Transpiler WIP by sbcgua · Pull Request #23 · sbcgua/text2tab

sbcgua · 2021-10-14T12:34:15Z

No description provided.

larshp · 2021-10-14T14:03:01Z

@abaplint/transpiler-cli instead of @abaplint/transpiler

larshp · 2021-10-14T15:17:27Z

raise conv_failed.

raising non class based exceptions is not implemented, and I dont think it will be anytime soon

suggest refactoring to a class based exception

sbcgua · 2021-10-14T15:53:07Z

conv_failed

Hmm, yeah, I don't remember why I did like that in the first place

sbcgua · 2021-10-14T16:01:16Z

Nice, it compiles !

sbcgua · 2021-10-14T17:08:21Z

Looks like an operator is missing ?

    ro_parser.get().mt_components.set(await abap.Classes['ZCL_TEXT2TAB_UTILS'].describe_struct({i_struc: ro_parser.get().mo_struc_descr, i_is_deep: abap.builtin.boolc(abap.compare.initial(i_deep_provider) === false), i_ignore_nonflat: i_ignore_nonflat}));
    if ((abap.compare.initial(i_amount_format)abap.compare.initial(i_amount_format.getOffset({offset: 1, length: 1})))) {
      ro_parser.get().mv_amount_format.set(i_amount_format);
    }

larshp · 2021-10-14T17:15:46Z

yea, looks like a simple upper/lower case problem

larshp · 2021-10-14T17:24:41Z

@@ -0,0 +1,25 @@
+{
+  "name": "text2tab",
+  "version": "1.0.0",


a trick: you can set private: true so that it cannot accidently be published to npm

larshp · 2021-10-15T04:45:05Z

next problem: Void type: RS38L_FNAM

Perhaps type as TFDIR-FUNCNAME instead, TFDIR will probably be added to open-abap sometime. RS38L_FNAM is a strange data element name

sbcgua · 2021-10-15T13:06:30Z

Actually funcname DE is a good replacement. Part of SABP pacakge

TFDIR refers RS38L_FNAM too in fact ...

sbcgua · 2023-07-29T09:57:42Z

@larshp Trying to do this again ... May I ask for some assistance ? The message says something wrong with /ui2/cl_json.clas which is not used anyhow ... :/

larshp · 2023-07-29T10:32:39Z

    "@abaplint/cli": "^2.79.29",
-    "@abaplint/runtime": "^1.6.47",
+    "@abaplint/runtime": "^1.8.40",
    "@abaplint/transpiler-cli": "^1.6.47"


try updating these to latest

larshp · 2023-07-29T10:37:11Z

anyhow, next problem to tackle is VERI_ALPHA, not really sure what to use instead

ideally something that is marked as released, https://abapedia.org/steampunk-2305-api/

sbcgua · 2023-07-29T10:45:17Z

update didn't help: same error:

$node transpiled/index.mjs
file:///.../text2tab/transpiled/%23ui2%23cl_json.clas.locals.mjs:281
lcl_parser.ty_data_tt = abap.types.TableFactory.construct(new abap.types.Structure({"parent": new abap.types.String({qualifiedName: "LCL_PARSER=>TY_DATA-PARENT"}), "name": new abap.types.String({qualifiedName: "LCL_PARSER=>TY_DATA-NAME"}), "full_name": new abap.types.String({qualifiedName: "LCL_PARSER=>TY_DATA-FULL_NAME"}), "value": new abap.types.String({qualifiedName: "LCL_PARSER=>TY_DATA-VALUE"}), "type": new abap.types.String({qualifiedName: "LCL_PARSER=>TY_DATA-TYPE"})}, "lcl_parser=>ty_data"), {"withHeader":false,"keyType":"DEFAULT","primaryKey":{"name":"primary_key","type":"STANDARD","isUnique":false,"keyFields":[]},"secondary":[]}, "lcl_parser=>ty_data_tt");
                                                ^

TypeError: Cannot read properties of undefined (reading 'construct')
    at file:///.../text2tab/transpiled/%23ui2%23cl_json.clas.locals.mjs:281:49

... And how do you see from it, that it is about VERI_ALPHA ?

sbcgua · 2023-07-29T10:50:27Z

:/ I guess the only way is to create own DOMA ...

larshp · 2023-07-29T10:55:34Z

ALPHA: the xml generation might be bugged and remove the conversion routine

package.json: latest @abaplint/runtime is 2.7.54

sbcgua · 2023-07-29T10:58:38Z

Added domain/dtel to "deps" but no effect - same error

larshp · 2023-07-29T11:08:42Z

try updating package.json to latest versions of the dependencies

sbcgua · 2023-07-29T12:09:48Z

try updating package.json to latest versions of the dependencies

Yes, it's done 957b209

sbcgua · 2023-07-29T12:14:41Z

Ah no, sorry (need to get rid of the package json in favor of script, it just makes things worse).
But there is now another error

$node transpiled/index.mjs
Void type: FLOAT

sbcgua · 2023-07-29T12:19:20Z

OK, now it is some issue with conv_exit

ZCL_TEXT2TAB_PARSER: running ltcl_text2tab_parser_test->create
ZCL_TEXT2TAB_PARSER: running ltcl_text2tab_parser_test->apply_conv_exit
<ref *1> kernel_cx_assert [Error]
    at cl_abap_unit_assert.assert_equals (file:///C:/Users/at/Documents/devs/Abap/text2tab/transpiled/cl_abap_unit_assert.clas.mjs:355:32)
    at async ltcl_text2tab_parser_test.apply_conv_exit (file:///C:/Users/at/Documents/devs/Abap/text2tab/transpiled/zcl_text2tab_parser.clas.testclasses.mjs:312:5)
    at async run (file:///C:/Users/at/Documents/devs/Abap/text2tab/transpiled/index.mjs:45:9) {
  me: ABAPObject {
    qualifiedName: undefined,
    RTTIName: undefined,
    value: [Circular *1]
  },
  previous: ABAPObject {
    qualifiedName: 'CX_ROOT',
    RTTIName: '\\CLASS=CX_ROOT',
    value: undefined
  },
  textid: Character {
    constant: false,
    length: 32,
    extra: {},
    value: '                                '
  },
  get_longtext: [AsyncFunction: if_message$get_longtext],
  get_text: [AsyncFunction: if_message$get_text],
  actual: String { value: '123', qualifiedName: 'STRING' },
  expected: String { value: '0000000123', qualifiedName: 'STRING' },
  msg: String {
    value: "Expected '0000000123', got '123'",
    qualifiedName: 'STRING'
  },
  EXTRA_CX: {
    INTERNAL_FILENAME: 'cl_abap_unit_assert.clas.abap',
    INTERNAL_LINE: 349
  }
}

larshp · 2023-07-29T12:31:11Z

reproduced in unit test in open-abap/open-abap-core#720

larshp · 2023-07-29T12:39:55Z

try again

larshp · 2023-07-29T12:59:21Z

Void type: SYS_CALLS, suggest using type abap_callstack instead

sbcgua · 2023-07-30T15:46:05Z

Hmmm

  actual: String { value: '', qualifiedName: 'STRING' },
  expected: String { value: '', qualifiedName: 'STRING' },

larshp · 2023-07-30T16:03:47Z

the code calls fail(), so there is no actual and expected,

suggest adding the error code to the text in:

larshp · 2023-08-09T06:04:38Z

the classes are not in open-abap yet, and there is a nice but to fix in the transpiler to give a better error message in this scenario

larshp · 2023-08-09T06:39:08Z

abaplint/transpiler#1295

larshp · 2023-08-09T06:45:25Z

open-abap/open-abap-core@937cd4b part of open-abap/open-abap-core#737

larshp · 2023-08-09T06:52:53Z

btw, added abaplint/abaplint@b9c8242 to https://rules.abaplint.org/function_module_recommendations/

larshp · 2023-08-09T07:03:02Z

add these two skips,

      { "object": "ZCL_TEXT2TAB_PARSER", "class": "ltcl_text2tab_parser_test", "method": "parse_df16",  "note": "TypeError: abap.types.typeTodoDecFloat16Type is not a constructor" },
      { "object": "ZCL_TEXT2TAB_PARSER", "class": "ltcl_text2tab_parser_test", "method": "parse_ignore_convexit",  "note": "CUNIT issue" },

not feeling like implementing DecFloat16 this week..

sbcgua · 2023-08-09T07:52:00Z

:)

Done, though parse_ignore_convexit should theoretically work ... because the idea is to don't run the exit ... but can be checked later

larshp · 2023-08-10T05:12:08Z

current problem reproduced in open-abap/open-abap-core@2afe78a

larshp · 2023-08-10T06:42:27Z

two places where DMBTR is used,

sbcgua · 2023-08-10T15:31:40Z

fixed.

P.S. is there any way to make the error output less cryptic? I can potentially contribute ... though ... I'm not sure I understand the output enough. E.g. these look the same

  actual: String {
    value: 'name: MANDT, type: [object Object], as_include: , suffix: ',
    qualifiedName: 'STRING'
  },
  expected: String {
    value: 'name: MANDT, type: [object Object], as_include: , suffix: ',
    qualifiedName: 'STRING'
  },

larshp · 2023-08-10T15:58:42Z

fixed.

P.S. is there any way to make the error output less cryptic? I can potentially contribute ... though ... I'm not sure I understand the output enough. E.g. these look the same
  actual: String {
    value: 'name: MANDT, type: [object Object], as_include: , suffix: ',
    qualifiedName: 'STRING'
  },
  expected: String {
    value: 'name: MANDT, type: [object Object], as_include: , suffix: ',
    qualifiedName: 'STRING'
  },

for the current issue, the commented testcases are in open-abap/open-abap-core#739

problem is the RTTI in open-abap doesnt return identical object references for identical types

I implemented cl_abap_weak_reference earlier today, think I'm gonna make the RTTI classes serializable, instantiate everything, and compare with a weak reference cache, but not sure, its difficult

a bit less cryptic, but still cryptic message fixed in https://github.com/open-abap/open-abap-core/pull/739/files#diff-c761d101268df6f7b95ee1c9f81a84d0c9bbc2d15a262804315573b258e13439

sbcgua · 2023-08-22T06:38:37Z

Hi Lars, any progress with that issue? Can I do anything from my side?

larshp · 2023-08-24T05:09:21Z

not really sure what is the best way to fix the issue, perhaps just disable those testcases for now

larshp mentioned this pull request Oct 14, 2021

enable use class based exceptions in validation? abaplint/transpiler#541

Closed

sbcgua force-pushed the transpiler branch from 949ae79 to d3a20f0 Compare October 14, 2021 15:59

larshp mentioned this pull request Oct 14, 2021

OR upper/lower case? abaplint/transpiler#543

Closed

larshp reviewed Oct 14, 2021

View reviewed changes

sbcgua force-pushed the transpiler branch 2 times, most recently from f71c8b9 to 9790a37 Compare July 29, 2023 09:48

larshp reviewed Jul 29, 2023

View reviewed changes

sbcgua added 16 commits November 7, 2023 21:08

initial config

fd59d9e

add actions

989ebe6

fix transpiler

ac1370d

adjust dir

e831b36

adjust ci

d26b2a7

update

9e02ba8

dep domain and dtel

e8a3e53

update again

515e522

more ci

96c7803

update

b387005

pkg update and sy-sybrc fixes

84eef06

next

e7a9267

skips

97b9d0e

npm updates

4fcc3b5

excludes

280a380

update packages

e938802

sbcgua force-pushed the transpiler branch from bb6b770 to e938802 Compare November 7, 2023 19:08

update pkg

f2d1220

Conversation

sbcgua commented Oct 14, 2021

Uh oh!

larshp commented Oct 14, 2021

Uh oh!

larshp commented Oct 14, 2021

Uh oh!

sbcgua commented Oct 14, 2021

Uh oh!

sbcgua commented Oct 14, 2021

Uh oh!

sbcgua commented Oct 14, 2021

Uh oh!

larshp commented Oct 14, 2021

Uh oh!

larshp Oct 14, 2021

Choose a reason for hiding this comment

Uh oh!

sbcgua Oct 14, 2021

Choose a reason for hiding this comment

Uh oh!

larshp commented Oct 15, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sbcgua commented Oct 15, 2021

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

larshp Jul 29, 2023

Choose a reason for hiding this comment

Uh oh!

larshp commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

larshp commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

larshp commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 29, 2023

Uh oh!

larshp commented Jul 29, 2023

Uh oh!

larshp commented Jul 29, 2023

Uh oh!

larshp commented Jul 29, 2023

Uh oh!

sbcgua commented Jul 30, 2023

Uh oh!

larshp commented Jul 30, 2023

Uh oh!

larshp commented Aug 9, 2023

Uh oh!

larshp commented Aug 9, 2023

Uh oh!

larshp commented Aug 9, 2023

Uh oh!

larshp commented Aug 9, 2023

Uh oh!

larshp commented Aug 9, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sbcgua commented Aug 9, 2023

Uh oh!

larshp commented Aug 10, 2023

Uh oh!

larshp commented Aug 10, 2023

Uh oh!

sbcgua commented Aug 10, 2023

Uh oh!

larshp commented Aug 10, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

larshp commented Oct 15, 2021 •

edited

Loading

larshp commented Aug 9, 2023 •

edited

Loading

larshp commented Aug 10, 2023 •

edited

Loading