diff options
Diffstat (limited to 'lib9/tokenize.c')
| -rw-r--r-- | lib9/tokenize.c | 54 |
1 files changed, 51 insertions, 3 deletions
diff --git a/lib9/tokenize.c b/lib9/tokenize.c index 1b222c52..790b7c2a 100644 --- a/lib9/tokenize.c +++ b/lib9/tokenize.c @@ -3,14 +3,14 @@ static char qsep[] = " \t\r\n"; static char* -qtoken(char *s) +qtoken(char *s, char *sep) { int quoting; char *t; quoting = 0; t = s; /* s is output string, t is input string */ - while(*t!='\0' && (quoting || utfrune(qsep, *t)==nil)){ + while(*t!='\0' && (quoting || utfrune(sep, *t)==nil)){ if(*t != '\''){ *s++ = *t++; continue; @@ -40,6 +40,54 @@ qtoken(char *s) return t; } +static char* +etoken(char *t, char *sep) +{ + int quoting; + + /* move to end of next token */ + quoting = 0; + while(*t!='\0' && (quoting || utfrune(sep, *t)==nil)){ + if(*t != '\''){ + t++; + continue; + } + /* *t is a quote */ + if(!quoting){ + quoting = 1; + t++; + continue; + } + /* quoting and we're on a quote */ + if(t[1] != '\''){ + /* end of quoted section; absorb closing quote */ + t++; + quoting = 0; + continue; + } + /* doubled quote; fold one quote into two */ + t += 2; + } + return t; +} + +int +gettokens(char *s, char **args, int maxargs, char *sep) +{ + int nargs; + + for(nargs=0; nargs<maxargs; nargs++){ + while(*s!='\0' && utfrune(sep, *s)!=nil) + *s++ = '\0'; + if(*s == '\0') + break; + args[nargs] = s; + s = etoken(s, sep); + } + + return nargs; +} + int tokenize(char *s, char **args, int maxargs) { @@ -51,7 +99,7 @@ tokenize(char *s, char **args, int maxargs) if(*s == '\0') break; args[nargs] = s; - s = qtoken(s); + s = qtoken(s, qsep); } return nargs; |