Dev-cpp模仿者

在网页上展示C/C++代码时，如果没有语法高亮，就会显得很平淡。我们需要分别为每一个保留字、符号、注释添加粗体、高亮等。但每一份代码都手动添加的话，费时又费力，因此必须使用程序来自动生成。Dev-C++自带了这个功能，只需在文件＞导出菜单中选择“到HTML”，就会提供一份当前源码的HTML副本。然而：

  1. 我当时竟然还没发现有这个工具（太惨了）
  2. 它不支持自动显示行号
  3. 它以span class=xxx的形式输出每一个标记，这样会使HTML文件占用的空间变得很大。现在已经没有浏览器还不支持CSS1的后代选择器了，我更愿意以自己的方式处理，如将i标签用作注释，em标签用于保留字，b标签用于标点符号，等等。这样子不是更加简洁吗？

另一个解决方案是，使用例如highlight.js之类的库。我不想那样做，因为

  1. 我还看不懂js代码，尤其是有些地方的判断用到了正则表达式，简直搞得我头晕脑胀的。对我来说，代码完全看不懂，开源了就跟没开似的。
  2. 个性化比较差，感觉网上到处都是这种类型的代码高亮，千篇一律。（虽然我也知道，改个CSS文件就行了吧。）
  3. 可能用到了一些高级特性，对旧浏览器的支持不好。
  4. 他们搞得太复杂啦！比如同时支持多语言。我暂时不需要那么多的功能，只想搞好我最迫切需要的的C语言，来个极简化的代码。

于是我用自己最擅长的C语言开发了一个程序，可以用HTML打印C++代码，和Dev-C++中一样丰富多彩。其实我早在四年以前就有了这个想法，今天总算是实现啦！和原版生成的HTML效果几乎一样，除了一些特殊的地方会有差别，例如在宏定义的中间穿插注释时（具体的见test-all.c，第26行与第66行那里），这算是Dev-cpp模仿者目前版本一个已知的bug，以后会完善的。

源码如下：

  1. /*
  2. * Description: Dev-cpp Imitator
  3. * Version: 0.1
  4. * Date: 2025-08-15
  5. * Author: Shurui Zhang
  6. * Contact: https://qowk.xyz/contact/
  7. */

  9. #include <stdio.h>

  10. #include <ctype.h>

  11. #include <string.h>

  13. char l[200];
  15. enum Line_State {
  16. none,
  17. string_literal,
  18. preprocessor_directive,
  19. single_line_comment,
  20. multiple_line_comment
  21. } line_state;
  23. typedef enum {
  24. NUMBER,
  25. KEYWORD,
  26. STRING,
  27. SYMBOL,
  28. COMMENT,
  29. PREPROCESSOR,
  30. IDENTIFIER
  31. } token_types;
  33. const char *opening_tag[] = {
  34. "<a>", "<em>", "<q>", "<b>", "<i>", "<p>", ""
  35. };
  37. const char *closing_tag[] = {
  38. "</a>", "</em>", "</q>", "</b>", "</i>", "</p>", ""
  39. };
  41. /* range: [start, end) */
  42. void outputHTML(token_types style, int start, int end)
  43. {
  44. static token_types last_style = IDENTIFIER;
  45. if (style != last_style) {
  46. /* merge consecutive elements */
  47. printf("%s%s", closing_tag[last_style], opening_tag[style]);
  48. last_style = style;
  49. }
  50. for (; start < end; start++) {
  51. /* check for escaped characters */
  52. if (l[start] == '<') printf("&lt;");
  53. else if (l[start] == '>') printf("&gt;");
  54. else if (l[start] == '&') printf("&amp;");
  55. else putchar(l[start]);
  56. }
  57. }
  59. int is_keyword(int start, int end)
  60. {
  61. const char *keywords[] = {
  62. "alignas", "alignof", "and", "and_eq", "asm", "auto", "bitand",
  63. "bitor", "bool", "break", "case", "catch", "char", "char16_t",
  64. "char32_t", "class", "compl", "const", "const_cast",
  65. "constexpr", "continue", "decltype", "default", "delete", "do",
  66. "double", "dynamic_cast", "else", "enum", "explicit", "export",
  67. "extern", "false", "float", "for", "friend", "goto", "if",
  68. "inline", "int", "long", "mutable", "namespace", "new",
  69. "noexcept", "not", "not_eq", "nullptr", "operator", "or",
  70. "or_eq", "private", "protected", "public", "register",
  71. "reinterpret_cast", "return", "short", "signed", "sizeof",
  72. "static", "static_assert", "static_cast", "struct", "switch",
  73. "template", "this", "thread_local", "throw", "true", "try",
  74. "typedef", "typeid", "typename", "union", "unsigned", "using",
  75. "virtual", "void", "volatile", "wchar_t", "while", "xor",
  76. "xor_eq"
  77. };
  78. char t;
  79. int i;
  80. t = l[end];
  81. l[end] = '\0';
  82. for (i = 0; i < sizeof keywords / sizeof *keywords; i++) {
  83. if (!strcmp(&l[start], keywords[i])) {
  84. l[end] = t;
  85. return 1;
  86. }
  87. }
  88. l[end] = t;
  89. return 0;
  90. }
  92. void process_line()
  93. {
  94. int head = 0, tail = 0;
  95. int len;
  97. /* #define DEBUG */

  98. #ifdef DEBUG

  99. for (; l[head]; head++) {
  100. printf("%x ", l[head]);
  101. }
  102. return;

  103. #endif

  105. len = strlen(l);
  107. if (len == sizeof l - 1) {
  108. fputs("line character buffer limit exceeded", stderr);
  109. return;
  110. }
  112. /* strip trivial LF and CR characters */
  113. if (l[len - 1] == '\n')
  114. l[--len] = '\0';
  116. if (l[len - 1] == '\r')
  117. l[--len] = '\0';
  119. switch (line_state)
  120. do {
  121. head = tail;
  123. default:
  124. if (isspace(l[tail])) {
  125. putchar(l[tail++]);
  126. continue;
  127. }
  129. if (isalpha(l[tail]) || l[tail] == '_') {
  130. do tail++;
  131. while (isalnum(l[tail]) || l[tail] == '_');
  132. outputHTML(is_keyword(head, tail) ?
  133. KEYWORD : IDENTIFIER, head, tail);
  134. continue;
  135. }
  137. if (l[tail] == '#') {
  138. case preprocessor_directive:
  139. /* repeat until meet comments */
  140. do tail++;
  141. while (l[tail] && (l[tail] != '/'
  142. || (l[tail + 1] != '/'
  143. && l[tail + 1] != '*')));
  144. if (!l[tail] && l[tail - 1] == '\\')
  145. line_state = preprocessor_directive;
  146. outputHTML(PREPROCESSOR, head, tail);
  147. if (!l[tail] && l[tail - 1] == '\\') return;
  148. continue;
  149. }
  151. if (l[tail] == '"') {
  152. tail++;
  153. case string_literal:
  154. /* When l[tail] = '\0', it is (maybe)
  155. a C++11 raw string literal, or syntax error */
  156. while (l[tail] != '"' && l[tail]) {
  157. if (l[tail] == '\\' && !l[++tail]) {
  158. /* line wrap */
  159. line_state = string_literal;
  160. outputHTML(STRING, head, tail);
  161. return;
  162. }
  163. tail++;
  164. }
  165. tail++;
  166. outputHTML(STRING, head, tail);
  167. continue;
  168. }
  170. if (l[tail] == '\'') {
  171. do {
  172. if (l[tail] == '\\' && l[tail + 1])
  173. tail++;
  174. tail++;
  175. } while (l[tail] != '\'' && l[tail]);
  176. tail++;
  177. outputHTML(IDENTIFIER, head, tail);
  178. continue;
  179. }
  181. if (l[tail] == '/' && l[tail + 1] == '/') {
  182. case single_line_comment:
  183. line_state = l[len - 1] == '\\' ?
  184. single_line_comment : none;
  185. outputHTML(COMMENT, head, len);
  186. return;
  187. }
  189. if (l[tail] == '/' && l[tail + 1] == '*') {
  190. tail += 2;
  191. case multiple_line_comment:
  192. /* repeat until the end of the multiple-line comment */
  193. while (l[tail] != '*' || l[tail + 1] != '/') {
  194. if (!l[tail]) {
  195. /* at the end of line, the multiple line comment will
  196. continue */
  197. line_state = multiple_line_comment;
  198. outputHTML(COMMENT, head, tail);
  199. return;
  200. }
  201. tail++;
  202. }
  203. tail += 2;
  204. outputHTML(COMMENT, head, tail);
  205. continue;
  206. }
  208. if (isdigit(l[tail]) ||
  209. (l[tail] == '.' && isdigit(l[tail + 1]))) {
  210. do {
  211. tail++;
  212. if (toupper(l[tail]) == 'E'
  213. && (l[tail + 1] == '+'
  214. || l[tail + 1] == '-')){
  215. tail += 2;
  216. }
  217. } while (isalnum(l[tail]) || l[tail] == '.');
  218. outputHTML(NUMBER, head, tail);
  219. continue;
  220. }
  222. if (ispunct(l[tail])) {
  223. tail++;
  224. outputHTML(SYMBOL, head, tail);
  225. continue;
  226. }
  228. if (l[tail]) {
  229. fputs("Unknown Error", stderr);
  230. return;
  231. }
  232. } while (l[tail]);
  234. line_state = none;
  235. }
  237. int main()
  238. {
  239. printf(
  240. "<!DOCTYPE HTML PUBLIC '-//W3C//DTD HTML 4.01//EN' "
  241. "'http://www.w3.org/TR/html4/strict.dtd'>"
  242. "<meta http-equiv=content-type "
  243. "content='text/html; charset=utf-8'>" /* charset=gb2312 */
  244. "<style type='text/css'>"
  245. "ol.devcppimitator i {color:#0078d7} "
  246. "ol.devcppimitator q "
  247. "{color:blue; font-weight:bold; quotes:none} "
  248. "ol.devcppimitator b {color:red} "
  249. "ol.devcppimitator em "
  250. "{color:black; font-weight:bold; font-style:normal} "
  251. "ol.devcppimitator a {color:purple} "
  252. "ol.devcppimitator p {display:inline; color:green} "
  253. "ol.devcppimitator li {white-space:pre} " /* ; tab-size:4 */
  254. /* tab-size is in CSS3 */
  255. );
  256. printf(
  257. "ol.devcppimitator {width:800px; margin:0 auto; "
  258. "padding-left:60px; border:solid 1px; "
  259. "font-family:Consolas,'Courier New',monospace}"
  260. "</style>"
  261. "<title>Dev-cpp Imitator</title>"
  262. "<div style='width:1000px; margin:0 auto; padding:10px;"
  263. "border:solid 1px'>"
  264. "<p>The sample is as follows:\n"
  265. "<ol class=devcppimitator>\n"
  266. );
  268. while (fgets(l, sizeof l, stdin)) {
  269. printf("<li>");
  270. process_line();
  272. /* flush buffer */
  273. outputHTML(IDENTIFIER, 0, 0);
  275. /* This is necessary, because an empty li element
  276. doesn't occupy a line, resulting in copy failures. */
  277. putchar('\n');
  278. }
  280. printf("</ol>\n</div>");
  281. return 0;
  282. }

编译与运行的方法：

先把上面的代码全部复制下来，保存为a.c。

在DOS/Windows上：

先用Dev-C++等工具直接编译，然后在cmd命令行中输入（使用重定向的方法）
a.exe < a.c > a.htm

如果你用PowerShell那么则是

Get-Content a.c | .\a.exe | Out-File a.htm -Encoding utf8

在Linux上：

gcc a.c -o a.out
./a.out < a.c > a.htm

再用浏览器打开a.htm文件，就可以看到被高亮的代码了。

最后将a.htm中的整个ol元素（每行代码都会在HTML文件当中占用一行，相当于从第二行复制到倒数第二行）粘贴到你网页的HTML代码中间去，记得在头部添加文档级样式表，或者以外部样式表的方式，就比如我网站的theme02.css。