statement.split(";") cannot be used with strings which contains symbo… #24
Conversation
… Charset and don't use default-value.
…l ";" like "vasya;petya". I think, it would be better to check for the end of line after ";" symbol. At least we would have an opportunity to use semicolons in text.
|
@olamy ? |
|
@stephenc @seblm @ajs6f @giejay @djarnis73 @Willem1987 @dmitrymurashenkov |
|
Looks good, but unfortunately I do not have the access rights to merge it |
|
This is not a fix in my opinion, it will quickly break with a cryptic error message (like today), I have a version that uses the cql lexer, so it should handle any wellformed cql. But I did not have time to polish it into a PR. The splitter code looks like this: protected static String[] splitStatementsUsingCqlLexer(String statements) {
ANTLRStringStream stream = new ANTLRStringStream(statements);
CqlLexer lexer = new CqlLexer(stream);
ArrayList<String> statementList = new ArrayList<String>();
StringBuffer currentStatement = new StringBuffer();
boolean inComment;
// Not the prettiest code i ever wrote, but it gets the job done.
for (Token token = lexer.nextToken(); token.getType() != Token.EOF; token = lexer.nextToken()) {
if (token.getText().equals(";")) {
// when we meet a ; terminate current statement and prepare the next
currentStatement.append(";");
statementList.add(currentStatement.toString());
currentStatement = new StringBuffer();
} else if (token.getType() == CqlLexer.STRING_LITERAL) {
// If we meet a string we should quote it and escape any enclosed ' as ''
currentStatement.append("'");
// TODO: There must be a cassandra util method somewhere that escapes a string for sql
currentStatement.append(token.getText().replaceAll("'", "''"));
currentStatement.append("'");
} else {
currentStatement.append(token.getText());
}
}
if (currentStatement.length() > 0 && currentStatement.toString().trim().length() > 0) {
statementList.add(currentStatement.toString());
}
return statementList.toArray(new String[statementList.size()]);
} |
|
Thanks. I will try to update my pull request with your code. |
This will use the cql lexer instead of blindly splitting on ;
|
I have pushed my changes to my clone, you can see full commit here: djarnis73@0ed68b2 i can make a PR for it if it looks good to you. |
|
@djarnis73 Your code much better but it is not finished yet. It does not work with my cql-code from my job for which I trying to adapt the plugin. And it still has problems with comments. I will try to fix it. I will make a pull request when I finish. |
Conflicts: src/main/java/org/codehaus/mojo/cassandra/AbstractCqlExecMojo.java
|
How about now? |
|
@stephenc @seblm @ajs6f @giejay @djarnis73 @Willem1987 @dmitrymurashenkov @olamy |
| * It is not enabled by default since has not been extensively tested. | ||
| * | ||
| * @parameter default-value=false | ||
| * @since 3.6 |
There was a problem hiding this comment.
since should be bumped to 3.7, since 3.6 is already released.
|
Fixed. |
|
@stephenc @seblm @ajs6f @giejay @djarnis73 @Willem1987 @dmitrymurashenkov @olamy |
|
I don't know why you think I can merge your work. I have no role in this project and have never contributed to it. |
|
I think It is better to fork this repo and produce my own version of this plugin. I have really no idea whom and how should I push to force him to merge this branch. |
…l ";" like "vasya;petya". I think, it would be better to check for the end of line after ";" symbol. At least we would have an opportunity to use semicolons in text.
Example for test:
https://github.com/urvanov-ru/cassandra-maven-plugin-semicolon-test