[CI][R][Python] R and Python integration's tests failed #39933

kou · 2024-02-04T04:33:50Z

Describe the bug, including details regarding any error messages, version, and platform.

https://github.com/apache/arrow/actions/runs/7760862085/job/21168058007#step:6:32228

══ Failed tests ════════════════════════════════════════════════════════════════
── Error ('test-python-flight.R:32:5'): flight_put ─────────────────────────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. └─arrow::flight_put(client, example_data, path = flight_obj) at test-python-flight.R:32:5
 2.   ├─reticulate::r_to_py(data)
 3.   └─arrow:::r_to_py.Table(data)
 4.     ├─reticulate::r_to_py(as_record_batch_reader(x), convert = FALSE)
 5.     └─arrow:::r_to_py.RecordBatchReader(...)
 6.       └─pa$lib$RecordBatchReader$`_import_from_c`(pyarrow_compatible_pointer(stream_ptr))
 7.         └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)
── Error ('test-python-flight.R:42:5'): flight_put with max_chunksize ──────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. └─arrow::flight_put(client, example_data, path = flight_obj, max_chunksize = 1) at test-python-flight.R:42:5
 2.   ├─reticulate::r_to_py(data)
 3.   └─arrow:::r_to_py.Table(data)
 3.   └─pa$Field$`_import_from_c`(pyarrow_compatible_pointer(schema_ptr))
 4.     └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)
── Error ('test-python.R:157:3'): RecordBatchReader to python ──────────────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. ├─reticulate::r_to_py(reader) at test-python.R:157:3
 2. └─arrow:::r_to_py.RecordBatchReader(reader)
 3.   └─pa$lib$RecordBatchReader$`_import_from_c`(pyarrow_compatible_pointer(stream_ptr))
 4.     └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)
── Error ('test-python.R:177:3'): RecordBatchReader from python ────────────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. ├─reticulate::r_to_py(reader) at test-python.R:177:3
 2. └─arrow:::r_to_py.RecordBatchReader(reader)
 3.   └─pa$lib$RecordBatchReader$`_import_from_c`(pyarrow_compatible_pointer(stream_ptr))
 4.     └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

The last success commit: 2721134
The first failed commit: b684028

Commits between them:

@jorisvandenbossche Could you take a look at this PyCapsule related error messages?

Component(s)

Continuous Integration, Python, R

The text was updated successfully, but these errors were encountered:

kou · 2024-02-05T23:48:01Z

#39942 may be happen by the same cause.

amoeba · 2024-02-06T00:58:34Z

The timing of this failure starting to show up is around the time version 1.35.0 of reticulate came out and it looks like 1.35.0 was used in during the above failures.

kou · 2024-02-06T01:34:24Z

Can we pin reticulate for now?
Could you report this to reticulate?

amoeba · 2024-02-06T01:42:47Z

Let me verify my guess, I'll report back.

amoeba · 2024-02-06T02:08:48Z

I ran the tests in ./r/tests/testthat/test-python-flight.R with reticulate 1.34.0 and 1.35.0 and get different failures so I think this needs a deeper investigation.

reticulate 1.34.0

==> Testing R file using 'testthat'

ℹ Loading arrow
[ FAIL 6 | WARN 0 | SKIP 1 | PASS 1 ]

── Error (test-python-flight.R:28:5): flight_path_exists ───────────────────────
<pyarrow.lib.ArrowInvalid/python.builtin.ValueError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_iterate(it, f)`: pyarrow.lib.ArrowInvalid: invalid literal for int() with base 10: ''. Detail: Python exception: ValueError
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. ├─testthat::expect_false(flight_obj %in% list_flights(client)) at test-python-flight.R:28:5
 2. │ └─testthat::quasi_label(enquo(object), label, arg = "object")
 3. │   └─rlang::eval_bare(expr, quo_get_env(quo))
 4. ├─flight_obj %in% list_flights(client)
 5. └─arrow::list_flights(client)
 6.   └─reticulate::iterate(generator, function(x) as.character(x$descriptor$path[[1]])) at r/R/flight.R:121:3
 7.     └─reticulate:::py_iterate(it, f)

── Error (test-python-flight.R:33:5): flight_put ───────────────────────────────
<pyarrow.lib.ArrowInvalid/python.builtin.ValueError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: pyarrow.lib.ArrowInvalid: invalid literal for int() with base 10: ''. Detail: Python exception: ValueError
Run `reticulate::py_last_error()` for details.
Backtrace:
     ▆
  1. ├─testthat::expect_true(flight_path_exists(client, flight_obj)) at test-python-flight.R:33:5
  2. │ └─testthat::quasi_label(enquo(object), label, arg = "object")
  3. │   └─rlang::eval_bare(expr, quo_get_env(quo))
  4. └─arrow::flight_path_exists(client, flight_obj)
  5.   ├─base::tryCatch(...) at r/R/flight.R:128:3
  6.   │ └─base (local) tryCatchList(expr, classes, parentenv, handlers)
  7.   │   └─base (local) tryCatchOne(expr, names, parentenv, handlers[[1L]])
  8.   │     └─base (local) doTryCatch(return(expr), name, parentenv, handler)
  9.   └─client$get_flight_info(descriptor_for_path(path)) at r/R/flight.R:130:7
 10.     └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:43:5): flight_put with max_chunksize ────────────
<pyarrow.lib.ArrowInvalid/python.builtin.ValueError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: pyarrow.lib.ArrowInvalid: invalid literal for int() with base 10: ''. Detail: Python exception: ValueError
Run `reticulate::py_last_error()` for details.
Backtrace:
     ▆
  1. ├─testthat::expect_true(flight_path_exists(client, flight_obj)) at test-python-flight.R:43:5
  2. │ └─testthat::quasi_label(enquo(object), label, arg = "object")
  3. │   └─rlang::eval_bare(expr, quo_get_env(quo))
  4. └─arrow::flight_path_exists(client, flight_obj)
  5.   ├─base::tryCatch(...) at r/R/flight.R:128:3
  6.   │ └─base (local) tryCatchList(expr, classes, parentenv, handlers)
  7.   │   └─base (local) tryCatchOne(expr, names, parentenv, handlers[[1L]])
  8.   │     └─base (local) doTryCatch(return(expr), name, parentenv, handler)
  9.   └─client$get_flight_info(descriptor_for_path(path)) at r/R/flight.R:130:7
 10.     └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:56:5): flight_get ───────────────────────────────
<pyarrow.lib.ArrowInvalid/python.builtin.ValueError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: pyarrow.lib.ArrowInvalid: invalid literal for int() with base 10: ''. Detail: Python exception: ValueError
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. ├─expect_equal_data_frame(flight_get(client, flight_obj), example_data) at test-python-flight.R:56:5
 2. │ ├─expect_equal(as.data.frame(x), as.data.frame(y), ...) at tests/testthat/helper-expectation.R:24:3
 3. │ │ └─base::inherits(object, "ArrowObject") at tests/testthat/helper-expectation.R:35:3
 4. │ └─base::as.data.frame(x)
 5. └─arrow::flight_get(client, flight_obj)
 6.   └─arrow:::flight_reader(client, path) at r/R/flight.R:96:3
 7.     └─client$get_flight_info(descriptor_for_path(path)) at r/R/flight.R:102:3
 8.       └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:62:5): flight_put with RecordBatch ──────────────
<pyarrow.lib.ArrowInvalid/python.builtin.ValueError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: pyarrow.lib.ArrowInvalid: invalid literal for int() with base 10: ''. Detail: Python exception: ValueError
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. ├─expect_equal_data_frame(flight_get(client, flight_obj2), example_data) at test-python-flight.R:62:5
 2. │ ├─expect_equal(as.data.frame(x), as.data.frame(y), ...) at tests/testthat/helper-expectation.R:24:3
 3. │ │ └─base::inherits(object, "ArrowObject") at tests/testthat/helper-expectation.R:35:3
 4. │ └─base::as.data.frame(x)
 5. └─arrow::flight_get(client, flight_obj2)
 6.   └─arrow:::flight_reader(client, path) at r/R/flight.R:96:3
 7.     └─client$get_flight_info(descriptor_for_path(path)) at r/R/flight.R:102:3
 8.       └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:66:5): flight_put with overwrite = FALSE ────────
<pyarrow.lib.ArrowInvalid/python.builtin.ValueError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: pyarrow.lib.ArrowInvalid: invalid literal for int() with base 10: ''. Detail: Python exception: ValueError
Run `reticulate::py_last_error()` for details.
Backtrace:
     ▆
  1. ├─testthat::expect_error(...) at test-python-flight.R:66:5
  2. │ └─testthat:::expect_condition_matching(...)
  3. │   └─testthat:::quasi_capture(...)
  4. │     ├─testthat (local) .capture(...)
  5. │     │ └─base::withCallingHandlers(...)
  6. │     └─rlang::eval_bare(quo_get_expr(.quo), quo_get_env(.quo))
  7. └─arrow::flight_put(...)
  8.   └─arrow::flight_path_exists(client, path) at r/R/flight.R:66:3
  9.     ├─base::tryCatch(...) at r/R/flight.R:128:3
 10.     │ └─base (local) tryCatchList(expr, classes, parentenv, handlers)
 11.     │   └─base (local) tryCatchOne(expr, names, parentenv, handlers[[1L]])
 12.     │     └─base (local) doTryCatch(return(expr), name, parentenv, handler)
 13.     └─client$get_flight_info(descriptor_for_path(path)) at r/R/flight.R:130:7
 14.       └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Skipped tests (1) ───────────────────────────────────────────────────────────────────────────────────────────────
• empty test (1): test-python-flight.R:75:3

[ FAIL 6 | WARN 0 | SKIP 1 | PASS 1 ]

Test complete

reticulate 1.35.0

==> Testing R file using 'testthat'

ℹ Loading arrow
[ FAIL 5 | WARN 0 | SKIP 1 | PASS 2 ]

── Error (test-python-flight.R:32:5): flight_put ───────────────────────────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. └─arrow::flight_put(client, example_data, path = flight_obj) at test-python-flight.R:32:5
 2.   ├─reticulate::r_to_py(data) at r/R/flight.R:73:3
 3.   └─arrow:::r_to_py.Table(data)
 4.     ├─reticulate::r_to_py(as_record_batch_reader(x), convert = FALSE) at r/R/python.R:111:3
 5.     └─arrow:::r_to_py.RecordBatchReader(...)
 6.       └─pa$lib$RecordBatchReader$`_import_from_c`(pyarrow_compatible_pointer(stream_ptr)) at r/R/python.R:227:3
 7.         └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:42:5): flight_put with max_chunksize ────────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. └─arrow::flight_put(client, example_data, path = flight_obj, max_chunksize = 1) at test-python-flight.R:42:5
 2.   ├─reticulate::r_to_py(data) at r/R/flight.R:73:3
 3.   └─arrow:::r_to_py.Table(data)
 4.     ├─reticulate::r_to_py(as_record_batch_reader(x), convert = FALSE) at r/R/python.R:111:3
 5.     └─arrow:::r_to_py.RecordBatchReader(...)
 6.       └─pa$lib$RecordBatchReader$`_import_from_c`(pyarrow_compatible_pointer(stream_ptr)) at r/R/python.R:227:3
 7.         └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:56:5): flight_get ───────────────────────────────
<pyarrow.lib.ArrowKeyError/python.builtin.KeyError/python.builtin.LookupError/pyarrow.lib.ArrowException/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: pyarrow.lib.ArrowKeyError: 'Flight not found.'. Detail: Python exception: KeyError
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. ├─expect_equal_data_frame(flight_get(client, flight_obj), example_data) at test-python-flight.R:56:5
 2. │ ├─expect_equal(as.data.frame(x), as.data.frame(y), ...) at tests/testthat/helper-expectation.R:24:3
 3. │ │ └─base::inherits(object, "ArrowObject") at tests/testthat/helper-expectation.R:35:3
 4. │ └─base::as.data.frame(x)
 5. └─arrow::flight_get(client, flight_obj)
 6.   └─arrow:::flight_reader(client, path) at r/R/flight.R:96:3
 7.     └─client$get_flight_info(descriptor_for_path(path)) at r/R/flight.R:102:3
 8.       └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:61:5): flight_put with RecordBatch ──────────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
    ▆
 1. └─arrow::flight_put(client, RecordBatch$create(example_data), path = flight_obj2) at test-python-flight.R:61:5
 2.   ├─reticulate::r_to_py(data) at r/R/flight.R:73:3
 3.   └─arrow:::r_to_py.RecordBatch(data)
 4.     └─pa$RecordBatch$`_import_from_c`(...) at r/R/python.R:84:3
 5.       └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Error (test-python-flight.R:66:5): flight_put with overwrite = FALSE ────────
<python.builtin.ValueError/python.builtin.Exception/python.builtin.BaseException/python.builtin.object/error/condition>
Error in `py_call_impl(callable, call_args$unnamed, call_args$named)`: ValueError: PyCapsule_GetPointer called with incorrect name
Run `reticulate::py_last_error()` for details.
Backtrace:
     ▆
  1. ├─testthat::expect_error(...) at test-python-flight.R:66:5
  2. │ └─testthat:::expect_condition_matching(...)
  3. │   └─testthat:::quasi_capture(...)
  4. │     ├─testthat (local) .capture(...)
  5. │     │ └─base::withCallingHandlers(...)
  6. │     └─rlang::eval_bare(quo_get_expr(.quo), quo_get_env(.quo))
  7. └─arrow::flight_put(...)
  8.   ├─reticulate::r_to_py(data) at r/R/flight.R:73:3
  9.   └─arrow:::r_to_py.Table(data)
 10.     ├─reticulate::r_to_py(as_record_batch_reader(x), convert = FALSE) at r/R/python.R:111:3
 11.     └─arrow:::r_to_py.RecordBatchReader(...)
 12.       └─pa$lib$RecordBatchReader$`_import_from_c`(pyarrow_compatible_pointer(stream_ptr)) at r/R/python.R:227:3
 13.         └─reticulate:::py_call_impl(callable, call_args$unnamed, call_args$named)

── Skipped tests (1) ───────────────────────────────────────────────────────────────────────────────────────────────
• empty test (1): test-python-flight.R:75:3

[ FAIL 5 | WARN 0 | SKIP 1 | PASS 2 ]

Test complete

paleolimbot · 2024-02-06T02:15:11Z

I believe reticulate 1.35.0 changed the "name" of the capsule. I am not sure what it was before and I am not sure what it is now...we started using external pointers on the R side in Arrow 7.0.0, which converted to Python capsules automatically, because it felt better than converting the pointer to an R double(), sending that to Python, then static casting that to a uintptr_t and calling it a pointer again.

One of the errors is along the lines of PyCapsule_GetPointer() called with incorrect name, so perhaps the name check was made stricter on the Python side when capsules started being used there.

On the R side, we could go back to calling _import_from_c with a floating-point value (although I would rather not). We also have the ability to call it with a string representation of the address, which is also lossless, but this is rejected by pyarrow (I just checked). I also just checked passing a floating point value and that results in the test passing for me.

The code on the R side that could change is here:

arrow/r/R/python.R

Lines 341 to 353 in 672238f

    
           pyarrow_compatible_pointer <- function(ptr) { 
        
             pa <- reticulate::import("pyarrow") 
        
             version_string <- pa$`__version__` 
        
             # remove trailing .devXXX because it won't work with package_version() 
        
             pyarrow_version <- package_version(gsub("\\.dev.*?$", "", version_string)) 
        
             # pyarrow pointers changed in version 7.0.0 
        
             if (pyarrow_version >= "7.0.0") { 
        
               return(ptr) 
        
             } else { 
        
               return(external_pointer_addr_double(ptr)) 
        
             } 
        
           }

amoeba · 2024-02-06T05:37:14Z

Thanks @paleolimbot. I just checked the diff between 1.34.0 and 1.35.0 and I think you're right on both counts. In rstudio/reticulate@9b2c8d4 the name changed from "NULL" to "r_extptr" and code was added to validate against the name (by calling PyCapsule_IsValid).

jorisvandenbossche · 2024-02-06T08:34:12Z

so perhaps the name check was made stricter on the Python side when capsules started being used there.

FWIW, the python.R code is not yet using capsules (it's still using _import_from_c(..) with raw pointers, as can also be seen in the error traceback), so this "name check" is happening somewhere in reticulate itself?
(just want to say that AFAIU this is not related to the capsule work on the python side)

paleolimbot · 2024-02-06T12:52:53Z

Got it!

It looks like capsules with a null name are accepted...perhaps it should be updated to accept a capsule with NULL or r_extptr?

arrow/python/pyarrow/types.pxi

Lines 104 to 105 in cd5faaf

    
               Integers are accepted as well as capsule objects with a NULL name. 
        
               (the latter for compatibility with raw pointers exported by reticulate)

paleolimbot · 2024-02-06T12:56:01Z

Or, alternatively, we could fix in R by generating the Python integer there:

py <- reticulate::import_builtins(convert = FALSE)
# addr <- external_pointer_addr_double(ptr)
addr <- "12345"
class(py$int(addr))
#> [1] "python.builtin.int"    "python.builtin.object"
py$int(addr)
#> 12345

^{Created on 2024-02-06 with reprex v2.0.2}

pitrou · 2024-02-06T13:43:04Z

In rstudio/reticulate@9b2c8d4 the name changed from "NULL" to "r_extptr" and code was added to validate against the name (by calling PyCapsule_IsValid).

This seems to be about storing R objects inside capsules. This is not what the R-PyArrow bridge is doing, is it?

Also, if the name is not r_extptr, this code doesn't seem to actually change anything.

jorisvandenbossche · 2024-02-06T13:49:24Z

My understanding is that the r_to_py will convert any extptr to such a capsule:

https://github.com/rstudio/reticulate/blob/950169382a92328bde39bff3b68f5014ac582986/src/python.cpp#L1971-L1974

And I suppose the array/schema pointers in the R package are "external pointers", and so reticulate automatically puts them in a capsule before passing to _import_from_c (since this is a python call, it converts any argument to it with r_to_py?)

pitrou · 2024-02-06T13:54:58Z

Ahah, I see. Then I guess it makes sense to make the check more tolerant on the PyArrow side.

paleolimbot · 2024-02-06T15:28:24Z

I will also submit a fix on the R side...it didn't occur to me at the time to convert to a Python integer using reticulate to work around the fact that R doesn't have 64-bit integers. If it's easy, it would probably also be a good idea to fix it in Python (or remove the ability to pass a pointer in that way, or fix the comment to reflect the new nature of how an external pointer gets converted using reticulate).

pitrou · 2024-02-06T15:44:35Z

We should do both, because ideally one can use an older R-Arrow with a newer PyArrow and vice-versa.

…39969) ### Rationale for this change The integration tests and documentation build is failing ### What changes are included in this PR? Instead of relying on how reticulate converts an R external pointer, use a Python integer instead. We can't use an R integer (because they're only 32 bits); we can't use an R double (because the static cast to/from uintptr_t is a bit iffy); however, we can use Python to convert a string to Python integer. This is probably how I should have written it the first time but it didn't occur to me at the time. ### Are these changes tested? Yes, covered by existing tests. ### Are there any user-facing changes? No * Closes: #39933 Lead-authored-by: Dewey Dunnington <[email protected]> Co-authored-by: Dewey Dunnington <[email protected]> Signed-off-by: Dewey Dunnington <[email protected]>

…ulate (apache#39969) ### Rationale for this change The integration tests and documentation build is failing ### What changes are included in this PR? Instead of relying on how reticulate converts an R external pointer, use a Python integer instead. We can't use an R integer (because they're only 32 bits); we can't use an R double (because the static cast to/from uintptr_t is a bit iffy); however, we can use Python to convert a string to Python integer. This is probably how I should have written it the first time but it didn't occur to me at the time. ### Are these changes tested? Yes, covered by existing tests. ### Are there any user-facing changes? No * Closes: apache#39933 Lead-authored-by: Dewey Dunnington <[email protected]> Co-authored-by: Dewey Dunnington <[email protected]> Signed-off-by: Dewey Dunnington <[email protected]>

…39969) ### Rationale for this change The integration tests and documentation build is failing ### What changes are included in this PR? Instead of relying on how reticulate converts an R external pointer, use a Python integer instead. We can't use an R integer (because they're only 32 bits); we can't use an R double (because the static cast to/from uintptr_t is a bit iffy); however, we can use Python to convert a string to Python integer. This is probably how I should have written it the first time but it didn't occur to me at the time. ### Are these changes tested? Yes, covered by existing tests. ### Are there any user-facing changes? No * Closes: #39933 Lead-authored-by: Dewey Dunnington <[email protected]> Co-authored-by: Dewey Dunnington <[email protected]> Signed-off-by: Dewey Dunnington <[email protected]>

…ulate (apache#39969) ### Rationale for this change The integration tests and documentation build is failing ### What changes are included in this PR? Instead of relying on how reticulate converts an R external pointer, use a Python integer instead. We can't use an R integer (because they're only 32 bits); we can't use an R double (because the static cast to/from uintptr_t is a bit iffy); however, we can use Python to convert a string to Python integer. This is probably how I should have written it the first time but it didn't occur to me at the time. ### Are these changes tested? Yes, covered by existing tests. ### Are there any user-facing changes? No * Closes: apache#39933 Lead-authored-by: Dewey Dunnington <[email protected]> Co-authored-by: Dewey Dunnington <[email protected]> Signed-off-by: Dewey Dunnington <[email protected]>

…39969) ### Rationale for this change The integration tests and documentation build is failing ### What changes are included in this PR? Instead of relying on how reticulate converts an R external pointer, use a Python integer instead. We can't use an R integer (because they're only 32 bits); we can't use an R double (because the static cast to/from uintptr_t is a bit iffy); however, we can use Python to convert a string to Python integer. This is probably how I should have written it the first time but it didn't occur to me at the time. ### Are these changes tested? Yes, covered by existing tests. ### Are there any user-facing changes? No * Closes: #39933 Lead-authored-by: Dewey Dunnington <[email protected]> Co-authored-by: Dewey Dunnington <[email protected]> Signed-off-by: Dewey Dunnington <[email protected]>

…ulate (apache#39969) ### Rationale for this change The integration tests and documentation build is failing ### What changes are included in this PR? Instead of relying on how reticulate converts an R external pointer, use a Python integer instead. We can't use an R integer (because they're only 32 bits); we can't use an R double (because the static cast to/from uintptr_t is a bit iffy); however, we can use Python to convert a string to Python integer. This is probably how I should have written it the first time but it didn't occur to me at the time. ### Are these changes tested? Yes, covered by existing tests. ### Are there any user-facing changes? No * Closes: apache#39933 Lead-authored-by: Dewey Dunnington <[email protected]> Co-authored-by: Dewey Dunnington <[email protected]> Signed-off-by: Dewey Dunnington <[email protected]>

kou added the Type: bug label Feb 4, 2024

github-actions bot added Component: R Component: Python Component: Continuous Integration labels Feb 4, 2024

Tom-Newton mentioned this issue Feb 4, 2024

GH-39621: [CI][Packaging] Update vcpkg to 2023.11.20 release #39622

Merged

kou mentioned this issue Feb 5, 2024

[CI][R][Docs] Docs jobs failing due to R Markdown error #39942

Closed

github-actions bot assigned paleolimbot Feb 6, 2024

github-actions bot mentioned this issue Feb 6, 2024

GH-39933: [R] Fix pointer conversion to Python for latest reticulate #39969

Merged

paleolimbot closed this as completed in #39969 Feb 7, 2024

paleolimbot added this to the 16.0.0 milestone Feb 7, 2024

assignUser modified the milestones: 16.0.0, 15.0.1 Feb 21, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[CI][R][Python] R and Python integration's tests failed #39933

[CI][R][Python] R and Python integration's tests failed #39933

kou commented Feb 4, 2024

kou commented Feb 5, 2024

amoeba commented Feb 6, 2024

kou commented Feb 6, 2024

amoeba commented Feb 6, 2024

amoeba commented Feb 6, 2024 •

edited

Loading

paleolimbot commented Feb 6, 2024

amoeba commented Feb 6, 2024

jorisvandenbossche commented Feb 6, 2024

paleolimbot commented Feb 6, 2024

paleolimbot commented Feb 6, 2024

pitrou commented Feb 6, 2024 •

edited

Loading

jorisvandenbossche commented Feb 6, 2024

pitrou commented Feb 6, 2024

paleolimbot commented Feb 6, 2024

pitrou commented Feb 6, 2024

[CI][R][Python] R and Python integration's tests failed #39933

[CI][R][Python] R and Python integration's tests failed #39933

Comments

kou commented Feb 4, 2024

Describe the bug, including details regarding any error messages, version, and platform.

Component(s)

kou commented Feb 5, 2024

amoeba commented Feb 6, 2024

kou commented Feb 6, 2024

amoeba commented Feb 6, 2024

amoeba commented Feb 6, 2024 • edited Loading

paleolimbot commented Feb 6, 2024

amoeba commented Feb 6, 2024

jorisvandenbossche commented Feb 6, 2024

paleolimbot commented Feb 6, 2024

paleolimbot commented Feb 6, 2024

pitrou commented Feb 6, 2024 • edited Loading

jorisvandenbossche commented Feb 6, 2024

pitrou commented Feb 6, 2024

paleolimbot commented Feb 6, 2024

pitrou commented Feb 6, 2024

amoeba commented Feb 6, 2024 •

edited

Loading

pitrou commented Feb 6, 2024 •

edited

Loading