syntax: Fix hex and octal int and errors highlight #2584

k33nice · 2019-11-13T22:38:55Z

According to language specification it is allowed to use underscore "_"
for readability in integer literals. Hence syntax matches for
goHexadecimalInt and goOctalInt must allow it. Also goOctalError regexp
mustn't match on accessing slice or array element by index.

bhcleek · 2019-11-14T02:00:39Z

This highlights valid hexadecimals goOctalError.

bhcleek · 2019-11-14T18:54:34Z

FWIW, I added some tests for verifying the syntax highlight groups of the numeric literals: #2587. You can rebase your branch off of master and use the tests to make sure your changes match as expected.

bhcleek · 2019-11-14T20:13:29Z

This does not highlight 07a as goOctalError.

k33nice · 2019-11-14T20:43:39Z

Hey @bcarrell! I'm currently trying to run the tests with docker but get some errors from gopls :(
About your question on 07a - no, it doesn't highlight it as an error right now. I tried to explain it in the commit message:

    The regular expression was changed that way to match all wrong sequences
    of characters. To do not make a false matching on other language structures
    the regexp should detect only falsy octal literals based on
    the allowed symbols: [0-7], "o", "O", "_". The exclusion is "8" and "9"
    because it might be very easy to make that kind of mistake.

    The regexp breakdown is described below:

    0                  |<-- "0" digit
    \(                 |
        [0-7oO_]*      |<-- 0+ [0-7] or "o" or "O" or "_" -----+
        \(             |                                       |
            [89]\+     |<-- 1+ "8" or "9" digit ----+          |
                \|     |                            |   +---+  |   +---+
            [oO]\{2,\} |<-- 2+ "o" or "O" character |<--| 1 |  |<--| 2 |
                \|     |                            |   +---+  |   +---+
            _\{2,\}    |<-- 2+ "_" character -------+          |
        \)             |                                       |
        [0-7oO_]*      |<-- 0+ [0-7] or "o" or "O" or "_" -----+
    \)\+               |
    \S*                |<-- 0+ "non space" characters

    1. One of the rules in this group should appear exactly once.
    2. All of the rules in this group should appear once or more.

TBO, as a user of vim-go I'm not sure what kind of sequences I want to be treated as an octal error. What do you think about it?

k33nice · 2019-11-14T21:23:02Z

I can change [89]\+ to [89A-Fa-f]\+ so that kind of sequences like 07a will be highlighted as goOctalError. But I'm wondering is that should be goOctalError or goHexadecimalError 🤔

k33nice · 2019-11-14T21:47:07Z

I've also added some changes to other regexp here 809c94c related to #2588

bhcleek · 2019-11-14T21:48:26Z

@k33nice you can ignore the gopls errors; they're due to some kind of regression in gopls that is already being addressed by the Go tools team.

Generally speaking, the regular expression changes you're making are going to fail to highlight some patterns that are correctly highlighted on master. Here's a few examples:

syn match       goDecimalInt        "\<-\=\d\+\%([Ee][-+]\=\d\+\)\=\>"

This is now going to regress and match as a decimal instead of an octdal when the number begins with a 0.

syn match       goOctalError        "\<-\=0\([0-7oO_]*\([89]\+\|[oO]\{2,\}\|_\{2,\}\)[0-7oO_]*\)\+\S*\>"

This treats 8 and 9 specially for some reason when there's really no need to. The scope of matching invalid characters should be much wider - as it already is on master.

It also introduces an inconsistency in how goOctalError handle more than one _ vs how goBinaryError and goHexadecimalError handle them.

I'd also strongly recommend that you rebase your branch onto master, because I added tests that will be useful for changing these highlight groups.

Try hard not to break existing behaviors. You're right that ideally _ would be handled in all numerics and that octals should also allow 0, 0o, and 0O as leading characters, but if we can't get those changes done in a way that doesn't negatively impact performance and that is relatively easy to read, then I'm ok with leaving master as-is.

k33nice · 2019-11-14T21:54:30Z

Thanks for your clarification! A quick note on goDecimalInt: that change https://github.com/fatih/vim-go/pull/2587/files#diff-6bbebb95ce25cd40d0072544e2dad91eR165 was broke highlighting single zero as a number.

bhcleek · 2019-11-14T21:59:37Z

Good catch. I'll fix that in a separate PR.

k33nice · 2019-11-14T22:12:04Z

I'll try to rework this pr a little bit. Also just my point of view, but I'm pretty convinced that much better to receive false negatives than positives for error highlighting. Especially for the current case where we have these syntax rules are enabled by default.

bhcleek · 2019-11-14T22:42:52Z

I agree that false negatives for errors is better than false positives. However, I see no reason to introduce false negatives where they don't exist now.

k33nice · 2019-11-15T16:49:49Z

@bhcleek I added support of "_" for decimal, hex, octal and binary and also some more test cases for it. I tried don't regress previous behavior. I'd be glad to get any suggestions or thoughts.

bhcleek

Why do the *DoublePrefix tests need to exist? The existing *ErrorLeading tests should cover that case fine.

This is getting closer to being mergeable. I have some questions and requests for changes in-line.

Can you also squash these change so that your merge commits aren't in the branch?

Finally, as an aside and a quick tip to help you get PRs merged more quickly in the future: try to make your commits tell a story of adding incremental and easily reviewable value. For instance, a commit that adds support for 0o and 0O for octal numbers and tests for those, then a commit that adds support and tests for the underscores. When the set of changes taken together is large, complex, or otherwise noisy, it's useful for reviewers to be able to review commit by commit. Reviewing regular expressions is hard. Difficult to review PRs are less likely to be merged because it's hard to assess their risk vs value.

autoload/go/highlight_test.vim

syntax/go.vim

bhcleek · 2019-11-15T18:44:29Z

syntax/go.vim

+syn match       goOctalInt          "\<-\=0[oO]\?_\?\(\o\+_\?\)\+\>"
+syn match       goOctalError        "\<-\=0\([0-7oO_]*\([^ \t0-7oOxX\]\}\_]\+\|[oO]\{2,\}\|_\{2,\}\)[0-7oO_]*\)\+\S*\>"
+syn match       goBinaryInt         "\<-\=0[bB]_\?\([01]\+_\?\)\+\>"
+syn match       goBinaryError       "\<-\=0[bB]_\?\([01]\+_\?\)*\([^ \t01_]\S*\|_\{2,\}\S*\|_\)\>"


for both goBinaryInt and goBinaryError, wouldn't [01_]\+ suffic instead of [01]\+_\??

[01_]\+ matches more than one sequential underscore, and that is syntaxly incorrect.

Yes, it's syntactically in incorrect, but that's what goBinaryError is trying to catch. Can goBinaryError be simplified to \<-\=0[bB]\([01_]*[^ \t01_]\S*\|[01]*__\S*\)\>

Gotcha. Technically no, cuz it doesn't match on 0b_01__01 or 0b_01_01_ but I'll try to simplify it.

bhcleek · 2019-11-15T18:55:07Z

syntax/go.vim

+syn match       goHexadecimalInt    "\<-\=0[xX]_\?\(\x\+_\?\)\+\>"
+syn match       goHexadecimalError  "\<-\=0[xX]_\?\(\x\+_\?\)*\(\([^ \t0-9A-Fa-f_]\|_\{2,\}\)\S*\|_\)\>"
+syn match       goOctalInt          "\<-\=0[oO]\?_\?\(\o\+_\?\)\+\>"
+syn match       goOctalError        "\<-\=0\([0-7oO_]*\([^ \t0-7oOxX\]\}\_]\+\|[oO]\{2,\}\|_\{2,\}\)[0-7oO_]*\)\+\S*\>"


I think this might end up regressing 69d5dad.

I changed it a bit already. But I cannot see any sequences that the previous regexp covered and the new does not. Of course, if not count those that it should not match (e.g. 0o1, 0_1 0123/5 , 0123+5and so on). Maybe I missed smth. Could you please elaborate.

69d5dad handles list indexes as reported in the issue that commit fixed.

Sigh. You're right, thanks! I've checked it on [0] and []int{0} but forgot about [0][0] and [][]int{{0},{0}}. I'll expand it.

According to language specification it is allowed to use underscore "_" for readability in integer literals. Hence syntax matches for goHexadecimalInt and goOctalInt must allow it. Also goOctalError regexp mustn't match on accessing slice or array element by index.

The regular expression was changed that way to match all wrong sequences of characters. To do not make a false matching on other language structures the regexp should detect only falsy octal literals based on the allowed symbols: [0-7], "o", "O", "_". The exclusion is "8" and "9" because it might be very easy to make that kind of mistake. The regexp breakdown is described below: 0 |<-- "0" digit \( | [0-7oO_]* |<-- 0+ [0-7] or "o" or "O" or "_" -----+ \( | | [89]\+ |<-- 1+ "8" or "9" digit ----+ | \| | | +---+ | +---+ [oO]\{2,\} |<-- 2+ "o" or "O" character |<--| 1 | |<--| 2 | \| | | +---+ | +---+ _\{2,\} |<-- 2+ "_" character -------+ | \) | | [0-7oO_]* |<-- 0+ [0-7] or "o" or "O" or "_" -----+ \)\+ | \S* |<-- 0+ "non space" characters 1. One of the rules in this group should appear exaclty once. 2. All of the rules in this group should appears once or more.

*Changes in regexp* - Added [A-Fa-f] to goOctalError to sequences like `07a` considered as errors - Added "_" to goHexadecimalError to be matched on wrong characters after correct groups (e.g. `0x_123_123_1yy`) - Added "_" to goBinaryInt and goBinaryError to correctly handle groups in binary integers

*Changes* - Add new goDecimalErrorInt hi - Add support of "_" in goDecimalInt - Fix goOctalError to support more wide patterns of errors - Correct support of goups by "_" in goBinaryError

*Fix* - goDecimalInt: add support for sequneces like `1_234567890` - goBinaryInt: add support for sequneces like `0b_01`

*Changes* - removed redundant characters from exclusion pattern - fixed tests so that test include only one testing error - added test case for explicitly test trailing "o" or "O"

k33nice · 2019-11-15T22:48:59Z

@bhcleek

Why do the *DoublePrefix tests need to exist? The existing *ErrorLeading tests should cover that case fine.

Yup, *ErrorLeading should cover that. I just added them as a common typo and since we need to explicitly check it in goOctalError. I got rid of them though.

I really appreciate your advice! I know that the review is the hard work especially for something tedious as regexp. I just decided to help little cuz errors started to annoyed during my work. I'll try to do it better and not so dreary for reviewers next time. Again, thanks a lot for all of your work and generally for maintaining this project!

bhcleek · 2019-11-16T18:31:13Z

syntax/go.vim

+syn match       goOctalInt          "\<-\=0[oO]\?_\?\(\o\+_\?\)\+\>"
+syn match       goOctalError        "\<-\=0\([0-7oO_]*\([^ \t0-7oOxX\]\}\_]\+\|[oO]\{2,\}\|_\{2,\}\)[0-7oO_]*\)\+\S*\>"
+syn match       goBinaryInt         "\<-\=0[bB]_\?\([01]\+_\?\)\+\>"
+syn match       goBinaryError       "\<-\=0[bB]_\?\([01]\+_\?\)*\([^ \t01_]\S*\|_\{2,\}\S*\|_\)\>"


Yes, it's syntactically in incorrect, but that's what goBinaryError is trying to catch. Can goBinaryError be simplified to \<-\=0[bB]\([01_]*[^ \t01_]\S*\|[01]*__\S*\)\>

bhcleek · 2019-11-16T18:33:53Z

autoload/go/highlight_test.vim

+        \ 'octalErrorDoubleUnderscore': {'group': 'goOctalError', 'value': '0o__1234567'},
+        \ 'octalErrorDoubleTrailingUnderscore': {'group': 'goOctalError', 'value': '0o_1234567__'},
+        \ 'octalErrorTrailingUnderscore': {'group': 'goOctalError', 'value': '0o_123456_7_'},
+        \ 'octalErrorTrailingO': {'group': 'goOctalError', 'value': '0o_123456_7o'},


What value does this add over octalErrorTrailing?

For checking a specific subgroup that matches on "oO". We need this group cuz we allow 'oO' in this subgroup [^ \t0-7oOxX_/]. And just realized that we need to expand that specific subgroup to [oOxX] to match on 0123x.

bhcleek · 2019-11-16T18:36:48Z

syntax/go.vim

+syn match       goHexadecimalInt    "\<-\=0[xX]_\?\(\x\+_\?\)\+\>"
+syn match       goHexadecimalError  "\<-\=0[xX]_\?\(\x\+_\?\)*\(\([^ \t0-9A-Fa-f_]\|_\{2,\}\)\S*\|_\)\>"
+syn match       goOctalInt          "\<-\=0[oO]\?_\?\(\o\+_\?\)\+\>"
+syn match       goOctalError        "\<-\=0\([0-7oO_]*\([^ \t0-7oOxX\]\}\_]\+\|[oO]\{2,\}\|_\{2,\}\)[0-7oO_]*\)\+\S*\>"


69d5dad handles list indexes as reported in the issue that commit fixed.

bhcleek

Why are there tests for double trailing underscores? Shouldn't those be tested already by virtue of the *TrailingUnderscore tests?

I really appreciate your advice! I know that the review is the hard work especially for something tedious as regexp. I just decided to help little cuz errors started to annoyed during my work. I'll try to do it better and not so dreary for reviewers next time. Again, thanks a lot for all of your work and generally for maintaining this project!

I appreciate your help on this. Most open source work happens because someone gets started resolving things that bother or annoy them.

I'll admit that part of me thinks we should just drop the error groups for numerics entirely. They're complicated and arguably don't add too much value.

bhcleek · 2019-11-16T18:43:15Z

Thanks for sticking with this. I know it's tedious to get through this review; you managed to pick up a piece of work for your first contribution that complex, high visibility, affects performance quite a bit, and that will be noticed by the community very quickly if it introduces false positives.

**Changes** - Added "]" and "}" to whitelist subgroup. To doesn't match on [0][0] and [][]int{{0},{0}}. - Expanded subgroup for trailing [oO] to [oOxX]. - Added test case for trailing 'x'.

k33nice · 2019-11-16T21:22:28Z

Why are there tests for double trailing underscores? Shouldn't those be tested already by virtue of the *TrailingUnderscore tests?

Agree. Them seems to be useless. Added them cuz was keeping in mind that case all the way.

Yup, can't disagree. That might be the best solution. I got the same thoughts when they started to falsy emerge during my working. Handling by regexp these grouped integers is way too tricky. Furthermore, AFAIK VSCode doesn't support them.

syntax/go.vim

bhcleek · 2019-11-16T23:21:38Z

Is that the last of the changes?

k33nice · 2019-11-17T15:47:14Z

Is that the last of the changes?

Yup, I think that's it. Not sure that you saw all of my responses but it seems like all your reviews were address. One more thing that I've found out that we have the same goDecimalInt, goHexadecimalInt and others in syntax/gotexttmpl.vim. I'm not sure that we want to change them here as well. It might be better to do in separate PR. What do you think?

bhcleek · 2019-11-17T18:21:45Z

Let's leave the gotexttmpl for another PR.

I'll merge this later today.

v1.22 - (January 30, 2020) BACKWARDS INCOMPATIBILITIES: * Drop support for Vim 7.4. The minimum required version of Vim is now 8.0.1453. [[GH-2495]](fatih/vim-go#2495) [[GH-2497]](fatih/vim-go#2497) * Drop support for `gometalinter` [[GH-2494]](fatih/vim-go#2494) IMPROVEMENTS: * Highlight the `go` keyword in go.mod files. [[GH-2473]](fatih/vim-go#2473) * Use echo functions consistently. [[GH-2458]](fatih/vim-go#2458) * Add support for managing goroutines in debugger. [[GH-2463]](fatih/vim-go#2463) [[GH-2527]](fatih/vim-go#2527) * Document `g:go_doc_popup_window`. [[GH-2506]](fatih/vim-go#2506) * Make `g:go_doc_popup_window=1` work for Neovim, too. [[GH-2451]](fatih/vim-go#2451) [[GH-2512]](fatih/vim-go#2512) * Handle errors jumping to a definition in a file open in another Vim process better. [[GH-2518]](fatih/vim-go#2518) * Improve the UX when the gopls binary is missing. [[GH-2522]](fatih/vim-go#2522) * Use gopls instead of guru for `:GoSameIds`. [[GH-2519]](fatih/vim-go#2519) * Use gopls instead of guru for `:GoReferrers`. [[GH-2535]](fatih/vim-go#2535) * Update documentation for `g:go_addtags_transform`. [[GH-2541]](fatih/vim-go#2541) * Install most helper tools in module aware mode. [[GH-2545]](fatih/vim-go#2545) * Add a new option, `g:go_referrers_mode` to allow the user to choose whether to use gopls or guru for finding references. [[GH-2566]](fatih/vim-go#2566) * Add options to control how gopls responds to completion requests. [[GH-2567]](fatih/vim-go#2567) [[GH-2568]](fatih/vim-go#2568) * Add syntax highlighting for binary literals. [[GH-2557]](fatih/vim-go#2557) * Improve highlighting of invalid numeric literals. [[GH-2571]](fatih/vim-go#2571) [[GH-2587]](fatih/vim-go#2587) [[GH-2589]](fatih/vim-go#2589) [[GH-2584]](fatih/vim-go#2584) [[GH-2597]](fatih/vim-go#2597) [[GH-2599]](fatih/vim-go#2599) * Add highlighting of sections reported by gopls diagnostics' errors and warnings. [[GH-2569]](fatih/vim-go#2569) [[GH-2643]](fatih/vim-go#2643) * Make the highlighting of fzf decls configurable. [[GH-2572]](fatih/vim-go#2572) [[GH-2579]](fatih/vim-go#2579) * Support renaming with gopls. [[GH-2577]](fatih/vim-go#2577) [[GH-2618]](fatih/vim-go#2618) * Add an option, `g:go_gopls_enabled`, to allow gopls integration to be disabled. [[GH-2605]](fatih/vim-go#2605) [[GH-2609]](fatih/vim-go#2609) [[GH-2638]](fatih/vim-go#2638) [[GH-2640]](fatih/vim-go#2640) * Add a buffer level option, `b:go_fmt_options`, to control formatting options per buffer. [[GH-2613]](fatih/vim-go#2613) * Use build tags when running `:GoVet`. [[GH-2615]](fatih/vim-go#2615) * Add new snippets for UltiSnips. [[GH-2623]](fatih/vim-go#2623) [[GH-2627]](fatih/vim-go#2627) * Expand completions as snippets when `g:go_gopls_use_placeholders` is set. [[GH-2624]](fatih/vim-go#2624) * Add a new function, `:GoDiagnostics` and an associated mapping for seeing `gopls` diagnostics. Because of the performance implications on large projects, `g:go_diagnostics_enabled` controls whether all diagnostics are processed or only the diagnostics for the current buffer. [[GH-2612]](fatih/vim-go#2612) * Explain how to find and detect multiple copies of vim-go in the FAQ. [[GH-2632]](fatih/vim-go#2632) * Update the issue template to ask for the gopls version and `:GoReportGitHubIssue` to provide it. [[GH-2630]](fatih/vim-go#2630) * Use text properties when possible for some highlighting cases. [[GH-2652]](fatih/vim-go#2652) [[GH-2662]](fatih/vim-go#2662) [[GH-2663]](fatih/vim-go#2663) [[GH-2672]](fatih/vim-go#2672) [[GH-2678]](fatih/vim-go#2678) BUG FIXES: * Fix removal of missing directories from gopls workspaces. [[GH-2507]](fatih/vim-go#2507) * Change to original window before trying to change directories when term job ends. [[GH-2508]](fatih/vim-go#2508) * Swallow errors when the hover info cannot be determined. [[GH-2515]](fatih/vim-go#2515) * Fix errors when trying to debug lsp and hover. [[GH-2516]](fatih/vim-go#2516) * Reset environment variables on Vim <= 8.0.1831 . [[GH-2523]](fatih/vim-go#2523) * Handle empty results from delve. [[GH-2526]](fatih/vim-go#2526) * Do not overwrite `updatetime` when `g:go_auto_sameids` or `g:go_auto_type_info` is set. [[GH-2529]](fatih/vim-go#2529) * Fix example for `g:go_debug_log_output` in docs. [[GH-2547]](fatih/vim-go#2547) * Use FileChangedShellPost instead of FileChangedShell so that reload messages are not hidden. [[GH-2549]](fatih/vim-go#2549) * Restore cwd after `:GoTest` when `g:go_term_enabled` is set. [[GH-2556]](fatih/vim-go#2556) * Expand struct variable correctly in the variables debug window. [[GH-2574]](fatih/vim-go#2574) * Show output from errcheck when there are failures other than problems it can report. [[GH-2667]](fatih/vim-go#2667) Signed-off-by: Patrick McCarty <patrick.mccarty@intel.com>

k33nice force-pushed the master branch from 520e57e to 20f21f2 Compare November 14, 2019 20:59

k33nice changed the title ~~syntax: Fix hex and octal int and errors highlight~~ [WIP] syntax: Fix hex and octal int and errors highlight Nov 14, 2019

k33nice changed the title ~~[WIP] syntax: Fix hex and octal int and errors highlight~~ syntax: Fix hex and octal int and errors highlight Nov 15, 2019

bhcleek reviewed Nov 15, 2019

View reviewed changes

k33nice added 8 commits November 15, 2019 23:41

syntax: Fix nubmers hilightings add more tests

c5ad443

*Changes* - Add new goDecimalErrorInt hi - Add support of "_" in goDecimalInt - Fix goOctalError to support more wide patterns of errors - Correct support of goups by "_" in goBinaryError

syntax: Support "_" after prefix or first char

3a23084

*Fix* - goDecimalInt: add support for sequneces like `1_234567890` - goBinaryInt: add support for sequneces like `0b_01`

Remove redundant test cases, fix type on test name

b30bb2d

syntax: Change decimal error to not match octal error

fb643bf

syntax: Fix goOctalError and test cases for it

f6ed8b6

*Changes* - removed redundant characters from exclusion pattern - fixed tests so that test include only one testing error - added test case for explicitly test trailing "o" or "O"

k33nice force-pushed the master branch from 2363cc3 to f6ed8b6 Compare November 15, 2019 22:49

bhcleek reviewed Nov 16, 2019

View reviewed changes

k33nice added 4 commits November 16, 2019 21:47

syntax: Simplify goBinaryError regexp

67e76b8

syntax: fix and simplify goOctalErro regexp, test

2c6b985

**Changes** - Added "]" and "}" to whitelist subgroup. To doesn't match on [0][0] and [][]int{{0},{0}}. - Expanded subgroup for trailing [oO] to [oOxX]. - Added test case for trailing 'x'.

syntax: Simplify hexadecimal, decimal error regexp

cf661f5

syntax:Get rid of test cases that don't give value

fc8cef8

bhcleek reviewed Nov 16, 2019

View reviewed changes

syntax/go.vim Show resolved Hide resolved

syntax: Add testing case to ensure overrdies works

cc54040

bhcleek merged commit 9a3808d into fatih:master Nov 18, 2019

bhcleek added a commit that referenced this pull request Nov 18, 2019

update CHANGELOG.md for #2584

7fadfcc

bhcleek mentioned this pull request Nov 18, 2019

Binary number literals with underscores appear as error blocks #2588

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

syntax: Fix hex and octal int and errors highlight #2584

syntax: Fix hex and octal int and errors highlight #2584

k33nice commented Nov 13, 2019

bhcleek commented Nov 14, 2019

bhcleek commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 14, 2019

k33nice commented Nov 14, 2019

k33nice commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 15, 2019

bhcleek left a comment

bhcleek Nov 15, 2019

k33nice Nov 15, 2019

bhcleek Nov 16, 2019

k33nice Nov 16, 2019

bhcleek Nov 15, 2019

k33nice Nov 15, 2019

bhcleek Nov 16, 2019

k33nice Nov 16, 2019 •

edited

Loading

k33nice commented Nov 15, 2019

bhcleek Nov 16, 2019

bhcleek Nov 16, 2019

k33nice Nov 16, 2019

bhcleek Nov 16, 2019

bhcleek left a comment

bhcleek commented Nov 16, 2019

k33nice commented Nov 16, 2019 •

edited

Loading

bhcleek commented Nov 16, 2019

k33nice commented Nov 17, 2019

bhcleek commented Nov 17, 2019

syntax: Fix hex and octal int and errors highlight #2584

syntax: Fix hex and octal int and errors highlight #2584

Conversation

k33nice commented Nov 13, 2019

bhcleek commented Nov 14, 2019

bhcleek commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 14, 2019

k33nice commented Nov 14, 2019

k33nice commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 14, 2019

bhcleek commented Nov 14, 2019

k33nice commented Nov 15, 2019

bhcleek 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

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

k33nice Nov 16, 2019 • edited Loading

Choose a reason for hiding this comment

k33nice commented Nov 15, 2019

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

bhcleek left a comment

Choose a reason for hiding this comment

bhcleek commented Nov 16, 2019

k33nice commented Nov 16, 2019 • edited Loading

bhcleek commented Nov 16, 2019

k33nice commented Nov 17, 2019

bhcleek commented Nov 17, 2019

k33nice Nov 16, 2019 •

edited

Loading

k33nice commented Nov 16, 2019 •

edited

Loading