Internal NO_BACKSLASH_ESCAPES support #702
Conversation
…l and no explicit statement terminator, like this one: select 'foo' Signed-off-by: Laurents Meyer <[email protected]>
…derived classes and MySqlParameter. Single quotes are now always escaped with an additional single quote instead of a leading backspace. This addresses everything except triggering/detecting the NO_BACKSLASH_ESCAPES from mysql-net#701 and PomeloFoundation/Pomelo.EntityFrameworkCore.MySql#827 Signed-off-by: Laurents Meyer <[email protected]>
…o the existing test. Signed-off-by: Laurents Meyer <[email protected]>
lauxjpn
force-pushed
the
no_backslash_escapes
branch
2 times, most recently
from
fbb913d to
c17ee05
Compare
|
I got the tests running on my machine now. I will therefor add a couple of test cases to the suite and remove the console test project from this PR. |
…quote instead of leading backslash). Signed-off-by: Laurents Meyer <[email protected]>
lauxjpn
force-pushed
the
no_backslash_escapes
branch
from
c17ee05 to
1618786
Compare
src/MySqlConnector/Core/SqlParser.cs
Outdated
| else if (state == State.SingleQuotedStringSingleQuote) | ||
| { | ||
| state = State.Statement; | ||
| } |
There was a problem hiding this comment.
This is a general class of error I can push a fix for (before this PR is merged).
There was a problem hiding this comment.
5060d9d; will merge to master soon.
|
The Travis CI tests time out for PRs. I haven't figured out why this is happening (they just seem to run uniformly slower, and I don't know why), but you can ignore them. Long-term, I plan to migrate to Azure Pipelines. Short-term, I suppose I could just disable Travis CI on PRs, since I know it'll fail. |
I'll take a quick look at it, if I can figure out why this is happening there. |
lauxjpn
force-pushed
the
no_backslash_escapes
branch
2 times, most recently
from
35abfaf to
cdb0c63
Compare
Signed-off-by: Laurents Meyer <[email protected]>
lauxjpn
force-pushed
the
no_backslash_escapes
branch
from
cdb0c63 to
00aea2a
Compare
|
The test suite just seems to take to long. Nothing really hangs, the individual tests are just getting executed extremely slow. I have very little experience with Travis CI, so I can't say if this is expected. The tests could be split into multiple test ENVs and then executed separately/parallel. But it's probably best to just make the switch to Azure DevOps, as @caleblloyd just did here: PomeloFoundation/Pomelo.EntityFrameworkCore.MySql#823 |
Signed-off-by: Laurents Meyer <[email protected]>
Signed-off-by: Laurents Meyer <[email protected]>
|
Okay, this came up in a search, regarding Travis CI: travis-ci/travis-ci#3049 (comment) |
lauxjpn
force-pushed
the
no_backslash_escapes
branch
3 times, most recently
from
b195329 to
1813fe9
Compare
|
Those are interesting links, but it seems like that problem has been ongoing for two (or more) years. From PR history, I know that Travis was working in April, but was timing out in May, so the change (for this project) happened in that timeframe. I need to review what Travis is testing that I haven't replicated on Azure Pipelines and either finish copying it, or decide to leave it untested. |
lauxjpn
force-pushed
the
no_backslash_escapes
branch
3 times, most recently
from
4e4b4c4 to
73bd33f
Compare
Signed-off-by: Laurents Meyer <[email protected]>
lauxjpn
force-pushed
the
no_backslash_escapes
branch
from
73bd33f to
e14a3d2
Compare
|
I ended up disabling most of the tests on Travis CI. |
|
The new tests look good--thanks! Do you have any preference on how I merge/squash/rebase this? |
|
Just squash it. |
|
Merged via d550a69. |
I included the test (console) project I used to test the functionality before implementing this PR against MySqlConnector and Connector/ODBC.
These need to be integrated into the actual test suite.
I just can't get this to run without VS 16.3 (which will be released with .NET Core 3.0 on Monday, i presume).Generally, all existing tests should work with
NO_BACKSLASH_ESCAPESenabled as well.This addresses #701 and PomeloFoundation/Pomelo.EntityFrameworkCore.MySql#827