState started) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ uInt nread = z->avail_in; Bytef *orig_in = z->next_in; bool done = FALSE; CURLcode result = CURLE_OK; /* Curl_client_write status */ char *decomp; /* Put the decompressed data here. */ /* Check state. */ if(zp->zlib_init != ZLIB_INIT && zp->zlib_init != ZLIB_INFLATING && zp->zlib_init != ZLIB_INIT_GZIP && zp->zlib_init != ZLIB_GZIP_INFLATING) return exit_zlib(data, z, &zp->zlib_init, CURLE_WRITE_ERROR); /* Dynamically allocate a buffer for decompression because it's uncommonly large to hold on the stack */ decomp = malloc(DSIZ); if(!decomp) return exit_zlib(data, z, &zp->zlib_init, CURLE_OUT_OF_MEMORY); /* because the buffer size is fixed, iteratively decompress and transfer to the client via downstream_write function. */ while(!done) { int status; /* zlib status */ done = TRUE; /* (re)set buffer for decompressed output for every iteration */ z->next_out = (Bytef *) decomp; z->avail_out = DSIZ; #ifdef Z_BLOCK /* Z_BLOCK is only available in zlib ver. >= 1.2.0.5 */ status = inflate(z, Z_BLOCK); #else /* fallback for zlib ver. < 1.2.0.5 */ status = inflate(z, Z_SYNC_FLUSH); #endif /* Flush output data if some. */ if(z->avail_out != DSIZ) { if(status == Z_OK || status == Z_STREAM_END) { zp->zlib_init = started; /* Data started. */ result = Curl_unencode_write(data, writer->downstream, decomp, DSIZ - z->avail_out); if(result) { exit_zlib(data, z, &zp->zlib_init, result); break; } } } /* Dispatch by inflate() status. */ switch(status) { case Z_OK: /* Always loop: there may be unflushed latched data in zlib state. */ done = FALSE; break; case Z_BUF_ERROR: /* No more data to flush: just exit loop. */ break; case Z_STREAM_END: result = process_trailer(data, zp); break; case Z_DATA_ERROR: /* some servers seem to not generate zlib headers, so this is an attempt to fix and continue anyway */ if(zp->zlib_init == ZLIB_INIT) { /* Do not use inflateReset2(): only available since zlib 1.2.3.4. */ (void) inflateEnd(z); /* don't care about the return code */ if(inflateInit2(z, -MAX_WBITS) == Z_OK) { z->next_in = orig_in; z->avail_in = nread; zp->zlib_init = ZLIB_INFLATING; zp->trailerlen = 4; /* Tolerate up to 4 unknown trailer bytes. */ done = FALSE; break; } zp->zlib_init = ZLIB_UNINIT; /* inflateEnd() already called. */ } result = exit_zlib(data, z, &zp->zlib_init, process_zlib_error(data, z)); break; default: result = exit_zlib(data, z, &zp->zlib_init, process_zlib_error(data, z)); break; } } free(decomp); /* We're about to leave this call so the `nread' data bytes won't be seen again. If we are in a state that would wrongly allow restart in raw mode at the next call, assume output has already started. */ if(nread && zp->zlib_init == ZLIB_INIT) zp->zlib_init = started; /* Cannot restart anymore. */ return result; } /* Deflate handler. */ static CURLcode deflate_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ if(!writer->downstream) return CURLE_WRITE_ERROR; /* Initialize zlib */ z->zalloc = (alloc_func) zalloc_cb; z->zfree = (free_func) zfree_cb; if(inflateInit(z) != Z_OK) return process_zlib_error(data, z); zp->zlib_init = ZLIB_INIT; return CURLE_OK; } static CURLcode deflate_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ /* Set the compressed input when this function is called */ z->next_in = (Bytef *) buf; z->avail_in = (uInt) nbytes; if(zp->zlib_init == ZLIB_EXTERNAL_TRAILER) return process_trailer(data, zp); /* Now uncompress the data */ return inflate_stream(data, writer, ZLIB_INFLATING); } static void deflate_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ exit_zlib(data, z, &zp->zlib_init, CURLE_OK); } static const struct content_encoding deflate_encoding = { "deflate", NULL, deflate_init_writer, deflate_unencode_write, deflate_close_writer, sizeof(struct zlib_writer) }; /* Gzip handler. */ static CURLcode gzip_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ if(!writer->downstream) return CURLE_WRITE_ERROR; /* Initialize zlib */ z->zalloc = (alloc_func) zalloc_cb; z->zfree = (free_func) zfree_cb; if(strcmp(zlibVersion(), "1.2.0.4") >= 0) { /* zlib ver. >= 1.2.0.4 supports transparent gzip decompressing */ if(inflateInit2(z, MAX_WBITS + 32) != Z_OK) { return process_zlib_error(data, z); } zp->zlib_init = ZLIB_INIT_GZIP; /* Transparent gzip decompress state */ } else { /* we must parse the gzip header and trailer ourselves */ if(inflateInit2(z, -MAX_WBITS) != Z_OK) { return process_zlib_error(data, z); } zp->trailerlen = 8; /* A CRC-32 and a 32-bit input size (RFC 1952, 2.2) */ zp->zlib_init = ZLIB_INIT; /* Initial call state */ } return CURLE_OK; } #ifdef OLD_ZLIB_SUPPORT /* Skip over the gzip header */ static enum { GZIP_OK, GZIP_BAD, GZIP_UNDERFLOW } check_gzip_header(unsigned char const *data, ssize_t len, ssize_t *headerlen) { int method, flags; const ssize_t totallen = len; /* The shortest header is 10 bytes */ if(len < 10) return GZIP_UNDERFLOW; if((data[0] != GZIP_MAGIC_0) || (data[1] != GZIP_MAGIC_1)) return GZIP_BAD; method = data[2]; flags = data[3]; if(method != Z_DEFLATED || (flags & RESERVED) != 0) { /* Can't handle this compression method or unknown flag */ return GZIP_BAD; } /* Skip over time, xflags, OS code and all previous bytes */ len -= 10; data += 10; if(flags & EXTRA_FIELD) { ssize_t extra_len; if(len < 2) return GZIP_UNDERFLOW; extra_len = (data[1] << 8) | data[0]; if(len < (extra_len + 2)) return GZIP_UNDERFLOW; len -= (extra_len + 2); data += (extra_len + 2); } if(flags & ORIG_NAME) { /* Skip over NUL-terminated file name */ while(len && *data) { --len; ++data; } if(!len || *data) return GZIP_UNDERFLOW; /* Skip over the NUL */ --len; ++data; } if(flags & COMMENT) { /* Skip over NUL-terminated comment */ while(len && *data) { --len; ++data; } if(!len || *data) return GZIP_UNDERFLOW; /* Skip over the NUL */ --len; } if(flags & HEAD_CRC) { if(len < 2) return GZIP_UNDERFLOW; len -= 2; } *headerlen = totallen - len; return GZIP_OK; } #endif static CURLcode gzip_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ if(zp->zlib_init == ZLIB_INIT_GZIP) { /* Let zlib handle the gzip decompression entirely */ z->next_in = (Bytef *) buf; z->avail_in = (uInt) nbytes; /* Now uncompress the data */ return inflate_stream(data, writer, ZLIB_INIT_GZIP); } #ifndef OLD_ZLIB_SUPPORT /* Support for old zlib versions is compiled away and we are running with an old version, so return an error. */ return exit_zlib(data, z, &zp->zlib_init, CURLE_WRITE_ERROR); #else /* This next mess is to get around the potential case where there isn't * enough data passed in to skip over the gzip header. If that happens, we * malloc a block and copy what we have then wait for the next call. If * there still isn't enough (this is definitely a worst-case scenario), we * make the block bigger, copy the next part in and keep waiting. * * This is only required with zlib versions < 1.2.0.4 as newer versions * can handle the gzip header themselves. */ switch(zp->zlib_init) { /* Skip over gzip header? */ case ZLIB_INIT: { /* Initial call state */ ssize_t hlen; switch(check_gzip_header((unsigned char *) buf, nbytes, &hlen)) { case GZIP_OK: z->next_in = (Bytef *) buf + hlen; z->avail_in = (uInt) (nbytes - hlen); zp->zlib_init = ZLIB_GZIP_INFLATING; /* Inflating stream state */ break; case GZIP_UNDERFLOW: /* We need more data so we can find the end of the gzip header. It's * possible that the memory block we malloc here will never be freed if * the transfer abruptly aborts after this point. Since it's unlikely * that circumstances will be right for this code path to be followed in * the first place, and it's even more unlikely for a transfer to fail * immediately afterwards, it should seldom be a problem. */ z->avail_in = (uInt) nbytes; z->next_in = malloc(z->avail_in); if(!z->next_in) { return exit_zlib(data, z, &zp->zlib_init, CURLE_OUT_OF_MEMORY); } memcpy(z->next_in, buf, z->avail_in); zp->zlib_init = ZLIB_GZIP_HEADER; /* Need more gzip header data state */ /* We don't have any data to inflate yet */ return CURLE_OK; case GZIP_BAD: default: return exit_zlib(data, z, &zp->zlib_init, process_zlib_error(data, z)); } } break; case ZLIB_GZIP_HEADER: { /* Need more gzip header data state */ ssize_t hlen; z->avail_in += (uInt) nbytes; z->next_in = Curl_saferealloc(z->next_in, z->avail_in); if(!z->next_in) { return exit_zlib(data, z, &zp->zlib_init, CURLE_OUT_OF_MEMORY); } /* Append the new block of data to the previous one */ memcpy(z->next_in + z->avail_in - nbytes, buf, nbytes); switch(check_gzip_header(z->next_in, z->avail_in, &hlen)) { case GZIP_OK: /* This is the zlib stream data */ free(z->next_in); /* Don't point into the malloced block since we just freed it */ z->next_in = (Bytef *) buf + hlen + nbytes - z->avail_in; z->avail_in = (uInt) (z->avail_in - hlen); zp->zlib_init = ZLIB_GZIP_INFLATING; /* Inflating stream state */ break; case GZIP_UNDERFLOW: /* We still don't have any data to inflate! */ return CURLE_OK; case GZIP_BAD: default: return exit_zlib(data, z, &zp->zlib_init, process_zlib_error(data, z)); } } break; case ZLIB_EXTERNAL_TRAILER: z->next_in = (Bytef *) buf; z->avail_in = (uInt) nbytes; return process_trailer(data, zp); case ZLIB_GZIP_INFLATING: default: /* Inflating stream state */ z->next_in = (Bytef *) buf; z->avail_in = (uInt) nbytes; break; } if(z->avail_in == 0) { /* We don't have any data to inflate; wait until next time */ return CURLE_OK; } /* We've parsed the header, now uncompress the data */ return inflate_stream(data, writer, ZLIB_GZIP_INFLATING); #endif } static void gzip_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ exit_zlib(data, z, &zp->zlib_init, CURLE_OK); } static const struct content_encoding gzip_encoding = { "gzip", "x-gzip", gzip_init_writer, gzip_unencode_write, gzip_close_writer, sizeof(struct zlib_writer) }; #endif /* HAVE_LIBZ */ #ifdef HAVE_BROTLI /* Brotli writer. */ struct brotli_writer { struct contenc_writer super; BrotliDecoderState *br; /* State structure for brotli. */ }; static CURLcode brotli_map_error(BrotliDecoderErrorCode be) { switch(be) { case BROTLI_DECODER_ERROR_FORMAT_EXUBERANT_NIBBLE: case BROTLI_DECODER_ERROR_FORMAT_EXUBERANT_META_NIBBLE: case BROTLI_DECODER_ERROR_FORMAT_SIMPLE_HUFFMAN_ALPHABET: case BROTLI_DECODER_ERROR_FORMAT_SIMPLE_HUFFMAN_SAME: case BROTLI_DECODER_ERROR_FORMAT_CL_SPACE: case BROTLI_DECODER_ERROR_FORMAT_HUFFMAN_SPACE: case BROTLI_DECODER_ERROR_FORMAT_CONTEXT_MAP_REPEAT: case BROTLI_DECODER_ERROR_FORMAT_BLOCK_LENGTH_1: case BROTLI_DECODER_ERROR_FORMAT_BLOCK_LENGTH_2: case BROTLI_DECODER_ERROR_FORMAT_TRANSFORM: case BROTLI_DECODER_ERROR_FORMAT_DICTIONARY: case BROTLI_DECODER_ERROR_FORMAT_WINDOW_BITS: case BROTLI_DECODER_ERROR_FORMAT_PADDING_1: case BROTLI_DECODER_ERROR_FORMAT_PADDING_2: #ifdef BROTLI_DECODER_ERROR_COMPOUND_DICTIONARY case BROTLI_DECODER_ERROR_COMPOUND_DICTIONARY: #endif #ifdef BROTLI_DECODER_ERROR_DICTIONARY_NOT_SET case BROTLI_DECODER_ERROR_DICTIONARY_NOT_SET: #endif case BROTLI_DECODER_ERROR_INVALID_ARGUMENTS: return CURLE_BAD_CONTENT_ENCODING; case BROTLI_DECODER_ERROR_ALLOC_CONTEXT_MODES: case BROTLI_DECODER_ERROR_ALLOC_TREE_GROUPS: case BROTLI_DECODER_ERROR_ALLOC_CONTEXT_MAP: case BROTLI_DECODER_ERROR_ALLOC_RING_BUFFER_1: case BROTLI_DECODER_ERROR_ALLOC_RING_BUFFER_2: case BROTLI_DECODER_ERROR_ALLOC_BLOCK_TYPE_TREES: return CURLE_OUT_OF_MEMORY; default: break; } return CURLE_WRITE_ERROR; } static CURLcode brotli_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct brotli_writer *bp = (struct brotli_writer *) writer; (void) data; if(!writer->downstream) return CURLE_WRITE_ERROR; bp->br = BrotliDecoderCreateInstance(NULL, NULL, NULL); return bp->br? CURLE_OK: CURLE_OUT_OF_MEMORY; } static CURLcode brotli_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { struct brotli_writer *bp = (struct brotli_writer *) writer; const uint8_t *src = (const uint8_t *) buf; char *decomp; uint8_t *dst; size_t dstleft; CURLcode result = CURLE_OK; BrotliDecoderResult r = BROTLI_DECODER_RESULT_NEEDS_MORE_OUTPUT; if(!bp->br) return CURLE_WRITE_ERROR; /* Stream already ended. */ decomp = malloc(DSIZ); if(!decomp) return CURLE_OUT_OF_MEMORY; while((nbytes || r == BROTLI_DECODER_RESULT_NEEDS_MORE_OUTPUT) && result == CURLE_OK) { dst = (uint8_t *) decomp; dstleft = DSIZ; r = BrotliDecoderDecompressStream(bp->br, &nbytes, &src, &dstleft, &dst, NULL); result = Curl_unencode_write(data, writer->downstream, decomp, DSIZ - dstleft); if(result) break; switch(r) { case BROTLI_DECODER_RESULT_NEEDS_MORE_OUTPUT: case BROTLI_DECODER_RESULT_NEEDS_MORE_INPUT: break; case BROTLI_DECODER_RESULT_SUCCESS: BrotliDecoderDestroyInstance(bp->br); bp->br = NULL; if(nbytes) result = CURLE_WRITE_ERROR; break; default: result = brotli_map_error(BrotliDecoderGetErrorCode(bp->br)); break; } } free(decomp); return result; } static void brotli_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct brotli_writer *bp = (struct brotli_writer *) writer; (void) data; if(bp->br) { BrotliDecoderDestroyInstance(bp->br); bp->br = NULL; } } static const struct content_encoding brotli_encoding = { "br", NULL, brotli_init_writer, brotli_unencode_write, brotli_close_writer, sizeof(struct brotli_writer) }; #endif #ifdef HAVE_ZSTD /* Zstd writer. */ struct zstd_writer { struct contenc_writer super; ZSTD_DStream *zds; /* State structure for zstd. */ void *decomp; }; static CURLcode zstd_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct zstd_writer *zp = (struct zstd_writer *) writer; (void)data; if(!writer->downstream) return CURLE_WRITE_ERROR; zp->zds = ZSTD_createDStream(); zp->decomp = NULL; return zp->zds ? CURLE_OK : CURLE_OUT_OF_MEMORY; } static CURLcode zstd_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { CURLcode result = CURLE_OK; struct zstd_writer *zp = (struct zstd_writer *) writer; ZSTD_inBuffer in; ZSTD_outBuffer out; size_t errorCode; if(!zp->decomp) { zp->decomp = malloc(DSIZ); if(!zp->decomp) return CURLE_OUT_OF_MEMORY; } in.pos = 0; in.src = buf; in.size = nbytes; for(;;) { out.pos = 0; out.dst = zp->decomp; out.size = DSIZ; errorCode = ZSTD_decompressStream(zp->zds, &out, &in); if(ZSTD_isError(errorCode)) { return CURLE_BAD_CONTENT_ENCODING; } if(out.pos > 0) { result = Curl_unencode_write(data, writer->downstream, zp->decomp, out.pos); if(result) break; } if((in.pos == nbytes) && (out.pos < out.size)) break; } return result; } static void zstd_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { struct zstd_writer *zp = (struct zstd_writer *) writer; (void)data; if(zp->decomp) { free(zp->decomp); zp->decomp = NULL; } if(zp->zds) { ZSTD_freeDStream(zp->zds); zp->zds = NULL; } } static const struct content_encoding zstd_encoding = { "zstd", NULL, zstd_init_writer, zstd_unencode_write, zstd_close_writer, sizeof(struct zstd_writer) }; #endif /* Identity handler. */ static CURLcode identity_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { (void) data; return writer->downstream? CURLE_OK: CURLE_WRITE_ERROR; } static CURLcode identity_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { return Curl_unencode_write(data, writer->downstream, buf, nbytes); } static void identity_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { (void) data; (void) writer; } static const struct content_encoding identity_encoding = { "identity", "none", identity_init_writer, identity_unencode_write, identity_close_writer, sizeof(struct contenc_writer) }; /* supported content encodings table. */ static const struct content_encoding * const encodings[] = { &identity_encoding, #ifdef HAVE_LIBZ &deflate_encoding, &gzip_encoding, #endif #ifdef HAVE_BROTLI &brotli_encoding, #endif #ifdef HAVE_ZSTD &zstd_encoding, #endif NULL }; /* Return a list of comma-separated names of supported encodings. */ char *Curl_all_content_encodings(void) { size_t len = 0; const struct content_encoding * const *cep; const struct content_encoding *ce; char *ace; for(cep = encodings; *cep; cep++) { ce = *cep; if(!strcasecompare(ce->name, CONTENT_ENCODING_DEFAULT)) len += strlen(ce->name) + 2; } if(!len) return strdup(CONTENT_ENCODING_DEFAULT); ace = malloc(len); if(ace) { char *p = ace; for(cep = encodings; *cep; cep++) { ce = *cep; if(!strcasecompare(ce->name, CONTENT_ENCODING_DEFAULT)) { strcpy(p, ce->name); p += strlen(p); *p++ = ','; *p++ = ' '; } } p[-2] = '\0'; } return ace; } /* Real client writer: no downstream. */ static CURLcode client_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { (void) data; return writer->downstream? CURLE_WRITE_ERROR: CURLE_OK; } static CURLcode client_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { struct SingleRequest *k = &data->req; (void) writer; if(!nbytes || k->ignorebody) return CURLE_OK; return Curl_client_write(data, CLIENTWRITE_BODY, (char *) buf, nbytes); } static void client_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { (void) data; (void) writer; } static const struct content_encoding client_encoding = { NULL, NULL, client_init_writer, client_unencode_write, client_close_writer, sizeof(struct contenc_writer) }; /* Deferred error dummy writer. */ static CURLcode error_init_writer(struct Curl_easy *data, struct contenc_writer *writer) { (void) data; return writer->downstream? CURLE_OK: CURLE_WRITE_ERROR; } static CURLcode error_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { char *all = Curl_all_content_encodings(); (void) writer; (void) buf; (void) nbytes; if(!all) return CURLE_OUT_OF_MEMORY; failf(data, "Unrecognized content encoding type. " "libcurl understands %s content encodings.", all); free(all); return CURLE_BAD_CONTENT_ENCODING; } static void error_close_writer(struct Curl_easy *data, struct contenc_writer *writer) { (void) data; (void) writer; } static const struct content_encoding error_encoding = { NULL, NULL, error_init_writer, error_unencode_write, error_close_writer, sizeof(struct contenc_writer) }; /* Create an unencoding writer stage using the given handler. */ static struct contenc_writer * new_unencoding_writer(struct Curl_easy *data, const struct content_encoding *handler, struct contenc_writer *downstream, int order) { struct contenc_writer *writer; DEBUGASSERT(handler->writersize >= sizeof(struct contenc_writer)); writer = (struct contenc_writer *) calloc(1, handler->writersize); if(writer) { writer->handler = handler; writer->downstream = downstream; writer->order = order; if(handler->init_writer(data, writer)) { free(writer); writer = NULL; } } return writer; } /* Write data using an unencoding writer stack. "nbytes" is not allowed to be 0. */ CURLcode Curl_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { if(!nbytes) return CURLE_OK; return writer->handler->unencode_write(data, writer, buf, nbytes); } /* Close and clean-up the connection's writer stack. */ void Curl_unencode_cleanup(struct Curl_easy *data) { struct SingleRequest *k = &data->req; struct contenc_writer *writer = k->writer_stack; while(writer) { k->writer_stack = writer->downstream; writer->handler->close_writer(data, writer); free(writer); writer = k->writer_stack; } } /* Find the content encoding by name. */ static const struct content_encoding *find_encoding(const char *name, size_t len) { const struct content_encoding * const *cep; for(cep = encodings; *cep; cep++) { const struct content_encoding *ce = *cep; if((strncasecompare(name, ce->name, len) && !ce->name[len]) || (ce->alias && strncasecompare(name, ce->alias, len) && !ce->alias[len])) return ce; } return NULL; } /* allow no more than 5 "chained" compression steps */ #define MAX_ENCODE_STACK 5 /* Set-up the unencoding stack from the Content-Encoding header value. * See RFC 7231 section 3.1.2.2. */ CURLcode Curl_build_unencoding_stack(struct Curl_easy *data, const char *enclist, int is_transfer) { struct SingleRequest *k = &data->req; unsigned int order = is_transfer? 2: 1; do { const char *name; size_t namelen; /* Parse a single encoding name. */ while(ISBLANK(*enclist) || *enclist == ',') enclist++; name = enclist; for(namelen = 0; *enclist && *enclist != ','; enclist++) if(!ISSPACE(*enclist)) namelen = enclist - name + 1; /* Special case: chunked encoding is handled at the reader level. */ if(is_transfer && namelen == 7 && strncasecompare(name, "chunked", 7)) { k->chunk = TRUE; /* chunks coming our way. */ Curl_httpchunk_init(data); /* init our chunky engine. */ } else if(namelen) { const struct content_encoding *encoding = find_encoding(name, namelen); struct contenc_writer *writer; if(!k->writer_stack) { k->writer_stack = new_unencoding_writer(data, &client_encoding, NULL, 0); if(!k->writer_stack) return CURLE_OUT_OF_MEMORY; } if(!encoding) encoding = &error_encoding; /* Defer error at stack use. */ if(k->writer_stack_depth++ >= MAX_ENCODE_STACK) { failf(data, "Reject response due to more than %u content encodings", MAX_ENCODE_STACK); return CURLE_BAD_CONTENT_ENCODING; } /* Stack the unencoding stage. */ if(order >= k->writer_stack->order) { writer = new_unencoding_writer(data, encoding, k->writer_stack, order); if(!writer) return CURLE_OUT_OF_MEMORY; k->writer_stack = writer; } else { struct contenc_writer *w = k->writer_stack; while(w->downstream && order < w->downstream->order) w = w->downstream; writer = new_unencoding_writer(data, encoding, w->downstream, order); if(!writer) return CURLE_OUT_OF_MEMORY; w->downstream = writer; } } } while(*enclist); return CURLE_OK; } #else /* Stubs for builds without HTTP. */ CURLcode Curl_build_unencoding_stack(struct Curl_easy *data, const char *enclist, int is_transfer) { (void) data; (void) enclist; (void) is_transfer; return CURLE_NOT_BUILT_IN; } CURLcode Curl_unencode_write(struct Curl_easy *data, struct contenc_writer *writer, const char *buf, size_t nbytes) { (void) data; (void) writer; (void) buf; (void) nbytes; return CURLE_NOT_BUILT_IN; } void Curl_unencode_cleanup(struct Curl_easy *data) { (void) data; } char *Curl_all_content_encodings(void) { return strdup(CONTENT_ENCODING_DEFAULT); /* Satisfy caller. */ } #endif /* CURL_DISABLE_HTTP */ 上边的是patch 下边的是源码： From 76f83f0db23846e254d940ec7fe141010077eb88 Mon Sep 17 00:00:00 2001 From: Daniel Stenberg <daniel@haxx.se> Date: Fri, 24 Jan 2025 11:13:24 +0100 Subject: [PATCH] content_encoding: drop support for zlib before 1.2.0.4 zlib 1.2.0.4 was released on 10 August 2003 Closes #16079 --- docs/INTERNALS.md | 2 +- lib/content_encoding.c | 276 ++++------------------------------------- 2 files changed, 25 insertions(+), 253 deletions(-) diff --git a/docs/INTERNALS.md b/docs/INTERNALS.md index ae77f0e54b05..4e42f4fd1015 100644 --- a/docs/INTERNALS.md +++ b/docs/INTERNALS.md @@ -23,7 +23,7 @@ versions of libs and build tools. - OpenSSL 0.9.7 - GnuTLS 3.1.10 - - zlib 1.1.4 + - zlib 1.2.0.4 - libssh2 1.0 - c-ares 1.16.0 - libidn2 2.0.0 diff --git a/lib/content_encoding.c b/lib/content_encoding.c index e19595d5ec42..d2b17297890d 100644 --- a/lib/content_encoding.c +++ b/lib/content_encoding.c @@ -55,33 +55,13 @@ #define DSIZ CURL_MAX_WRITE_SIZE /* buffer size for decompressed data */ - #ifdef HAVE_LIBZ -/* Comment this out if zlib is always going to be at least ver. 1.2.0.4 - (doing so will reduce code size slightly). */ -#define OLD_ZLIB_SUPPORT 1 - -#define GZIP_MAGIC_0 0x1f -#define GZIP_MAGIC_1 0x8b - -/* gzip flag byte */ -#define CURL_GZIPFLAG_ASCII 0x01 /* bit 0 set: file probably ASCII - text */ -#define CURL_GZIPFLAG_HEAD_CRC 0x02 /* bit 1 set: header CRC present */ -#define CURL_GZIPFLAG_EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ -#define CURL_GZIPFLAG_ORIG_NAME 0x08 /* bit 3 set: original filename - present */ -#define CURL_GZIPFLAG_COMMENT 0x10 /* bit 4 set: file comment present */ -#define CURL_GZIPFLAG_RESERVED 0xE0 /* bits 5..7: reserved */ - typedef enum { ZLIB_UNINIT, /* uninitialized */ ZLIB_INIT, /* initialized */ ZLIB_INFLATING, /* inflating started. */ ZLIB_EXTERNAL_TRAILER, /* reading external trailer */ - ZLIB_GZIP_HEADER, /* reading gzip header */ - ZLIB_GZIP_INFLATING, /* inflating gzip stream */ ZLIB_INIT_GZIP /* initialized in transparent gzip mode */ } zlibInitState; @@ -139,9 +119,6 @@ static CURLcode exit_zlib(struct Curl_easy *data, z_stream *z, zlibInitState *zlib_init, CURLcode result) { - if(*zlib_init == ZLIB_GZIP_HEADER) - Curl_safefree(z->next_in); - if(*zlib_init != ZLIB_UNINIT) { if(inflateEnd(z) != Z_OK && result == CURLE_OK) result = process_zlib_error(data, z); @@ -190,8 +167,7 @@ static CURLcode inflate_stream(struct Curl_easy *data, /* Check state. */ if(zp->zlib_init != ZLIB_INIT && zp->zlib_init != ZLIB_INFLATING && - zp->zlib_init != ZLIB_INIT_GZIP && - zp->zlib_init != ZLIB_GZIP_INFLATING) + zp->zlib_init != ZLIB_INIT_GZIP) return exit_zlib(data, z, &zp->zlib_init, CURLE_WRITE_ERROR); /* Dynamically allocate a buffer for decompression because it is uncommonly @@ -280,7 +256,7 @@ static CURLcode inflate_stream(struct Curl_easy *data, /* Deflate handler. */ static CURLcode deflate_do_init(struct Curl_easy *data, - struct Curl_cwriter *writer) + struct Curl_cwriter *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ @@ -296,8 +272,8 @@ static CURLcode deflate_do_init(struct Curl_easy *data, } static CURLcode deflate_do_write(struct Curl_easy *data, - struct Curl_cwriter *writer, int type, - const char *buf, size_t nbytes) + struct Curl_cwriter *writer, int type, + const char *buf, size_t nbytes) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ @@ -317,7 +293,7 @@ static CURLcode deflate_do_write(struct Curl_easy *data, } static void deflate_do_close(struct Curl_easy *data, - struct Curl_cwriter *writer) + struct Curl_cwriter *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ @@ -337,124 +313,34 @@ static const struct Curl_cwtype deflate_encoding = { /* Gzip handler. */ static CURLcode gzip_do_init(struct Curl_easy *data, - struct Curl_cwriter *writer) + struct Curl_cwriter *writer) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ + const char *v = zlibVersion(); /* Initialize zlib */ z->zalloc = (alloc_func) zalloc_cb; z->zfree = (free_func) zfree_cb; - if(strcmp(zlibVersion(), "1.2.0.4") >= 0) { - /* zlib ver. >= 1.2.0.4 supports transparent gzip decompressing */ + if(strcmp(v, "1.2.0.4") >= 0) { + /* zlib version >= 1.2.0.4 supports transparent gzip decompressing */ if(inflateInit2(z, MAX_WBITS + 32) != Z_OK) { return process_zlib_error(data, z); } zp->zlib_init = ZLIB_INIT_GZIP; /* Transparent gzip decompress state */ } else { - /* we must parse the gzip header and trailer ourselves */ - if(inflateInit2(z, -MAX_WBITS) != Z_OK) { - return process_zlib_error(data, z); - } - zp->trailerlen = 8; /* A CRC-32 and a 32-bit input size (RFC 1952, 2.2) */ - zp->zlib_init = ZLIB_INIT; /* Initial call state */ + failf(data, "too old zlib version: %s", v); + return CURLE_FAILED_INIT; } return CURLE_OK; } -#ifdef OLD_ZLIB_SUPPORT -/* Skip over the gzip header */ -typedef enum { - GZIP_OK, - GZIP_BAD, - GZIP_UNDERFLOW -} gzip_status; - -static gzip_status check_gzip_header(unsigned char const *data, ssize_t len, - ssize_t *headerlen) -{ - int method, flags; - const ssize_t totallen = len; - - /* The shortest header is 10 bytes */ - if(len < 10) - return GZIP_UNDERFLOW; - - if((data[0] != GZIP_MAGIC_0) || (data[1] != GZIP_MAGIC_1)) - return GZIP_BAD; - - method = data[2]; - flags = data[3]; - - if(method != Z_DEFLATED || (flags & CURL_GZIPFLAG_RESERVED) != 0) { - /* cannot handle this compression method or unknown flag */ - return GZIP_BAD; - } - - /* Skip over time, xflags, OS code and all previous bytes */ - len -= 10; - data += 10; - - if(flags & CURL_GZIPFLAG_EXTRA_FIELD) { - ssize_t extra_len; - - if(len < 2) - return GZIP_UNDERFLOW; - - extra_len = (data[1] << 8) | data[0]; - - if(len < (extra_len + 2)) - return GZIP_UNDERFLOW; - - len -= (extra_len + 2); - data += (extra_len + 2); - } - - if(flags & CURL_GZIPFLAG_ORIG_NAME) { - /* Skip over NUL-terminated filename */ - while(len && *data) { - --len; - ++data; - } - if(!len || *data) - return GZIP_UNDERFLOW; - - /* Skip over the NUL */ - --len; - ++data; - } - - if(flags & CURL_GZIPFLAG_COMMENT) { - /* Skip over NUL-terminated comment */ - while(len && *data) { - --len; - ++data; - } - if(!len || *data) - return GZIP_UNDERFLOW; - - /* Skip over the NUL */ - --len; - } - - if(flags & CURL_GZIPFLAG_HEAD_CRC) { - if(len < 2) - return GZIP_UNDERFLOW; - - len -= 2; - } - - *headerlen = totallen - len; - return GZIP_OK; -} -#endif - static CURLcode gzip_do_write(struct Curl_easy *data, - struct Curl_cwriter *writer, int type, - const char *buf, size_t nbytes) + struct Curl_cwriter *writer, int type, + const char *buf, size_t nbytes) { struct zlib_writer *zp = (struct zlib_writer *) writer; z_stream *z = &zp->z; /* zlib state structure */ @@ -470,117 +356,8 @@ static CURLcode gzip_do_write(struct Curl_easy *data, return inflate_stream(data, writer, type, ZLIB_INIT_GZIP); } -#ifndef OLD_ZLIB_SUPPORT - /* Support for old zlib versions is compiled away and we are running with - an old version, so return an error. */ + /* We are running with an old version: return error. */ return exit_zlib(data, z, &zp->zlib_init, CURLE_WRITE_ERROR); - -#else - /* This next mess is to get around the potential case where there is not - * enough data passed in to skip over the gzip header. If that happens, we - * malloc a block and copy what we have then wait for the next call. If - * there still is not enough (this is definitely a worst-case scenario), we - * make the block bigger, copy the next part in and keep waiting. - * - * This is only required with zlib versions < 1.2.0.4 as newer versions - * can handle the gzip header themselves. - */ - - switch(zp->zlib_init) { - /* Skip over gzip header? */ - case ZLIB_INIT: - { - /* Initial call state */ - ssize_t hlen; - - switch(check_gzip_header((unsigned char *) buf, nbytes, &hlen)) { - case GZIP_OK: - z->next_in = (Bytef *) buf + hlen; - z->avail_in = (uInt) (nbytes - hlen); - zp->zlib_init = ZLIB_GZIP_INFLATING; /* Inflating stream state */ - break; - - case GZIP_UNDERFLOW: - /* We need more data so we can find the end of the gzip header. it is - * possible that the memory block we malloc here will never be freed if - * the transfer abruptly aborts after this point. Since it is unlikely - * that circumstances will be right for this code path to be followed in - * the first place, and it is even more unlikely for a transfer to fail - * immediately afterwards, it should seldom be a problem. - */ - z->avail_in = (uInt) nbytes; - z->next_in = malloc(z->avail_in); - if(!z->next_in) { - return exit_zlib(data, z, &zp->zlib_init, CURLE_OUT_OF_MEMORY); - } - memcpy(z->next_in, buf, z->avail_in); - zp->zlib_init = ZLIB_GZIP_HEADER; /* Need more gzip header data state */ - /* We do not have any data to inflate yet */ - return CURLE_OK; - - case GZIP_BAD: - default: - return exit_zlib(data, z, &zp->zlib_init, process_zlib_error(data, z)); - } - - } - break; - - case ZLIB_GZIP_HEADER: - { - /* Need more gzip header data state */ - ssize_t hlen; - z->avail_in += (uInt) nbytes; - z->next_in = Curl_saferealloc(z->next_in, z->avail_in); - if(!z->next_in) { - return exit_zlib(data, z, &zp->zlib_init, CURLE_OUT_OF_MEMORY); - } - /* Append the new block of data to the previous one */ - memcpy(z->next_in + z->avail_in - nbytes, buf, nbytes); - - switch(check_gzip_header(z->next_in, (ssize_t)z->avail_in, &hlen)) { - case GZIP_OK: - /* This is the zlib stream data */ - free(z->next_in); - /* Do not point into the malloced block since we just freed it */ - z->next_in = (Bytef *) buf + hlen + nbytes - z->avail_in; - z->avail_in = z->avail_in - (uInt)hlen; - zp->zlib_init = ZLIB_GZIP_INFLATING; /* Inflating stream state */ - break; - - case GZIP_UNDERFLOW: - /* We still do not have any data to inflate! */ - return CURLE_OK; - - case GZIP_BAD: - default: - return exit_zlib(data, z, &zp->zlib_init, process_zlib_error(data, z)); - } - - } - break; - - case ZLIB_EXTERNAL_TRAILER: - z->next_in = (Bytef *) buf; - z->avail_in = (uInt) nbytes; - return process_trailer(data, zp); - - case ZLIB_GZIP_INFLATING: - default: - /* Inflating stream state */ - z->next_in = (Bytef *) buf; - z->avail_in = (uInt) nbytes; - break; - } - - if(z->avail_in == 0) { - /* We do not have any data to inflate; wait until next time */ - return CURLE_OK; - } - - /* We have parsed the header, now uncompress the data */ - return inflate_stream(data, writer, type, ZLIB_GZIP_INFLATING); -#endif } static void gzip_do_close(struct Curl_easy *data, @@ -603,7 +380,6 @@ static const struct Curl_cwtype gzip_encoding = { #endif /* HAVE_LIBZ */ - #ifdef HAVE_BROTLI /* Brotli writer. */ struct brotli_writer { @@ -650,7 +426,7 @@ static CURLcode brotli_map_error(BrotliDecoderErrorCode be) } static CURLcode brotli_do_init(struct Curl_easy *data, - struct Curl_cwriter *writer) + struct Curl_cwriter *writer) { struct brotli_writer *bp = (struct brotli_writer *) writer; (void) data; @@ -660,8 +436,8 @@ static CURLcode brotli_do_init(struct Curl_easy *data, } static CURLcode brotli_do_write(struct Curl_easy *data, - struct Curl_cwriter *writer, int type, - const char *buf, size_t nbytes) + struct Curl_cwriter *writer, int type, + const char *buf, size_t nbytes) { struct brotli_writer *bp = (struct brotli_writer *) writer; const uint8_t *src = (const uint8_t *) buf; @@ -733,7 +509,6 @@ static const struct Curl_cwtype brotli_encoding = { }; #endif - #ifdef HAVE_ZSTD /* Zstd writer. */ struct zstd_writer { @@ -757,7 +532,7 @@ static void Curl_zstd_free(void *opaque, void *address) #endif static CURLcode zstd_do_init(struct Curl_easy *data, - struct Curl_cwriter *writer) + struct Curl_cwriter *writer) { struct zstd_writer *zp = (struct zstd_writer *) writer; @@ -778,8 +553,8 @@ static CURLcode zstd_do_init(struct Curl_easy *data, } static CURLcode zstd_do_write(struct Curl_easy *data, - struct Curl_cwriter *writer, int type, - const char *buf, size_t nbytes) + struct Curl_cwriter *writer, int type, + const char *buf, size_t nbytes) { CURLcode result = CURLE_OK; struct zstd_writer *zp = (struct zstd_writer *) writer; @@ -810,7 +585,7 @@ static CURLcode zstd_do_write(struct Curl_easy *data, } if(out.pos > 0) { result = Curl_cwriter_write(data, writer->next, type, - zp->decomp, out.pos); + zp->decomp, out.pos); if(result) break; } @@ -848,7 +623,6 @@ static const struct Curl_cwtype zstd_encoding = { }; #endif - /* Identity handler. */ static const struct Curl_cwtype identity_encoding = { "identity", @@ -859,7 +633,6 @@ static const struct Curl_cwtype identity_encoding = { sizeof(struct Curl_cwriter) }; - /* supported general content decoders. */ static const struct Curl_cwtype * const general_unencoders[] = { &identity_encoding, @@ -923,7 +696,7 @@ void Curl_all_content_encodings(char *buf, size_t blen) /* Deferred error dummy writer. */ static CURLcode error_do_init(struct Curl_easy *data, - struct Curl_cwriter *writer) + struct Curl_cwriter *writer) { (void)data; (void)writer; @@ -931,8 +704,8 @@ static CURLcode error_do_init(struct Curl_easy *data, } static CURLcode error_do_write(struct Curl_easy *data, - struct Curl_cwriter *writer, int type, - const char *buf, size_t nbytes) + struct Curl_cwriter *writer, int type, + const char *buf, size_t nbytes) { (void) writer; (void) buf; @@ -1107,5 +880,4 @@ void Curl_all_content_encodings(char *buf, size_t blen) strcpy(buf, CONTENT_ENCODING_DEFAULT); } - #endif /* CURL_DISABLE_HTTP */ 现在就是想做的就是根据源码相应的函数的行数来修改patch中要修复的行数，不修改源码
09-12
@@ -337,126 +314,36 @@ static const struct content_encoding deflate_encoding = { /* Gzip handler. */ static CURLcode gzip_init_writer(struct Curl_easy *data, - struct contenc_writer *writer) + ...
# rate_set/rate_sync.py import json import os import re import logging import sys from pathlib import Path from utils import resource_path from datetime import datetime from typing import Dict, List, Tuple, Any # ------------------------------- # 日志配置 # ------------------------------- PROJECT_ROOT = Path(__file__).parent.parent.resolve() LOG_DIR = PROJECT_ROOT / "output" / "log" LOG_DIR.mkdir(parents=True, exist_ok=True) LOG_FILE = LOG_DIR / f"rate_sync_{datetime.now().strftime('%Y%m%d_%H%M%S')}.log" class RateSetSynchronizer: MAX_ENUM_PER_LINE = 4 # enum 每行最多几个宏 MAX_DATA_ITEMS_PER_LINE = 4 # data 数组每行最多几个值 MAX_INDEX_ITEMS_PER_LINE = 15 # index 数组每行最多几个值 def __init__(self, c_file_path=None, dry_run=False, config_path="config/config.json"): self.logger = logging.getLogger(f"{__name__}.RateSetSynchronizer") # 加载配置 self.config_file_path = resource_path(config_path) if not os.path.exists(self.config_file_path): raise FileNotFoundError(f"配置文件不存在: {self.config_file_path}") with open(self.config_file_path, 'r', encoding='utf-8') as f: self.config = json.load(f) self.dry_run = dry_run # C 文件路径 if c_file_path is None: internal_c_path = self.config["target_c_file"] self.c_file_path = resource_path(internal_c_path) else: self.c_file_path = Path(c_file_path) if not self.c_file_path.exists(): raise FileNotFoundError(f"找不到 C 源文件: {self.c_file_path}") # === 单一锚点标记 === self.block_start = self.config["STR_RATE_SET_DATA"] self.block_end = self.config["END_RATE_SET_DATA"] # 数组与枚举名 self.data_array_name = "rate_sets_2g_20m" self.index_array_name = "rate_sets_index_2g_20m" self.enum_name = "rate_set_2g_20m" # 扫描所有子目录中的 .c 文件（排除自身） self.rate_set_dir = Path(__file__).parent self.rate_files = [ f for f in self.rate_set_dir.rglob("*.c") # 递归匹配所有 .c 文件 if f.is_file() and f.name != "rate_sync.py" ] # 加载文件名和结构映射 self.target_map = self.config.get("rate_set_map") if not isinstance(self.target_map, dict): raise ValueError("config.json 中缺少 'rate_set_map' 字段或格式错误") self._validate_target_map() # ← 添加一致性校验 def _validate_target_map(self): """验证 rate_set_map 是否一致，防止多个 full_key 映射到同一数组""" seen_data = {} seen_index = {} seen_enum = {} for key, cfg in self.target_map.items(): d = cfg["data"] i = cfg["index"] e = cfg["enum"] if d in seen_data: raise ValueError(f"data 数组冲突: '{d}' 被 '{seen_data[d]}' 和 '{key}' 同时使用") if i in seen_index: raise ValueError(f"index 数组冲突: '{i}' 被 '{seen_index[i]}' 和 '{key}' 同时使用") if e in seen_enum: raise ValueError(f"enum 名称冲突: '{e}' 被 '{seen_enum[e]}' 和 '{key}' 同时使用") seen_data[d] = key seen_index[i] = key seen_enum[e] = key def parse_filename(self, filename: str) -> str: """ 从文件名提取 band_bw_ext 类型键，用于查找 rate_set_map 示例： 2G_20M_rate_set.c → 2G_20M_BASE 2G_20M_EXT_rate_set.c → 2G_20M_EXT 5G_80M_EXT4_rate_set.c → 5G_80M_EXT4 """ match = re.match(r'^([A-Z0-9]+)_([0-9]+M)(?:_(EXT\d*))?_rate_set\.c$', filename, re.I) if not match: raise ValueError(f"无法识别的文件名格式: {filename}") band, bw, ext = match.groups() ext_type = ext.upper() if ext else "BASE" return f"{band.upper()}_{bw.upper()}_{ext_type}" def extract_sub_rate_sets(self, content: str) -> List[Dict[str, Any]]: """ 提取 /*NAME*/ N, 后续多行 WL_RATE_xxx 列表 支持跨行、缩进、逗号、空行、注释干扰等 使用“按行扫描 + 状态机”方式，避免正则越界 """ self.logger.info("开始提取速率集...") self.logger.info("...") sub_sets = [] lines = [line.rstrip('\r\n') for line in content.splitlines()] # 保留原始行尾 i = 0 # 匹配 /*NAME*/ N, 的开头 header_pattern = re.compile(r'/\*\s*([A-Za-z0-9_]+)\s*\*/\s*(\d+)\s*,?') while i < len(lines): line = lines[i] stripped = line.strip() # 跳过空行和纯注释 if not stripped or stripped.startswith("//"): i += 1 continue # 查找头: /*NAME*/ N, match = header_pattern.search(stripped) if not match: i += 1 continue name = match.group(1) try: count = int(match.group(2)) except ValueError: self.logger.warning(f"⚠️ 计数无效，跳过: {name} = '{match.group(2)}'") i += 1 continue self.logger.info(f"🔍 发现块: {name}, 预期数量={count}") # 开始收集 body 内容（保留原始带缩进的行） body_lines = [] j = i + 1 max_lines_to_read = 200 while j < len(lines) and len(body_lines) < max_lines_to_read: ln = lines[j].strip() # 终止条件：遇到新 block / 结构结束 if ln.startswith("/*") or ln.startswith("}") or ln.startswith("enum"): break if ln and not ln.startswith("//"): body_lines.append(lines[j]) # ← 原样保存（含缩进） else: body_lines.append(lines[j]) # 也保留注释或空行（保持格式一致） j += 1 # 提取宏名用于校验（但不再用于生成数据） body_text = "\n".join(body_lines) all_macros = re.findall(r'WL_RATE_[A-Za-z0-9_]+', body_text) rate_items = all_macros[:count] if len(rate_items) < count: self.logger.warning(f"[{name}] 条目不足: 需要 {count}, 实际 {len(rate_items)}") # 构建结果：增加 raw_header 和 raw_body（关键改动） sub_sets.append({ "name": name, "count": count, "rates": rate_items, "raw_header": line, # ← 原始头行（如 /*...*/ 4,） "raw_body": body_lines, # ← 原始 body 行列表 "start_line": i, "end_line": j - 1 }) self.logger.debug(f"✅ 提取成功: {name} → {len(rate_items)} 个速率") i = j # 跳到下一个 block self.logger.info(f" 共提取 {len(sub_sets)} 个有效子集") return sub_sets def parse_all_structures(self, full_content: str) -> Dict: """ 直接从完整 C 文件中解析 enum/data/index 结构 """ self.logger.info("开始解析所有结构...") self.logger.info("...") result = { 'existing_enum': {}, 'data_entries': [], 'index_values': [], 'data_len': 0 } # === 解析 enum === enum_pattern = rf'enum\s+{re.escape(self.enum_name)}\s*\{{([^}}]+)\}};' enum_match = re.search(enum_pattern, full_content, re.DOTALL) if enum_match: body = enum_match.group(1) entries = re.findall(r'(RATE_SET_[^=,\s]+)\s*=\s*(\d+)', body) result['existing_enum'] = {k: int(v) for k, v in entries} self.logger.info(f"解析出 {len(entries)} 个已有枚举项") else: self.logger.warning(f"未找到 enum 定义: {self.enum_name}") # === 解析 data 数组 === data_pattern = rf'static const unsigned char {re.escape(self.data_array_name)}\[\] = \{{([^}}]+)\}};' data_match = re.search(data_pattern, full_content, re.DOTALL) if not data_match: raise ValueError(f"未找到 data 数组: {self.data_array_name}") data_code = data_match.group(1) result['data_entries'] = [item.strip() for item in re.split(r'[,\n]+', data_code) if item.strip()] result['data_len'] = len(result['data_entries']) # === 解析 index 数组 === index_pattern = rf'static const unsigned short {re.escape(self.index_array_name)}\[\] = \{{([^}}]+)\}};' index_match = re.search(index_pattern, full_content, re.DOTALL) if not index_match: raise ValueError(f"未找到 index 数组: {self.index_array_name}") index_code = index_match.group(1) result['index_values'] = [int(x.strip()) for x in re.split(r'[,\n]+', index_code) if x.strip()] return result def build_injection_with_format(self, new_subsets: List[Dict], existing_enum: Dict[str, int]) -> List[Dict]: """ 返回需要注入的原始块列表（包含 raw_header + raw_body） 不再返回 new_data/new_indices/new_enums """ valid_blocks = [] next_enum_value = max(existing_enum.values(), default=-1) + 1 self.logger.info(f"开始构建注入内容，当前最大枚举值 = {next_enum_value}") for subset in new_subsets: enum_name = subset["name"] if enum_name in existing_enum: self.logger.info(f"跳过已存在的枚举项: {enum_name} = {existing_enum[enum_name]}") continue # 只保存必要信息，不计算偏移 valid_blocks.append({ "enum_name": enum_name, "raw_header": subset["raw_header"], "raw_body": subset["raw_body"], "count": subset["count"], # 用于计算 data 占用空间 "enum_value": next_enum_value, # ✅ 必须存在！ }) self.logger.debug(f" 准备注入: {enum_name}") self.logger.info(f"新增条目: {enum_name} enum={next_enum_value}") next_enum_value += 1 self.logger.info(f"构建完成：共 {len(valid_blocks)} 个新条目（保留原始格式）") return valid_blocks def format_list(self, items: List[str], indent: str = " ", width: int = 8) -> str: """格式化数组为多行字符串""" lines = [] for i in range(0, len(items), width): chunk = items[i:i + width] lines.append(indent + ", ".join(chunk) + ",") return "\n".join(lines).rstrip(",") def _safe_write_back(self, old_content: str, new_content: str) -> bool: """安全写回文件，带备份""" if old_content == new_content: self.logger.info("主文件内容无变化，无需写入") return False if self.dry_run: self.logger.info("DRY-RUN 模式启用，跳过实际写入") print("[DRY RUN] 差异预览（前 20 行）:") diff = new_content.splitlines()[:20] for line in diff: print(f" {line}") return True # 创建备份 timestamp = datetime.now().strftime("%Y%m%d_%H%M%S") backup = self.c_file_path.with_name(f"{self.c_file_path.stem}_{timestamp}.c.bak") try: self.c_file_path.rename(backup) self.logger.info(f"原文件已备份为: {backup.name}") except Exception as e: self.logger.error(f"备份失败: {e}") raise # 写入新内容 try: self.c_file_path.write_text(new_content, encoding='utf-8') self.logger.info(f"✅ 成功写入更新后的文件: {self.c_file_path.name}") return True except Exception as e: self.logger.error(f"写入失败: {e}", exc_info=True) raise def inject_new_data(self) -> bool: try: full_content = self.c_file_path.read_text(encoding='utf-8') except Exception as e: self.logger.error(f"读取主 C 文件失败: {e}") raise self.logger.info(f"正在处理 C 文件: {self.c_file_path.name}") start_pos = full_content.find(self.block_start) end_pos = full_content.find(self.block_end) if start_pos == -1: raise ValueError(f"未找到起始锚点: {self.block_start}") if end_pos == -1: raise ValueError(f"未找到结束锚点: {self.block_end}") if end_pos <= start_pos: raise ValueError("结束锚点位于起始锚点之前") inner_start = start_pos + len(self.block_start) block_content = full_content[inner_start:end_pos].strip() all_changes_made = False # === 遍历每一个 rate set 子文件 === for file_path in self.rate_files: try: self.logger.info(f"→ 处理子文件: {file_path.name}") # --- 1. 解析文件名得到 full_key --- try: full_key = self.parse_filename(file_path.name) self.logger.debug(f" ├─ 解析出 key: {full_key}") except ValueError as ve: self.logger.warning(f" └─ 跳过无效文件名: {ve}") continue # --- 2. 查找 rate_set_map 映射 --- target = self.target_map.get(full_key) if not target: self.logger.warning(f" └─ 未在 config.json 中定义映射关系: {full_key}，跳过") continue # --- 3. 动态设置当前注入目标 --- self.data_array_name = target["data"] self.index_array_name = target["index"] self.enum_name = target["enum"] self.logger.debug(f" ├─ 绑定目标:") self.logger.debug(f" data: {self.data_array_name}") self.logger.debug(f" index: {self.index_array_name}") self.logger.debug(f" enum: {self.enum_name}") # --- 4. 解析主文件中的当前结构 --- try: parsed = self.parse_all_structures(full_content) except Exception as e: self.logger.error(f" └─ 解析主文件结构失败: {e}") continue # --- 5. 提取该子文件中的 rate sets --- file_content = file_path.read_text(encoding='utf-8') subsets = self.extract_sub_rate_sets(file_content) if not subsets: self.logger.info(f" └─ 无有效子集数据") continue # --- 6. 构建要注入的内容 --- valid_blocks = self.build_injection_with_format( subsets, existing_enum=parsed['existing_enum'] ) if not valid_blocks: self.logger.info(f" └─ 无需更新") continue # --- 7. 写回新内容（精准插入）--- updated_content = self._write_back_in_blocks( full_content, parsed, valid_blocks ) if updated_content != full_content: all_changes_made = True full_content = updated_content # 更新内存内容供后续文件使用 self.logger.info(f"✅ 成功注入 {len(subsets)} 条目到 {self.enum_name}") except Exception as e: self.logger.warning(f"❌ 处理文件失败 [{file_path.name}]: {e}") if self.logger.isEnabledFor(logging.DEBUG): self.logger.debug("详细堆栈:", exc_info=True) continue # 最终写回磁盘 if all_changes_made: try: return self._safe_write_back(self.c_file_path.read_text(encoding='utf-8'), full_content) except Exception as e: self.logger.error(f"写入最终文件失败: {e}") raise else: self.logger.info("没有需要更新的内容") return False def _format_with_inline_fallback( self, lines: List[str], new_items: List[str], max_per_line: int, indent_marker: str = " ", item_separator: str = ", ", line_suffix: str = "", # 注意：现在我们不在这里加逗号！ extract_func=None, align_eq_col: bool = False, detect_spacing_from_last_line: bool = True, ) -> str: if not lines: lines = [""] last_line = lines[-1].rstrip() indent_match = re.match(r'^(\s*)', last_line) line_indent = indent_match.group(1) if indent_match else indent_marker clean_last = re.sub(r'//.*|/\*.*?\*/', '', last_line).strip() existing_items = extract_func(clean_last) if extract_func else re.findall(r'\w+', clean_last) current_count = len(existing_items) space_left = max(0, max_per_line - current_count) to_append_inline = new_items[:space_left] to_append_newline = new_items[space_left:] # === 检测真实分隔符 === actual_sep = item_separator if detect_spacing_from_last_line and len(existing_items) >= 2: first = re.escape(existing_items[0]) second = re.escape(existing_items[1]) match = re.search(f"({first})(\\s+)({second})", last_line) if match: actual_sep = match.group(2) # === 对齐等号列：关键修复 → 对齐后再加逗号 === formatted_new_items = [] for item in new_items: raw_item = item.rstrip(',') # 去掉可能已有的逗号避免重复 if align_eq_col: m = re.match(r'(\w+)\s*=\s*(\d+)', raw_item) if m: name, val = m.groups() # 计算目标列位置 target_eq_col = None for i in range(len(lines) - 1, -1, -1): ln = lines[i] eq_match = re.search(r'=\s*\d+', ln) if eq_match: raw_before = ln[:eq_match.start()] expanded_before = raw_before.expandtabs(4) target_eq_col = len(expanded_before) break if target_eq_col is None: target_eq_col = 30 padding = max(1, target_eq_col - len(name.replace('\t', ' ').expandtabs(4))) spaces = ' ' * padding aligned_item = f"{name}{spaces}= {val}" formatted_new_items.append(aligned_item) else: formatted_new_items.append(raw_item) else: formatted_new_items.append(raw_item) # 现在统一加逗号：每个 item 都要加！ # 注意：是否加逗号应该由调用者或此函数控制，不要混合 final_formatted_items = [f"{item}," for item in formatted_new_items] to_append_inline = final_formatted_items[:space_left] to_append_newline = final_formatted_items[space_left:] # === 构建结果 === result_lines = lines[:-1] # 保留前面所有行 final_main_line = lines[-1].rstrip() # 添加 inline 项 if to_append_inline: joined_inline = actual_sep.join(to_append_inline) if final_main_line.strip(): final_main_line += actual_sep + joined_inline else: final_main_line = joined_inline result_lines.append(final_main_line) # 添加新行（每行最多 max_per_line 个） if to_append_newline: for i in range(0, len(to_append_newline), max_per_line): chunk = to_append_newline[i:i + max_per_line] joined = actual_sep.join(chunk) result_lines.append(f"{line_indent}{joined}") return '\n'.join(result_lines) def _write_back_in_blocks(self, full_content: str, parsed: Dict, valid_blocks: List[Dict]) -> str: """ 使用局部块操作策略：只在 /* START */ ... /* END */ 范围内修改内容 关键改进：直接插入 raw_header + raw_body，保留原始格式 """ self.logger.info("开始执行局部块写入操作...") self.logger.info("...") # 在 _write_back_in_blocks 最上方添加： base_data_offset = parsed['data_len'] current_new_data_size = 0 # 动态记录已写入的新 data 大小 start_pos = full_content.find(self.block_start) end_pos = full_content.find(self.block_end) if start_pos == -1 or end_pos == -1: raise ValueError(f"未找到锚点标记: {self.block_start} 或 {self.block_end}") if end_pos <= start_pos: raise ValueError("结束锚点位于起始锚点之前") inner_start = start_pos + len(self.block_start) block_content = full_content[inner_start:end_pos] replacements = [] # (start_in_block, end_in_block, replacement) # === Step 2: 更新 ENUM === if valid_blocks: # 提取函数：从字符串中提取 RATE_SET_xxx extract_enum = lambda s: re.findall(r'RATE_SET_[A-Z0-9_]+', s) enum_pattern = rf'(enum\s+{re.escape(self.enum_name)}\s*\{{)([^}}]*)\}}\s*;' match = re.search(enum_pattern, block_content, re.DOTALL | re.IGNORECASE) if not match: raise ValueError(f"未找到枚举定义: {self.enum_name}") header = match.group(1) # "enum rate_set_2g_20m {" body_content = match.group(2) lines = [ln.rstrip() for ln in body_content.splitlines() if ln.strip()] # 计算新值 new_macros = [] for block in valid_blocks: name = block["enum_name"] value = block["enum_value"] # ✅ 来自 build_injection_with_format 的正确值 new_macros.append(f"{name} = {value}") # === 关键：获取标准缩进 === indent_match = re.match(r'^(\s*)', lines[0] if lines else "") standard_indent = indent_match.group(1) if indent_match else " " # 格式化新 body new_body = self._format_with_inline_fallback( lines=lines, new_items=new_macros, max_per_line=self.MAX_ENUM_PER_LINE, indent_marker=standard_indent, item_separator=" ", line_suffix="", extract_func=extract_enum, align_eq_col=True, detect_spacing_from_last_line=True, ) # 关键修复：确保每行都有缩进（包括第一行） formatted_lines = [] for line in new_body.splitlines(): stripped = line.strip() if stripped: formatted_lines.append(f"{standard_indent}{stripped}") else: formatted_lines.append(line) final_body = '\n'.join(formatted_lines) # 关键：header 单独占一行，新 body 换行开始 new_enum_code = f"{header}\n{final_body}\n}};" replacements.append((match.start(), match.end(), new_enum_code)) self.logger.debug(f"更新 enum: 添加 {len(valid_blocks)} 项") # === Step 3: 更新 DATA 数组 === if valid_blocks: data_pattern = rf'(static const unsigned char {re.escape(self.data_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(data_pattern, block_content, re.DOTALL) if not match: raise ValueError(f"未找到 data 数组: {self.data_array_name}") header = match.group(1) body_content = match.group(2).strip() footer = match.group(3) lines = body_content.splitlines() last_line = lines[-1] if lines else "" indent_match = re.match(r'^(\s*)', last_line) line_indent = indent_match.group(1) if indent_match else " " new_body = body_content.rstrip() if not new_body.endswith(','): new_body += ',' for block in valid_blocks: # 插入头行（如 /*...*/ 4,） new_body += f"\n{line_indent}{block['raw_header'].strip()}" # 插入每行 body（保持原始缩进） for raw_line in block["raw_body"]: new_body += f"\n{line_indent}{raw_line}" new_data_code = f"{header}{new_body}\n{footer}" replacements.append((match.start(), match.end(), new_data_code)) self.logger.debug(f"计划更新 data 数组: 添加 {len(valid_blocks)} 个原始块") # === Step 2: 更新 INDEX 数组 === if valid_blocks: index_pattern = rf'(static const unsigned short {re.escape(self.index_array_name)}\[\]\s*=\s*\{{)([^}}]*)(\}}\s*;)' match = re.search(index_pattern, block_content, re.DOTALL) if not match: raise ValueError(f"未找到 index 数组: {self.index_array_name}") header = match.group(1) body_content = match.group(2) footer = match.group(3).strip() lines = [ln.rstrip() for ln in body_content.splitlines()] non_empty_lines = [ln for ln in lines if ln.strip()] # 获取标准缩进（与 enum 一致） if non_empty_lines: indent_match = re.match(r'^(\s*)', non_empty_lines[0]) standard_indent = indent_match.group(1) if indent_match else " " else: standard_indent = " " # 生成新索引值 # 正确计算 index values：基于 data 偏移 + 每个 block 的实际大小 current_offset = parsed['data_len'] # 初始偏移 = 原 data 长度 new_index_values = [] for block in valid_blocks: # 添加当前 block 的起始偏移 new_index_values.append(str(current_offset)) # 偏移 += 当前 block 的数据条数 current_offset += block["count"]+1 # ← 使用 block 自带的 count！ self.logger.info(f"生成新的 index values: {new_index_values}") # 格式化 index body formatted_body = self._format_with_inline_fallback( lines=non_empty_lines, new_items=new_index_values, max_per_line=self.MAX_INDEX_ITEMS_PER_LINE, indent_marker=standard_indent, item_separator=" ", line_suffix="", extract_func=lambda s: re.findall(r'\d+', s), detect_spacing_from_last_line=True, align_eq_col=False, ) # 统一添加缩进 final_lines = [] for line in formatted_body.splitlines(): stripped = line.strip() if stripped: final_lines.append(f"{standard_indent}{stripped}") else: final_lines.append("") final_body = '\n'.join(final_lines) new_index_code = f"{header}\n{final_body}\n{footer}" replacements.append((match.start(), match.end(), new_index_code)) self.logger.debug(f"更新 index 数组: 添加 {len(valid_blocks)} 个索引") # === Step 5: 倒序应用所有替换 === if not replacements: self.logger.info("无任何变更需要写入") return full_content for start, end, r in sorted(replacements, key=lambda x: x[0], reverse=True): block_content = block_content[:start] + r + block_content[end:] # === Step 6: 拼接回完整文件 === final_content = ( full_content[:inner_start] + block_content + full_content[end_pos:] ) self.logger.info(f"成功构建新内容，总长度变化: {len(full_content)} → {len(final_content)}") return final_content def run(self): self.logger.info("开始同步 RATE_SET 数据...") try: changed = self.inject_new_data() if changed: print(" 同步完成") else: print(" 无新数据，无需更新") return { "success": True, "changed": changed, "file": str(self.c_file_path), "backup": f"{self.c_file_path.stem}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.c.bak" if changed and not self.dry_run else None } except Exception as e: self.logger.error(f"同步失败: {e}", exc_info=True) print("❌ 同步失败，详见日志。") return {"success": False, "error": str(e)} def main(): logging.basicConfig( level=logging.INFO, format='%(asctime)s [%(levelname)s] %(name)s: %(message)s', handlers=[ logging.FileHandler(LOG_FILE, encoding='utf-8'), logging.StreamHandler(sys.stdout) ], force=True ) dry_run = False # 设置为 True 可进行试运行 try: sync = RateSetSynchronizer(dry_run=dry_run) sync.run() print("同步完成！") except FileNotFoundError as e: logging.error(f"文件未找到: {e}") print(" 文件错误，请检查路径。") sys.exit(1) except PermissionError as e: logging.error(f"权限错误: {e}") print(" 权限不足，请关闭编辑器或以管理员运行。") sys.exit(1) except Exception as e: logging.error(f"程序异常退出: {e}", exc_info=True) print(" 同步失败，详见日志。") sys.exit(1) if __name__ == '__main__': main() 加一个最终打印变化
10-30
### 1️⃣ 修改 `RateSetSynchronizer` 类，增加 `self.summary` 在 `__init__` 初始化时添加： ```python # 变更摘要 self.summary = { "total_new_sets": 0, "per_band": {"2g": 0, "5g": 0, "6g": 0}, ...
OCF相关（二）
u012385733的专栏
07-20 613
贴上一个源码，关于OCF测试 cryptotest.c  /*- * Copyright (c) 2004 Sam Leffler, Errno Consulting * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification...
全局变量，局部变量，static，const，数组变量的地址打印及地址空间总结
qq_35535992的博客
10-30 1573
在Linux中写一个程序，里面全局变量，局部变量，static，const，数组，输出这些变量的地址（包括赋值的和没有赋值的，各类数据类型） #include #define MAX_SIZE 10 int a; int b = 1; static int st1; static int st2 = 1; const int e = 1; char src1[MAX_SIZE];
'LoadLibraryW' : cannot convert parameter 1 from 'const char *' to 'LPCWSTR' 解决方案
小龙王2010
01-12 4971
原因描述  当前项目工程只支持UNICODE字符  解决方案一 工程属性->配置属性–>常规—>字符集—->使用多字节符字符集 即可。 解决方案二  也就是宽字符的问题，所以这行代码，应该编译有错误 hinst=LoadLibrary(“InTheHand.Net.Personal.dll”);  报错： cannot convert parameter 1 fro
基于Python3开发的插件化可扩展网络安全漏洞自动化扫描与检测框架_集成多种常见服务未授权访问漏洞检测如MongoDB_Redis_Memcached_Zookeeper以及高危.zip
01-05
基于Python3开发的插件化可扩展网络安全漏洞自动化扫描与检测框架_集成多种常见服务未授权访问漏洞检测如MongoDB_Redis_Memcached_Zookeeper以及高危.zip
WIFI热点连接搜索-下载即用.zip
最新发布
01-05
下载方式：https://pan.quark.cn/s/e2e4dfd55716 该安卓应用源码旨在实现搜索并扫描周边WIFI网络，进而获取热点信息并支持直接连接至WIFI热点的功能。 用户可通过该应用直接在界面中开启WIFI功能，无需进入系统设置菜单进行操作。 该项目包含一个util类，该类提供了对安卓设备连接包括WEP、WPA、WEP2及无密码等不同类型热点的功能支持。 代码已附带详尽的注释说明，并经过必要的修正，确保整个项目能够稳定且无误地运行。 项目采用GBK编码，默认编译环境为版本4.1.2，最低系统兼容版本为4.0及以上。
基于Scrapy网络爬虫框架实时抓取各大招聘网站与行业数据平台信息的毕业生人才需求分析系统_结合Flask轻量级Web框架构建后端API接口与数据处理逻辑_集成Echarts数据可.zip
01-05
基于Scrapy网络爬虫框架实时抓取各大招聘网站与行业数据平台信息的毕业生人才需求分析系统_结合Flask轻量级Web框架构建后端API接口与数据处理逻辑_集成Echarts数据可.zip
JPMorgan-Chase-Co-量化研究职位模拟项目_通过参与摩根大通量化研究项目深入了解量化研究团队工作流程与内容运用时间序列分析与机器学习算法结合Python进行数据清.zip
01-05
JPMorgan-Chase-Co-量化研究职位模拟项目_通过参与摩根大通量化研究项目深入了解量化研究团队工作流程与内容运用时间序列分析与机器学习算法结合Python进行数据清.zip
MATLAB中处理心电图和加速度计信号，以可视化心率和呼吸次数随时间的变化。.zip
01-05
1.版本：matlab2014a/2019b/2024b 2.附赠案例数据可直接运行。 3.代码特点：参数化编程、参数可方便更改、代码编程思路清晰、注释明细。 4.适用对象：计算机，电子信息工程、数学等专业的大学生课程设计、期末大作业和毕业设计。
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
红包 添加红包
表情包 插入表情
表情包 代码片
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付：10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明：

1.余额是钱包充值的虚拟货币，按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载，可以购买VIP、付费专栏及课程。

余额充值