[geany/geany] 756344: Rust: Parse character literals.

[SiegeLord]

Branch:      refs/heads/master
Author:      SiegeLord <slabode at aim.com>
Committer:   SiegeLord <slabode at aim.com>
Date:        Tue, 29 Jul 2014 14:51:36 UTC
Commit:      756344d901afba9e5a51ae099aa7c1fe9ecef019
             https://github.com/geany/geany/commit/756344d901afba9e5a51ae099aa7c1fe9ecef019

Log Message:
-----------
Rust: Parse character literals.

These were omitted by mistake. Caused bugs like '"' being interpreted as a
start/end of a string, '}' as the end of a block etc.


Modified Paths:
--------------
    tagmanager/ctags/rust.c

Modified: tagmanager/ctags/rust.c
40 lines changed, 40 insertions(+), 0 deletions(-)
===================================================================
@@ -292,6 +292,41 @@ static void scanRawString (lexerState *lexer)
 	}
 }
 
+/* This deals with character literals: 'n', '\n', '\uFFFF'; and lifetimes:
+ * 'lifetime. We'll use this approximate regexp for the literals:
+ * \' \\ [^']+ \' or \' [^'] \' or \' \\ \' \'. Either way, we'll treat this
+ * token as a string, so it gets preserved as is for function signatures with
+ * lifetimes. */
+static void scanCharacterOrLifetime (lexerState *lexer)
+{
+	vStringClear(lexer->token_str);
+	advanceAndStoreChar(lexer);
+
+	if (lexer->cur_c == '\\')
+	{
+		advanceAndStoreChar(lexer);
+		/* The \' \\ \' \' (literally '\'') case */
+		if (lexer->cur_c == '\'' && lexer->next_c == '\'')
+		{
+			advanceAndStoreChar(lexer);
+			advanceAndStoreChar(lexer);
+		}
+		/* The \' \\ [^']+ \' case */
+		else
+		{
+			while (lexer->cur_c != EOF && lexer->cur_c != '\'')
+				advanceAndStoreChar(lexer);
+		}
+	}
+	/* The \' [^'] \' case */
+	else if (lexer->cur_c != '\'' && lexer->next_c == '\'')
+	{
+		advanceAndStoreChar(lexer);
+		advanceAndStoreChar(lexer);
+	}
+	/* Otherwise it is malformed, or a lifetime */
+}
+
 /* Advances the parser one token, optionally skipping whitespace
  * (otherwise it is concatenated and returned as a single whitespace token).
  * Whitespace is needed to properly render function signatures. Unrecognized
@@ -334,6 +369,11 @@ static int advanceToken (lexerState *lexer, boolean skip_whitspace)
 			scanRawString(lexer);
 			return lexer->cur_token = TOKEN_STRING;
 		}
+		else if (lexer->cur_c == '\'')
+		{
+			scanCharacterOrLifetime(lexer);
+			return lexer->cur_token = TOKEN_STRING;
+		}
 		else if (isIdentifierStart(lexer->cur_c))
 		{
 			scanIdentifier(lexer);



--------------
This E-Mail was brought to you by github_commit_mail.py (Source: https://github.com/geany/infrastructure).