doxysearch.cpp

doxygen(一个自动从源代码生成文档的工具)的源代码

        int rank=(int)(d->rank*100+0.5);
        if (rank==0) rank++;
        printf("        <tr><td align=\"right\"><b><font color=#%2x00%2x>%d</font></b>"
               "</td><td><a href=\"%s/%s\">%s</a></td></tr>\n",
               rank*2+55, 255-rank*2, rank, 
               d->fileInfo->url, htmlName, linkName);
        pageEntries--;
        if (d->fileInfo->f != 0)
        {
          fclose(d->fileInfo->f);
          d->fileInfo->f = 0;
        }
      }
      else 
      {
        skipEntries--;
      }
    }
    printf("      </table>\n"
           "    </td>\n"
           "  </tr>\n"
           "  <tr><td colspan=2 bgcolor=\"#4040c0\"><center>&nbsp;");
    
    if (page>0)
    {
      printf("<a href=\"%s?page=%d&which=%s&what=%s&query=%s\">"
             "<font color=\"ffffff\">prev</font></a>&nbsp;",
              cgiBin, page-1, 
              wordOnly?"word":"doc", 
              nameOnly?"name":"string",
              encQueryInput);
    }
    int startPage = page-5 < 0 ? 0 : page-5;
    int endPage   = page+6 > numPages ? numPages : page+5;
    if (endPage-startPage>1)
    {
      for (i=startPage;i<endPage;i++)
      {
        if (i!=page)
        {
          printf("<a href=\"%s?page=%d&which=%s&what=%s&query=%s\">"
              "<font color=\"ffffff\">%d</font></a>&nbsp;",
              cgiBin, i, 
              wordOnly?"word":"doc", 
              nameOnly?"name":"string", 
              encQueryInput, i+1);
        }
        else
        {
          printf("<b><font color=\"ffffff\">%d</font></b></a>&nbsp;",i+1);
        }
      }
    }
    if (page<numPages-1)
    {
      printf("<a href=\"%s?page=%d&which=%s&what=%s&query=%s\">"
             "<font color=\"ffffff\">next</font></a>",
              cgiBin, page+1, 
              wordOnly?"word":"doc", 
              nameOnly?"name":"string",
              encQueryInput);
    }
    printf("  &nbsp;</center></td></tr>\n"
           "  <tr bgcolor=#ffffff>\n"
           "    <td colspan=2><p>\n"
           "       Occurrence count: ");
    WordInfo *wi=wordList.first;
    while (wi)
    {
      printf("<b><font color=#ff0000>%s</font></b> ",wi->word);
      if (wi->freq>0) printf("(%d)",wi->freq); else printf("(ignored)");
      wi=wi->next;
      if (wi) 
      {
        if (wi->next) printf(", "); else printf(" and ");
      }
    }
    printf("\n"
           "    </td>\n"
           "  </tr>\n");
  }
  printf("</table>\n</center>\n</form>\n");
  printFooter();
  free(docPtrArray);
}

//----------------------------------------------------------------------------

void generateHelpPage()
{
  printHeader();
  printf("<h3>Doxysearch help page</h3>\n"
         "<b>Basic search instructions</b>\n"
         "<ul>\n"
         "Just type in one or more words or fragments of words and press the "
         "search button. Doxysearch will return a list of functions and class names, "
         "whose documentation or name matches one or more of the words. "
         "Documents containing more matches will appear earlier in the list. "
         "\n"
         "</ul>\n"
         "<b>Search modes</b>\n"
         "<ul>\n"
         "Doxysearch has four search modes. From least restrictive to most "
         "restrictive they are:\n"
         "  <ul><p>\n"
         "    <li><em>Search for strings in the documentation:</em>\n"
         "      The words you type will be searched in the documentation as substrings." 
         "      Typing the word &quot;<b>the</b>&quot; for example, will find "
         "      documents containing the words &quot;<b>the</b>&quot;, "
         "      &quot;<b>the</b>re&quot;, and &quot;fea<b>the</b>r&quot;.<p>\n"
         "    <li><em>Search for words in the documentation:</em>\n"
         "      The words you type will be searched in the documentation as words."
         "      Typing the word &quot;<b>the</b>&quot; for example, will look "
         "      for documents containing the word &quot;<b>the</b>&quot; and not for"
         "      documents containing the word &quot;<b>the</b>re&quot;.<p>\n"
         "    <li><em>Search for strings in the function and class names:</em>\n"
         "      The words you type will be searched in the function and class names "
         "      as substrings. Typing the word &quot;<b>set</b>&quot; for example, will "
         "      result in a list of links to the documentation of all functions and classes "
         "      that contain the word &quot;<b>set</b>&quot; in the name, such as the function "
         "      <code><b>set</b>Cursor</code>.<p>\n"
         "    <li><em>Search for words in the function and class names:</em>\n"
         "      The words you type will be matched against the function and class names."
         "      The result will be a list of links to the documentation of all function "
         "      and class names that are equal to one of the words.\n"
         "  </ul>\n"
         "</ul>\n"
         "<b>Requiring/Excluding words</b>\n"
         "<ul>\n"
         "  Often you will know a word that will be guaranteed to appear in a document "
         "  for which you are searching. If this is the case, require that the word appears "
         "  in all of the results by attaching a &quot;<b>+</b>&quot; to the beginning of "
         "  the word. You can quickly reject results by adding a word that appears only "
         "  in unwanted documents with a &quot;<b>-</b>&quot; before it."
         "</ul>\n"
        );
         
  printFooter();
}

//----------------------------------------------------------------------------

void mergeSearchResults(SearchResults *totalResults,
                        SearchResults *sr,int mode)
{
  SearchDoc *otd = 0,
            *td  = totalResults->docList, 
            *d   = sr->docList;
  totalResults->totalFreq += sr->totalFreq;
  if (!d) return; // nothing to add
  while (td && d)
  {
    int otdih = otd ? otd->fileInfo->index : -1,
        otdil = otd ? otd->index : -1, 
        tdih  = td->fileInfo->index,
        tdil  = td->index, 
        dih   = d->fileInfo->index,
        dil   = d->index;
    if (tdih==dih && tdil==dil) // combine results
    {
      if (mode != EXCLUDE)
      {
        td->rank += d->rank;
        td->rank *= 2;          // put extra emphasis on multiple word matches
        td->freq += d->freq;
        d = d->next;
        otd = td; td = td->next;
      }
      else // mode == EXCLUDE => remove entry from results
      {
        SearchDoc *tmp=td;
        td=td->next;
        if (otd) otd->next=td; else totalResults->docList = td;
        delete tmp;
      }
    }
    else if ((otdih<dih || (otdih==dih && otdil<dil)) &&
             (tdih>dih  || (tdih==dih && tdil>dil))) // insert
    {
      if (mode == NORMAL)
      {
        SearchDoc *nd = new SearchDoc(*d);
        if (otd) otd->next = nd; else totalResults->docList = nd;
        nd->next = td;
        td = nd;
      }
      d = d->next;
    }
    else if (tdih<dih || (tdih==dih && tdil<dil)) // remove or skip
    {
      if (mode == INCLUDE)
      {
        SearchDoc *tmp=td;
        td=td->next;
        if (otd) otd->next=td; else totalResults->docList = td;
        delete tmp;
      }
      else // mode == EXCLUDE or mode == NORMAL
      {
        otd = td; td = td->next;
      }
    }
  }
  if (td==0 && d && mode==NORMAL) // append rest of sr to totalResults
  {
    while (d)
    {
      SearchDoc *nd = new SearchDoc(*d);
      if (otd) otd->next = nd; else totalResults->docList = nd;
      nd->next = 0;
      d = d->next;
      otd = nd;
    } 
  }
  if (td && d==0 && mode==INCLUDE) // delete the rest of the results
  {
    while (td)
    {
      SearchDoc *tmp=td;
      td=td->next;
      if (otd) otd->next = td; else totalResults->docList = td;
      delete tmp;
    }
  }
}

//----------------------------------------------------------------------------

int asciiToHex(char c)
{
  char l=tolower(c);
  if (l>='0' && l<='9') 
    return l-'0';
  else if (l>='a' && l<='f')
    return l+10-'a';
  else // invalid hex char
    return 0;
}

//----------------------------------------------------------------------------

void fileToBuf(const char *name, char **buf)
{
  FILE *f;
  struct stat file_stats;
  if ((f=fopen(name,"rb"))==NULL) return;
  if (stat(name,&file_stats)==-1)
  {
    message("Error: could not fstat file %s\n",name);
    exit(1);
  }
  unsigned int len=file_stats.st_size;
  if ((*buf=(char *)malloc(len+1))==NULL) 
  {
    message("Error: out of memory\n");
    exit(1);
  }
  if (fread(*buf,1,len,f)!=len)
  {
    message("Error: could not read file %s\n",name);
    exit(1);
  }
  (*buf)[len]='\0';
  fclose(f);
}

//----------------------------------------------------------------------------

void getConfig(const char *s)
{
  int l;
  char configFile[MAXSTRLEN];
  strcpy(configFile,s);
  strcat(configFile,"/search.cfg");

  FILE *f;
  if ((f=fopen(configFile,"r"))==NULL)
  {
    message("Error: could not open config file %s\n",configFile);
    exit(1);
  }
    
  // get the URL to the documentation
  fgets(firstDocUrl,MAXSTRLEN,f);
  l=strlen(firstDocUrl)-1;
  if (firstDocUrl[l]=='\n') firstDocUrl[l]='\0';
  l=strlen(firstDocUrl);
  if (firstDocUrl[l]=='/') firstDocUrl[l]='\0';
  
  // get the URL to the cgi script
  fgets(cgiBin,MAXSTRLEN,f);
  l=strlen(cgiBin)-1;
  if (cgiBin[l]=='\n') cgiBin[l]='\0';
  
  fclose(f);
  
  char headerFile[MAXSTRLEN];
  strcpy(headerFile,s);
#if defined(_WIN32)
  strcat(headerFile,"\\header.html");
#else
  strcat(headerFile,"/header.html");
#endif
  fileToBuf(headerFile,&headerBuf);
  
  char footerFile[MAXSTRLEN];
  strcpy(footerFile,s);
#if defined(_WIN32)
  strcat(footerFile,"\\footer.html");
#else
  strcat(footerFile,"/footer.html");
#endif
  fileToBuf(footerFile,&footerBuf);
  
}

//----------------------------------------------------------------------------
// copy and convert string to lower case

void strlowercpy(char *d,const char *s)
{
  while (*s!='\0') *d++=tolower(*s++); *d='\0';
}

//----------------------------------------------------------------------------

int main(int argc,char **argv)
{
#ifdef PROFILING
  struct timeval tv_start,tv_end;
  gettimeofday(&tv_start,0);
#endif
  char *argString=getenv("QUERY_STRING");
  if (argc==1)
  {
    message("Error: invalid number of arguments. "
            "Usage: %s doc_path [doc_path ...]",argv[0]);
    exit(1);
  }
  // read the configuration file for this instance of the search engine
  getConfig(argv[1]);

  if (!argString) 
  {
    printHeader();
    printSearchPage();
    printFooter();
    exit(1);
  }
  
  SearchResults tsr;

  // parse cgi arguments
  char *arg=strtok(argString,"&");
  char *query = 0;
  char *what  = 0;
  char *which = 0;
  char *help  = 0;
  page = 0;
  while (arg)
  {
    int namelen=strcspn(arg,"=");
    if      (!strncmp(arg,"query", namelen)) query = &arg[namelen+1];
    else if (!strncmp(arg,"what",  namelen)) what  = &arg[namelen+1];
    else if (!strncmp(arg,"which", namelen)) which = &arg[namelen+1];
    else if (!strncmp(arg,"help",  namelen)) help  = &arg[namelen+1];
    else if (!strncmp(arg,"page",  namelen)) page  = atoi(&arg[namelen+1]);
    arg=strtok(0,"&");
  }

  wordOnly=which ? strcmp(which,"word")==0 : FALSE;
  nameOnly=what  ? strcmp(what, "name")==0 : FALSE;
  helpOnly=help  ? strcmp(help, "on"  )==0 : FALSE;
  
  // store encoded query string
  if (query) strcpy(encQueryInput,query); else encQueryInput[0]='\0';

  // convert query string to original input
  char *s=query,*d=queryInput;
  if (s)
  {
    while (*s!='\0')
    {
      char c=*s++;
      if (c=='+') 
        *d++=' ';
      else if (c=='%')
      {
        *d=asciiToHex(*s++)*16;
        *d+=asciiToHex(*s++);
        d++;
      }
      else
        *d++=c;
    }
  }
  *d='\0';

  if (helpOnly) 
  {
    generateHelpPage();
    exit(1);
  }

  // read search index files
  int argIndex=0;
  for (argIndex=1;argIndex<argc;argIndex++)
  {
    char configFile[MAXSTRLEN];
    strcpy(configFile,argv[argIndex]);
#if defined(_WIN32)
    strcat(configFile,"\\search.cfg");
#else
    strcat(configFile,"/search.cfg");
#endif

    char indexFile[MAXSTRLEN];
    strcpy(indexFile,argv[argIndex]);
#if defined(_WIN32)
    strcat(indexFile,"\\search.idx");
#else
    strcat(indexFile,"/search.idx");
#endif

    FileInfo *fi=fileList.add();
    FILE *g;
    
    strcpy(fi->name,indexFile);
    if ((fi->f=fopen(indexFile,"rb"))==NULL)
    {
      message("Error: could not open index file %s\n",indexFile);
      exit(1);
    } 
    if ((g=fopen(configFile,"r"))==NULL)
    {
      message("Error: could not open config file %s\n",configFile);
      exit(1);
    }
    
    // get URL to the documentation
    char tmp[MAXSTRLEN];
    fgets(tmp,MAXSTRLEN,g);
    if (tmp[strlen(tmp)-1]=='\n') tmp[strlen(tmp)-1]='\0';
    if (tmp[strlen(tmp)-1]=='/')  tmp[strlen(tmp)-1]='\0';
    fi->url = new char[strlen(tmp)+1];
    strcpy(fi->url,tmp);
    fclose(g);
    
    // read & check the format of the search index file
    fseek(fi->f,0,SEEK_SET);
    char header[5];
    if (fread(header,1,4,fi->f)!=4) 
    {
      message("Error: Couldn't read header of the index file %s\n",indexFile);
      exit(1);
    }
    header[4]='\0';
    if (strcmp(header,"DOXI"))
    {
      message("Error: Index file %s has an unknown format\n",indexFile);
      exit(1);
    }
    // read and store the offset to the link index
    fi->refOffset=readInt(fi->f);
    fclose(fi->f);
    fi->f = 0;
  }

  char *word;
  char wordString[MAXSTRLEN];

  // search for included words
  strlowercpy(wordString,queryInput);
  word=strtok(wordString," "); 
  bool first=TRUE;
  while (word)
  {
    if (word[0]=='+') // + character => include
    {
      SearchResults sr;
      searchIndex(&word[1],&sr);
      if (first)
        mergeSearchResults(&tsr,&sr,NORMAL);
      else
        mergeSearchResults(&tsr,&sr,INCLUDE);
      first=FALSE;
    }
    word=strtok(0," ");
  }

  // search for normal words
  strlowercpy(wordString,queryInput);
  word=strtok(wordString," "); 
  while (word)
  {
    if (word[0]!='-' && word[0]!='+') // normal word
    {
      SearchResults sr;
      searchIndex(word,&sr);
      mergeSearchResults(&tsr,&sr,NORMAL);
    }
    word=strtok(0," ");
  }

  // search for excluded words
  strlowercpy(wordString,queryInput);
  word=strtok(wordString," "); 
  while (word)
  {
    if (word[0]=='-') // - character => exclude
    {
      SearchResults sr;
      searchIndex(&word[1],&sr);
      mergeSearchResults(&tsr,&sr,EXCLUDE);
    }
    word=strtok(0," ");
  }

  // write results to HTML page
  generateResults(&tsr);

  free(headerBuf);
  free(footerBuf);

#ifdef PROFILING
  gettimeofday(&tv_end,0);
  printf("processing time %3.3f msec\n",
      ((tv_end.tv_sec-tv_start.tv_sec)*1000000+
        tv_end.tv_usec-tv_start.tv_usec)/1000.0
        );
#endif
  return 0;
}