Allow setting database name and user when using JDBC URL #594

manikmagar · 2018-02-26T05:10:49Z

This PR implements #566.

Extracts out JDBC Url manipulations to a separate class - ConnectionUrl.
Adds an overloaded method JdbcDatabaseContainerProvider.newInstance(ConnectionUrl), with default implementation delegating to the existing newInstance(tag) method.
Adds an implementation of newInstance(ConnectionUrl) in MySQLContainerProvider to use Database Name, user, and password from JDBC URL while creating new Container.
Added/Updated Test cases for new functionality.
Bug Fix: Query Parameters are not used when a connection is created from a cached container. (see this comment) Fixes db extra connection params are only passed to the first connection in a pool #610

Note: Other Containers Providers such as PostgreSQL will remain unaffected and continue to use earlier implementation of newInstance(tag) due to default delegation, until another implementation is provided.

rnorth · 2018-02-26T08:53:20Z

Ah, interesting. This might solve #357 / #345 as well, so seems good. Will review properly later.

…ainers#593) * Mark DockerMachineClientProviderStrategy as not persistable * Update CHANGELOG.md

manikmagar · 2018-02-26T14:55:58Z

I did not see those two earlier. For filtering out TC parameters, we could do this -

This line where we extract out queryString, we could call getQueryParameters() function.
The Query Param Matching Pattern, could be modified to exclude anything that starts with TC_, which are TestContainers parameters (any possibility of that prefix conflicting with any real params?!).
Join map entries returned by getQueryParameters method to build new queryString.

WDYT?

That makes me think if TC_PARAM_MATCHING_PATTERN should only consider Pattern to include parameters starting with TC_?

manikmagar · 2018-02-28T16:19:43Z

@rnorth Codacy is complaining to have atleast one JUnit assert but ignoring the VisibleAssertions that I have in test cases. I hope that is ok ;).

…ainers-java into feature/TC_ISSUE_566

manikmagar · 2018-03-03T05:02:07Z

The latest update to this PR fixes #596, #357 / #345.

manikmagar · 2018-03-08T16:28:49Z

@rnorth @kiview When you get a chance, could you please review this? Thank you!

rnorth · 2018-03-12T14:18:17Z

This PR has a lot of Spock changes showing when I view the diff. Has something strange happened with the branches? 😄

rnorth · 2018-03-12T14:20:36Z

modules/jdbc-test/src/test/java/org/testcontainers/jdbc/JDBCDriverTest.java

+              assertEquals("Database name from URL String is used.", "databasename", resultDB);
+              return true;
+          });
+


This is just a trivial comment, but please could you reformat using 4 spaces for indentation? IntelliJ/Eclipse defaults ought to be reasonably close. I'll add automatic formatting soon...

Sure, I'll do that. Thanks for reviewing!

I have updated formatting in recent commit. Thanks!

manikmagar · 2018-03-12T14:29:28Z

Hm, could it be because I merged latest upstream changes from the master? I thought that would make it easy to merge (if appropriate) :)

rnorth · 2018-03-12T15:01:36Z

Yeah, it looks like that’s what’s happened. Very odd. I’ll try and review properly later today!

rnorth

This is good work - thank you for doing this, and I apologise that you've had to refactor quite a few bits along the way. I have a few comments but nothing massive!

rnorth · 2018-03-13T20:13:06Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java

+    @Getter
+    private Optional<String> queryString;
+
+    private ConnectionUrl() {


I think this is redundant, given that there is a public constructor with a parameter.

This has been refactored along with adding a static factory method to instantiate the class.

rnorth · 2018-03-13T20:15:26Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java

+        return url.startsWith("jdbc:tc:");
+    }
+
+    public void parseUrl() {


It looks like this is only used together with the constructor. Could we instead have a static factory method that instantiates, parses and returns?

Makes sense. I have added a static method newInstance(final String url) that returns ready to use instance of ConnectionUrl.

rnorth · 2018-03-13T20:16:27Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java

+        }
+
+        Matcher daemonMatcher = Patterns.DAEMON_MATCHING_PATTERN.matcher(this.getUrl());
+        inDaemonMode = daemonMatcher.matches() ? Boolean.parseBoolean(daemonMatcher.group(2)) : false;


IntelliJ is telling me that this can be simplified to:

inDaemonMode = daemonMatcher.matches() && Boolean.parseBoolean(daemonMatcher.group(2));

This has been refactored now.

rnorth · 2018-03-13T20:22:35Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java

+     *
+     * @return {@link Map}
+     */
+    public Map<String, String> getQueryParameters() {


It looks like this method mutates the queryString field as a side effect of constructing the Map. If someone calls getQueryString too early, they'll get nothing.

Please could we avoid this? Maybe we could make all fields be final (which a static factory method and maybe Lombok could help with)?

I have modified the query and container parameter extraction method to make them immutable. Now, queryParameters and containerParameters have getters that returns unmodifiable maps and queryString is instantiated only once during initial parsing. Please check.

rnorth · 2018-03-13T20:23:52Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java

+    public interface Patterns {
+        final Pattern URL_MATCHING_PATTERN = Pattern.compile("jdbc:tc:([a-z]+)(:([^:]+))?://([^\\?]+)(\\?.*)?");
+
+        final Pattern ORACLE_URL_MATCHING_PATTERN = Pattern.compile("jdbc:tc:([a-z]+)(:([^(thin:)]+))?:thin:@([^\\?]+)(\\?.*)?");


I apologise for putting you through the pain of working with these regular expressions
😂

Actually, I liked it. I never used these many and types of regex pattern, got to learn new things ;), So Thank you!

rnorth · 2018-03-13T20:25:56Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ContainerDatabaseDriver.java

@@ -80,51 +69,34 @@ public synchronized Connection connect(String url, final Properties info) throws
        if (!acceptsURL(url)) {
            return null;
        }
+
+        ConnectionUrl cUrl = new ConnectionUrl(url);


Please could you call this parameter connectionUrl? Maybe it's just me, but every time I see cUrl I think of this. In general we're quite OK with medium-verbosity names for things in this project 😄

Changed now 😄

rnorth · 2018-03-13T20:28:10Z

modules/mysql/src/main/java/org/testcontainers/containers/MySQLContainerProvider.java

 /**
 * Factory for MySQL containers.
 */
 public class MySQLContainerProvider extends JdbcDatabaseContainerProvider {
+
+    /**
+     * Groups URL of format "jdbc:tc:(databaseType):(optinal_image_tag)//(hostanme)(optional :(numeric_port))/(databasename)(?parameters)"


A couple of typos in here (optinal and hostanme)

Typos are corrected now.

rnorth · 2018-03-13T20:31:10Z

modules/mysql/src/main/java/org/testcontainers/containers/MySQLContainerProvider.java

+      Matcher urlMatcher = MYSQL_URL_MATCHING_PATTERN.matcher(url.getUrl());
+
+      if(!urlMatcher.matches()) {
+        //TODO: Is this necessary?


Probably not necessary if the ConnectionUrl.parseQuery() method has been called..?

I agree. I have removed this matcher related code.

rnorth · 2018-03-13T20:31:47Z

modules/spock/README.md

@@ -0,0 +1,48 @@
+# TestContainers-Spock


No idea why these Spock files are showing up in the PR - looking at the branch locally I don't see this.

I am not sure what should I do for this one. Maybe it can be ignored while merging, if appropriate! I don't have any changes for Spock 😄

rnorth · 2018-03-13T20:34:39Z

CHANGELOG.md

@@ -9,6 +9,7 @@ All notable changes to this project will be documented in this file.
 ### Changed
 - Abstracted and changed database init script functionality to support use of SQL-like scripts with non-JDBC connections. ([\#551](https://github.com/testcontainers/testcontainers-java/pull/551))
 - Added `JdbcDatabaseContainer(Future)` constructor. ([\#543](https://github.com/testcontainers/testcontainers-java/issues/543))
+- Mark DockerMachineClientProviderStrategy as not persistable ([\#593](https://github.com/testcontainers/testcontainers-java/pull/593))



Please could you update the changelog to describe this change?

I have updated Changelog with items that I feel related to this PR. Please take a look and let me know/feel free to modify as you seem right.

VladRassokhin · 2018-03-13T21:49:03Z

JFYI Unless this PR integrated there's problem in master branch with queryParameter not passed to actual jdbc driver on second and subsequent calls of ContainerDatabaseDriver#connect since it's set to non-empty string only if there's no container yet.
Could you please add test for that?

manikmagar · 2018-03-13T22:25:07Z

Thank you very much for the review @rnorth. I will check these comments and make appropriate changes.

…

On Mar 13, 2018 at 4:33 PM, <Richard North ***@***.***)> wrote: @rnorth requested changes on this pull request. This is good work - thank you for doing this, and I apologise that you've had to refactor quite a few bits along the way. I have a few comments but nothing massive! In modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java (#594 (comment)): > + @getter + private Optional<Integer> databasePort = Optional.empty(); + + @getter + private Optional<String> databaseName = Optional.empty(); + + @getter + private Optional<String> initScriptPath = Optional.empty(); + + @getter + private Optional<InitFunctionDef> initFunction = Optional.empty(); + + @getter + private Optional<String> queryString; + + private ConnectionUrl() { I think this is redundant, given that there is a public constructor with a parameter. In modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java (#594 (comment)): > + + /** + * This is a part of the connection string that may specify host:port/databasename. + * It may vary for different clients and so clients can parse it as needed. + * + * @return + */ + public String getDbHostString() { + return Objects.requireNonNull(this.dbHostString, "Database Host String cannot be null. Have you called parseUrl() method?"); + } + + public static boolean accepts(final String url) { + return url.startsWith("jdbc:tc:"); + } + + public void parseUrl() { It looks like this is only used together with the constructor. Could we instead have a static factory method that instantiates, parses and returns? In modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java (#594 (comment)): > + + queryString = Optional.ofNullable(urlMatcher.group(5)); + getQueryParameters(); + + Matcher matcher = Patterns.INITSCRIPT_MATCHING_PATTERN.matcher(this.getUrl()); + if (matcher.matches()) { + initScriptPath = Optional.ofNullable(matcher.group(2)); + } + + Matcher funcMatcher = Patterns.INITFUNCTION_MATCHING_PATTERN.matcher(this.getUrl()); + if (funcMatcher.matches()) { + initFunction = Optional.of(new InitFunctionDef(funcMatcher.group(2), funcMatcher.group(4))); + } + + Matcher daemonMatcher = Patterns.DAEMON_MATCHING_PATTERN.matcher(this.getUrl()); + inDaemonMode = daemonMatcher.matches() ? Boolean.parseBoolean(daemonMatcher.group(2)) : false; IntelliJ is telling me that this can be simplified to: inDaemonMode = daemonMatcher.matches() && Boolean.parseBoolean(daemonMatcher.group(2)); In modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java (#594 (comment)): > + Matcher matcher = Patterns.TC_PARAM_MATCHING_PATTERN.matcher(this.getUrl()); + while (matcher.find()) { + String key = matcher.group(1); + String value = matcher.group(2); + results.put(key, value); + } + + return results; + } + + /** + * Get all Query paramters specified in the Connection URL after ?. This also includes TestContainers parameters. + * + * @return ***@***.*** Map} + */ + public Map<String, String> getQueryParameters() { It looks like this method mutates the queryString field as a side effect of constructing the Map. If someone calls getQueryString too early, they'll get nothing. Please could we avoid this? Maybe we could make all fields be final (which a static factory method and maybe Lombok could help with)? In modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java (#594 (comment)): > + + @OverRide + public boolean equals(Object obj) { + if (Objects.isNull(obj) || !(obj instanceof ConnectionUrl)) return false; + return this.getUrl().equals(((ConnectionUrl) obj).getUrl()); + } + + /** + * This interface defines the Regex Patterns used by ***@***.*** ConnectionUrl}. + * + * @author manikmagar + */ + public interface Patterns { + final Pattern URL_MATCHING_PATTERN = Pattern.compile("jdbc:tc:([a-z]+)(:([^:]+))?://([^\\?]+)(\\?.*)?"); + + final Pattern ORACLE_URL_MATCHING_PATTERN = Pattern.compile("jdbc:tc:([a-z]+)(:([^(thin:)]+))?:thin:@([^\\?]+)(\\?.*)?"); I apologise for putting you through the pain of working with these regular expressions 😂 In modules/jdbc/src/main/java/org/testcontainers/jdbc/ContainerDatabaseDriver.java (#594 (comment)): > @@ -80,51 +69,34 @@ public synchronized Connection connect(String url, final Properties info) throws if (!acceptsURL(url)) { return null; } + + ConnectionUrl cUrl = new ConnectionUrl(url); Please could you call this parameter connectionUrl? Maybe it's just me, but every time I see cUrl I think of this (https://curl.haxx.se/). In general we're quite OK with medium-verbosity names for things in this project 😄 In modules/mysql/src/main/java/org/testcontainers/containers/MySQLContainerProvider.java (#594 (comment)): > /** * Factory for MySQL containers. */ public class MySQLContainerProvider extends JdbcDatabaseContainerProvider { + + /** + * Groups URL of format "jdbc:tc:(databaseType):(optinal_image_tag)//(hostanme)(optional :(numeric_port))/(databasename)(?parameters)" A couple of typos in here (optinal and hostanme) In modules/mysql/src/main/java/org/testcontainers/containers/MySQLContainerProvider.java (#594 (comment)): > @@ -13,4 +30,26 @@ public boolean supports(String databaseType) { public JdbcDatabaseContainer newInstance(String tag) { return new MySQLContainer(MySQLContainer.IMAGE + ":" + tag); } + + @OverRide + public JdbcDatabaseContainer newInstance(ConnectionUrl url) { + Objects.requireNonNull(url, "Connection URL cannot be null"); + + Matcher urlMatcher = MYSQL_URL_MATCHING_PATTERN.matcher(url.getUrl()); + + if(!urlMatcher.matches()) { + //TODO: Is this necessary? Probably not necessary if the ConnectionUrl.parseQuery() method has been called..? In modules/spock/README.md (#594 (comment)): > @@ -0,0 +1,48 @@ +# TestContainers-Spock No idea why these Spock files are showing up in the PR - looking at the branch locally I don't see this. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub (#594 (review)), or mute the thread (https://github.com/notifications/unsubscribe-auth/AA1i5hZU68S_58BNYdOIZ_ENf63Wmqs_ks5teC0JgaJpZM4SSr4s).

manikmagar · 2018-03-14T19:22:18Z

@VladRassokhin Thanks for that info. I did add a test and verified behavior. It is working as expected with changes in this PR. Should there be a Bug entry created and referenced here? Anyways, I'll update the PR description too.

VladRassokhin · 2018-03-14T19:40:17Z

@manikmagar I think no bug entry needed if this PR would be merged prior to any new release

manikmagar · 2018-03-15T03:13:07Z

@rnorth I have made changes to take care of review comments, please take a look when you get chance. Thanks for reviewing.

@VladRassokhin I did make changes to a test case that would fail if query parameters are not used when creating connection from cached container - Please See this.

jeacott · 2018-03-15T12:12:01Z

could use a formatter run over this. lots if apparent diffs are just formatting changes, and the rest aren't consistent. makes it hard to see what has really changed.

manikmagar · 2018-03-15T12:55:00Z

Thanks @jeacott for looking through these changes. I tried to use the same formatting as rest of the code, which is pretty much standard intellij code formatting. Could you please point me to some code where you see it is not appropriate? I'll be happy to revisit and correct appropriately.

kiview · 2018-03-15T13:08:08Z

I think there is something off with this PR, showing a lot of changes that aren't changes but have been introduced simply by rebasing to master or something (i.e. the changes in the travis.yml file). I'm not sure where this is coming from?

Oh, I just see @rnorth has already mentioned this :D
Maybe it's easier to simply open a new PR with the current changes?

jeacott1 · 2018-03-15T23:00:54Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ContainerDatabaseDriver.java

    @Override
    public boolean acceptsURL(String url) throws SQLException {
-        return url.startsWith("jdbc:tc:");
+      return url.startsWith("jdbc:tc:");


formatting in this file is inconsistent with the other 4 space indents etc.

jeacott1 · 2018-03-15T23:01:28Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ContainerDatabaseDriver.java

-                    if (candidateContainerType.supports(databaseType)) {
-                        container = candidateContainerType.newInstance(tag);
+                    if (candidateContainerType.supports(connectionUrl.getDatabaseType())) {
+                        container = candidateContainerType.newInstance(connectionUrl);


is this mod to connectionUrl instead of 'tag' intentional?

Yes, it is intentional and part of the implementation. Please see PR Description for details.

jeacott1 · 2018-03-15T23:10:54Z

modules/jdbc/src/main/java/org/testcontainers/jdbc/ConnectionUrl.java

+            databaseName = Optional.of(dbInstanceMatcher.group(4));
+        }
+
+        queryParameters = Collections.unmodifiableMap(


this doesnt allow for duplicate params which might be very handy if for example I want to set jdbc://...?TC_INITSCRIPT=schema.sql&TC_INITSCRIPT=data.sql

regular expressions seem like a hard work way to parse a jdbc url.

I see what you mean but I don't think it is possible even with the existing implementation. I would rather think of that as a separate feature request, maybe @rnorth can weigh in.

Please let's see if we can tackle this without regular expressions separately. I'd be very happy if we could do that and make this code easier to read.

Just making sure I understand it correctly. Do You mean, 'Allowing multiple TC_INITSCRIPT like parameter support' should be implemented in separate PR?

On other thought, maybe a comma delimited list of files is accepted for single TC_INITSCRIPT parameter and executed in the sequence they are provided. Eg. TC_INITSCRIPT=Script1.sql,Script2.sql,Script3.sql.

jeacott1 · 2018-03-15T23:20:21Z

core/src/main/java/org/testcontainers/dockerclient/DockerClientProviderStrategy.java

@@ -93,7 +97,10 @@ public static DockerClientProviderStrategy getFirstValidStrategy(List<DockerClie
                                        LOGGER.warn("Can't instantiate a strategy from {}", it, e);
                                        return Stream.empty();
                                    }
-                                }),
+                                })


is the formatting in this Stream.concat... block just 'idea' being weird? doesnt seem to follow the 4 space indents everywhere else.

This class is not modified and related for this PR, probably upstream has changed and so you see the difference. I would leave it as-is for now.

~~Please let's see if we can tackle this without regular expressions separately. I'd be very happy if we could do that and make this code easier to read.~~
Response to wrong comment

jeacott1

I originally thought this looked worse than it is.

manikmagar · 2018-03-17T20:50:00Z

Thanks for taking a look @kiview . I am not sure, why those Spock related changes are showing up as a difference.
@rnorth @kiview Do you guys really think this should be added as a separate PR?

rnorth · 2018-03-18T21:35:02Z

@manikmagar could you please try rebasing onto master and force-push this branch to see if this clears the diff noise?

I've pulled the branch locally and the commits look fine, so I find it a bit disconcerting that we're all seeing phantom changes in the PR. I'm not sure I trust Github to merge it well if it can't show the diff correctly...

manikmagar · 2018-03-18T21:51:11Z

Ok @rnorth , let me try rebasing onto master. In the worst case, I can create a new PR with related changes only.

manikmagar · 2018-03-20T21:07:20Z

I think something is really weird happened with this branch, I will try to create a new branch and re-submit the clean PR on the latest upstream master.

manikmagar · 2018-03-21T12:58:47Z

@rnorth @kiview I have made a new PR #617 with all these changes, please take a look at that when you get a chance. If that is merged then this PR should be closed without merging. Thanks!

kiview · 2018-03-23T12:54:35Z

@manikmagar Thanks a lot for the new PR, I'll close this one, so we don't get confused.

manikmagar added 2 commits February 26, 2018 00:01

Added feature to set database name and user from JDBC URL

63343fe

Updated to ignore Eclipse metadata files

6a1726a

Mark DockerMachineClientProviderStrategy as not persistable (testcont…

6f151da

…ainers#593) * Mark DockerMachineClientProviderStrategy as not persistable * Update CHANGELOG.md

manikmagar mentioned this pull request Feb 26, 2018

Allow setting database name when using JDBC URL #566

Closed

kiview added 3 commits February 27, 2018 17:54

Add spock module

6daa5b0

Remove LICENSE block from spock module README

8e313f1

Fix problems with Travis gradle cache

b3d290a

manikmagar added 6 commits March 2, 2018 19:46

Added feature to set database name and user from JDBC URL

ffb84f4

Updated to ignore Eclipse metadata files

4006d4f

Allow Oracle JDBC URL to be parsed

c10f4b6

Allow Oracle JDBC URL to be parsed

d807546

Allow Oracle JDBC URL to be parsed

1be1c16

Merge branch 'feature/TC_ISSUE_566' of github.com:manikmagar/testcont…

d6dcc89

…ainers-java into feature/TC_ISSUE_566

Codacy issue fixes

5d8f12a

rnorth reviewed Mar 12, 2018

View reviewed changes

Code formatting and javadoc addition

f162081

rnorth requested changes Mar 13, 2018

View reviewed changes

rnorth reviewed Mar 13, 2018

View reviewed changes

rnorth mentioned this pull request Mar 13, 2018

Apply simple filtering of the query string passed through to real database drivers #357

Closed

manikmagar added 2 commits March 14, 2018 22:57

Incorporated Code Review Changes

8fde60e

Removed unused imports

e29a339

This was referenced Mar 15, 2018

db extra connection params are only passed to the first connection in a pool #610

Closed

fixes bug where extra connection params were only passed to the first connection in a pool #609

Closed

bsideup added type/feature modules/jdbc labels Mar 15, 2018

jeacott1 reviewed Mar 15, 2018

View reviewed changes

manikmagar mentioned this pull request Mar 21, 2018

Enhancements and Fixes around JDBC URL Based Containers #617

Merged

kiview closed this Mar 23, 2018

kiview added the resolution/duplicate label Mar 23, 2018

manikmagar deleted the feature/TC_ISSUE_566 branch June 15, 2018 02:18

Allow setting database name and user when using JDBC URL #594

Allow setting database name and user when using JDBC URL #594

Conversation

manikmagar commented Feb 26, 2018 • edited

rnorth commented Feb 26, 2018

manikmagar commented Feb 26, 2018

manikmagar commented Feb 28, 2018

manikmagar commented Mar 3, 2018

manikmagar commented Mar 8, 2018

rnorth commented Mar 12, 2018

Choose a reason for hiding this comment

Choose a reason for hiding this comment

manikmagar Mar 13, 2018 • edited

Choose a reason for hiding this comment

manikmagar commented Mar 12, 2018

rnorth commented Mar 12, 2018 via email

rnorth left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

manikmagar Mar 14, 2018 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

VladRassokhin commented Mar 13, 2018 • edited

manikmagar commented Mar 13, 2018 via email

manikmagar commented Mar 14, 2018

VladRassokhin commented Mar 14, 2018

manikmagar commented Mar 15, 2018

jeacott commented Mar 15, 2018

manikmagar commented Mar 15, 2018

kiview commented Mar 15, 2018 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jeacott1 Mar 15, 2018 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

manikmagar Mar 16, 2018 • edited

Choose a reason for hiding this comment

rnorth Mar 18, 2018 • edited

Choose a reason for hiding this comment

jeacott1 left a comment

Choose a reason for hiding this comment

manikmagar commented Mar 17, 2018

rnorth commented Mar 18, 2018

manikmagar commented Mar 18, 2018

manikmagar commented Mar 20, 2018

manikmagar commented Mar 21, 2018 • edited

kiview commented Mar 23, 2018

manikmagar commented Feb 26, 2018 •

edited

manikmagar Mar 13, 2018 •

edited

manikmagar Mar 14, 2018 •

edited

VladRassokhin commented Mar 13, 2018 •

edited

kiview commented Mar 15, 2018 •

edited

jeacott1 Mar 15, 2018 •

edited

manikmagar Mar 16, 2018 •

edited

rnorth Mar 18, 2018 •

edited

manikmagar commented Mar 21, 2018 •

edited