C語(yǔ)言容易被忽視的函數(shù)設(shè)計(jì)原則基礎(chǔ)
一、函數(shù)設(shè)計(jì)原則
- 函數(shù)從意義上應(yīng)該是一個(gè)獨(dú)立的功能模塊
- 函數(shù)名要在一定程度上反映函數(shù)的功能
- 函數(shù)參數(shù)名要能夠體現(xiàn)參數(shù)的意義
- 盡量避免在函數(shù)中使用全局變量
- 當(dāng)函數(shù)參數(shù)不應(yīng)該在函數(shù)體內(nèi)部被修改時(shí),應(yīng)加上 const 聲明
- 如果參數(shù)是指針,且僅作輸入?yún)?shù),則應(yīng)加上 const 聲明,如下:
- 不能省略返回值的類型
- 如果函數(shù)沒(méi)有返回值,那么應(yīng)聲明為 void
- 類型對(duì)參數(shù)進(jìn)行有效性檢查
- 對(duì)于指針參數(shù)的檢查尤為重要
- 不要返回指向“棧內(nèi)存”的指針
- 棧內(nèi)存在函數(shù)體結(jié)束時(shí)被自動(dòng)釋放
- 函數(shù)體的規(guī)模要小,盡量控制在 80 行代碼之內(nèi)
- 相同的輸入對(duì)應(yīng)相同的輸出,避免函數(shù)帶有“記憶”功能
- 避免函數(shù)有過(guò)多的參數(shù),參數(shù)個(gè)數(shù)盡量控制在 4 個(gè)以內(nèi)
- 有時(shí)候函數(shù)不需要返回值,但為了增加靈活性,如支持鏈?zhǔn)奖磉_(dá),可以附加返回值
- 函數(shù)名與返回值類型在語(yǔ)義上不可沖突
下面來(lái)欣賞一份優(yōu)秀的代碼:
/******************************************************************************* * Copyright (c) 2000, 2005 IBM Corporation and others. * All rights reserved. This program and the accompanying materials * are made available under the terms of the Eclipse Public License v1.0 * which accompanies this distribution, and is available at * http://www.eclipse.org/legal/epl-v10.html * * Contributors: * IBM Corporation - initial API and implementation * Kevin Cornell (Rational Software Corporation) *******************************************************************************/ /* Eclipse Launcher Utility Methods */ #include "eclipseOS.h" #include "eclipseCommon.h" #include "eclipseUtil.h" #include <string.h> #include <stdlib.h> #include <stdio.h> #include <sys/stat.h> #ifdef _WIN32 #include <direct.h> #else #include <unistd.h> #include <strings.h> #endif #define MAX_LINE_LENGTH 256 /* Is the given VM J9 */ int isJ9VM( _TCHAR* vm ) { _TCHAR * ch = NULL, *ch2 = NULL; int res = 0; if (vm == NULL) return 0; ch = lastDirSeparator( vm ); if (isVMLibrary(vm)) { /* a library, call it j9 if the parent dir is j9vm */ if(ch == NULL) return 0; ch[0] = 0; ch2 = lastDirSeparator(vm); if(ch2 != NULL) { res = (_tcsicmp(ch2 + 1, _T_ECLIPSE("j9vm")) == 0); } ch[0] = dirSeparator; return res; } else { if (ch == NULL) ch = vm; else ch++; return (_tcsicmp( ch, _T_ECLIPSE("j9") ) == 0); } } int checkProvidedVMType( _TCHAR* vm ) { _TCHAR* ch = NULL; struct _stat stats; if (vm == NULL) return VM_NOTHING; if (_tstat(vm, &stats) == 0 && (stats.st_mode & S_IFDIR) != 0) { /* directory */ return VM_DIRECTORY; } ch = _tcsrchr( vm, _T_ECLIPSE('.') ); if(ch == NULL) return VM_OTHER; #ifdef _WIN32 if (_tcsicmp(ch, _T_ECLIPSE(".dll")) == 0) #else if ((_tcsicmp(ch, _T_ECLIPSE(".so")) == 0) || (_tcsicmp(ch, _T_ECLIPSE(".jnilib")) == 0) || (_tcsicmp(ch, _T_ECLIPSE(".dylib")) == 0)) #endif { return VM_LIBRARY; } if (_tcsicmp(ch, _T_ECLIPSE(".ee")) == 0) return VM_EE_PROPS; return VM_OTHER; } /* * pathList is a pathSeparator separated list of paths, run each through * checkPath and recombine the results. * New memory is always allocated for the result */ _TCHAR * checkPathList( _TCHAR* pathList, _TCHAR* programDir, int reverseOrder) { _TCHAR * c1, *c2; _TCHAR * checked, *result; size_t checkedLength = 0, resultLength = 0; size_t bufferLength = _tcslen(pathList); result = malloc(bufferLength * sizeof(_TCHAR)); c1 = pathList; while (c1 != NULL && *c1 != _T_ECLIPSE('\0')) { c2 = _tcschr(c1, pathSeparator); if (c2 != NULL) *c2 = 0; checked = checkPath(c1, programDir, reverseOrder); checkedLength = _tcslen(checked); if (resultLength + checkedLength + 1> bufferLength) { bufferLength += checkedLength + 1; result = realloc(result, bufferLength * sizeof(_TCHAR)); } if(resultLength > 0) { result[resultLength++] = pathSeparator; result[resultLength] = _T_ECLIPSE('\0'); } _tcscpy(result + resultLength, checked); resultLength += checkedLength; if(checked != c1) free(checked); if(c2 != NULL) *(c2++) = pathSeparator; c1 = c2; } return result; } _TCHAR * concatStrings(_TCHAR**strs) { return concatPaths(strs, 0); } _TCHAR * concatPaths(_TCHAR** strs, _TCHAR separator) { _TCHAR separatorString[] = { separator, 0 }; _TCHAR * result; int i = -1; size_t length = 0; /* first count how large a buffer we need */ while (strs[++i] != NULL) { length += _tcslen(strs[i]) + (separator != 0 ? 1 : 0); } result = malloc((length + 1) * sizeof(_TCHAR)); result[0] = 0; i = -1; while (strs[++i] != NULL) { result = _tcscat(result, strs[i]); if (separator != 0) result = _tcscat(result, separatorString); } return result; } /* * buffer contains a pathSeparator separated list of paths, check * that it contains all the paths given. Each path is expected to be * terminated with a pathSeparator character. */ int containsPaths(_TCHAR * str, _TCHAR** paths) { _TCHAR * buffer; _TCHAR * c; int i; /* terminate the string with a pathSeparator */ buffer = malloc((_tcslen(str) + 2) * sizeof(_TCHAR)); _stprintf(buffer, _T_ECLIPSE("%s%c"), str, pathSeparator); for (i = 0; paths[i] != NULL; i++) { c = _tcsstr(buffer, paths[i]); if ( c == NULL || !(c == buffer || *(c - 1) == pathSeparator)) { /* entry not found */ free(buffer); return 0; } } free(buffer); return 1; } int isVMLibrary( _TCHAR* vm ) { _TCHAR *ch = NULL; if (vm == NULL) return 0; ch = _tcsrchr( vm, '.' ); if(ch == NULL) return 0; #ifdef _WIN32 return (_tcsicmp(ch, _T_ECLIPSE(".dll")) == 0); #else return (_tcsicmp(ch, _T_ECLIPSE(".so")) == 0) || (_tcsicmp(ch, _T_ECLIPSE(".jnilib")) == 0) || (_tcsicmp(ch, _T_ECLIPSE(".dylib")) == 0); #endif } #ifdef AIX #include <sys/types.h> #include <time.h> /* Return the JVM version in the format x.x.x */ char* getVMVersion( char *vmPath ) { char cmd[MAX_LINE_LENGTH]; char lineString[MAX_LINE_LENGTH]; char* firstChar; char fileName[MAX_LINE_LENGTH]; time_t curTime; FILE* fp; int numChars = 0; char* version = NULL; /* Define a unique filename for the java output. */ (void) time(&curTime); (void) sprintf(fileName, "/tmp/tmp%ld.txt", curTime); /* Write java -version output to a temp file */ (void) sprintf(cmd,"%s -version 2> %s", vmPath, fileName); (void) system(cmd); fp = fopen(fileName, "r"); if (fp != NULL) { /* Read java -version output from a temp file */ if (fgets(lineString, MAX_LINE_LENGTH, fp) == NULL) lineString[0] = '\0'; fclose(fp); unlink(fileName); /* Extract version number */ firstChar = (char *) (strchr(lineString, '"') + 1); if (firstChar != NULL) numChars = (int) (strrchr(lineString, '"') - firstChar); /* Allocate a buffer and copy the version string into it. */ if (numChars > 0) { version = malloc( numChars + 1 ); strncpy(version, firstChar, numChars); version[numChars] = '\0'; } } return version; } /* Compare JVM Versions of the form "x.x.x..." * * Returns -1 if ver1 < ver2 * Returns 0 if ver1 = ver2 * Returns 1 if ver1 > ver2 */ int versionCmp(char *ver1, char *ver2) { char* dot1; char* dot2; int num1; int num2; dot1 = strchr(ver1, '.'); dot2 = strchr(ver2, '.'); num1 = atoi(ver1); num2 = atoi(ver2); if (num1 > num2) return 1; if (num1 < num2) return -1; if (dot1 && !dot2) /* x.y > x */ return 1; if (!dot1 && dot2) /* x < x.y */ return -1; if (!dot1 && !dot2) /* x == x */ return 0; return versionCmp((char*)(dot1 + 1), (char*)(dot2 + 1) ); } #endif /* AIX */
二、總結(jié)
- C 語(yǔ)言的學(xué)習(xí)需要勤思考勤動(dòng)手才能得到
- 提高難點(diǎn)部分為指針的學(xué)習(xí)
- 指針的本質(zhì),指針的運(yùn)算,指針和數(shù)組
- 學(xué)習(xí)過(guò)程可以采用各個(gè)擊破的方法
- 在一個(gè)特定的時(shí)間段只重點(diǎn)學(xué)習(xí)和練習(xí)某個(gè)主題
- 在熟練掌握C語(yǔ)言的各個(gè)特性后再進(jìn)行項(xiàng)目練習(xí)
至此,C 語(yǔ)言進(jìn)階剖析學(xué)完,完結(jié)撒花。
但是,以后路還很遠(yuǎn),還需要多加總結(jié),才能真正掌握 C 語(yǔ)言。
到此這篇關(guān)于C語(yǔ)言容易被忽視的函數(shù)設(shè)計(jì)原則基礎(chǔ)的文章就介紹到這了,更多相關(guān)C語(yǔ)言函數(shù)設(shè)計(jì)原則內(nèi)容請(qǐng)搜索腳本之家以前的文章或繼續(xù)瀏覽下面的相關(guān)文章希望大家以后多多支持腳本之家!
相關(guān)文章
C語(yǔ)言深入分析整形數(shù)據(jù)存儲(chǔ)
C語(yǔ)言中,我們經(jīng)常使用數(shù)據(jù)類型,那么整形數(shù)據(jù)在內(nèi)存中如何存儲(chǔ)?存儲(chǔ)方式是什么?如果你對(duì)這些內(nèi)容不太了解的話,相信看完這篇博客后,你會(huì)對(duì)整形數(shù)據(jù)的存儲(chǔ)有一個(gè)新的認(rèn)識(shí)。話不多說(shuō),我們進(jìn)入正題2022-08-08C語(yǔ)言中自動(dòng)隱式轉(zhuǎn)換與類型強(qiáng)制轉(zhuǎn)換實(shí)例分析
這篇文章主要介紹了C語(yǔ)言中自動(dòng)隱式轉(zhuǎn)換與類型強(qiáng)制轉(zhuǎn)換實(shí)例分析,需要的朋友可以參考下2014-07-07C++小知識(shí):C/C++中不要按值傳遞數(shù)組
今天小編就為大家分享一篇關(guān)于C++小知識(shí):C/C++中不要按值傳遞數(shù)組,小編覺(jué)得內(nèi)容挺不錯(cuò)的,現(xiàn)在分享給大家,具有很好的參考價(jià)值,需要的朋友一起跟隨小編來(lái)看看吧2019-01-01C語(yǔ)言使用深度優(yōu)先搜索算法解決迷宮問(wèn)題(堆棧)
這篇文章主要介紹了C語(yǔ)言使用深度優(yōu)先搜索算法解決迷宮問(wèn)題,涉及C語(yǔ)言堆棧的使用與深度優(yōu)先算法解決迷宮問(wèn)題的相關(guān)操作技巧,需要的朋友可以參考下2017-09-09C++11智能指針unique_ptr用法使用場(chǎng)景分析
unique_ptr 是 C++ 11 提供的用于防止內(nèi)存泄漏的智能指針中的一種實(shí)現(xiàn),即使在異常發(fā)生時(shí)也可幫助避免資源泄露。這篇文章主要介紹了C++11智能指針unique_ptr用法介紹,需要的朋友可以參考下2021-08-08C語(yǔ)言指針學(xué)習(xí)經(jīng)驗(yàn)總結(jié)淺談
指針是C語(yǔ)言的難點(diǎn)和重點(diǎn),但指針也是C語(yǔ)言的靈魂 。2013-03-03可能是全網(wǎng)最詳細(xì)的Qt連接MySQL數(shù)據(jù)庫(kù)教程
QT眾所周知是一個(gè)開(kāi)源的,以C++為底層的可視化工具庫(kù),下面這篇文章主要給大家介紹了關(guān)于最詳細(xì)的Qt連接MySQL數(shù)據(jù)庫(kù)教程的相關(guān)資料,文中通過(guò)圖文介紹的非常詳細(xì),需要的朋友可以參考下2023-04-04