DEV: Sync libasr with LFortran #2827

czgdp1807 · 2025-02-26T06:39:26Z

Steps to complete,

Comment out tests to pass the integration tests.
Then uncomment dict, set, list ones and keep the changes which are needed to pass them.
Rest of them keep commented out.
Merge.

certik · 2025-02-26T07:05:32Z

src/libasr/codegen/asr_to_llvm.cpp

-        strings_to_be_deallocated.p = strings_to_be_deallocated_copy;
-    }
-
-    void visit_ForEach(const ASR::ForEach_t &x) {


Not sure if this function / node was removed? It might be a specific node in LPython.

certik · 2025-02-26T07:07:31Z

I think there was work in LPython to get various dicts and lists operations working, so it would be good to keep that if we can.

But I think it is more important to sync, and keep it synced after that, so if it is difficult, we can temporarily remove it, and try to add it back later (the proper way: in LFortran first, then update a submodule).

assem2002 · 2025-02-26T07:19:24Z

src/libasr/ASR.asdl

@@ -45,23 +45,23 @@ stmt
    | GoToTarget(int id, identifier name)
    | If(expr test, stmt* body, stmt* orelse)
    | IfArithmetic(expr test, int lt_label, int eq_label, int gt_label)
-    | Print(expr* values, expr? separator, expr? end)


I think end and separator are used in python syntax.

They definitely are used in python syntax, but the ASR is supposed to be language agnostic, so I'm not really sure if we should've programming language details into ASR. Handle them in the frontend, and if it isn't possible, then have it in ASR.asdl of LFortran as well.

assem2002 · 2025-02-26T07:32:25Z

src/libasr/codegen/asr_to_c.cpp

-        if (x.m_separator) {
-            this->visit_expr(*x.m_separator);
-            separator = src;
+        separator = "\" \"";
+        //HACKISH way to handle print refactoring (always using stringformat).
+        // TODO : Implement stringformat visitor.


let's keep combination of this as we should keep separator and end memebers.

assem2002 · 2025-02-26T07:33:24Z

src/libasr/codegen/asr_to_c.cpp

-        if (x.m_end) {
-            this->visit_expr(*x.m_end);
-            tmp_gen += "\%s\"";
-            v.push_back(src);
-        } else {
-            tmp_gen += "\\n\"";
-        }


as well as keeping this.

czgdp1807 · 2025-03-25T06:29:14Z

If you merge it then do squash and merge.

Pranavchiku · 2025-03-25T14:21:55Z

src/libasr/asr_base_visitor.h

@@ -0,0 +1,551 @@
+#pragma once


this file should be automatically generated. Please add it to .gitignore.

Pranavchiku · 2025-03-25T14:25:51Z

src/libasr/asr_deserialization_visitor.h

@@ -0,0 +1,4688 @@
+#pragma once


same, this has to be generated automatically

czgdp1807 · 2025-03-25T14:32:09Z

I would be busy. Please push the changes directly to my branch.

Pranavchiku

I'm OK with these changes. I think sync will be complete once we get these tests one-by-one working.

gxyd · 2025-03-25T13:49:33Z

integration_tests/array_expr_05.py

@@ -1,4 +1,4 @@
-from lpython import u8, u16, u32, u64
+from lpython import u8, u16, u32, u64, i8


Are the changes in this file needed for tests to pass?

gxyd · 2025-03-25T15:54:07Z

src/libasr/asr_expr_base_replacer_visitor.h

+
+// Generated by grammar/asdl_cpp.py
+
+#include <libasr/alloc.h>


I'm commenting here to make sure that we don't commit this either. So that one of us can undo it.

gxyd · 2025-03-25T15:54:49Z

src/libasr/asr_expr_call_replacer_visitor.h

this one as well.

gxyd · 2025-03-25T15:55:38Z

src/libasr/asr_expr_stmt_duplicator_visitor.h

this one as well.

gxyd · 2025-03-25T15:55:59Z

src/libasr/asr_expr_type_visitor.h

this one as well.

gxyd · 2025-03-25T16:00:43Z

src/libasr/asr_lookup_name_visitor.h

maybe this as well.

gxyd · 2025-03-25T16:01:45Z

src/libasr/asr_pass_walk_visitor.h

maybe this as well.

gxyd · 2025-03-25T16:05:41Z

src/libasr/asr_pickle_visitor.h

maybe this as well

gxyd · 2025-03-25T16:08:02Z

src/libasr/asr_serialization_visitor.h

maybe this as well.

gxyd · 2025-03-25T16:09:29Z

src/libasr/asr_stmt_base_replacer_visitor.h

@@ -0,0 +1,569 @@
+#pragma once


maybe this as well.

certik reviewed Feb 26, 2025

View reviewed changes

assem2002 reviewed Feb 26, 2025

View reviewed changes

czgdp1807 force-pushed the libasr_sync branch 3 times, most recently from ce689a9 to ede6e7f Compare February 27, 2025 13:59

czgdp1807 added 2 commits March 9, 2025 18:18

DEV: Sync libasr with LFortran

fcd554f

TEST: Comment out tests

dd2f4cc

czgdp1807 force-pushed the libasr_sync branch from ede6e7f to dd2f4cc Compare March 13, 2025 14:02

czgdp1807 added 14 commits March 13, 2025 19:34

Merge main

8badcba

comment out tests

51be85a

Fixed warnings

ee1d451

Merge main

dd04cb2

Comment out failing tests

a571fd3

Comment out c_py tests

16c5d47

Comment out tests

8bad3df

Comment out fast tests

7a05684

Update tests

4a8dcde

Updated reference tests

6717e4d

Comment out test

fec8a12

Comment out tests

f0ef435

Comment out more tests

b3eb353

Comment out tests

cae555e

czgdp1807 requested a review from certik March 24, 2025 17:31

Pranavchiku reviewed Mar 25, 2025

View reviewed changes

src/libasr/asr_deserialization_visitor.h

@@ -0,0 +1,4688 @@

#pragma once

Copy link

Collaborator

Pranavchiku Mar 25, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same, this has to be generated automatically

Pranavchiku reviewed Mar 25, 2025

View reviewed changes

gxyd reviewed Mar 25, 2025

View reviewed changes

certik mentioned this pull request Mar 26, 2025

Finish LPython sync lfortran/lfortran#6739

Open

5 tasks

assem2002 mentioned this pull request Mar 27, 2025

enh: add LFortran as submodule with sparse checkout for src/libasr #2832

Closed

certik merged commit cae555e into lcompilers:main Mar 30, 2025
18 checks passed

		@@ -1,4 +1,4 @@
		from lpython import u8, u16, u32, u64
		from lpython import u8, u16, u32, u64, i8


		// Generated by grammar/asdl_cpp.py

		#include <libasr/alloc.h>

DEV: Sync libasr with LFortran #2827

DEV: Sync libasr with LFortran #2827

Uh oh!

Conversation

czgdp1807 commented Feb 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

certik commented Feb 26, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

czgdp1807 commented Mar 25, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

czgdp1807 commented Mar 25, 2025

Uh oh!

Pranavchiku left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

czgdp1807 commented Feb 26, 2025 •

edited

Loading