Fix bug with parsing UART output

[XingMicrochip]

This PR fixes issue #317.

Problematic code snippet in GDBTargetDebugSession.ts:

            const eolChar: string =
                uart.eolCharacter === 'CRLF' ? '\r\n' : '\n';

            let tcpUartData = '';
            this.socket.on('data', (data: string) => {
                for (const char of data) {
                    if (char === eolChar) {
                        this.sendEvent(
                            new OutputEvent(tcpUartData + os.EOL, 'Socket')
                        );
                        tcpUartData = '';
                    } else {
                        tcpUartData += char;
                    }
                }
            });

If data is CoreTimer timer interrupt.\r\n - for example - without setting eolCharacter in launch configuration, the adapter uses default \n to parse data. It accumulates CoreTimer timer interrupt.\r in variable tcpUartData and then appends os.EOL at the end. On Windows, the final tcpUartData becomes CoreTimer timer interrupt.\r\r\n. The extra new line comes from here. To fix this, the adapter appends \n instead of \r\n because \r is already in tcpUartData.

I tried setting eolCharacter to CRLF in launch configuration. This didn't work. char is one character in length, but eolChar is two in length, so char === eolChar is never true. The adapter also can't parse two characters at a time because one burst of data might be CoreTimer timer interrupt.\r and the next one \n. If this happens, \r\n will never be detected together.

I have tested this change fixes the problem I described in issue #317.

[jonahgraham]

This change looks fine to me, but it would be nice to have a test that demonstrates it and prevents it regressing.

[XingMicrochip]

I will add a test.

[jonahgraham]

I will add a test.

Great! Seeing as there are already some failing tests with the change you provided it may be worth looking into whether there is already a test that just needs a modification.

[XingMicrochip]

Can you re-run the tests? @jonahgraham I don't know what the one that's failing is about.

[XingMicrochip]

All the tests are passing now. Can you release a new version on Open VSX for this change? Thanks.

[jonahgraham]

All the tests are passing now. Can you release a new version on Open VSX for this change? Thanks.

I'll try to look at it soon, but it may have to wait until Monday.

[XingMicrochip]

That works. Thank you.