fork download
copy 
  1. /* get_src.c */
  2. #include <stdio.h>
  3. #include <string.h>
  4. #include <ctype.h>
  5. #include <stdlib.h>
  6.  
  7. #define GET_SOURCE
  8. //#include "get_src.h"
  9.  
  10. /* get_src.h */
  11.  
  12. #ifndef GET_SRC_INCLUDED
  13. #define GET_SRC_INCLUDED
  14.  
  15. #include <stdio.h>
  16.  
  17. #ifdef __cplusplus
  18. extern "C" {
  19. #endif
  20.  
  21. /* This is the size of the largest token we'll attempt to deal with.  If
  22.  * you want to deal with bigger tokens, change this, and recompile
  23.  * get_src.c.  Note that an entire comment is treated as a single token,
  24.  * so long comments could overflow this.  In case of an overflow, the
  25.  * entire comment will be read as a single token, but the part larger
  26.  * than this will not be stored.
  27.  */
  28. #define MAX_TOKEN_SIZE 8192
  29.  
  30. /* `last_token' will contain the text of the most recently read token.
  31.  */
  32. extern char last_token[];
  33.  
  34. /* This is the maximum number of characters that can be put back into a
  35.  * file opened with parse_fopen or parse_fdopen.
  36.  */
  37. #define MAX_UNGETS 5
  38.  
  39. #include <limits.h>
  40. #include <stdio.h>
  41.  
  42. typedef struct {
  43.     FILE *file;
  44.     char peeks[MAX_UNGETS];
  45.     int last_peek;
  46. } PFILE;
  47.  
  48. /* Some codes we return to indicate having found various items in the
  49.  * source code.  ERROR is returned to indicate a newline found in the
  50.  * middle of a character or string literal or if a file ends inside a
  51.  * comment, or if a character literal contains more than two characters.
  52.  *
  53.  * Note that this starts at INT_MIN, the most negative number available
  54.  * in an int.  This keeps these symbols from conflicting with any
  55.  * characters read from the file.  However, one of these could
  56.  * theoretically conflict with EOF.  EOF usually -1, and these are far
  57.  * more negative than that.  However, officially EOF can be any value
  58.  * less than 0...
  59.  */
  60. enum {
  61.     ERROR = INT_MIN,
  62.     COMMENT,
  63.     CHAR_LIT,
  64.     STR_LIT,
  65.     IDENT,  /* This is for any word except those below...       */
  66.     CASE,
  67.     DEFAULT,
  68.     DO,
  69.     ELSE,
  70.     IF,
  71.     SWITCH,
  72.     WHILE,
  73.     INCLUDE,
  74.     DEFINE
  75. };
  76.  
  77. /* Opens a file for parsing and returns a pointer to a structure which
  78.  * can be passed to the other functions in the parser/lexer to identify
  79.  * the file being worked with.
  80.  */
  81. PFILE *parse_fopen(char const *name);
  82.  
  83. /* This corresponds closely to fdopen - it takes a FILE * as its
  84.  * only parameter, creates a PFILE structure identifying that file, and
  85.  * returns a pointer to that structure.
  86.  */
  87. PFILE *parse_fdopen(FILE *stream);
  88.  
  89. /* Corresponds to fclose.
  90.  */
  91. int parse_fclose(PFILE *stream);
  92.  
  93. /* returns characters from `stream' read as C source code.  String
  94.  * literals, characters literals and comments are each returned as a
  95.  * single code from those above.  All strings of any kind of whitespace
  96.  * are returned as a single space character.
  97.  */
  98. int get_source(PFILE *stream);
  99.  
  100. /* As above, but adds classification of some C keywords as well.  The
  101.  * keywords recognized are mostly for flow control and are those listed
  102.  * in the enumeration above, following IDENT.  All identifiers and
  103.  * unrecognized keywords are returned as IDENT.
  104.  */
  105. int get_token(PFILE *stream);
  106.  
  107. /* If called with a value of 0, turns off recognition of C++ digraphs
  108.  * and single line comments.  If called with a non-zero value, turns
  109.  * on recognition of same.  Default is 1.
  110.  */
  111. //void read_CPP(int status);
  112.  
  113. /* Basically, these two work just like the normal versions of the same,
  114.  * with the minor exception that unget_character can unget more than one
  115.  * character.
  116.  */
  117. int get_character(PFILE *stream);
  118. void unget_character(int ch, PFILE *stream);
  119.  
  120. #ifdef __cplusplus
  121. }
  122. #endif
  123.  
  124. #endif
  125.  
  126. #if 1
  127. /* These are the keywords we recognize - those that involve flow control.
  128.  * recognition of all keywords simply involves adding them to this list,
  129.  * and adding matching identifiers to the enumeration in get_src.h.  The
  130.  * matching enumerators for these keywords start after IDENT in the
  131.  * enumeration, but MUST be maintained in the same order as the keywords
  132.  * appear here.  E.g. if "case" remains the first keyword here, `CASE'
  133.  * must follow immediately after IDENT in the enumeration.  Any types
  134.  * added to the enumeration that do not have matching keywords should
  135.  * precede `IDENT'.
  136.  */
  137. static char *keys[] = {
  138.     "case",
  139.     "default",
  140.     "do",
  141.     "else",
  142.     "if",
  143.     "switch",
  144.     "while",
  145.     "#include",
  146.     "#define"
  147. };
  148.  
  149. #define elems(x) (sizeof(x) / sizeof(x[0]))
  150. #endif
  151.  
  152. static size_t current = 0;
  153.  
  154. char last_token[MAX_TOKEN_SIZE];
  155.  
  156. PFILE *parse_fopen(char const *name) {
  157.  
  158.     PFILE *temp = malloc(sizeof(PFILE));
  159.  
  160.     if ( NULL != temp ) {
  161.         temp->file = fopen(name, "r");
  162.         memset(temp->peeks, 0, sizeof(temp->peeks));
  163.         temp->last_peek = 0;
  164.     }
  165.     return temp;
  166. }
  167.  
  168. PFILE *parse_fdopen(FILE *file) {
  169.  
  170.     PFILE *temp = malloc(sizeof(PFILE));
  171.  
  172.     if ( NULL != temp) {
  173.         temp->file = file;
  174.         memset(temp->peeks, 0, sizeof(temp->peeks));
  175.         temp->last_peek = 0;
  176.     }
  177.     return temp;
  178. }
  179.  
  180. int parse_fclose(PFILE *stream) {
  181.  
  182.     int retval = fclose(stream->file);
  183.  
  184.     free(stream);
  185.     return retval;
  186. }
  187.  
  188. static void addchar(int ch) {
  189. /* adds the passed character to the end of `last_token' */
  190.  
  191.     if ( current < sizeof(last_token) -1 )
  192.         last_token[current++] = (char)ch;
  193.  
  194.     if ( current == sizeof(last_token)-1 )
  195.         last_token[current] = '\0';
  196. }
  197.  
  198. static void clear(void) {
  199. /* clears the previous token and starts building a new one. */
  200.     current = 0;
  201. }
  202.  
  203. static int read_char(PFILE *stream) {
  204.     if ( stream->last_peek > 0 )
  205.         return stream->peeks[--stream->last_peek];
  206.     return fgetc(stream->file);
  207. }
  208.  
  209. void unget_character(int ch, PFILE * stream) {
  210.     if ( stream->last_peek < sizeof(stream->peeks) )
  211.         stream->peeks[stream->last_peek++] = ch;
  212. }
  213.  
  214. /* Here's where we start getting into sort of sophisticated stuff.
  215.  */
  216.  
  217. static int check_trigraph(PFILE *stream) {
  218. /* Checks for trigraphs and returns the equivalant character if there
  219.  * is one.  Expects that the leading '?' of the trigraph has already
  220.  * been read before this is called.
  221.  */
  222.  
  223.     int ch;
  224.  
  225.     if ( '?' != (ch=read_char(stream))) {
  226.         unget_character(ch, stream);
  227.         return '?';
  228.     }
  229.  
  230.     ch = read_char(stream);
  231.  
  232.     switch( ch ) {
  233.         case '(':	return '[';
  234.         case ')':   return ']';
  235.         case '/':   return '\\';
  236.         case '\'':  return '^';
  237.         case '<':   return '{';
  238.         case '>':   return '}';
  239.         case '!':   return '|';
  240.         case '-':   return '~';
  241.         case '=':   return '#';
  242.         default:
  243.             unget_character('?', stream);
  244.             unget_character(ch, stream);
  245.             return '?';
  246.     }
  247. }
  248.  
  249. static int check_digraph(PFILE *stream, int first) {
  250. /* Checks for a digraph.  The first character of the digraph is
  251.  * transmitted as the second parameter, as there are several possible
  252.  * first characters of a digraph.
  253.  */
  254.  
  255.     int ch = read_char(stream);
  256.  
  257.     switch(first) {
  258.         case '<':
  259.             if ( '%' == ch )
  260.                 return '{';
  261.             if ( ':' == ch )
  262.                 return '[';
  263.             break;
  264.         case ':':
  265.             if ( '>' == ch )
  266.                 return ']';
  267.             break;
  268.         case '%':
  269.             if ( '>' == ch )
  270.                 return '}';
  271.             if ( ':' == ch )
  272.                 return '#';
  273.             break;
  274.     }
  275.  
  276. /* If it's not one of the specific combos above, return the characters
  277.  * separately and unchanged by putting the second one back into the
  278.  * stream, and returning the first one as-is.
  279.  */
  280.     unget_character(ch, stream);
  281.     return first;
  282. }
  283.  
  284. static int get_char(PFILE *stream) {
  285. /* Gets a single character from the stream with any trigraphs ( and if
  286.  * C++ support is turned on, digraphs ) converted to the single character
  287.  * represented.
  288.  */
  289.     int ch = read_char(stream);
  290.  
  291.     if ( ch == '?' )
  292.         return check_trigraph(stream);
  293.  
  294.     if (( ch == '<' || ch == ':' || ch == '%' ))
  295.         return check_digraph(stream, ch);
  296.  
  297.     return ch;
  298. }
  299.  
  300. int get_character(PFILE *stream) {
  301. /* gets a character from `stream'.  Any amount of any kind of whitespace
  302.  * is returned as a single space. Escaped new-lines are "eaten" here as well.
  303.  */
  304.     int ch;
  305.  
  306.     if ( !isspace(ch=get_char(stream)) && ch != '\\')
  307.         return ch;
  308.  
  309. 	// handle line-slicing
  310. 	if (ch == '\\') {
  311. 		ch = get_char(stream);
  312. 		if (ch == '\n') 
  313. 			ch = get_char(stream);
  314. 		else {
  315. 			unget_character(ch, stream);
  316. 			return ch;
  317. 		}
  318. 	}
  319.  
  320.     /* If it's a space, skip over consecutive white-space */
  321.     while (isspace(ch) && ('\n' != ch))
  322.         ch = get_char(stream);
  323.  
  324.     if ('\n' == ch)
  325.         return ch;
  326.  
  327.     /* Then put the non-ws character back */
  328.     unget_character(ch, stream);
  329.  
  330.     /* and return a single space character... */
  331.     return ' ';
  332. }
  333.  
  334. static int read_char_lit(PFILE *stream) {
  335. /* This is used internally by `get_source' (below) - it expects the
  336.  * opening quote of a character literal to have already been read and
  337.  * returns CHAR_LIT or ERROR if there's a newline before a close
  338.  * quote is found, or if the character literal contains more than two
  339.  * characters after escapes are taken into account.
  340.  */
  341.  
  342.     int ch;
  343.     int i;
  344.  
  345.  
  346.     clear();
  347.     addchar('\'');
  348.  
  349.     for (i=0; i<2 && ('\'' != ( ch = read_char(stream))); i++) {
  350.  
  351.         addchar(ch);
  352.  
  353.     	if ( ch == '\n' )
  354.             return ERROR;
  355.  
  356.         if (ch == '\\' ) {
  357.             ch = get_char(stream);
  358.             addchar(ch);
  359.         }
  360.     }
  361.     addchar('\'');
  362.     addchar('\0');
  363.  
  364.     if ( i > 2 )
  365.         return ERROR;
  366.  
  367.     return CHAR_LIT;
  368. }
  369.  
  370. static int read_str_lit(PFILE *stream) {
  371. /* Used internally by get_source.  Expects the opening quote of a string
  372.  * literal to have already been read.  Returns STR_LIT, or ERROR if a
  373.  * un-escaped newline is found before the close quote.
  374.  */
  375.  
  376.     int ch;
  377.  
  378.     clear();
  379.     addchar('"');
  380.  
  381.     while ( '"' != ( ch = get_char(stream))) {
  382.  
  383.         if ( '\n' == ch || EOF == ch )
  384.             return ERROR;
  385.  
  386.         addchar(ch);
  387.  
  388.         if( ch == '\\' ) {
  389.             ch = read_char(stream);
  390.             addchar(ch);
  391.         }
  392.  
  393.     }
  394.  
  395.     addchar('"');
  396.     addchar('\0');
  397.  
  398.     return STR_LIT;
  399. }
  400.  
  401. static int read_comment(PFILE *stream) {
  402. /* Skips over a comment in stream.  Assumes the leading '/' has already
  403.  * been read and skips over the body.  If we're reading C++ source, skips
  404.  * C++ single line comments as well as normal C comments.
  405.  */
  406.     int ch;
  407.  
  408.     clear();
  409.  
  410.     ch = get_char(stream);
  411.  
  412.     /* Handle a single line comment.
  413.      */
  414.     if ('/' == ch) {
  415.         addchar('/');
  416.         addchar('/');
  417.  
  418.         while ( '\n' != ( ch = get_char(stream))) 
  419.             addchar(ch);       
  420.  
  421.         addchar('\0');
  422.         return COMMENT;
  423.     }
  424.  
  425.     if ('*' != ch ) {
  426.         unget_character(ch, stream);
  427.         return '/';
  428.     }
  429.  
  430.     addchar('/');
  431.  
  432.     do {
  433.         addchar(ch);
  434.         while ('*' !=(ch = get_char(stream)))
  435.             if (EOF == ch)
  436.                 return ERROR;
  437.             else
  438.                 addchar(ch);
  439.         addchar(ch);
  440.     } while ( '/' != (ch=get_char(stream)));
  441.  
  442.     addchar('/');
  443.     addchar('\0');
  444.  
  445.     return COMMENT;
  446. }
  447.  
  448. int get_source(PFILE *stream) {
  449. /* reads and returns a single "item" from the stream.  An "item" is a
  450.  * comment, a literal or a single character after trigraph and possible
  451.  * digraph substitution has taken place.
  452.  */
  453.  
  454.     int ch = get_character(stream);
  455.  
  456.     switch(ch) {
  457.         case '\'':
  458.             return read_char_lit(stream);
  459.         case '"':
  460.             return read_str_lit(stream);
  461.         case '/':
  462.             return read_comment(stream);
  463.         default:
  464.             return ch;
  465.     }
  466. }
  467. #if TOKEN
  468. int get_token(PFILE *stream) {
  469. /* This gets a single token from the input stream and places the text
  470.  * of the token in last_token, and returns an identifier of the type of
  471.  * the token.  Only flow control keywords are recognized individually.
  472.  * All other keywords are simply returned as IDENT's, just like other
  473.  * identifiers.
  474.  */
  475.  
  476.     int ch;
  477.     int i;
  478.  
  479.     ch = get_source(stream);
  480.  
  481.     /* If we've got an identifier, read as many characters as can
  482.      * possibly constitute the identifier ( maximal munch ) and build
  483.      * up the complete identifier in `last_token'
  484.      */
  485.     if ( ch > 0 && ('_' == ch || isalpha(ch))) {
  486.         clear();
  487.         while(ch > 0 && (isalpha(ch) || isdigit(ch) || '_' == ch )) {
  488. 	        addchar(ch);
  489.             ch = get_source(stream);
  490.         }
  491.         unget_character(ch,stream);
  492.  
  493.         addchar('\0');
  494.  
  495.         /* Now we look in our table to see if we've got a keyword
  496.          * we recognize, or some random identifier.
  497.          */
  498.         for (i=0;i<elems(keys);i++) {
  499.             if ( 0 == strcmp(last_token, keys[i]))
  500.                 return IDENT+i+1;
  501.         }
  502.  
  503.         /* we didn't recognize it - it must be a normal identifier. */
  504.         return IDENT;
  505.     }
  506.  
  507.     /* it's not an identifier - just return it as a character. */
  508.     return ch;
  509. }
  510. #endif
  511. #ifdef TEST
  512.  
  513. int main(int argc, char **argv)  {
  514. 	PFILE *f;
  515. 	int ch;
  516.  
  517. 	if (argc != 2) {
  518. 		fprintf(stderr, "Usage: get_src <filename>\n");
  519. 		return EXIT_FAILURE;
  520. 	}
  521.  
  522. 	if (NULL==(f= parse_fopen(argv[1]))) {
  523. 		fprintf(stderr, "Unable to open: %s\n", argv[1]);
  524. 		return EXIT_FAILURE;
  525. 	}
  526.  
  527. 	while (EOF!=(ch=get_source(f))) 
  528. 		if (ch < 0) 
  529. 			printf("\n%s\n", last_token);
  530. 		else
  531. 			printf("%c", ch);
  532. 	parse_fclose(f);
  533. 	return 0;		
  534. }
  535.  
  536. #endif
  537.  
Compilation error #stdin compilation error #stdout 0s 0KB
comments (?)
stdin
copy 
Standard input is empty
compilation info 
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp:147:1: warning: deprecated conversion from string constant to ‘char*’ [-Wwrite-strings]
prog.cpp: In function ‘PFILE* parse_fopen(const char*)’:
prog.cpp:158:39: error: invalid conversion from ‘void*’ to ‘PFILE*’ [-fpermissive]
prog.cpp: In function ‘PFILE* parse_fdopen(FILE*)’:
prog.cpp:170:39: error: invalid conversion from ‘void*’ to ‘PFILE*’ [-fpermissive]
prog.cpp: In function ‘void unget_character(int, PFILE*)’:
prog.cpp:210:50: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
prog.cpp: At global scope:
prog.cpp:137:14: warning: ‘keys’ defined but not used [-Wunused-variable]
stdout
copy 
Standard output is empty
https://ideone.com/bJaXSB
language:
C++14 (gcc 8.3)
created:
12 years ago
visibility:
public

Share or Embed source code

Discover > Sphere Engine API

The brand new service which powers Ideone!

Discover > IDE Widget

Widget for compiling and running the source code in a web browser!