Apache
/
olingo-odata4
mirror of https://github.com/apache/olingo-odata4.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

catch for bad characters in server URI parsing 

Change-Id: I8bff20757e4cb0060d7b613e8d79f9580fc3a267

Signed-off-by: Christian Amend <chrisam@apache.org>
This commit is contained in:
Klaus Straubinger 2014-10-07 15:11:42 +02:00 committed by Christian Amend
parent 9c60663523
commit 03475605eb
3 changed files with 35 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

47
lib/server-core/src/main/antlr4/org/apache/olingo/server/core/uri/antlr/UriLexer.g4
View File

@ -19,9 +19,9 @@
lexer grammar UriLexer;
//;==============================================================================
// Mode "DEFAULT_MODE": Processes everything bevor the first '?' char
// On '?' the next mode "MODE_QUERY" is used
// The percent encoding rules a defined in RFC3986 ABNF rule "path-rootless" apply
// Mode "DEFAULT_MODE": Processes everything before the first '?' char.
// On '?' the next mode "MODE_QUERY" is used.
// The percent encoding rules are defined in RFC3986; ABNF rule "path-rootless" applies.
//;==============================================================================
QM : '?' -> pushMode(MODE_QUERY); //first query parameter
AMP : '&' -> pushMode(MODE_QUERY); //more query parameters
@ -99,7 +99,7 @@ fragment DAY : '0' '1'..'9' | ('1'|'2') DIGIT | '3' ('0'|'1');
fragment MONTH : '0' ONE_TO_NINE | '1' ( '0' | '1' | '2' );
fragment YEAR : ('-')? ( '0' DIGIT DIGIT DIGIT | ONE_TO_NINE DIGIT DIGIT DIGIT );
//tags start with $
//tags starting with $
BATCH : '$batch';
ENTITY : '$entity';
METADATA : '$metadata';
@ -116,15 +116,13 @@ TOP_I : '$top' -> type(TOP);
SKIP_I : '$skip' -> type(SKIP);
FILTER_I : '$filter' -> type(FILTER);
ORDERBY_I: '$orderby' -> type(ORDERBY);
SELECT_I: '$select' -> type(SELECT);
EXPAND_I: '$expand' -> type(EXPAND);
LEVELS_I: '$levels' -> type(LEVELS);
SELECT_I : '$select' -> type(SELECT);
EXPAND_I : '$expand' -> type(EXPAND);
LEVELS_I : '$levels' -> type(LEVELS);
MAX: 'max';
ROOT : '$root/';
//rest
NULLVALUE : 'null';
@ -188,7 +186,7 @@ MINUS :'-';
IT : '$it';
ITSLASH : '$it/';
LEVELS : '$levels';
LEVELS : '$levels';
CONTAINS_WORD : 'contains(';
STARTSWITH_WORD : 'startswith(';
@ -237,13 +235,15 @@ LINK : '$link';
DELETED_LINK : '$deletedLink';
DELTA : '$delta';
//ODI
ODATAIDENTIFIER : ODI_LEADINGCHARACTER (ODI_CHARACTER)*;
ODATAIDENTIFIER : ODI_LEADINGCHARACTER (ODI_CHARACTER)*;
//handle characters that failed to match any other token
ERROR_CHARACTER : .;
//;==============================================================================
// Mode "QUERY": Processes everything between the first '?' and the '#' char
// On '?' the next mode "FRAGMENT" is used
// The percent encoding rules a defined in RFC3986 ABNF rule "query" apply
// Mode "QUERY": Processes everything between the first '?' and the '#' char.
// On '?' the next mode "FRAGMENT" is used.
// The percent encoding rules are defined in RFC3986; ABNF rule "query" applies.
mode MODE_QUERY;
//;==============================================================================
@ -270,6 +270,9 @@ AT_Q : '@' -> pushMode(DEFAULT_MODE);
CUSTOMNAME : ~[&=@$] ~[&=]* -> pushMode(MODE_SYSTEM_QUERY_REST);
//handle characters that failed to match any other token
ERROR_CHARACTER_q : .;
//;==============================================================================
mode MODE_SYSTEM_QUERY_PCHAR;
//;==============================================================================
@ -287,18 +290,19 @@ fragment UNRESERVED_sqp : ALPHA_sqp | DIGIT_sqp | '-' |'.' | '_' | '~';
fragment PCHAR : UNRESERVED_sqp | PCT_ENCODED_sqp | SUB_DELIMS_sqp | ':' | '@';
fragment PCHARSTART : UNRESERVED_sqp | PCT_ENCODED_sqp | '$' | /*'&' |*/ '\'' | OTHER_DELIMS_sqp | ':' | '@';
ATOM : [Aa][Tt][Oo][Mm];
JSON : [Jj][Ss][Oo][Nn];
XML : [Xx][Mm][Ll];
PCHARS : PCHARSTART PCHAR*;
SLASH_sqp : '/' -> type(SLASH);
EQ_sqp : '=' -> type(EQ);
FRAGMENT_sqp : '#' -> type(FRAGMENT), pushMode(MODE_FRAGMENT);
//handle characters that failed to match any other token
ERROR_CHARACTER_sqp : .;
//;==============================================================================
mode MODE_FRAGMENT;
// Read the remaining characters of a URI queryparameter up to an & or #
@ -336,19 +340,18 @@ QUOTATION_MARK_sqc : '\u0022' | '%22';
SEARCHWORD : ('a'..'z'|'A'..'Z')+;
SEARCHPHRASE : QUOTATION_MARK_sqc ~["]* QUOTATION_MARK_sqc -> popMode;
//;==============================================================================
mode MODE_STRING;
// Read the remaining characters up to an ' character.
// An "'" character inside a string are expressed as double ''
// Reads the remaining characters up to an ' character.
// Any "'" characters inside a string are expressed as double "''".
//;==============================================================================
STRING_s : ('\'\'' | ~[\u0027] )* '\'' -> type(STRING), popMode;
//;==============================================================================
mode MODE_JSON_STRING;
// Read the remaining characters up to an " character.
// An "'" character inside a string are expressed excaped \"
// Reads the remaining characters up to an " character.
// Any """ characters inside a string are escaped with "\".
//;==============================================================================
STRING_IN_JSON : ('\\"' | ~[\u0022] )* ('"' | '%22') -> popMode;

14
lib/server-core/src/main/antlr4/org/apache/olingo/server/core/uri/antlr/UriParser.g4
View File

@ -20,16 +20,16 @@
grammar UriParser;
//------------------------------------------------------------------------------
// This grammer refers to the "odata-abnf-construction-rules.txt" Revision 517.
// This grammar refers to the "odata-abnf-construction-rules.txt" Revision 517.
// URL: https://tools.oasis-open.org/version-control/browse/wsvn/odata/trunk/spec/ABNF/odata-abnf-construction-rules.txt?rev=517
// While contructing this grammer we tried to keep it close to the ABNF. However this is not really possible
// in order to support
// While contructing this grammar we tried to keep it close to the ABNF.
// However this is not really possible in order to support
// - percent decoding
// - operator precedence
// - have context free grammar ( without out java snipples to add context)
// - generate the parser in diffend target languages
// Currently not supported
// - having a context free grammar ( without java snipplets to add context)
// - generating the parser in different target languages
// Currently not supported are
// - $search
// - geometry data
// - json data in url
@ -54,7 +54,7 @@ metadataEOF : METADATA EOF;
//;------------------------------------------------------------------------------
//; 1. Resource Path
//;------------------------------------------------------------------------------
//resourcePathEOF : vlPS=pathSegments EOF;
crossjoinEOF : CROSSJOIN OPEN WSP? vlODI+=odataIdentifier WSP? ( COMMA WSP? vlODI+=odataIdentifier WSP?)* CLOSE EOF;

3
lib/server-test/src/test/java/org/apache/olingo/server/core/uri/antlr/TestFullResourcePath.java
View File

@ -2638,6 +2638,9 @@ public class TestFullResourcePath {
testUri.runEx("$entity/olingo.odata.test1.ETKeyNav/$ref")
.isExSyntax(UriParserSyntaxException.MessageKeys.MUST_BE_LAST_SEGMENT);
testUri.runEx("$wrong").isExSyntax(UriParserSyntaxException.MessageKeys.SYNTAX);
testUri.runEx("", "$wrong").isExSyntax(UriParserSyntaxException.MessageKeys.UNKNOWN_SYSTEM_QUERY_OPTION);
testUri.run("ESKeyNav")
.isKind(UriInfoKind.resource)
.goPath().first()