Add workaround for displaying ellipsis on windows #43

jankatins · 2016-04-13T19:40:30Z

On windows, ellipses are usually outside the chars which can be displayed in our
way to display plain text /with capture.output(print(...)) in the current
locale. The leads to strangely escaped chars and we would have such a escaped
chars each time we display a dataframe.

This problem is probably not going away soon:
#28 (comment)

Work around this by simple using three dots.

Fix the unittests on windows

jankatins · 2016-04-13T19:40:58Z

Lets see if my changes to the unittests let this pass... on windows, they don't :-(

flying-sheep · 2016-04-14T10:11:09Z

tests/testthat/test_array_manipulation.r

-  ellip.h <- '\u22EF'
-
+  # copy of the function repr_matrix_df.r
+  .char_fallback <- function(char, default) {


no duplicate code. please put it into utils.r.

and btw, why can’t the tests just use the ellip.{hvd} from repr_matrix_df?

aren’t tests executed in the package namespace and have private access?

I've no clue, but the tests are anyway failing for me on windows (and I've no clue why). I tried putting the function in only once as a top level function in the test file, but that didn't work either and I don't want to export the function just to make the tests happy...

what do you mean with “work” here?

didn't work = the test failed even more... (actually I was walking in the dark with this: I've no clue what and why test failed on windows) If you tell me that adding the function on top of the file will still let it pass on travis, that's fine with me.

ok, the fact that you don’t know is weird: shouldn’t the test runner tell you?

It tells me that the test is broken, but it didn't show me what specific was broken (basically everything but the matrix comparison in the last testcase were :-()

I haven't worked with test_that yet, so no clue how to get better output or how to debug this stuff :-/

something’s fucky. you should see sth like here.

how do you run the tests?

R CMD check ...

as you can see in the link, that’s what travis uses to produce that output, as well. hmm.

anyway: the testthat package has a function to run the tests, you can try that one.

jankatins · 2016-04-19T20:33:54Z

These are the failures (now via Rstudio):

==> devtools::test()

Loading repr
Loading required package: testthat
Loading required namespace: Cairo
Testing repr
Array and vector truncation: ..............
Attaching package: 'dplyr'

The following objects are masked from 'package:data.table':

    between, last

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

................................WWWW.123WWWW.456WWWW.789
LaTeX and HTML escaping: ..............abcd....efFF

Warnings -------------------------------------------------------------------
1. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#318) - duplicated levels in factors are deprecated

2. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#319) - duplicated levels in factors are deprecated

3. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#320) - duplicated levels in factors are deprecated

4. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#325) - duplicated levels in factors are deprecated

5. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#340) - duplicated levels in factors are deprecated

6. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#341) - duplicated levels in factors are deprecated

7. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#342) - duplicated levels in factors are deprecated

8. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#348) - duplicated levels in factors are deprecated

9. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#363) - duplicated levels in factors are deprecated

10. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#364) - duplicated levels in factors are deprecated

11. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#365) - duplicated levels in factors are deprecated

12. ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#372) - duplicated levels in factors are deprecated

Failed ---------------------------------------------------------------------
1. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#333) 
`limited_df` not equal to `expected_df`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


2. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#334) 
`limited_dt` not equal to `expected_dt`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


3. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#335) 
`limited_tbl` not equal to `expected_tbl`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


4. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#356) 
`limited_df` not equal to `expected_df`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


5. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#357) 
`limited_dt` not equal to `expected_dt`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


6. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#358) 
`limited_tbl` not equal to `expected_tbl`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


7. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#380) 
`limited_df` not equal to `expected_df`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


8. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#381) 
`limited_dt` not equal to `expected_dt`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


9. Failure: ellip.limit.arr limits arrays that are long and wide (@test_array_manipulation.r#382) 
`limited_tbl` not equal to `expected_tbl`.
Component "...": Attributes: < Component "levels": Lengths (2, 3) differ (string compare on first 2) >
Component "...": Attributes: < Component "levels": 1 string mismatch >


10. Failure: LaTeX escaping in vectors works (@test_escaping.r#24) -------------
repr_latex("[") not equal to "'{[}'".
1/1 mismatches
x[1]: "\"{[}\""
y[1]: "'{[}'"


11. Failure: LaTeX escaping in vectors works (@test_escaping.r#25) -------------
repr_latex(c("[", "|")) not equal to "\\begin{enumerate*}\n\\item '{[}'\n\\item '\\textbar{}'\n\\end{enumerate*}\n".
1/1 mismatches
x[1]: "\\begin{enumerate*}\n\\item \"{[}\"\n\\item \"\\textbar{}\"\n\\end{enumer
x[1]: ate*}\n"
y[1]: "\\begin{enumerate*}\n\\item '{[}'\n\\item '\\textbar{}'\n\\end{enumerate*
y[1]: }\n"


12. Failure: HTML escaping in vectors works (@test_escaping.r#34) --------------
repr_html("<") not equal to "'&lt;'".
1/1 mismatches
x[1]: "\"&lt;\""
y[1]: "'&lt;'"


13. Failure: HTML escaping in vectors works (@test_escaping.r#35) --------------
repr_html(c("<", "&")) not equal to "<ol class=list-inline>\n\t<li>'&lt;'</li>\n\t<li>'&amp;'</li>\n</ol>\n".
1/1 mismatches
x[1]: "<ol class=list-inline>\n\t<li>\"&lt;\"</li>\n\t<li>\"&amp;\"</li>\n</ol>\
x[1]: n"
y[1]: "<ol class=list-inline>\n\t<li>'&lt;'</li>\n\t<li>'&amp;'</li>\n</ol>\n"


14. Failure: LaTeX escaping in lists works (@test_escaping.r#80) ---------------
repr_latex(list(lbr = "[")) not equal to "\\textbf{\\$lbr} = '{[}'".
1/1 mismatches
x[1]: "\\textbf{\\$lbr} = \"{[}\""
y[1]: "\\textbf{\\$lbr} = '{[}'"


15. Error: LaTeX escaping in lists works (@test_escaping.r#81) -----------------
too few arguments
1: expect_equal(repr_latex(list(`&` = "%")), "\\textbf{\\$`\\&`} = '\\%'") at C:\data\external\R\repr/tests/testthat/test_escaping.r:81
2: compare(object, expected, ...)
3: compare.character(object, expected, ...)
4: difference(format(mismatches, max_diffs = max_diffs, max_lines = max_lines, width = width))
5: comparison(FALSE, sprintf(...))
6: stopifnot(is.character(message), length(message) == 1)
7: sprintf(...)

DONE =======================================================================

Not sure if I executed the test in the right way...

jankatins · 2016-04-21T15:09:41Z

SO, IMO the unittest failures on win and the ggplot error logging are the only blocker for a release.

jankatins · 2016-04-21T15:44:48Z

And it is green on windows... Lets see what travis tells us...

flying-sheep · 2016-04-21T17:43:14Z

tests/testthat/test_array_manipulation.r

@@ -5,6 +5,17 @@ library(data.table)
 library(dplyr)
 options('stringsAsFactors' = FALSE)

+# copy of the function repr_matrix_df.r
+.char_fallback <- function(char, default) {


can you really not use the one from the package? maybe do ellip.h <- IRkernel:::ellip.h` (and so on)?

done

Doing this all in RStudio is make this much easier...

On windows, ellipses are usually outside the chars which can be displayed in our way to display plain text /with `capture.output(print(...))` in the current locale. The leads to strangely escaped chars and we would have such a escaped chars each time we display a dataframe. This problem is probably not going away soon: IRkernel#28 (comment) Work around this by simple using three dots.

The tests were failing because for unknown reasons, the repr_xxx was returning inline quotes signs as " instead of the expected '. The easiest fix is to substitute that in a special expect_equivalent_string function.

jankatins · 2016-04-21T19:57:35Z

travis is green, all redundant functions and fields were remove. If @flying-sheep doesn't find any style problems, it's IMO ready to be merged.

flying-sheep reviewed Apr 14, 2016
View reviewed changes

jankatins force-pushed the ellipsis branch from 8da0b6b to 1e07e20 Compare April 21, 2016 15:44

jankatins force-pushed the ellipsis branch from 1e07e20 to 39d8155 Compare April 21, 2016 15:50

jankatins mentioned this pull request Apr 21, 2016

broken unicode uses <U+884C> which needs to be escaped #21

Open

flying-sheep reviewed Apr 21, 2016
View reviewed changes

jankatins added 2 commits April 21, 2016 21:35

Fix failing tests on windows

e11624d

The tests were failing because for unknown reasons, the repr_xxx was returning inline quotes signs as " instead of the expected '. The easiest fix is to substitute that in a special expect_equivalent_string function.

jankatins force-pushed the ellipsis branch from 39d8155 to e11624d Compare April 21, 2016 19:35

flying-sheep merged commit 31dc972 into IRkernel:master Apr 21, 2016

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add workaround for displaying ellipsis on windows #43

Add workaround for displaying ellipsis on windows #43

jankatins commented Apr 13, 2016 •

edited

Loading

jankatins commented Apr 13, 2016

flying-sheep Apr 14, 2016

jankatins Apr 14, 2016

flying-sheep Apr 14, 2016

jankatins Apr 14, 2016 •

edited

Loading

flying-sheep Apr 14, 2016

jankatins Apr 14, 2016 •

edited

Loading

flying-sheep Apr 15, 2016

jankatins Apr 15, 2016

flying-sheep Apr 17, 2016

jankatins commented Apr 19, 2016 •

edited

Loading

jankatins commented Apr 21, 2016

jankatins commented Apr 21, 2016

flying-sheep Apr 21, 2016

jankatins Apr 21, 2016 •

edited

Loading

jankatins commented Apr 21, 2016

Add workaround for displaying ellipsis on windows #43

Add workaround for displaying ellipsis on windows #43

Conversation

jankatins commented Apr 13, 2016 • edited Loading

jankatins commented Apr 13, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jankatins Apr 14, 2016 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jankatins Apr 14, 2016 • edited Loading

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

jankatins commented Apr 19, 2016 • edited Loading

jankatins commented Apr 21, 2016

jankatins commented Apr 21, 2016

Choose a reason for hiding this comment

jankatins Apr 21, 2016 • edited Loading

Choose a reason for hiding this comment

jankatins commented Apr 21, 2016

jankatins commented Apr 13, 2016 •

edited

Loading

jankatins Apr 14, 2016 •

edited

Loading

jankatins Apr 14, 2016 •

edited

Loading

jankatins commented Apr 19, 2016 •

edited

Loading

jankatins Apr 21, 2016 •

edited

Loading