[MOD] Drop old Parser API, add new ID parser

src/Parser.c

@@ -24,149 +24,47 @@
 
 #include <Parser.h>
 
-#include <Cytoplasm/Memory.h>
 #include <Cytoplasm/Str.h>
-#include <Cytoplasm/Log.h>
 
 #include <string.h>
-#include <sys/types.h>
-
-struct Parser {
-    char *string;
-    size_t idx;
-
-    ssize_t mark;
-
-    int eof;
-};
-
-Parser *
-ParserCreate(char * str)
-{
-    Parser * ret;
-    if (!str)
-    {
-        return NULL;
-    }
-
-    ret = Malloc(sizeof(Parser));
-
-    ret->string = StrDuplicate(str);
-    ret->idx = 0;
-    ret->eof = 0;
-    ret->mark = -1;
-
-    return ret;
-}
-
-void
-ParserStartCopy(Parser * p)
-{
-    if (!p)
-    {
-        return;
-    }
-    p->mark = p->idx;
-}
-char *
-ParserEndCopy(Parser * p)
-{
-    char *ret;
-    size_t len;
-    if (!p || p->mark < 0)
-    {
-        return NULL;
-    }
-
-    len = p->idx - p->mark;
-
-    ret = Malloc(len + 1);
-    memcpy(ret, p->string + p->mark, len);
-    ret[len] = '\0';
-
-    return ret;
-}
 
 int
-ParserExcept(Parser * p, char c)
+ParseCommonID(char *str, CommonID *id)
 {
-    char strc;
+    char sigil;
-    if (!p || p->eof)
+    char *servstart;
+
+    if (!str || !id)
     {
         return 0;
     }
     
-    strc = p->string[p->idx++];
+    /* There must at least be 2 chararacters: the sigil and a string.*/
-    if (!strc)
+    if (strlen(str) < 2)
-    {
-        p->eof = 1;
-        p->idx--; /* Avoid overflows */
-    }
-    return strc == c;
-}
-int
-ParserExceptOneOf(Parser * p, const char * str)
-{
-    char strc;
-    if (!p || !str || p->eof)
     {
         return 0;
     }
 
-    strc = p->string[p->idx++];
+    sigil = *str++;
-    if (!strc)
+    /* Some sigils have the following restriction:
+     * > MUST NOT exceed 255 bytes (including the # sigil and the domain).
+     */
+    if ((sigil == '#' || sigil == '@') && strlen(str) > 255)
     {
-        p->eof = 1;
+        return 0;
-        p->idx--; /* Avoid overflows */
     }
-    if (strchr(str, strc))
+    id->sigil = sigil;
-    {
-        return strc;
-    }
-    return 0;
-}
 
-int
+    /* Find the position of a ':' if there is one. */
-ParserGetc(Parser * p)
+    if ((servstart = strchr(str, ':')))
-{
-    int c;
-    if (!p || p->eof)
     {
-        return EOF;
+        id->local   = StrSubstr(str, 1, (char) (servstart - str));
-    }
+        id->server  = StrDuplicate(servstart + 1);
-    c = p->string[p->idx++];
+        return 1;
-    if (!c)
-    {
-        p->eof = 1;
-        p->idx--; /* Avoid overflows */
-        c = EOF;
-    }
-    return c;
-}
-void
-ParserUndo(Parser * p)
-{
-    if (!p)
-    {
-        return;
     }
+    /* Otherwise, just take the rest of the string. */
+    id->local   = StrDuplicate(str + 1) ;
+    id->server  = NULL                  ;
     
-    if (!p->idx)
+    return 1;
-    {
-        return;
-    }
-    p->eof = 0;
-    p->idx--;
 }
-
-void
-ParserEnd(Parser * p)
-{
-    if (!p)
-    {
-        return;
-    }
-    Free(p->string);
-    Free(p);
-}
-

src/User.c

@@ -31,10 +31,6 @@
 #include <Cytoplasm/Int64.h>
 #include <Cytoplasm/UInt64.h>
 
-#include <Parser.h>
-
-#include <ctype.h>
-
 #include <string.h>
 
 struct User
@@ -886,89 +882,51 @@ finish:
     return arr;
 }
 
-static int
-UserIdParseLocal(Parser *p, char **ret)
-{
-    int c;
-    if (!p || !ret)
-    {
-        return 0;
-    }
-
-    ParserStartCopy(p);
-    /* Always assume the general case here: every ASCII char, 
-     * minus ':'. */
-    while (isascii((c = ParserGetc(p))) && c != ':')
-    {
-        /* Do nothing. */
-    }
-    if (c != EOF)
-    {
-        ParserUndo(p);
-    }
-    *ret = ParserEndCopy(p);
-    return 1;
-}
-static int
-UserIdParseServer(Parser *p, char **ret)
-{
-    int c;
-    if (!p || !ret)
-    {
-        return 0;
-    }
-
-    ParserStartCopy(p);
-    /* Allow *every* ASCII character. */
-    while (isascii((c = ParserGetc(p))))
-    {
-        /* Do nothing. */
-    }
-    if (c != EOF)
-    {
-        ParserUndo(p);
-    }
-    *ret = ParserEndCopy(p);
-    return 1;
-}
-
 UserId *
 UserIdParse(char *id, char *defaultServer)
 {
     UserId *userId;
-    Parser *p;
 
     if (!id)
     {
         return NULL;
     }
 
-    p = ParserCreate(id);
+    id = StrDuplicate(id);
-    userId = Malloc(sizeof(UserId));
+    if (!id)
-    userId->localpart = NULL;
+    {
-    userId->server = NULL;
+        return NULL;
+    }
 
+    userId = Malloc(sizeof(UserId));
     if (!userId)
     {
         goto finish;
     }
 
-    /* Fully-qualified user ID.
+    /* Fully-qualified user ID */
-     * TODO: Generalise this for parsing namespaced IDs, 
+    if (*id == '@')
-     * like room aliases and IDs. */
-    if (ParserExcept(p, '@'))
     {
-        if (!UserIdParseLocal(p, &userId->localpart)    || 
+        char *localStart = id + 1;
-            !ParserExcept(p, ':')                       || 
+        char *serverStart = localStart;
-            !UserIdParseServer(p, &userId->server))
+
+        while (*serverStart != ':' && *serverStart != '\0')
+        {
+            serverStart++;
+        }
+
+        if (*serverStart == '\0')
         {
-            if (userId->localpart)
-            {
-                Free(userId->localpart);
-            }
             Free(userId);
             userId = NULL;
+            goto finish;
         }
+
+        *serverStart = '\0';
+        serverStart++;
+
+        userId->localpart = StrDuplicate(localStart);
+        userId->server = StrDuplicate(serverStart);
     }
     else
     {
@@ -984,7 +942,7 @@ UserIdParse(char *id, char *defaultServer)
     }
 
 finish:
-    ParserEnd(p);
+    Free(id);
     return userId;
 }
 

src/include/Parser.h

@@ -26,73 +26,31 @@
 
 /***
  * @Nm Parser
- * @Nd Functions for dealing with simple grammars.
+ * @Nd Functions for dealing with grammars found in Matrix
- * @Dd November 24 2023
+ * @Dd November 25 2023
  * @Xr User
  *
  * The
  * .Nm
- * API provides a simple parsing interface for basic parsers.
+ * API provides an interface for parsing grammars within the 
+ * Matrix specification
  */
 
 /**
- * A lot of functions here use this opaque structure.
+ * A common identifier in the form '&local[:server]', where 
+ * & determines the *type* of the identifier.
  */
-typedef struct Parser Parser;
+typedef struct CommonID {
+    char sigil;
+    char *local;
+    char *server; /* Might be NULL for some sigils(e.g: room IDs >= v3) */
+} CommonID;
 
 /** 
- * Create a new parser with an input string.
+ * Parses a common identifier, as per the Common Identifier Format as defined 
+ * by the [matrix] specification.
  */
-extern Parser * ParserCreate(char *);
+extern int ParseCommonID(char *, CommonID *);
 
-/**
- * Returns true if the character at the current position in 
- * the parsing process is the one given in, and advances by 
- * one character eitherway.
- */
-extern int ParserExcept(Parser *, char);
-
-/**
- * Gets the currently pointed at character in the parser, then 
- * advances it by one, unless if hit by an EOF.
- */
-extern int ParserGetc(Parser *);
-
-/**
- * Rolls back the currently pointed character by one(unless already 
- * at the start.)
- */
-extern void ParserUndo(Parser *);
-
-/**
- * Returns a non-zero value if the character at the current position 
- * in the parsing process is one of the characters in the inputted 
- * string, and advances by one character eitherway. 
- *
- * If non-zero, the return value is the actual character.
- */
-extern int ParserExceptOneOf(Parser *, const char *);
-
-/**
- * Starts a clipping session. This stores an index into the 
- * parser, until 
- * .Fn ParserEndCopy 
- * is called.
- */
-extern void ParserStartCopy(Parser *);
-/**
- * Ends a clipping station created by 
- * .Fn ParserStartCopy 
- * if it was called beforehand into a new buffer to be freed 
- * with
- * .Fn Free .
- */
-extern char * ParserEndCopy(Parser *);
-
-/**
- * Frees a parser created with 
- * .Fn ParserCreate .
- */
-extern void ParserEnd(Parser *);
 
 #endif                             /* TELODENDRIA_PARSER_H */